From d8b5beb0b0a5cb3ec3ea20e9fff415057dcf25f6 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 28 Jan 2022 08:18:01 +0100 Subject: [PATCH 001/661] Fix2 `select_device()` for Multi-GPU (#6461) * Fix2 select_device() for Multi-GPU * Cleanup * Cleanup * Simplify error message * Improve assert * Update torch_utils.py --- utils/datasets.py | 6 +++--- utils/torch_utils.py | 8 ++++---- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/utils/datasets.py b/utils/datasets.py index 4eb444087860..07f6321e0285 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -29,13 +29,12 @@ from utils.augmentations import Albumentations, augment_hsv, copy_paste, letterbox, mixup, random_perspective from utils.general import (LOGGER, NUM_THREADS, check_dataset, check_requirements, check_yaml, clean_str, segments2boxes, xyn2xy, xywh2xyxy, xywhn2xyxy, xyxy2xywhn) -from utils.torch_utils import device_count, torch_distributed_zero_first +from utils.torch_utils import torch_distributed_zero_first # Parameters HELP_URL = 'https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' IMG_FORMATS = ['bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp'] # include image suffixes VID_FORMATS = ['asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'wmv'] # include video suffixes -DEVICE_COUNT = max(device_count(), 1) # number of CUDA devices # Get orientation exif tag for orientation in ExifTags.TAGS.keys(): @@ -110,7 +109,8 @@ def create_dataloader(path, imgsz, batch_size, stride, single_cls=False, hyp=Non prefix=prefix) batch_size = min(batch_size, len(dataset)) - nw = min([os.cpu_count() // DEVICE_COUNT, batch_size if batch_size > 1 else 0, workers]) # number of workers + nd = torch.cuda.device_count() # number of CUDA devices + nw = min([os.cpu_count() // max(nd, 1), batch_size if batch_size > 1 else 0, workers]) # number of workers sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle) loader = DataLoader if image_weights else InfiniteDataLoader # only DataLoader allows for attribute updates return loader(dataset, diff --git a/utils/torch_utils.py b/utils/torch_utils.py index d958a8951074..2b51821a3b62 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -54,7 +54,8 @@ def git_describe(path=Path(__file__).parent): # path must be a directory def device_count(): - # Returns number of CUDA devices available. Safe version of torch.cuda.device_count(). + # Returns number of CUDA devices available. Safe version of torch.cuda.device_count(). Only works on Linux. + assert platform.system() == 'Linux', 'device_count() function only works on Linux' try: cmd = 'nvidia-smi -L | wc -l' return int(subprocess.run(cmd, shell=True, capture_output=True, check=True).stdout.decode().split()[-1]) @@ -70,10 +71,9 @@ def select_device(device='', batch_size=0, newline=True): if cpu: os.environ['CUDA_VISIBLE_DEVICES'] = '-1' # force torch.cuda.is_available() = False elif device: # non-cpu device requested - nd = device_count() # number of CUDA devices - assert nd > int(max(device.split(','))), f'Invalid `--device {device}` request, valid devices are 0 - {nd - 1}' os.environ['CUDA_VISIBLE_DEVICES'] = device # set environment variable - must be before assert is_available() - assert torch.cuda.is_available(), 'CUDA is not available, use `--device cpu` or do not pass a --device' + assert torch.cuda.is_available() and torch.cuda.device_count() >= len(device.replace(',', '')), \ + f"Invalid CUDA '--device {device}' requested, use '--device cpu' or pass valid CUDA device(s)" cuda = not cpu and torch.cuda.is_available() if cuda: From 7539cd75c3a6c06d00848617f6265f39a765ccea Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 28 Jan 2022 20:23:17 +0100 Subject: [PATCH 002/661] Add Product Hunt social media icon (#6464) * Social media icons update * fix URL * Update README.md --- README.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/README.md b/README.md index a73ba2797b1b..f9947b98557d 100644 --- a/README.md +++ b/README.md @@ -27,6 +27,10 @@ + + + + @@ -282,6 +286,10 @@ professional support requests please visit [https://ultralytics.com/contact](htt + + + + From 6445a8137e87f67cf3275c70e3585f634260417b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 1 Feb 2022 15:54:51 +0100 Subject: [PATCH 003/661] Resolve dataset paths (#6489) --- utils/general.py | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/utils/general.py b/utils/general.py index e9f5ec2ac128..86e3b3c1c54b 100755 --- a/utils/general.py +++ b/utils/general.py @@ -394,12 +394,15 @@ def check_dataset(data, autodownload=True): with open(data, errors='ignore') as f: data = yaml.safe_load(f) # dictionary - # Parse yaml - path = extract_dir or Path(data.get('path') or '') # optional 'path' default to '.' + # Resolve paths + path = Path(extract_dir or data.get('path') or '') # optional 'path' default to '.' + if not path.is_absolute(): + path = (ROOT / path).resolve() for k in 'train', 'val', 'test': if data.get(k): # prepend path data[k] = str(path / data[k]) if isinstance(data[k], str) else [str(path / x) for x in data[k]] + # Parse yaml assert 'nc' in data, "Dataset 'nc' key missing." if 'names' not in data: data['names'] = [f'class{i}' for i in range(data['nc'])] # assign class names if missing From b884ea36c469d8501aa4016bf76cccfc3168ccd9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 1 Feb 2022 21:17:56 +0100 Subject: [PATCH 004/661] Simplify TF normalized to pixels (#6494) --- models/common.py | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/models/common.py b/models/common.py index 346fa37ae2d0..557163310e74 100644 --- a/models/common.py +++ b/models/common.py @@ -446,10 +446,7 @@ def forward(self, im, augment=False, visualize=False, val=False): if int8: scale, zero_point = output['quantization'] y = (y.astype(np.float32) - zero_point) * scale # re-scale - y[..., 0] *= w # x - y[..., 1] *= h # y - y[..., 2] *= w # w - y[..., 3] *= h # h + y[..., :4] *= [w, h, w, h] # xywh normalized to pixels y = torch.tensor(y) if isinstance(y, np.ndarray) else y return (y, []) if val else y From 5e4ff195b21816d96b1fe0a94a9670a7e2ad34e2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 1 Feb 2022 22:06:29 +0100 Subject: [PATCH 005/661] Improved `export.py` usage examples (#6495) * Improved `export.py` usage examples * Cleanup --- export.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/export.py b/export.py index 589b381e035a..bb17703821e8 100644 --- a/export.py +++ b/export.py @@ -469,10 +469,10 @@ def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path' f = [str(x) for x in f if x] # filter out '' and None LOGGER.info(f'\nExport complete ({time.time() - t:.2f}s)' f"\nResults saved to {colorstr('bold', file.parent.resolve())}" - f"\nVisualize with https://netron.app" - f"\nDetect with `python detect.py --weights {f[-1]}`" - f" or `model = torch.hub.load('ultralytics/yolov5', 'custom', '{f[-1]}')" - f"\nValidate with `python val.py --weights {f[-1]}`") + f"\nDetect: python detect.py --weights {f[-1]}" + f"\nPyTorch Hub: model = torch.hub.load('ultralytics/yolov5', 'custom', '{f[-1]}')" + f"\nValidate: python val.py --weights {f[-1]}" + f"\nVisualize: https://netron.app") return f # return list of exported files/dirs From 77977e07912768738ef7ca46f44f19b6959206d9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 1 Feb 2022 22:34:15 +0100 Subject: [PATCH 006/661] CoreML inference fix `list()` -> `sorted()` (#6496) --- models/common.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/common.py b/models/common.py index 557163310e74..29d02e741e17 100644 --- a/models/common.py +++ b/models/common.py @@ -427,7 +427,7 @@ def forward(self, im, augment=False, visualize=False, val=False): conf, cls = y['confidence'].max(1), y['confidence'].argmax(1).astype(np.float) y = np.concatenate((box, conf.reshape(-1, 1), cls.reshape(-1, 1)), 1) else: - y = y[list(y)[-1]] # last output + y = y[sorted(y)[-1]] # last output else: # TensorFlow (SavedModel, GraphDef, Lite, Edge TPU) im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) if self.saved_model: # SavedModel From 842d049e1bbe5db87ad36f4ba86e1a9c2b6e413a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 1 Feb 2022 22:59:26 +0100 Subject: [PATCH 007/661] Suppress `torch.jit.TracerWarning` on export (#6498) * Suppress torch.jit.TracerWarning TracerWarnings can be safely ignored. * Cleanup --- export.py | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/export.py b/export.py index bb17703821e8..8666f3de63e0 100644 --- a/export.py +++ b/export.py @@ -45,6 +45,7 @@ import subprocess import sys import time +import warnings from pathlib import Path import torch @@ -508,8 +509,10 @@ def parse_opt(): def main(opt): - for opt.weights in (opt.weights if isinstance(opt.weights, list) else [opt.weights]): - run(**vars(opt)) + with warnings.catch_warnings(): + warnings.filterwarnings(action='ignore', category=torch.jit.TracerWarning) # suppress TracerWarning + for opt.weights in (opt.weights if isinstance(opt.weights, list) else [opt.weights]): + run(**vars(opt)) if __name__ == "__main__": From 4c409332667477560200958b513b958bb8fdef71 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 1 Feb 2022 23:52:50 +0100 Subject: [PATCH 008/661] Suppress export.run() TracerWarnings (#6499) Suppresses warnings when calling export.run() directly, not just CLI python export.py. Also adds Requirements examples for CPU and GPU backends --- export.py | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/export.py b/export.py index 8666f3de63e0..09c50baa415a 100644 --- a/export.py +++ b/export.py @@ -16,6 +16,10 @@ TensorFlow Edge TPU | `edgetpu` | yolov5s_edgetpu.tflite TensorFlow.js | `tfjs` | yolov5s_web_model/ +Requirements: + $ pip install -r requirements.txt coremltools onnx onnx-simplifier onnxruntime openvino-dev tensorflow-cpu # CPU + $ pip install -r requirements.txt coremltools onnx onnx-simplifier onnxruntime-gpu openvino-dev tensorflow # GPU + Usage: $ python path/to/export.py --weights yolov5s.pt --include torchscript onnx openvino engine coreml tflite ... @@ -437,6 +441,7 @@ def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path' # Exports f = [''] * 10 # exported filenames + warnings.filterwarnings(action='ignore', category=torch.jit.TracerWarning) # suppress TracerWarning if 'torchscript' in include: f[0] = export_torchscript(model, im, file, optimize) if 'engine' in include: # TensorRT required before ONNX @@ -509,10 +514,8 @@ def parse_opt(): def main(opt): - with warnings.catch_warnings(): - warnings.filterwarnings(action='ignore', category=torch.jit.TracerWarning) # suppress TracerWarning - for opt.weights in (opt.weights if isinstance(opt.weights, list) else [opt.weights]): - run(**vars(opt)) + for opt.weights in (opt.weights if isinstance(opt.weights, list) else [opt.weights]): + run(**vars(opt)) if __name__ == "__main__": From b73c62ebc5180d1fa3b412e55ab831d8285e1673 Mon Sep 17 00:00:00 2001 From: Ayush Chaurasia Date: Thu, 3 Feb 2022 15:59:52 +0530 Subject: [PATCH 009/661] W&B: Remember batchsize on resuming (#6512) * log best.pt metrics at train end * update * Update __init__.py * flush callbacks when using evolve * remember batch size on resuming * Update train.py Co-authored-by: Glenn Jocher --- train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train.py b/train.py index 510377e1178e..2a973fb7164b 100644 --- a/train.py +++ b/train.py @@ -96,7 +96,7 @@ def train(hyp, # path/to/hyp.yaml or hyp dictionary if loggers.wandb: data_dict = loggers.wandb.data_dict if resume: - weights, epochs, hyp = opt.weights, opt.epochs, opt.hyp + weights, epochs, hyp, batch_size = opt.weights, opt.epochs, opt.hyp, opt.batch_size # Register actions for k in methods(loggers): From 19e0208fc9e33010717e066f9bd65c27db7c2b5c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 3 Feb 2022 12:15:13 +0100 Subject: [PATCH 010/661] Update hyp.scratch-high.yaml (#6525) Update `lrf: 0.1`, tested on YOLOv5x6 to 55.0 mAP@0.5:0.95, slightly higher than current. --- data/hyps/hyp.scratch-high.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/data/hyps/hyp.scratch-high.yaml b/data/hyps/hyp.scratch-high.yaml index 5a586cc63fae..123cc8407413 100644 --- a/data/hyps/hyp.scratch-high.yaml +++ b/data/hyps/hyp.scratch-high.yaml @@ -4,7 +4,7 @@ # See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials lr0: 0.01 # initial learning rate (SGD=1E-2, Adam=1E-3) -lrf: 0.2 # final OneCycleLR learning rate (lr0 * lrf) +lrf: 0.1 # final OneCycleLR learning rate (lr0 * lrf) momentum: 0.937 # SGD momentum/Adam beta1 weight_decay: 0.0005 # optimizer weight decay 5e-4 warmup_epochs: 3.0 # warmup epochs (fractions ok) From cb40c9afda52a149b49d5e8d06100c60f6cd1614 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 3 Feb 2022 18:11:28 +0100 Subject: [PATCH 011/661] TODO issues exempt from stale action (#6530) --- .github/workflows/stale.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/stale.yml b/.github/workflows/stale.yml index be2b0d97d5e7..7a83950c17b7 100644 --- a/.github/workflows/stale.yml +++ b/.github/workflows/stale.yml @@ -34,5 +34,5 @@ jobs: stale-pr-message: 'This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions YOLOv5 🚀 and Vision AI ⭐.' days-before-stale: 30 days-before-close: 5 - exempt-issue-labels: 'documentation,tutorial' + exempt-issue-labels: 'documentation,tutorial,TODO' operations-per-run: 100 # The maximum number of operations per run, used to control rate limiting. From c3e599cfda112455d69da0fea64faadfaeaedcf2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 3 Feb 2022 19:09:24 +0100 Subject: [PATCH 012/661] Update val_batch*.jpg for Chinese fonts (#6526) * Update plots for Chinese fonts * make is_chinese() non-str safe * Add global FONT * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update general.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/general.py | 71 ++++++++++++++++++++++++++++-------------------- utils/plots.py | 23 +++++++--------- 2 files changed, 52 insertions(+), 42 deletions(-) diff --git a/utils/general.py b/utils/general.py index 86e3b3c1c54b..fce5e38c6c9e 100755 --- a/utils/general.py +++ b/utils/general.py @@ -37,6 +37,7 @@ ROOT = FILE.parents[1] # YOLOv5 root directory NUM_THREADS = min(8, max(1, os.cpu_count() - 1)) # number of YOLOv5 multiprocessing threads VERBOSE = str(os.getenv('YOLOv5_VERBOSE', True)).lower() == 'true' # global verbose mode +FONT = 'Arial.ttf' # https://ultralytics.com/assets/Arial.ttf torch.set_printoptions(linewidth=320, precision=5, profile='long') np.set_printoptions(linewidth=320, formatter={'float_kind': '{:11.5g}'.format}) # format short g, %precision=5 @@ -55,6 +56,21 @@ def is_kaggle(): return False +def is_writeable(dir, test=False): + # Return True if directory has write permissions, test opening a file with write permissions if test=True + if test: # method 1 + file = Path(dir) / 'tmp.txt' + try: + with open(file, 'w'): # open file with write permissions + pass + file.unlink() # remove file + return True + except OSError: + return False + else: # method 2 + return os.access(dir, os.R_OK) # possible issues on Windows + + def set_logging(name=None, verbose=VERBOSE): # Sets level and returns logger if is_kaggle(): @@ -68,6 +84,22 @@ def set_logging(name=None, verbose=VERBOSE): LOGGER = set_logging('yolov5') # define globally (used in train.py, val.py, detect.py, etc.) +def user_config_dir(dir='Ultralytics', env_var='YOLOV5_CONFIG_DIR'): + # Return path of user configuration directory. Prefer environment variable if exists. Make dir if required. + env = os.getenv(env_var) + if env: + path = Path(env) # use environment variable + else: + cfg = {'Windows': 'AppData/Roaming', 'Linux': '.config', 'Darwin': 'Library/Application Support'} # 3 OS dirs + path = Path.home() / cfg.get(platform.system(), '') # OS-specific config dir + path = (path if is_writeable(path) else Path('/tmp')) / dir # GCP and AWS lambda fix, only /tmp is writeable + path.mkdir(exist_ok=True) # make if required + return path + + +CONFIG_DIR = user_config_dir() # Ultralytics settings dir + + class Profile(contextlib.ContextDecorator): # Usage: @Profile() decorator or 'with Profile():' context manager def __enter__(self): @@ -152,34 +184,6 @@ def get_latest_run(search_dir='.'): return max(last_list, key=os.path.getctime) if last_list else '' -def user_config_dir(dir='Ultralytics', env_var='YOLOV5_CONFIG_DIR'): - # Return path of user configuration directory. Prefer environment variable if exists. Make dir if required. - env = os.getenv(env_var) - if env: - path = Path(env) # use environment variable - else: - cfg = {'Windows': 'AppData/Roaming', 'Linux': '.config', 'Darwin': 'Library/Application Support'} # 3 OS dirs - path = Path.home() / cfg.get(platform.system(), '') # OS-specific config dir - path = (path if is_writeable(path) else Path('/tmp')) / dir # GCP and AWS lambda fix, only /tmp is writeable - path.mkdir(exist_ok=True) # make if required - return path - - -def is_writeable(dir, test=False): - # Return True if directory has write permissions, test opening a file with write permissions if test=True - if test: # method 1 - file = Path(dir) / 'tmp.txt' - try: - with open(file, 'w'): # open file with write permissions - pass - file.unlink() # remove file - return True - except OSError: - return False - else: # method 2 - return os.access(dir, os.R_OK) # possible issues on Windows - - def is_docker(): # Is environment a Docker container? return Path('/workspace').exists() # or Path('/.dockerenv').exists() @@ -207,7 +211,7 @@ def is_ascii(s=''): def is_chinese(s='人工智能'): # Is string composed of any Chinese characters? - return re.search('[\u4e00-\u9fff]', s) + return True if re.search('[\u4e00-\u9fff]', str(s)) else False def emojis(str=''): @@ -378,6 +382,15 @@ def check_file(file, suffix=''): return files[0] # return file +def check_font(font=FONT): + # Download font to CONFIG_DIR if necessary + font = Path(font) + if not font.exists() and not (CONFIG_DIR / font.name).exists(): + url = "https://ultralytics.com/assets/" + font.name + LOGGER.info(f'Downloading {url} to {CONFIG_DIR / font.name}...') + torch.hub.download_url_to_file(url, str(font), progress=False) + + def check_dataset(data, autodownload=True): # Download and/or unzip dataset if not found locally # Usage: https://github.com/ultralytics/yolov5/releases/download/v1.0/coco128_with_yaml.zip diff --git a/utils/plots.py b/utils/plots.py index 74868403edc0..be70ac8a030f 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -17,12 +17,11 @@ import torch from PIL import Image, ImageDraw, ImageFont -from utils.general import (LOGGER, Timeout, check_requirements, clip_coords, increment_path, is_ascii, is_chinese, - try_except, user_config_dir, xywh2xyxy, xyxy2xywh) +from utils.general import (CONFIG_DIR, FONT, LOGGER, Timeout, check_font, check_requirements, clip_coords, + increment_path, is_ascii, is_chinese, try_except, xywh2xyxy, xyxy2xywh) from utils.metrics import fitness # Settings -CONFIG_DIR = user_config_dir() # Ultralytics settings dir RANK = int(os.getenv('RANK', -1)) matplotlib.rc('font', **{'size': 11}) matplotlib.use('Agg') # for writing to files only @@ -49,16 +48,14 @@ def hex2rgb(h): # rgb order (PIL) colors = Colors() # create instance for 'from utils.plots import colors' -def check_font(font='Arial.ttf', size=10): +def check_pil_font(font=FONT, size=10): # Return a PIL TrueType Font, downloading to CONFIG_DIR if necessary font = Path(font) font = font if font.exists() else (CONFIG_DIR / font.name) try: return ImageFont.truetype(str(font) if font.exists() else font.name, size) except Exception as e: # download if missing - url = "https://ultralytics.com/assets/" + font.name - LOGGER.info(f'Downloading {url} to {font}...') - torch.hub.download_url_to_file(url, str(font), progress=False) + check_font(font) try: return ImageFont.truetype(str(font), size) except TypeError: @@ -67,7 +64,7 @@ def check_font(font='Arial.ttf', size=10): class Annotator: if RANK in (-1, 0): - check_font() # download TTF if necessary + check_pil_font() # download TTF if necessary # YOLOv5 Annotator for train/val mosaics and jpgs and detect/hub inference annotations def __init__(self, im, line_width=None, font_size=None, font='Arial.ttf', pil=False, example='abc'): @@ -76,8 +73,8 @@ def __init__(self, im, line_width=None, font_size=None, font='Arial.ttf', pil=Fa if self.pil: # use PIL self.im = im if isinstance(im, Image.Image) else Image.fromarray(im) self.draw = ImageDraw.Draw(self.im) - self.font = check_font(font='Arial.Unicode.ttf' if is_chinese(example) else font, - size=font_size or max(round(sum(self.im.size) / 2 * 0.035), 12)) + self.font = check_pil_font(font='Arial.Unicode.ttf' if is_chinese(example) else font, + size=font_size or max(round(sum(self.im.size) / 2 * 0.035), 12)) else: # use cv2 self.im = im self.lw = line_width or max(round(sum(im.shape) / 2 * 0.003), 2) # line width @@ -89,10 +86,10 @@ def box_label(self, box, label='', color=(128, 128, 128), txt_color=(255, 255, 2 if label: w, h = self.font.getsize(label) # text width, height outside = box[1] - h >= 0 # label fits outside box - self.draw.rectangle([box[0], + self.draw.rectangle((box[0], box[1] - h if outside else box[1], box[0] + w + 1, - box[1] + 1 if outside else box[1] + h + 1], fill=color) + box[1] + 1 if outside else box[1] + h + 1), fill=color) # self.draw.text((box[0], box[1]), label, fill=txt_color, font=self.font, anchor='ls') # for PIL>8.0 self.draw.text((box[0], box[1] - h if outside else box[1]), label, fill=txt_color, font=self.font) else: # cv2 @@ -210,7 +207,7 @@ def plot_images(images, targets, paths=None, fname='images.jpg', names=None, max # Annotate fs = int((h + w) * ns * 0.01) # font size - annotator = Annotator(mosaic, line_width=round(fs / 10), font_size=fs, pil=True) + annotator = Annotator(mosaic, line_width=round(fs / 10), font_size=fs, pil=True, example=names) for i in range(i + 1): x, y = int(w * (i // ns)), int(h * (i % ns)) # block origin annotator.rectangle([x, y, x + w, y + h], None, (255, 255, 255), width=2) # borders From a82292ec5376cd7ff07fc6e85b731c09cdaeff4f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 3 Feb 2022 19:55:19 +0100 Subject: [PATCH 013/661] Social icons after text (#6473) * Social icons after text * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update README.md Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- README.md | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/README.md b/README.md index f9947b98557d..7bfea7c24e8f 100644 --- a/README.md +++ b/README.md @@ -13,7 +13,13 @@ Open In Kaggle Join Forum +
+

+YOLOv5 🚀 is a family of object detection architectures and models pretrained on the COCO dataset, and represents Ultralytics + open-source research into future vision AI methods, incorporating lessons learned and best practices evolved over thousands of hours of research and development. +

+ -
-

-YOLOv5 🚀 is a family of object detection architectures and models pretrained on the COCO dataset, and represents Ultralytics - open-source research into future vision AI methods, incorporating lessons learned and best practices evolved over thousands of hours of research and development. -

- From 63ddb6f0d06f6309aa42bababd08c859197a27af Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 26 Feb 2022 19:15:12 +0100 Subject: [PATCH 073/661] Update autoanchor.py (#6794) * Update autoanchor.py * Update autoanchor.py --- utils/autoanchor.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/utils/autoanchor.py b/utils/autoanchor.py index 27d6fb68bb38..51d4de306efd 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -57,9 +57,10 @@ def metric(k): # compute metric anchors = torch.tensor(anchors, device=m.anchors.device).type_as(m.anchors) m.anchors[:] = anchors.clone().view_as(m.anchors) / m.stride.to(m.anchors.device).view(-1, 1, 1) # loss check_anchor_order(m) - LOGGER.info(f'{PREFIX}New anchors saved to model. Update model *.yaml to use these anchors in the future.') + s = f'{PREFIX}Done ✅ (optional: update model *.yaml to use these anchors in the future)' else: - LOGGER.info(f'{PREFIX}Original anchors better than new anchors. Proceeding with original anchors.') + s = f'{PREFIX}Done ⚠️ (original anchors better than new anchors, proceeding with original anchors)' + LOGGER.info(emojis(s)) def kmean_anchors(dataset='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen=1000, verbose=True): @@ -120,7 +121,7 @@ def print_results(k, verbose=True): # Filter i = (wh0 < 3.0).any(1).sum() if i: - LOGGER.info(f'{PREFIX}WARNING: Extremely small objects found. {i} of {len(wh0)} labels are < 3 pixels in size.') + LOGGER.info(f'{PREFIX}WARNING: Extremely small objects found: {i} of {len(wh0)} labels are < 3 pixels in size') wh = wh0[(wh0 >= 2.0).any(1)] # filter > 2 pixels # wh = wh * (npr.rand(wh.shape[0], 1) * 0.9 + 0.1) # multiply by random scale 0-1 From bcc92e2169a233c3e974db40ddc9b496d9c29ec8 Mon Sep 17 00:00:00 2001 From: Louis Combaldieu Date: Fri, 4 Mar 2022 09:39:23 +0100 Subject: [PATCH 074/661] Update sweep.yaml (#6825) * Update sweep.yaml Changed focal loss gamma search range between 1 and 4 * Update sweep.yaml lowered the min value to match default --- utils/loggers/wandb/sweep.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/loggers/wandb/sweep.yaml b/utils/loggers/wandb/sweep.yaml index c7790d75f6b2..688b1ea0285f 100644 --- a/utils/loggers/wandb/sweep.yaml +++ b/utils/loggers/wandb/sweep.yaml @@ -88,7 +88,7 @@ parameters: fl_gamma: distribution: uniform min: 0.0 - max: 0.1 + max: 4.0 hsv_h: distribution: uniform min: 0.0 From 601dbb83f01b58355211f2565cfa4eecb48b1220 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 4 Mar 2022 10:32:18 +0100 Subject: [PATCH 075/661] AutoAnchor improved initialization robustness (#6854) * Update AutoAnchor * Update AutoAnchor * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/autoanchor.py | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/utils/autoanchor.py b/utils/autoanchor.py index 51d4de306efd..a631c21a3b26 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -125,15 +125,17 @@ def print_results(k, verbose=True): wh = wh0[(wh0 >= 2.0).any(1)] # filter > 2 pixels # wh = wh * (npr.rand(wh.shape[0], 1) * 0.9 + 0.1) # multiply by random scale 0-1 - # Kmeans calculation - LOGGER.info(f'{PREFIX}Running kmeans for {n} anchors on {len(wh)} points...') - s = wh.std(0) # sigmas for whitening - k = kmeans(wh / s, n, iter=30)[0] * s # points - if len(k) != n: # kmeans may return fewer points than requested if wh is insufficient or too similar - LOGGER.warning(f'{PREFIX}WARNING: scipy.cluster.vq.kmeans returned only {len(k)} of {n} requested points') + # Kmeans init + try: + LOGGER.info(f'{PREFIX}Running kmeans for {n} anchors on {len(wh)} points...') + assert n <= len(wh) # apply overdetermined constraint + s = wh.std(0) # sigmas for whitening + k = kmeans(wh / s, n, iter=30)[0] * s # points + assert n == len(k) # kmeans may return fewer points than requested if wh is insufficient or too similar + except Exception: + LOGGER.warning(f'{PREFIX}WARNING: switching strategies from kmeans to random init') k = np.sort(npr.rand(n * 2)).reshape(n, 2) * img_size # random init - wh = torch.tensor(wh, dtype=torch.float32) # filtered - wh0 = torch.tensor(wh0, dtype=torch.float32) # unfiltered + wh, wh0 = (torch.tensor(x, dtype=torch.float32) for x in (wh, wh0)) k = print_results(k, verbose=False) # Plot From 8a66ebad44e8ecf90c7d27757c832579398d4baf Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 4 Mar 2022 14:10:13 +0100 Subject: [PATCH 076/661] Add `*.ts` to `VID_FORMATS` (#6859) --- utils/datasets.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/datasets.py b/utils/datasets.py index e132e04f6d9d..c325b9910ed3 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -33,8 +33,8 @@ # Parameters HELP_URL = 'https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' -IMG_FORMATS = ['bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp'] # include image suffixes -VID_FORMATS = ['asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'wmv'] # include video suffixes +IMG_FORMATS = 'bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp' # include image suffixes +VID_FORMATS = 'asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'ts', 'wmv' # include video suffixes # Get orientation exif tag for orientation in ExifTags.TAGS.keys(): From 47288407450f83ccbdbd2e950bf339e30e67a181 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 6 Mar 2022 16:16:17 +0100 Subject: [PATCH 077/661] Update `--cache disk` deprecate `*_npy/` dirs (#6876) * Updates * Updates * Updates * Updates * Updates * Updates * Updates * Updates * Updates * Updates * Cleanup * Cleanup --- utils/datasets.py | 76 +++++++++++++++--------------- utils/loggers/wandb/wandb_utils.py | 2 +- val.py | 2 +- 3 files changed, 40 insertions(+), 40 deletions(-) diff --git a/utils/datasets.py b/utils/datasets.py index c325b9910ed3..6a2dc58dd6cd 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -407,19 +407,19 @@ def __init__(self, path, img_size=640, batch_size=16, augment=False, hyp=None, r # f += [p.parent / x.lstrip(os.sep) for x in t] # local to global path (pathlib) else: raise Exception(f'{prefix}{p} does not exist') - self.img_files = sorted(x.replace('/', os.sep) for x in f if x.split('.')[-1].lower() in IMG_FORMATS) + self.im_files = sorted(x.replace('/', os.sep) for x in f if x.split('.')[-1].lower() in IMG_FORMATS) # self.img_files = sorted([x for x in f if x.suffix[1:].lower() in IMG_FORMATS]) # pathlib - assert self.img_files, f'{prefix}No images found' + assert self.im_files, f'{prefix}No images found' except Exception as e: raise Exception(f'{prefix}Error loading data from {path}: {e}\nSee {HELP_URL}') # Check cache - self.label_files = img2label_paths(self.img_files) # labels + self.label_files = img2label_paths(self.im_files) # labels cache_path = (p if p.is_file() else Path(self.label_files[0]).parent).with_suffix('.cache') try: cache, exists = np.load(cache_path, allow_pickle=True).item(), True # load dict assert cache['version'] == self.cache_version # same version - assert cache['hash'] == get_hash(self.label_files + self.img_files) # same hash + assert cache['hash'] == get_hash(self.label_files + self.im_files) # same hash except Exception: cache, exists = self.cache_labels(cache_path, prefix), False # cache @@ -437,7 +437,7 @@ def __init__(self, path, img_size=640, batch_size=16, augment=False, hyp=None, r labels, shapes, self.segments = zip(*cache.values()) self.labels = list(labels) self.shapes = np.array(shapes, dtype=np.float64) - self.img_files = list(cache.keys()) # update + self.im_files = list(cache.keys()) # update self.label_files = img2label_paths(cache.keys()) # update n = len(shapes) # number of images bi = np.floor(np.arange(n) / batch_size).astype(np.int) # batch index @@ -466,7 +466,7 @@ def __init__(self, path, img_size=640, batch_size=16, augment=False, hyp=None, r s = self.shapes # wh ar = s[:, 1] / s[:, 0] # aspect ratio irect = ar.argsort() - self.img_files = [self.img_files[i] for i in irect] + self.im_files = [self.im_files[i] for i in irect] self.label_files = [self.label_files[i] for i in irect] self.labels = [self.labels[i] for i in irect] self.shapes = s[irect] # wh @@ -485,24 +485,20 @@ def __init__(self, path, img_size=640, batch_size=16, augment=False, hyp=None, r self.batch_shapes = np.ceil(np.array(shapes) * img_size / stride + pad).astype(np.int) * stride # Cache images into RAM/disk for faster training (WARNING: large datasets may exceed system resources) - self.imgs, self.img_npy = [None] * n, [None] * n + self.ims = [None] * n + self.npy_files = [Path(f).with_suffix('.npy') for f in self.im_files] if cache_images: - if cache_images == 'disk': - self.im_cache_dir = Path(Path(self.img_files[0]).parent.as_posix() + '_npy') - self.img_npy = [self.im_cache_dir / Path(f).with_suffix('.npy').name for f in self.img_files] - self.im_cache_dir.mkdir(parents=True, exist_ok=True) gb = 0 # Gigabytes of cached images - self.img_hw0, self.img_hw = [None] * n, [None] * n - results = ThreadPool(NUM_THREADS).imap(self.load_image, range(n)) + self.im_hw0, self.im_hw = [None] * n, [None] * n + fcn = self.cache_images_to_disk if cache_images == 'disk' else self.load_image + results = ThreadPool(NUM_THREADS).imap(fcn, range(n)) pbar = tqdm(enumerate(results), total=n) for i, x in pbar: if cache_images == 'disk': - if not self.img_npy[i].exists(): - np.save(self.img_npy[i].as_posix(), x[0]) - gb += self.img_npy[i].stat().st_size + gb += self.npy_files[i].stat().st_size else: # 'ram' - self.imgs[i], self.img_hw0[i], self.img_hw[i] = x # im, hw_orig, hw_resized = load_image(self, i) - gb += self.imgs[i].nbytes + self.ims[i], self.im_hw0[i], self.im_hw[i] = x # im, hw_orig, hw_resized = load_image(self, i) + gb += self.ims[i].nbytes pbar.desc = f'{prefix}Caching images ({gb / 1E9:.1f}GB {cache_images})' pbar.close() @@ -512,8 +508,8 @@ def cache_labels(self, path=Path('./labels.cache'), prefix=''): nm, nf, ne, nc, msgs = 0, 0, 0, 0, [] # number missing, found, empty, corrupt, messages desc = f"{prefix}Scanning '{path.parent / path.stem}' images and labels..." with Pool(NUM_THREADS) as pool: - pbar = tqdm(pool.imap(verify_image_label, zip(self.img_files, self.label_files, repeat(prefix))), - desc=desc, total=len(self.img_files)) + pbar = tqdm(pool.imap(verify_image_label, zip(self.im_files, self.label_files, repeat(prefix))), + desc=desc, total=len(self.im_files)) for im_file, lb, shape, segments, nm_f, nf_f, ne_f, nc_f, msg in pbar: nm += nm_f nf += nf_f @@ -530,8 +526,8 @@ def cache_labels(self, path=Path('./labels.cache'), prefix=''): LOGGER.info('\n'.join(msgs)) if nf == 0: LOGGER.warning(f'{prefix}WARNING: No labels found in {path}. See {HELP_URL}') - x['hash'] = get_hash(self.label_files + self.img_files) - x['results'] = nf, nm, ne, nc, len(self.img_files) + x['hash'] = get_hash(self.label_files + self.im_files) + x['results'] = nf, nm, ne, nc, len(self.im_files) x['msgs'] = msgs # warnings x['version'] = self.cache_version # cache version try: @@ -543,7 +539,7 @@ def cache_labels(self, path=Path('./labels.cache'), prefix=''): return x def __len__(self): - return len(self.img_files) + return len(self.im_files) # def __iter__(self): # self.count = -1 @@ -622,17 +618,15 @@ def __getitem__(self, index): img = img.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB img = np.ascontiguousarray(img) - return torch.from_numpy(img), labels_out, self.img_files[index], shapes + return torch.from_numpy(img), labels_out, self.im_files[index], shapes def load_image(self, i): - # loads 1 image from dataset index 'i', returns (im, original hw, resized hw) - im = self.imgs[i] + # Loads 1 image from dataset index 'i', returns (im, original hw, resized hw) + im, f, fn = self.ims[i], self.im_files[i], self.npy_files[i], if im is None: # not cached in RAM - npy = self.img_npy[i] - if npy and npy.exists(): # load npy - im = np.load(npy) + if fn.exists(): # load npy + im = np.load(fn) else: # read image - f = self.img_files[i] im = cv2.imread(f) # BGR assert im is not None, f'Image Not Found {f}' h0, w0 = im.shape[:2] # orig hw @@ -643,7 +637,13 @@ def load_image(self, i): interpolation=cv2.INTER_LINEAR if (self.augment or r > 1) else cv2.INTER_AREA) return im, (h0, w0), im.shape[:2] # im, hw_original, hw_resized else: - return self.imgs[i], self.img_hw0[i], self.img_hw[i] # im, hw_original, hw_resized + return self.ims[i], self.im_hw0[i], self.im_hw[i] # im, hw_original, hw_resized + + def cache_images_to_disk(self, i): + # Saves an image as an *.npy file for faster loading + f = self.npy_files[i] + if not f.exists(): + np.save(f.as_posix(), cv2.imread(self.im_files[i])) def load_mosaic(self, index): # YOLOv5 4-mosaic loader. Loads 1 image + 3 random images into a 4-image mosaic @@ -777,16 +777,16 @@ def load_mosaic9(self, index): @staticmethod def collate_fn(batch): - img, label, path, shapes = zip(*batch) # transposed + im, label, path, shapes = zip(*batch) # transposed for i, lb in enumerate(label): lb[:, 0] = i # add target image index for build_targets() - return torch.stack(img, 0), torch.cat(label, 0), path, shapes + return torch.stack(im, 0), torch.cat(label, 0), path, shapes @staticmethod def collate_fn4(batch): img, label, path, shapes = zip(*batch) # transposed n = len(shapes) // 4 - img4, label4, path4, shapes4 = [], [], path[:n], shapes[:n] + im4, label4, path4, shapes4 = [], [], path[:n], shapes[:n] ho = torch.tensor([[0.0, 0, 0, 1, 0, 0]]) wo = torch.tensor([[0.0, 0, 1, 0, 0, 0]]) @@ -800,13 +800,13 @@ def collate_fn4(batch): else: im = torch.cat((torch.cat((img[i], img[i + 1]), 1), torch.cat((img[i + 2], img[i + 3]), 1)), 2) lb = torch.cat((label[i], label[i + 1] + ho, label[i + 2] + wo, label[i + 3] + ho + wo), 0) * s - img4.append(im) + im4.append(im) label4.append(lb) for i, lb in enumerate(label4): lb[:, 0] = i # add target image index for build_targets() - return torch.stack(img4, 0), torch.cat(label4, 0), path4, shapes4 + return torch.stack(im4, 0), torch.cat(label4, 0), path4, shapes4 # Ancillary functions -------------------------------------------------------------------------------------------------- @@ -999,12 +999,12 @@ def hub_ops(f, max_dim=1920): 'image_stats': {'total': dataset.n, 'unlabelled': int(np.all(x == 0, 1).sum()), 'per_class': (x > 0).sum(0).tolist()}, 'labels': [{str(Path(k).name): round_labels(v.tolist())} for k, v in - zip(dataset.img_files, dataset.labels)]} + zip(dataset.im_files, dataset.labels)]} if hub: im_dir = hub_dir / 'images' im_dir.mkdir(parents=True, exist_ok=True) - for _ in tqdm(ThreadPool(NUM_THREADS).imap(hub_ops, dataset.img_files), total=dataset.n, desc='HUB Ops'): + for _ in tqdm(ThreadPool(NUM_THREADS).imap(hub_ops, dataset.im_files), total=dataset.n, desc='HUB Ops'): pass # Profile diff --git a/utils/loggers/wandb/wandb_utils.py b/utils/loggers/wandb/wandb_utils.py index 3835436543d2..786e58a19972 100644 --- a/utils/loggers/wandb/wandb_utils.py +++ b/utils/loggers/wandb/wandb_utils.py @@ -403,7 +403,7 @@ def create_dataset_table(self, dataset: LoadImagesAndLabels, class_to_id: Dict[i # TODO: Explore multiprocessing to slpit this loop parallely| This is essential for speeding up the the logging artifact = wandb.Artifact(name=name, type="dataset") img_files = tqdm([dataset.path]) if isinstance(dataset.path, str) and Path(dataset.path).is_dir() else None - img_files = tqdm(dataset.img_files) if not img_files else img_files + img_files = tqdm(dataset.im_files) if not img_files else img_files for img_file in img_files: if Path(img_file).is_dir(): artifact.add_dir(img_file, name='data/images') diff --git a/val.py b/val.py index 78abbda8231a..8bde37bd5dc7 100644 --- a/val.py +++ b/val.py @@ -297,7 +297,7 @@ def run(data, pred = anno.loadRes(pred_json) # init predictions api eval = COCOeval(anno, pred, 'bbox') if is_coco: - eval.params.imgIds = [int(Path(x).stem) for x in dataloader.dataset.img_files] # image IDs to evaluate + eval.params.imgIds = [int(Path(x).stem) for x in dataloader.dataset.im_files] # image IDs to evaluate eval.evaluate() eval.accumulate() eval.summarize() From 7e98b4801a2f3e607aa2636a4346e2482f961596 Mon Sep 17 00:00:00 2001 From: vnekat <92971065+vnekat@users.noreply.github.com> Date: Mon, 7 Mar 2022 00:50:01 +0530 Subject: [PATCH 078/661] Update yolov5s.yaml (#6865) * Update yolov5s.yaml * Update yolov5s.yaml Co-authored-by: Glenn Jocher From 596de6d5a00223dc4be86377dfba6df4341b76b1 Mon Sep 17 00:00:00 2001 From: DavidB Date: Mon, 7 Mar 2022 03:21:16 +0700 Subject: [PATCH 079/661] Default FP16 TensorRT export (#6798) * Assert engine precision #6777 * Default to FP32 inputs for TensorRT engines * Default to FP16 TensorRT exports #6777 * Remove wrong line #6777 * Automatically adjust detect.py input precision #6777 * Automatically adjust val.py input precision #6777 * Add missing colon * Cleanup * Cleanup * Remove default trt_fp16_input definition * Experiment * Reorder detect.py if statement to after half checks * Update common.py * Update export.py * Cleanup Co-authored-by: Glenn Jocher --- detect.py | 4 ++++ export.py | 5 ++--- models/common.py | 3 +++ val.py | 4 ++++ 4 files changed, 13 insertions(+), 3 deletions(-) diff --git a/detect.py b/detect.py index 76f67bea1b90..ba43ed9e1eed 100644 --- a/detect.py +++ b/detect.py @@ -97,6 +97,10 @@ def run(weights=ROOT / 'yolov5s.pt', # model.pt path(s) half &= (pt or jit or onnx or engine) and device.type != 'cpu' # FP16 supported on limited backends with CUDA if pt or jit: model.model.half() if half else model.model.float() + elif engine and model.trt_fp16_input != half: + LOGGER.info('model ' + ( + 'requires' if model.trt_fp16_input else 'incompatible with') + ' --half. Adjusting automatically.') + half = model.trt_fp16_input # Dataloader if webcam: diff --git a/export.py b/export.py index 286df623d252..7a5205d55ee6 100644 --- a/export.py +++ b/export.py @@ -233,9 +233,8 @@ def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=F for out in outputs: LOGGER.info(f'{prefix}\toutput "{out.name}" with shape {out.shape} and dtype {out.dtype}') - half &= builder.platform_has_fast_fp16 - LOGGER.info(f'{prefix} building FP{16 if half else 32} engine in {f}') - if half: + LOGGER.info(f'{prefix} building FP{16 if builder.platform_has_fast_fp16 else 32} engine in {f}') + if builder.platform_has_fast_fp16: config.set_flag(trt.BuilderFlag.FP16) with builder.build_engine(network, config) as engine, open(f, 'wb') as t: t.write(engine.serialize()) diff --git a/models/common.py b/models/common.py index 0dae0244e932..70ee7105abfc 100644 --- a/models/common.py +++ b/models/common.py @@ -338,6 +338,7 @@ def __init__(self, weights='yolov5s.pt', device=None, dnn=False, data=None): import tensorrt as trt # https://developer.nvidia.com/nvidia-tensorrt-download check_version(trt.__version__, '7.0.0', hard=True) # require tensorrt>=7.0.0 Binding = namedtuple('Binding', ('name', 'dtype', 'shape', 'data', 'ptr')) + trt_fp16_input = False logger = trt.Logger(trt.Logger.INFO) with open(w, 'rb') as f, trt.Runtime(logger) as runtime: model = runtime.deserialize_cuda_engine(f.read()) @@ -348,6 +349,8 @@ def __init__(self, weights='yolov5s.pt', device=None, dnn=False, data=None): shape = tuple(model.get_binding_shape(index)) data = torch.from_numpy(np.empty(shape, dtype=np.dtype(dtype))).to(device) bindings[name] = Binding(name, dtype, shape, data, int(data.data_ptr())) + if model.binding_is_input(index) and dtype == np.float16: + trt_fp16_input = True binding_addrs = OrderedDict((n, d.ptr) for n, d in bindings.items()) context = model.create_execution_context() batch_size = bindings['images'].shape[0] diff --git a/val.py b/val.py index 8bde37bd5dc7..dfbfa3935210 100644 --- a/val.py +++ b/val.py @@ -144,6 +144,10 @@ def run(data, model.model.half() if half else model.model.float() elif engine: batch_size = model.batch_size + if model.trt_fp16_input != half: + LOGGER.info('model ' + ( + 'requires' if model.trt_fp16_input else 'incompatible with') + ' --half. Adjusting automatically.') + half = model.trt_fp16_input else: half = False batch_size = 1 # export.py models default to batch-size 1 From c8a589920e877016c8a9be00fd0077005dc68f51 Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Mon, 7 Mar 2022 13:48:59 +0100 Subject: [PATCH 080/661] Bump actions/setup-python from 2 to 3 (#6880) Bumps [actions/setup-python](https://github.com/actions/setup-python) from 2 to 3. - [Release notes](https://github.com/actions/setup-python/releases) - [Commits](https://github.com/actions/setup-python/compare/v2...v3) --- updated-dependencies: - dependency-name: actions/setup-python dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> --- .github/workflows/ci-testing.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 5cf1613ab0cd..10fab276f8f2 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -27,7 +27,7 @@ jobs: steps: - uses: actions/checkout@v2 - name: Set up Python ${{ matrix.python-version }} - uses: actions/setup-python@v2 + uses: actions/setup-python@v3 with: python-version: ${{ matrix.python-version }} From a5a1760ea6d1c172b91fa5b0606434c8379b45f0 Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Mon, 7 Mar 2022 13:49:27 +0100 Subject: [PATCH 081/661] Bump actions/checkout from 2 to 3 (#6881) Bumps [actions/checkout](https://github.com/actions/checkout) from 2 to 3. - [Release notes](https://github.com/actions/checkout/releases) - [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md) - [Commits](https://github.com/actions/checkout/compare/v2...v3) --- updated-dependencies: - dependency-name: actions/checkout dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> --- .github/workflows/ci-testing.yml | 2 +- .github/workflows/codeql-analysis.yml | 2 +- .github/workflows/rebase.yml | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 10fab276f8f2..f2096ce17a17 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -25,7 +25,7 @@ jobs: # Timeout: https://stackoverflow.com/a/59076067/4521646 timeout-minutes: 60 steps: - - uses: actions/checkout@v2 + - uses: actions/checkout@v3 - name: Set up Python ${{ matrix.python-version }} uses: actions/setup-python@v3 with: diff --git a/.github/workflows/codeql-analysis.yml b/.github/workflows/codeql-analysis.yml index 67f51f0e8bce..8bc88e957a36 100644 --- a/.github/workflows/codeql-analysis.yml +++ b/.github/workflows/codeql-analysis.yml @@ -22,7 +22,7 @@ jobs: steps: - name: Checkout repository - uses: actions/checkout@v2 + uses: actions/checkout@v3 # Initializes the CodeQL tools for scanning. - name: Initialize CodeQL diff --git a/.github/workflows/rebase.yml b/.github/workflows/rebase.yml index a4db1efb2971..75c57546166b 100644 --- a/.github/workflows/rebase.yml +++ b/.github/workflows/rebase.yml @@ -11,7 +11,7 @@ jobs: runs-on: ubuntu-latest steps: - name: Checkout the latest code - uses: actions/checkout@v2 + uses: actions/checkout@v3 with: token: ${{ secrets.ACTIONS_TOKEN }} fetch-depth: 0 # otherwise, you will fail to push refs to dest repo From acc58c1dcfba054ef936ee1458a8ff74a088ee74 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 7 Mar 2022 13:52:53 +0100 Subject: [PATCH 082/661] Fix TRT `max_workspace_size` deprecation notice (#6856) * Fix TRT `max_workspace_size` deprecation notice * Update export.py * Update export.py --- export.py | 1 + 1 file changed, 1 insertion(+) diff --git a/export.py b/export.py index 7a5205d55ee6..1e3d3e2f2e71 100644 --- a/export.py +++ b/export.py @@ -218,6 +218,7 @@ def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=F builder = trt.Builder(logger) config = builder.create_builder_config() config.max_workspace_size = workspace * 1 << 30 + # config.set_memory_pool_limit(trt.MemoryPoolType.WORKSPACE, workspace << 30) # fix TRT 8.4 deprecation notice flag = (1 << int(trt.NetworkDefinitionCreationFlag.EXPLICIT_BATCH)) network = builder.create_network(flag) From e6e36aac109794999f1dafab244b9ec4887a33d1 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 7 Mar 2022 19:26:37 +0100 Subject: [PATCH 083/661] Update bytes to GB with bitshift (#6886) --- utils/__init__.py | 7 +++---- utils/autobatch.py | 7 ++++--- utils/general.py | 5 +++-- utils/torch_utils.py | 2 +- 4 files changed, 11 insertions(+), 10 deletions(-) diff --git a/utils/__init__.py b/utils/__init__.py index 4658ed6473cd..a63c473a4340 100644 --- a/utils/__init__.py +++ b/utils/__init__.py @@ -21,14 +21,13 @@ def notebook_init(verbose=True): if is_colab(): shutil.rmtree('/content/sample_data', ignore_errors=True) # remove colab /sample_data directory + # System info if verbose: - # System info - # gb = 1 / 1000 ** 3 # bytes to GB - gib = 1 / 1024 ** 3 # bytes to GiB + gb = 1 << 30 # bytes to GiB (1024 ** 3) ram = psutil.virtual_memory().total total, used, free = shutil.disk_usage("/") display.clear_output() - s = f'({os.cpu_count()} CPUs, {ram * gib:.1f} GB RAM, {(total - free) * gib:.1f}/{total * gib:.1f} GB disk)' + s = f'({os.cpu_count()} CPUs, {ram / gb:.1f} GB RAM, {(total - free) / gb:.1f}/{total / gb:.1f} GB disk)' else: s = '' diff --git a/utils/autobatch.py b/utils/autobatch.py index cb94f041e95d..e53b4787b87d 100644 --- a/utils/autobatch.py +++ b/utils/autobatch.py @@ -34,11 +34,12 @@ def autobatch(model, imgsz=640, fraction=0.9, batch_size=16): LOGGER.info(f'{prefix}CUDA not detected, using default CPU batch-size {batch_size}') return batch_size + gb = 1 << 30 # bytes to GiB (1024 ** 3) d = str(device).upper() # 'CUDA:0' properties = torch.cuda.get_device_properties(device) # device properties - t = properties.total_memory / 1024 ** 3 # (GiB) - r = torch.cuda.memory_reserved(device) / 1024 ** 3 # (GiB) - a = torch.cuda.memory_allocated(device) / 1024 ** 3 # (GiB) + t = properties.total_memory / gb # (GiB) + r = torch.cuda.memory_reserved(device) / gb # (GiB) + a = torch.cuda.memory_allocated(device) / gb # (GiB) f = t - (r + a) # free inside reserved LOGGER.info(f'{prefix}{d} ({properties.name}) {t:.2f}G total, {r:.2f}G reserved, {a:.2f}G allocated, {f:.2f}G free') diff --git a/utils/general.py b/utils/general.py index d1594a8b5cea..36c180fe4cf2 100755 --- a/utils/general.py +++ b/utils/general.py @@ -223,11 +223,12 @@ def emojis(str=''): def file_size(path): # Return file/dir size (MB) + mb = 1 << 20 # bytes to MiB (1024 ** 2) path = Path(path) if path.is_file(): - return path.stat().st_size / 1E6 + return path.stat().st_size / mb elif path.is_dir(): - return sum(f.stat().st_size for f in path.glob('**/*') if f.is_file()) / 1E6 + return sum(f.stat().st_size for f in path.glob('**/*') if f.is_file()) / mb else: return 0.0 diff --git a/utils/torch_utils.py b/utils/torch_utils.py index c11d2a4269ef..2e6fba06626a 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -86,7 +86,7 @@ def select_device(device='', batch_size=0, newline=True): space = ' ' * (len(s) + 1) for i, d in enumerate(devices): p = torch.cuda.get_device_properties(i) - s += f"{'' if i == 0 else space}CUDA:{d} ({p.name}, {p.total_memory / 1024 ** 2:.0f}MiB)\n" # bytes to MB + s += f"{'' if i == 0 else space}CUDA:{d} ({p.name}, {p.total_memory / (1 << 20):.0f}MiB)\n" # bytes to MB else: s += 'CPU\n' From 6dd82c025298d219a1eb1fe8e486fb99d5324d34 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 9 Mar 2022 18:22:53 +0100 Subject: [PATCH 084/661] Move `git_describe()` to general.py (#6918) * Move `git_describe()` to general.py * Move `git_describe()` to general.py --- utils/general.py | 21 +++++++++++++++++++++ utils/torch_utils.py | 21 ++------------------- 2 files changed, 23 insertions(+), 19 deletions(-) diff --git a/utils/general.py b/utils/general.py index 36c180fe4cf2..a7891cbccbab 100755 --- a/utils/general.py +++ b/utils/general.py @@ -15,6 +15,7 @@ import signal import time import urllib +from datetime import datetime from itertools import repeat from multiprocessing.pool import ThreadPool from pathlib import Path @@ -221,6 +222,18 @@ def emojis(str=''): return str.encode().decode('ascii', 'ignore') if platform.system() == 'Windows' else str +def file_age(path=__file__): + # Return days since last file update + dt = (datetime.now() - datetime.fromtimestamp(Path(path).stat().st_mtime)) # delta + return dt.days # + dt.seconds / 86400 # fractional days + + +def file_update_date(path=__file__): + # Return human-readable file modification date, i.e. '2021-3-26' + t = datetime.fromtimestamp(Path(path).stat().st_mtime) + return f'{t.year}-{t.month}-{t.day}' + + def file_size(path): # Return file/dir size (MB) mb = 1 << 20 # bytes to MiB (1024 ** 2) @@ -243,6 +256,14 @@ def check_online(): return False +def git_describe(path=ROOT): # path must be a directory + # Return human-readable git description, i.e. v5.0-5-g3e25f1e https://git-scm.com/docs/git-describe + try: + return check_output(f'git -C {path} describe --tags --long --always', shell=True).decode()[:-1] + except Exception: + return '' + + @try_except @WorkingDirectory(ROOT) def check_git_status(): diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 2e6fba06626a..efcacc9ca735 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -3,7 +3,6 @@ PyTorch utils """ -import datetime import math import os import platform @@ -12,14 +11,13 @@ import warnings from contextlib import contextmanager from copy import deepcopy -from pathlib import Path import torch import torch.distributed as dist import torch.nn as nn import torch.nn.functional as F -from utils.general import LOGGER +from utils.general import LOGGER, file_update_date, git_describe try: import thop # for FLOPs computation @@ -40,21 +38,6 @@ def torch_distributed_zero_first(local_rank: int): dist.barrier(device_ids=[0]) -def date_modified(path=__file__): - # Return human-readable file modification date, i.e. '2021-3-26' - t = datetime.datetime.fromtimestamp(Path(path).stat().st_mtime) - return f'{t.year}-{t.month}-{t.day}' - - -def git_describe(path=Path(__file__).parent): # path must be a directory - # Return human-readable git description, i.e. v5.0-5-g3e25f1e https://git-scm.com/docs/git-describe - s = f'git -C {path} describe --tags --long --always' - try: - return subprocess.check_output(s, shell=True, stderr=subprocess.STDOUT).decode()[:-1] - except subprocess.CalledProcessError: - return '' # not a git repository - - def device_count(): # Returns number of CUDA devices available. Safe version of torch.cuda.device_count(). Only works on Linux. assert platform.system() == 'Linux', 'device_count() function only works on Linux' @@ -67,7 +50,7 @@ def device_count(): def select_device(device='', batch_size=0, newline=True): # device = 'cpu' or '0' or '0,1,2,3' - s = f'YOLOv5 🚀 {git_describe() or date_modified()} torch {torch.__version__} ' # string + s = f'YOLOv5 🚀 {git_describe() or file_update_date()} torch {torch.__version__} ' # string device = str(device).strip().lower().replace('cuda:', '') # to string, 'cuda:0' to '0' cpu = device == 'cpu' if cpu: From d3d9cbce221b2ced46dde374f24fde72c8e71c37 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 10 Mar 2022 12:41:06 +0100 Subject: [PATCH 085/661] PyTorch 1.11.0 compatibility updates (#6932) Resolves `AttributeError: 'Upsample' object has no attribute 'recompute_scale_factor'` first raised in https://github.com/ultralytics/yolov5/issues/5499 --- models/experimental.py | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) diff --git a/models/experimental.py b/models/experimental.py index 463e5514a06e..01bdfe72db4f 100644 --- a/models/experimental.py +++ b/models/experimental.py @@ -94,21 +94,22 @@ def attempt_load(weights, map_location=None, inplace=True, fuse=True): model = Ensemble() for w in weights if isinstance(weights, list) else [weights]: ckpt = torch.load(attempt_download(w), map_location=map_location) # load - if fuse: - model.append(ckpt['ema' if ckpt.get('ema') else 'model'].float().fuse().eval()) # FP32 model - else: - model.append(ckpt['ema' if ckpt.get('ema') else 'model'].float().eval()) # without layer fuse + ckpt = (ckpt['ema'] or ckpt['model']).float() # FP32 model + model.append(ckpt.fuse().eval() if fuse else ckpt.eval()) # fused or un-fused model in eval mode # Compatibility updates for m in model.modules(): - if type(m) in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6, nn.SiLU, Detect, Model]: - m.inplace = inplace # pytorch 1.7.0 compatibility - if type(m) is Detect: + t = type(m) + if t in (nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6, nn.SiLU, Detect, Model): + m.inplace = inplace # torch 1.7.0 compatibility + if t is Detect: if not isinstance(m.anchor_grid, list): # new Detect Layer compatibility delattr(m, 'anchor_grid') setattr(m, 'anchor_grid', [torch.zeros(1)] * m.nl) - elif type(m) is Conv: - m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility + elif t is nn.Upsample: + m.recompute_scale_factor = None # torch 1.11.0 compatibility + elif t is Conv: + m._non_persistent_buffers_set = set() # torch 1.6.0 compatibility if len(model) == 1: return model[-1] # return model From 055e72af5b887832d5e7267ac9226c825d498cd2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 10 Mar 2022 12:58:41 +0100 Subject: [PATCH 086/661] Optimize PyTorch 1.11.0 compatibility update (#6933) --- models/experimental.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/models/experimental.py b/models/experimental.py index 01bdfe72db4f..782ecbeface9 100644 --- a/models/experimental.py +++ b/models/experimental.py @@ -106,10 +106,10 @@ def attempt_load(weights, map_location=None, inplace=True, fuse=True): if not isinstance(m.anchor_grid, list): # new Detect Layer compatibility delattr(m, 'anchor_grid') setattr(m, 'anchor_grid', [torch.zeros(1)] * m.nl) - elif t is nn.Upsample: - m.recompute_scale_factor = None # torch 1.11.0 compatibility elif t is Conv: m._non_persistent_buffers_set = set() # torch 1.6.0 compatibility + elif t is nn.Upsample and not hasattr(m, 'recompute_scale_factor'): + m.recompute_scale_factor = None # torch 1.11.0 compatibility if len(model) == 1: return model[-1] # return model From caf7ad0500f8fc58567a7aa01ca91d5ee77691d6 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 10 Mar 2022 18:41:47 +0100 Subject: [PATCH 087/661] Allow 3-point segments (#6938) May resolve https://github.com/ultralytics/yolov5/issues/6931 --- utils/datasets.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/datasets.py b/utils/datasets.py index 6a2dc58dd6cd..00d0d94e0847 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -906,7 +906,7 @@ def verify_image_label(args): nf = 1 # label found with open(lb_file) as f: lb = [x.split() for x in f.read().strip().splitlines() if len(x)] - if any([len(x) > 8 for x in lb]): # is segment + if any(len(x) > 6 for x in lb): # is segment classes = np.array([x[0] for x in lb], dtype=np.float32) segments = [np.array(x[1:], dtype=np.float32).reshape(-1, 2) for x in lb] # (cls, xy1...) lb = np.concatenate((classes.reshape(-1, 1), segments2boxes(segments)), 1) # (cls, xywh) From 84efa62b2d0a619309a7437aa82cebdfc4de1bed Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 11 Mar 2022 16:18:40 +0100 Subject: [PATCH 088/661] Fix PyTorch Hub export inference shapes (#6949) May resolve https://github.com/ultralytics/yolov5/issues/6947 --- models/common.py | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/models/common.py b/models/common.py index 70ee7105abfc..ac3af20d533e 100644 --- a/models/common.py +++ b/models/common.py @@ -544,10 +544,9 @@ def forward(self, imgs, size=640, augment=False, profile=False): g = (size / max(s)) # gain shape1.append([y * g for y in s]) imgs[i] = im if im.data.contiguous else np.ascontiguousarray(im) # update - shape1 = [make_divisible(x, self.stride) for x in np.stack(shape1, 0).max(0)] # inference shape - x = [letterbox(im, new_shape=shape1 if self.pt else size, auto=False)[0] for im in imgs] # pad - x = np.stack(x, 0) if n > 1 else x[0][None] # stack - x = np.ascontiguousarray(x.transpose((0, 3, 1, 2))) # BHWC to BCHW + shape1 = [make_divisible(x, self.stride) if self.pt else size for x in np.array(shape1).max(0)] # inf shape + x = [letterbox(im, new_shape=shape1, auto=False)[0] for im in imgs] # pad + x = np.ascontiguousarray(np.array(x).transpose((0, 3, 1, 2))) # stack and BHWC to BCHW x = torch.from_numpy(x).to(p.device).type_as(p) / 255 # uint8 to fp16/32 t.append(time_sync()) From b94b59e199047aa8bf2cdd4401ae9f5f42b929e6 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 11 Mar 2022 16:31:52 +0100 Subject: [PATCH 089/661] DetectMultiBackend() `--half` handling (#6945) * DetectMultiBackend() `--half` handling * CI fixes * rename .half to .fp16 to avoid conflict * warmup fix * val update * engine update * engine update --- detect.py | 17 ++++------------- models/common.py | 13 ++++++++----- val.py | 25 +++++++++---------------- 3 files changed, 21 insertions(+), 34 deletions(-) diff --git a/detect.py b/detect.py index ba43ed9e1eed..ccb9fbf5103f 100644 --- a/detect.py +++ b/detect.py @@ -89,19 +89,10 @@ def run(weights=ROOT / 'yolov5s.pt', # model.pt path(s) # Load model device = select_device(device) - model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data) - stride, names, pt, jit, onnx, engine = model.stride, model.names, model.pt, model.jit, model.onnx, model.engine + model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data, fp16=half) + stride, names, pt = model.stride, model.names, model.pt imgsz = check_img_size(imgsz, s=stride) # check image size - # Half - half &= (pt or jit or onnx or engine) and device.type != 'cpu' # FP16 supported on limited backends with CUDA - if pt or jit: - model.model.half() if half else model.model.float() - elif engine and model.trt_fp16_input != half: - LOGGER.info('model ' + ( - 'requires' if model.trt_fp16_input else 'incompatible with') + ' --half. Adjusting automatically.') - half = model.trt_fp16_input - # Dataloader if webcam: view_img = check_imshow() @@ -114,12 +105,12 @@ def run(weights=ROOT / 'yolov5s.pt', # model.pt path(s) vid_path, vid_writer = [None] * bs, [None] * bs # Run inference - model.warmup(imgsz=(1 if pt else bs, 3, *imgsz), half=half) # warmup + model.warmup(imgsz=(1 if pt else bs, 3, *imgsz)) # warmup dt, seen = [0.0, 0.0, 0.0], 0 for path, im, im0s, vid_cap, s in dataset: t1 = time_sync() im = torch.from_numpy(im).to(device) - im = im.half() if half else im.float() # uint8 to fp16/32 + im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 im /= 255 # 0 - 255 to 0.0 - 1.0 if len(im.shape) == 3: im = im[None] # expand for batch dim diff --git a/models/common.py b/models/common.py index ac3af20d533e..251463525392 100644 --- a/models/common.py +++ b/models/common.py @@ -277,7 +277,7 @@ def forward(self, x): class DetectMultiBackend(nn.Module): # YOLOv5 MultiBackend class for python inference on various backends - def __init__(self, weights='yolov5s.pt', device=None, dnn=False, data=None): + def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, data=None, fp16=False): # Usage: # PyTorch: weights = *.pt # TorchScript: *.torchscript @@ -297,6 +297,7 @@ def __init__(self, weights='yolov5s.pt', device=None, dnn=False, data=None): pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs = self.model_type(w) # get backend stride, names = 64, [f'class{i}' for i in range(1000)] # assign defaults w = attempt_download(w) # download if not local + fp16 &= (pt or jit or onnx or engine) and device.type != 'cpu' # FP16 if data: # data.yaml path (optional) with open(data, errors='ignore') as f: names = yaml.safe_load(f)['names'] # class names @@ -305,11 +306,13 @@ def __init__(self, weights='yolov5s.pt', device=None, dnn=False, data=None): model = attempt_load(weights if isinstance(weights, list) else w, map_location=device) stride = max(int(model.stride.max()), 32) # model stride names = model.module.names if hasattr(model, 'module') else model.names # get class names + model.half() if fp16 else model.float() self.model = model # explicitly assign for to(), cpu(), cuda(), half() elif jit: # TorchScript LOGGER.info(f'Loading {w} for TorchScript inference...') extra_files = {'config.txt': ''} # model metadata model = torch.jit.load(w, _extra_files=extra_files) + model.half() if fp16 else model.float() if extra_files['config.txt']: d = json.loads(extra_files['config.txt']) # extra_files dict stride, names = int(d['stride']), d['names'] @@ -338,11 +341,11 @@ def __init__(self, weights='yolov5s.pt', device=None, dnn=False, data=None): import tensorrt as trt # https://developer.nvidia.com/nvidia-tensorrt-download check_version(trt.__version__, '7.0.0', hard=True) # require tensorrt>=7.0.0 Binding = namedtuple('Binding', ('name', 'dtype', 'shape', 'data', 'ptr')) - trt_fp16_input = False logger = trt.Logger(trt.Logger.INFO) with open(w, 'rb') as f, trt.Runtime(logger) as runtime: model = runtime.deserialize_cuda_engine(f.read()) bindings = OrderedDict() + fp16 = False # default updated below for index in range(model.num_bindings): name = model.get_binding_name(index) dtype = trt.nptype(model.get_binding_dtype(index)) @@ -350,7 +353,7 @@ def __init__(self, weights='yolov5s.pt', device=None, dnn=False, data=None): data = torch.from_numpy(np.empty(shape, dtype=np.dtype(dtype))).to(device) bindings[name] = Binding(name, dtype, shape, data, int(data.data_ptr())) if model.binding_is_input(index) and dtype == np.float16: - trt_fp16_input = True + fp16 = True binding_addrs = OrderedDict((n, d.ptr) for n, d in bindings.items()) context = model.create_execution_context() batch_size = bindings['images'].shape[0] @@ -458,11 +461,11 @@ def forward(self, im, augment=False, visualize=False, val=False): y = torch.tensor(y) if isinstance(y, np.ndarray) else y return (y, []) if val else y - def warmup(self, imgsz=(1, 3, 640, 640), half=False): + def warmup(self, imgsz=(1, 3, 640, 640)): # Warmup model by running inference once if self.pt or self.jit or self.onnx or self.engine: # warmup types if isinstance(self.device, torch.device) and self.device.type != 'cpu': # only warmup GPU models - im = torch.zeros(*imgsz).to(self.device).type(torch.half if half else torch.float) # input image + im = torch.zeros(*imgsz).to(self.device).type(torch.half if self.fp16 else torch.float) # input image self.forward(im) # warmup @staticmethod diff --git a/val.py b/val.py index dfbfa3935210..64c4d4ff9dae 100644 --- a/val.py +++ b/val.py @@ -125,7 +125,6 @@ def run(data, training = model is not None if training: # called by train.py device, pt, jit, engine = next(model.parameters()).device, True, False, False # get model device, PyTorch model - half &= device.type != 'cpu' # half precision only supported on CUDA model.half() if half else model.float() else: # called directly @@ -136,23 +135,17 @@ def run(data, (save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir # Load model - model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data) - stride, pt, jit, onnx, engine = model.stride, model.pt, model.jit, model.onnx, model.engine + model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data, fp16=half) + stride, pt, jit, engine = model.stride, model.pt, model.jit, model.engine imgsz = check_img_size(imgsz, s=stride) # check image size - half &= (pt or jit or onnx or engine) and device.type != 'cpu' # FP16 supported on limited backends with CUDA - if pt or jit: - model.model.half() if half else model.model.float() - elif engine: + half = model.fp16 # FP16 supported on limited backends with CUDA + if engine: batch_size = model.batch_size - if model.trt_fp16_input != half: - LOGGER.info('model ' + ( - 'requires' if model.trt_fp16_input else 'incompatible with') + ' --half. Adjusting automatically.') - half = model.trt_fp16_input else: - half = False - batch_size = 1 # export.py models default to batch-size 1 - device = torch.device('cpu') - LOGGER.info(f'Forcing --batch-size 1 square inference shape(1,3,{imgsz},{imgsz}) for non-PyTorch backends') + device = model.device + if not pt or jit: + batch_size = 1 # export.py models default to batch-size 1 + LOGGER.info(f'Forcing --batch-size 1 square inference (1,3,{imgsz},{imgsz}) for non-PyTorch models') # Data data = check_dataset(data) # check @@ -166,7 +159,7 @@ def run(data, # Dataloader if not training: - model.warmup(imgsz=(1 if pt else batch_size, 3, imgsz, imgsz), half=half) # warmup + model.warmup(imgsz=(1 if pt else batch_size, 3, imgsz, imgsz)) # warmup pad = 0.0 if task in ('speed', 'benchmark') else 0.5 rect = False if task == 'benchmark' else pt # square inference for benchmarks task = task if task in ('train', 'val', 'test') else 'val' # path to train/val/test images From c6b4f84fd1ce03496d64db4d4b1e5895ca5c879b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 12 Mar 2022 00:45:07 +0100 Subject: [PATCH 090/661] Update Dockerfile `torch==1.11.0+cu113` (#6954) --- Dockerfile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Dockerfile b/Dockerfile index 489dd04ce5c9..896751d50d2d 100644 --- a/Dockerfile +++ b/Dockerfile @@ -11,7 +11,7 @@ COPY requirements.txt . RUN python -m pip install --upgrade pip RUN pip uninstall -y torch torchvision torchtext RUN pip install --no-cache -r requirements.txt albumentations wandb gsutil notebook \ - torch==1.10.2+cu113 torchvision==0.11.3+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html + torch==1.11.0+cu113 torchvision==0.12.0+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html # RUN pip install --no-cache -U torch torchvision # Create working directory From c84dd27d62d979bf4a97472808a7ef8747d64491 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 12 Mar 2022 12:57:08 +0100 Subject: [PATCH 091/661] New val.py `cuda` variable (#6957) * New val.py `cuda` variable Fix for ONNX GPU val. * Update val.py --- val.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/val.py b/val.py index 64c4d4ff9dae..8f2119531949 100644 --- a/val.py +++ b/val.py @@ -143,7 +143,7 @@ def run(data, batch_size = model.batch_size else: device = model.device - if not pt or jit: + if not (pt or jit): batch_size = 1 # export.py models default to batch-size 1 LOGGER.info(f'Forcing --batch-size 1 square inference (1,3,{imgsz},{imgsz}) for non-PyTorch models') @@ -152,6 +152,7 @@ def run(data, # Configure model.eval() + cuda = device.type != 'cpu' is_coco = isinstance(data.get('val'), str) and data['val'].endswith('coco/val2017.txt') # COCO dataset nc = 1 if single_cls else int(data['nc']) # number of classes iouv = torch.linspace(0.5, 0.95, 10).to(device) # iou vector for mAP@0.5:0.95 @@ -177,7 +178,7 @@ def run(data, pbar = tqdm(dataloader, desc=s, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar for batch_i, (im, targets, paths, shapes) in enumerate(pbar): t1 = time_sync() - if pt or jit or engine: + if cuda: im = im.to(device, non_blocking=True) targets = targets.to(device) im = im.half() if half else im.float() # uint8 to fp16/32 From 52c1399fdc6c3db550123e47a2cdcb6dc951e211 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 12 Mar 2022 13:16:29 +0100 Subject: [PATCH 092/661] DetectMultiBackend() return `device` update (#6958) Fixes ONNX validation that returns outputs on CPU. --- models/common.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/models/common.py b/models/common.py index 251463525392..48cf55795dd4 100644 --- a/models/common.py +++ b/models/common.py @@ -458,7 +458,8 @@ def forward(self, im, augment=False, visualize=False, val=False): y = (y.astype(np.float32) - zero_point) * scale # re-scale y[..., :4] *= [w, h, w, h] # xywh normalized to pixels - y = torch.tensor(y) if isinstance(y, np.ndarray) else y + if isinstance(y, np.ndarray): + y = torch.tensor(y, device=self.device) return (y, []) if val else y def warmup(self, imgsz=(1, 3, 640, 640)): From 701e1177ac5cfec2f10552e55766d184ca760e12 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 12 Mar 2022 14:00:48 +0100 Subject: [PATCH 093/661] Tensor initialization on device improvements (#6959) * Update common.py speed improvements Eliminate .to() ops where possible for reduced data transfer overhead. Primarily affects warmup and PyTorch Hub inference. * Updates * Updates * Update detect.py * Update val.py --- models/common.py | 2 +- val.py | 6 +++--- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/models/common.py b/models/common.py index 48cf55795dd4..83aecb7569d6 100644 --- a/models/common.py +++ b/models/common.py @@ -466,7 +466,7 @@ def warmup(self, imgsz=(1, 3, 640, 640)): # Warmup model by running inference once if self.pt or self.jit or self.onnx or self.engine: # warmup types if isinstance(self.device, torch.device) and self.device.type != 'cpu': # only warmup GPU models - im = torch.zeros(*imgsz).to(self.device).type(torch.half if self.fp16 else torch.float) # input image + im = torch.zeros(*imgsz, dtype=torch.half if self.fp16 else torch.float, device=self.device) # input self.forward(im) # warmup @staticmethod diff --git a/val.py b/val.py index 8f2119531949..2dd2aec679f9 100644 --- a/val.py +++ b/val.py @@ -87,7 +87,7 @@ def process_batch(detections, labels, iouv): matches = matches[np.unique(matches[:, 1], return_index=True)[1]] # matches = matches[matches[:, 2].argsort()[::-1]] matches = matches[np.unique(matches[:, 0], return_index=True)[1]] - matches = torch.Tensor(matches).to(iouv.device) + matches = torch.from_numpy(matches).to(iouv.device) correct[matches[:, 1].long()] = matches[:, 2:3] >= iouv return correct @@ -155,7 +155,7 @@ def run(data, cuda = device.type != 'cpu' is_coco = isinstance(data.get('val'), str) and data['val'].endswith('coco/val2017.txt') # COCO dataset nc = 1 if single_cls else int(data['nc']) # number of classes - iouv = torch.linspace(0.5, 0.95, 10).to(device) # iou vector for mAP@0.5:0.95 + iouv = torch.linspace(0.5, 0.95, 10, device=device) # iou vector for mAP@0.5:0.95 niou = iouv.numel() # Dataloader @@ -196,7 +196,7 @@ def run(data, loss += compute_loss([x.float() for x in train_out], targets)[1] # box, obj, cls # NMS - targets[:, 2:] *= torch.Tensor([width, height, width, height]).to(device) # to pixels + targets[:, 2:] *= torch.tensor((width, height, width, height), device=device) # to pixels lb = [targets[targets[:, 0] == i, 1:] for i in range(nb)] if save_hybrid else [] # for autolabelling t3 = time_sync() out = non_max_suppression(out, conf_thres, iou_thres, labels=lb, multi_label=True, agnostic=single_cls) From c13d4ce7ef30acc78e3dbdd9aa4f17e01ed34521 Mon Sep 17 00:00:00 2001 From: paradigm Date: Sat, 12 Mar 2022 16:15:09 +0100 Subject: [PATCH 094/661] EdgeTPU optimizations (#6808) * removed transpose op for better edgetpu support * fix for training case * enabled experimental new quantizer flag * precalculate add and mul ops at compile time Co-authored-by: Glenn Jocher --- export.py | 2 +- models/tf.py | 10 ++++++---- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/export.py b/export.py index 1e3d3e2f2e71..7dd06433fe36 100644 --- a/export.py +++ b/export.py @@ -331,7 +331,7 @@ def export_tflite(keras_model, im, file, int8, data, ncalib, prefix=colorstr('Te converter.target_spec.supported_types = [] converter.inference_input_type = tf.uint8 # or tf.int8 converter.inference_output_type = tf.uint8 # or tf.int8 - converter.experimental_new_quantizer = False + converter.experimental_new_quantizer = True f = str(file).replace('.pt', '-int8.tflite') tflite_model = converter.convert() diff --git a/models/tf.py b/models/tf.py index 74681e403afd..728907f8fb47 100644 --- a/models/tf.py +++ b/models/tf.py @@ -222,19 +222,21 @@ def call(self, inputs): x.append(self.m[i](inputs[i])) # x(bs,20,20,255) to x(bs,3,20,20,85) ny, nx = self.imgsz[0] // self.stride[i], self.imgsz[1] // self.stride[i] - x[i] = tf.transpose(tf.reshape(x[i], [-1, ny * nx, self.na, self.no]), [0, 2, 1, 3]) + x[i] = tf.reshape(x[i], [-1, ny * nx, self.na, self.no]) if not self.training: # inference y = tf.sigmoid(x[i]) - xy = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy - wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] + grid = tf.transpose(self.grid[i], [0, 2, 1, 3]) - 0.5 + anchor_grid = tf.transpose(self.anchor_grid[i], [0, 2, 1, 3]) * 4 + xy = (y[..., 0:2] * 2 + grid) * self.stride[i] # xy + wh = y[..., 2:4] ** 2 * anchor_grid # Normalize xywh to 0-1 to reduce calibration error xy /= tf.constant([[self.imgsz[1], self.imgsz[0]]], dtype=tf.float32) wh /= tf.constant([[self.imgsz[1], self.imgsz[0]]], dtype=tf.float32) y = tf.concat([xy, wh, y[..., 4:]], -1) z.append(tf.reshape(y, [-1, self.na * ny * nx, self.no])) - return x if self.training else (tf.concat(z, 1), x) + return tf.transpose(x, [0, 2, 1, 3]) if self.training else (tf.concat(z, 1), x) @staticmethod def _make_grid(nx=20, ny=20): From 2d45de617e0e80fb96424425587b6ce123aa0012 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 14 Mar 2022 10:54:51 +0100 Subject: [PATCH 095/661] Model `ema` key backward compatibility fix (#6972) Fix for older model loading issue in https://github.com/ultralytics/yolov5/commit/d3d9cbce221b2ced46dde374f24fde72c8e71c37#commitcomment-68622388 --- models/experimental.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/experimental.py b/models/experimental.py index 782ecbeface9..1230f4656c8f 100644 --- a/models/experimental.py +++ b/models/experimental.py @@ -94,7 +94,7 @@ def attempt_load(weights, map_location=None, inplace=True, fuse=True): model = Ensemble() for w in weights if isinstance(weights, list) else [weights]: ckpt = torch.load(attempt_download(w), map_location=map_location) # load - ckpt = (ckpt['ema'] or ckpt['model']).float() # FP32 model + ckpt = (ckpt.get('ema') or ckpt['model']).float() # FP32 model model.append(ckpt.fuse().eval() if fuse else ckpt.eval()) # fused or un-fused model in eval mode # Compatibility updates From 99de551f979f6aca1f817504831c821cff64b5fd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 14 Mar 2022 12:41:06 +0100 Subject: [PATCH 096/661] pt model to cpu on TF export --- export.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/export.py b/export.py index 7dd06433fe36..c50de15cf0b8 100644 --- a/export.py +++ b/export.py @@ -494,7 +494,7 @@ def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path' if int8 or edgetpu: # TFLite --int8 bug https://github.com/ultralytics/yolov5/issues/5707 check_requirements(('flatbuffers==1.12',)) # required before `import tensorflow` assert not (tflite and tfjs), 'TFLite and TF.js models must be exported separately, please pass only one type.' - model, f[5] = export_saved_model(model, im, file, dynamic, tf_nms=nms or agnostic_nms or tfjs, + model, f[5] = export_saved_model(model.cpu(), im, file, dynamic, tf_nms=nms or agnostic_nms or tfjs, agnostic_nms=agnostic_nms or tfjs, topk_per_class=topk_per_class, topk_all=topk_all, conf_thres=conf_thres, iou_thres=iou_thres) # keras model if pb or tfjs: # pb prerequisite to tfjs From 932dc78496ca532a41780335468589ad7f0147f7 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 14 Mar 2022 15:07:13 +0100 Subject: [PATCH 097/661] YOLOv5 Export Benchmarks for GPU (#6963) * Add benchmarks.py GPU support * Updates * Updates * Updates * Updates * Add --half * Add TRT requirements * Cleanup * Add TF to warmup types * Update export.py * Update export.py * Update benchmarks.py --- export.py | 24 ++++++++++++------------ models/common.py | 7 ++++--- utils/benchmarks.py | 18 +++++++++++++++--- 3 files changed, 31 insertions(+), 18 deletions(-) diff --git a/export.py b/export.py index c50de15cf0b8..d4f980fdb993 100644 --- a/export.py +++ b/export.py @@ -75,18 +75,18 @@ def export_formats(): # YOLOv5 export formats - x = [['PyTorch', '-', '.pt'], - ['TorchScript', 'torchscript', '.torchscript'], - ['ONNX', 'onnx', '.onnx'], - ['OpenVINO', 'openvino', '_openvino_model'], - ['TensorRT', 'engine', '.engine'], - ['CoreML', 'coreml', '.mlmodel'], - ['TensorFlow SavedModel', 'saved_model', '_saved_model'], - ['TensorFlow GraphDef', 'pb', '.pb'], - ['TensorFlow Lite', 'tflite', '.tflite'], - ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite'], - ['TensorFlow.js', 'tfjs', '_web_model']] - return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix']) + x = [['PyTorch', '-', '.pt', True], + ['TorchScript', 'torchscript', '.torchscript', True], + ['ONNX', 'onnx', '.onnx', True], + ['OpenVINO', 'openvino', '_openvino_model', False], + ['TensorRT', 'engine', '.engine', True], + ['CoreML', 'coreml', '.mlmodel', False], + ['TensorFlow SavedModel', 'saved_model', '_saved_model', True], + ['TensorFlow GraphDef', 'pb', '.pb', True], + ['TensorFlow Lite', 'tflite', '.tflite', False], + ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False], + ['TensorFlow.js', 'tfjs', '_web_model', False]] + return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix', 'GPU']) def export_torchscript(model, im, file, optimize, prefix=colorstr('TorchScript:')): diff --git a/models/common.py b/models/common.py index 83aecb7569d6..4ad040fcd7f1 100644 --- a/models/common.py +++ b/models/common.py @@ -464,10 +464,11 @@ def forward(self, im, augment=False, visualize=False, val=False): def warmup(self, imgsz=(1, 3, 640, 640)): # Warmup model by running inference once - if self.pt or self.jit or self.onnx or self.engine: # warmup types - if isinstance(self.device, torch.device) and self.device.type != 'cpu': # only warmup GPU models + if any((self.pt, self.jit, self.onnx, self.engine, self.saved_model, self.pb)): # warmup types + if self.device.type != 'cpu': # only warmup GPU models im = torch.zeros(*imgsz, dtype=torch.half if self.fp16 else torch.float, device=self.device) # input - self.forward(im) # warmup + for _ in range(2 if self.jit else 1): # + self.forward(im) # warmup @staticmethod def model_type(p='path/to/model.pt'): diff --git a/utils/benchmarks.py b/utils/benchmarks.py index 962df812a9d3..bdbbdc43b639 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -19,6 +19,7 @@ Requirements: $ pip install -r requirements.txt coremltools onnx onnx-simplifier onnxruntime openvino-dev tensorflow-cpu # CPU $ pip install -r requirements.txt coremltools onnx onnx-simplifier onnxruntime-gpu openvino-dev tensorflow # GPU + $ pip install -U nvidia-tensorrt --index-url https://pypi.ngc.nvidia.com # TensorRT Usage: $ python utils/benchmarks.py --weights yolov5s.pt --img 640 @@ -41,20 +42,29 @@ import val from utils import notebook_init from utils.general import LOGGER, print_args +from utils.torch_utils import select_device def run(weights=ROOT / 'yolov5s.pt', # weights path imgsz=640, # inference size (pixels) batch_size=1, # batch size data=ROOT / 'data/coco128.yaml', # dataset.yaml path + device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu + half=False, # use FP16 half-precision inference ): y, t = [], time.time() formats = export.export_formats() - for i, (name, f, suffix) in formats.iterrows(): # index, (name, file, suffix) + device = select_device(device) + for i, (name, f, suffix, gpu) in formats.iterrows(): # index, (name, file, suffix, gpu-capable) try: - w = weights if f == '-' else export.run(weights=weights, imgsz=[imgsz], include=[f], device='cpu')[-1] + if device.type != 'cpu': + assert gpu, f'{name} inference not supported on GPU' + if f == '-': + w = weights # PyTorch format + else: + w = export.run(weights=weights, imgsz=[imgsz], include=[f], device=device, half=half)[-1] # all others assert suffix in str(w), 'export failed' - result = val.run(data, w, batch_size, imgsz=imgsz, plots=False, device='cpu', task='benchmark') + result = val.run(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) metrics = result[0] # metrics (mp, mr, map50, map, *losses(box, obj, cls)) speeds = result[2] # times (preprocess, inference, postprocess) y.append([name, metrics[3], speeds[1]]) # mAP, t_inference @@ -78,6 +88,8 @@ def parse_opt(): parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='inference size (pixels)') parser.add_argument('--batch-size', type=int, default=1, help='batch size') parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path') + parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') opt = parser.parse_args() print_args(FILE.stem, opt) return opt From c09fb2aa95b6ca86c460aa106e2308805649feb9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 15 Mar 2022 16:32:56 +0100 Subject: [PATCH 098/661] Update TQDM bar format (#6988) --- utils/autoanchor.py | 2 +- utils/datasets.py | 7 ++++--- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/utils/autoanchor.py b/utils/autoanchor.py index a631c21a3b26..6cd2267a375a 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -152,7 +152,7 @@ def print_results(k, verbose=True): # Evolve f, sh, mp, s = anchor_fitness(k), k.shape, 0.9, 0.1 # fitness, generations, mutation prob, sigma - pbar = tqdm(range(gen), desc=f'{PREFIX}Evolving anchors with Genetic Algorithm:') # progress bar + pbar = tqdm(range(gen), bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar for _ in pbar: v = np.ones(sh) while (v == 1).all(): # mutate until a change occurs (prevent duplicates) diff --git a/utils/datasets.py b/utils/datasets.py index 00d0d94e0847..5ce6d607fb7a 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -35,6 +35,7 @@ HELP_URL = 'https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' IMG_FORMATS = 'bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp' # include image suffixes VID_FORMATS = 'asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'ts', 'wmv' # include video suffixes +BAR_FORMAT = '{l_bar}{bar:10}{r_bar}{bar:-10b}' # tqdm bar format # Get orientation exif tag for orientation in ExifTags.TAGS.keys(): @@ -427,7 +428,7 @@ def __init__(self, path, img_size=640, batch_size=16, augment=False, hyp=None, r nf, nm, ne, nc, n = cache.pop('results') # found, missing, empty, corrupt, total if exists: d = f"Scanning '{cache_path}' images and labels... {nf} found, {nm} missing, {ne} empty, {nc} corrupt" - tqdm(None, desc=prefix + d, total=n, initial=n) # display cache results + tqdm(None, desc=prefix + d, total=n, initial=n, bar_format=BAR_FORMAT) # display cache results if cache['msgs']: LOGGER.info('\n'.join(cache['msgs'])) # display warnings assert nf > 0 or not augment, f'{prefix}No labels in {cache_path}. Can not train without labels. See {HELP_URL}' @@ -492,7 +493,7 @@ def __init__(self, path, img_size=640, batch_size=16, augment=False, hyp=None, r self.im_hw0, self.im_hw = [None] * n, [None] * n fcn = self.cache_images_to_disk if cache_images == 'disk' else self.load_image results = ThreadPool(NUM_THREADS).imap(fcn, range(n)) - pbar = tqdm(enumerate(results), total=n) + pbar = tqdm(enumerate(results), total=n, bar_format=BAR_FORMAT) for i, x in pbar: if cache_images == 'disk': gb += self.npy_files[i].stat().st_size @@ -509,7 +510,7 @@ def cache_labels(self, path=Path('./labels.cache'), prefix=''): desc = f"{prefix}Scanning '{path.parent / path.stem}' images and labels..." with Pool(NUM_THREADS) as pool: pbar = tqdm(pool.imap(verify_image_label, zip(self.im_files, self.label_files, repeat(prefix))), - desc=desc, total=len(self.im_files)) + desc=desc, total=len(self.im_files), bar_format=BAR_FORMAT) for im_file, lb, shape, segments, nm_f, nf_f, ne_f, nc_f, msg in pbar: nm += nm_f nf += nf_f From 3f634d43c8ecea14aa9037e2fd28ded0433d491d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 16 Mar 2022 15:33:54 +0100 Subject: [PATCH 099/661] Conditional `Timeout()` by OS (disable on Windows) (#7013) * Conditional `Timeout()` by OS (disable on Windows) * Update general.py --- utils/general.py | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/utils/general.py b/utils/general.py index a7891cbccbab..e8b3b05c5fe1 100755 --- a/utils/general.py +++ b/utils/general.py @@ -123,13 +123,15 @@ def _timeout_handler(self, signum, frame): raise TimeoutError(self.timeout_message) def __enter__(self): - signal.signal(signal.SIGALRM, self._timeout_handler) # Set handler for SIGALRM - signal.alarm(self.seconds) # start countdown for SIGALRM to be raised + if platform.system() != 'Windows': # not supported on Windows + signal.signal(signal.SIGALRM, self._timeout_handler) # Set handler for SIGALRM + signal.alarm(self.seconds) # start countdown for SIGALRM to be raised def __exit__(self, exc_type, exc_val, exc_tb): - signal.alarm(0) # Cancel SIGALRM if it's scheduled - if self.suppress and exc_type is TimeoutError: # Suppress TimeoutError - return True + if platform.system() != 'Windows': + signal.alarm(0) # Cancel SIGALRM if it's scheduled + if self.suppress and exc_type is TimeoutError: # Suppress TimeoutError + return True class WorkingDirectory(contextlib.ContextDecorator): From 7c6a33564a84a0e78ec19da66ea6016d51c32e0a Mon Sep 17 00:00:00 2001 From: Max Strobel Date: Thu, 17 Mar 2022 16:37:09 +0100 Subject: [PATCH 100/661] fix: add default PIL font as fallback (#7010) * fix: add default font as fallback Add default font as fallback if the downloading of the Arial.ttf font fails for some reason, e.g. no access to public internet. * Update plots.py Co-authored-by: Maximilian Strobel Co-authored-by: Glenn Jocher --- utils/plots.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/utils/plots.py b/utils/plots.py index 6c3f5bcaef37..90f3f241cc5a 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -7,6 +7,7 @@ import os from copy import copy from pathlib import Path +from urllib.error import URLError import cv2 import matplotlib @@ -55,11 +56,13 @@ def check_pil_font(font=FONT, size=10): try: return ImageFont.truetype(str(font) if font.exists() else font.name, size) except Exception: # download if missing - check_font(font) try: + check_font(font) return ImageFont.truetype(str(font), size) except TypeError: check_requirements('Pillow>=8.4.0') # known issue https://github.com/ultralytics/yolov5/issues/5374 + except URLError: # not online + return ImageFont.load_default() class Annotator: From 4effd064b169fc049b4a4bca401b120bf2e93c14 Mon Sep 17 00:00:00 2001 From: Mrinal Jain Date: Fri, 18 Mar 2022 07:29:24 -0400 Subject: [PATCH 101/661] Consistent saved_model output format (#7032) --- export.py | 2 +- models/common.py | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/export.py b/export.py index d4f980fdb993..2d4a68e62f89 100644 --- a/export.py +++ b/export.py @@ -275,7 +275,7 @@ def export_saved_model(model, im, file, dynamic, m = m.get_concrete_function(spec) frozen_func = convert_variables_to_constants_v2(m) tfm = tf.Module() - tfm.__call__ = tf.function(lambda x: frozen_func(x), [spec]) + tfm.__call__ = tf.function(lambda x: frozen_func(x)[0], [spec]) tfm.__call__(im) tf.saved_model.save( tfm, diff --git a/models/common.py b/models/common.py index 4ad040fcd7f1..5561d92ecb73 100644 --- a/models/common.py +++ b/models/common.py @@ -441,7 +441,7 @@ def forward(self, im, augment=False, visualize=False, val=False): else: # TensorFlow (SavedModel, GraphDef, Lite, Edge TPU) im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) if self.saved_model: # SavedModel - y = (self.model(im, training=False) if self.keras else self.model(im)[0]).numpy() + y = (self.model(im, training=False) if self.keras else self.model(im)).numpy() elif self.pb: # GraphDef y = self.frozen_func(x=self.tf.constant(im)).numpy() else: # Lite or Edge TPU From b0ba101ac0aa898a4e4b867d377e140af8d4258a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 01:04:48 +0100 Subject: [PATCH 102/661] `ComputeLoss()` indexing/speed improvements (#7048) * device as class attribute * Update loss.py * Update loss.py * improve zeros * tensor split --- utils/loss.py | 37 +++++++++++++++++++------------------ 1 file changed, 19 insertions(+), 18 deletions(-) diff --git a/utils/loss.py b/utils/loss.py index 5aa9f017d2af..0f0137817955 100644 --- a/utils/loss.py +++ b/utils/loss.py @@ -89,9 +89,10 @@ def forward(self, pred, true): class ComputeLoss: + sort_obj_iou = False + # Compute losses def __init__(self, model, autobalance=False): - self.sort_obj_iou = False device = next(model.parameters()).device # get model device h = model.hyp # hyperparameters @@ -111,26 +112,28 @@ def __init__(self, model, autobalance=False): self.balance = {3: [4.0, 1.0, 0.4]}.get(det.nl, [4.0, 1.0, 0.25, 0.06, 0.02]) # P3-P7 self.ssi = list(det.stride).index(16) if autobalance else 0 # stride 16 index self.BCEcls, self.BCEobj, self.gr, self.hyp, self.autobalance = BCEcls, BCEobj, 1.0, h, autobalance + self.device = device for k in 'na', 'nc', 'nl', 'anchors': setattr(self, k, getattr(det, k)) - def __call__(self, p, targets): # predictions, targets, model - device = targets.device - lcls, lbox, lobj = torch.zeros(1, device=device), torch.zeros(1, device=device), torch.zeros(1, device=device) + def __call__(self, p, targets): # predictions, targets + lcls = torch.zeros(1, device=self.device) # class loss + lbox = torch.zeros(1, device=self.device) # box loss + lobj = torch.zeros(1, device=self.device) # object loss tcls, tbox, indices, anchors = self.build_targets(p, targets) # targets # Losses for i, pi in enumerate(p): # layer index, layer predictions b, a, gj, gi = indices[i] # image, anchor, gridy, gridx - tobj = torch.zeros_like(pi[..., 0], device=device) # target obj + tobj = torch.zeros(pi.shape[:4], device=self.device) # target obj n = b.shape[0] # number of targets if n: - ps = pi[b, a, gj, gi] # prediction subset corresponding to targets + pxy, pwh, _, pcls = pi[b, a, gj, gi].tensor_split((2, 4, 5), dim=1) # target-subset of predictions # Regression - pxy = ps[:, :2].sigmoid() * 2 - 0.5 - pwh = (ps[:, 2:4].sigmoid() * 2) ** 2 * anchors[i] + pxy = pxy.sigmoid() * 2 - 0.5 + pwh = (pwh.sigmoid() * 2) ** 2 * anchors[i] pbox = torch.cat((pxy, pwh), 1) # predicted box iou = bbox_iou(pbox.T, tbox[i], x1y1x2y2=False, CIoU=True) # iou(prediction, target) lbox += (1.0 - iou).mean() # iou loss @@ -144,9 +147,9 @@ def __call__(self, p, targets): # predictions, targets, model # Classification if self.nc > 1: # cls loss (only if multiple classes) - t = torch.full_like(ps[:, 5:], self.cn, device=device) # targets + t = torch.full_like(pcls, self.cn, device=self.device) # targets t[range(n), tcls[i]] = self.cp - lcls += self.BCEcls(ps[:, 5:], t) # BCE + lcls += self.BCEcls(pcls, t) # BCE # Append targets to text file # with open('targets.txt', 'a') as file: @@ -170,15 +173,15 @@ def build_targets(self, p, targets): # Build targets for compute_loss(), input targets(image,class,x,y,w,h) na, nt = self.na, targets.shape[0] # number of anchors, targets tcls, tbox, indices, anch = [], [], [], [] - gain = torch.ones(7, device=targets.device) # normalized to gridspace gain - ai = torch.arange(na, device=targets.device).float().view(na, 1).repeat(1, nt) # same as .repeat_interleave(nt) + gain = torch.ones(7, device=self.device) # normalized to gridspace gain + ai = torch.arange(na, device=self.device).float().view(na, 1).repeat(1, nt) # same as .repeat_interleave(nt) targets = torch.cat((targets.repeat(na, 1, 1), ai[:, :, None]), 2) # append anchor indices g = 0.5 # bias off = torch.tensor([[0, 0], [1, 0], [0, 1], [-1, 0], [0, -1], # j,k,l,m # [1, 1], [1, -1], [-1, 1], [-1, -1], # jk,jm,lk,lm - ], device=targets.device).float() * g # offsets + ], device=self.device).float() * g # offsets for i in range(self.nl): anchors = self.anchors[i] @@ -206,14 +209,12 @@ def build_targets(self, p, targets): offsets = 0 # Define - b, c = t[:, :2].long().T # image, class - gxy = t[:, 2:4] # grid xy - gwh = t[:, 4:6] # grid wh + bc, gxy, gwh, a = t.unsafe_chunk(4, dim=1) # (image, class), grid xy, grid wh, anchors + a, (b, c) = a.long().view(-1), bc.long().T # anchors, image, class gij = (gxy - offsets).long() - gi, gj = gij.T # grid xy indices + gi, gj = gij.T # grid indices # Append - a = t[:, 6].long() # anchor indices indices.append((b, a, gj.clamp_(0, gain[3] - 1), gi.clamp_(0, gain[2] - 1))) # image, anchor, grid indices tbox.append(torch.cat((gxy - gij, gwh), 1)) # box anch.append(anchors[a]) # anchors From 9ebec7885fb461993cf7123b36abf61ffd5dfd95 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 13:51:22 +0100 Subject: [PATCH 103/661] Update Dockerfile to `git clone` instead of `COPY` (#7053) Resolves git command errors that currently happen in image, i.e.: ```bash root@382ae64aeca2:/usr/src/app# git pull Warning: Permanently added the ECDSA host key for IP address '140.82.113.3' to the list of known hosts. git@github.com: Permission denied (publickey). fatal: Could not read from remote repository. Please make sure you have the correct access rights and the repository exists. ``` --- Dockerfile | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/Dockerfile b/Dockerfile index 896751d50d2d..304e8b2801a9 100644 --- a/Dockerfile +++ b/Dockerfile @@ -19,7 +19,8 @@ RUN mkdir -p /usr/src/app WORKDIR /usr/src/app # Copy contents -COPY . /usr/src/app +RUN git clone https://github.com/ultralytics/yolov5 /usr/src/app +# COPY . /usr/src/app # Downloads to user config dir ADD https://ultralytics.com/assets/Arial.ttf /root/.config/Ultralytics/ From 6843ea5d7f9c5d4b8132d00ba17fb296dc81d867 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 13:55:32 +0100 Subject: [PATCH 104/661] Create SECURITY.md (#7054) * Create SECURITY.md Resolves https://github.com/ultralytics/yolov5/issues/7052 * Move into ./github * Update SECURITY.md --- .github/SECURITY.md | 7 +++++++ 1 file changed, 7 insertions(+) create mode 100644 .github/SECURITY.md diff --git a/.github/SECURITY.md b/.github/SECURITY.md new file mode 100644 index 000000000000..aa3e8409da6b --- /dev/null +++ b/.github/SECURITY.md @@ -0,0 +1,7 @@ +# Security Policy + +We aim to make YOLOv5 🚀 as secure as possible! If you find potential vulnerabilities or have any concerns please let us know so we can investigate and take corrective action if needed. + +### Reporting a Vulnerability + +To report vulnerabilities please email us at hello@ultralytics.com or visit https://ultralytics.com/contact. Thank you! From 9f4d11379bb931586c1f51c1d85c6fac9fc37eda Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 14:18:05 +0100 Subject: [PATCH 105/661] Fix incomplete URL substring sanitation (#7056) Resolves code scanning alert in https://github.com/ultralytics/yolov5/issues/7055 --- utils/datasets.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/utils/datasets.py b/utils/datasets.py index 5ce6d607fb7a..8627344af7b4 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -15,6 +15,7 @@ from multiprocessing.pool import Pool, ThreadPool from pathlib import Path from threading import Thread +from urllib.parse import urlparse from zipfile import ZipFile import cv2 @@ -301,7 +302,7 @@ def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True): for i, s in enumerate(sources): # index, source # Start thread to read frames from video stream st = f'{i + 1}/{n}: {s}... ' - if 'youtube.com/' in s or 'youtu.be/' in s: # if source is YouTube video + if urlparse(s).hostname in ('youtube.com', 'youtu.be'): # if source is YouTube video check_requirements(('pafy', 'youtube_dl==2020.12.2')) import pafy s = pafy.new(s).getbest(preftype="mp4").url # YouTube URL From 529fbc1814f899eab2df8146944c23d0e168610e Mon Sep 17 00:00:00 2001 From: Philip Gutjahr Date: Sun, 20 Mar 2022 15:46:29 +0100 Subject: [PATCH 106/661] Use PIL to eliminate chroma subsampling in crops (#7008) * use pillow to save higher-quality jpg (w/o color subsampling) * Cleanup and doc issue Co-authored-by: Glenn Jocher --- utils/plots.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/utils/plots.py b/utils/plots.py index 90f3f241cc5a..a30c0faf962a 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -458,7 +458,7 @@ def profile_idetection(start=0, stop=0, labels=(), save_dir=''): plt.savefig(Path(save_dir) / 'idetection_profile.png', dpi=200) -def save_one_box(xyxy, im, file='image.jpg', gain=1.02, pad=10, square=False, BGR=False, save=True): +def save_one_box(xyxy, im, file=Path('im.jpg'), gain=1.02, pad=10, square=False, BGR=False, save=True): # Save image crop as {file} with crop size multiple {gain} and {pad} pixels. Save and/or return crop xyxy = torch.tensor(xyxy).view(-1, 4) b = xyxy2xywh(xyxy) # boxes @@ -470,5 +470,7 @@ def save_one_box(xyxy, im, file='image.jpg', gain=1.02, pad=10, square=False, BG crop = im[int(xyxy[0, 1]):int(xyxy[0, 3]), int(xyxy[0, 0]):int(xyxy[0, 2]), ::(1 if BGR else -1)] if save: file.parent.mkdir(parents=True, exist_ok=True) # make directory - cv2.imwrite(str(increment_path(file).with_suffix('.jpg')), crop) + f = str(increment_path(file).with_suffix('.jpg')) + # cv2.imwrite(f, crop) # https://github.com/ultralytics/yolov5/issues/7007 chroma subsampling issue + Image.fromarray(cv2.cvtColor(crop, cv2.COLOR_BGR2RGB)).save(f, quality=95, subsampling=0) return crop From f327eee614384583a93e6f5374280e78b80a250d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 16:27:51 +0100 Subject: [PATCH 107/661] Fix `check_anchor_order()` in pixel-space not grid-space (#7060) * Update `check_anchor_order()` Use mean area per output layer for added stability. * Check in pixel-space not grid-space fix --- models/yolo.py | 2 +- utils/autoanchor.py | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index f659a04545b9..2f4bbe0f71d1 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -110,8 +110,8 @@ def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, anchors=None): # model, i s = 256 # 2x min stride m.inplace = self.inplace m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))]) # forward + check_anchor_order(m) # must be in pixel-space (not grid-space) m.anchors /= m.stride.view(-1, 1, 1) - check_anchor_order(m) self.stride = m.stride self._initialize_biases() # only run once diff --git a/utils/autoanchor.py b/utils/autoanchor.py index 6cd2267a375a..7eb46af91195 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -17,7 +17,7 @@ def check_anchor_order(m): # Check anchor order against stride order for YOLOv5 Detect() module m, and correct if necessary - a = m.anchors.prod(-1).view(-1) # anchor area + a = m.anchors.prod(-1).mean(-1).view(-1) # mean anchor area per output layer da = a[-1] - a[0] # delta a ds = m.stride[-1] - m.stride[0] # delta s if da.sign() != ds.sign(): # same order From d5e363f29d7619f2a186678eb6d61672f49b11f1 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 18:02:05 +0100 Subject: [PATCH 108/661] Update detect.py non-inplace with `y.tensor_split()` (#7062) --- models/yolo.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index 2f4bbe0f71d1..09215101e8a0 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -62,9 +62,10 @@ def forward(self, x): y[..., 0:2] = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953 - xy = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy - wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh - y = torch.cat((xy, wh, y[..., 4:]), -1) + xy, wh, conf = y.tensor_split((2, 4), 4) + xy = (xy * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy + wh = (wh * 2) ** 2 * self.anchor_grid[i] # wh + y = torch.cat((xy, wh, conf), 4) z.append(y.view(bs, -1, self.no)) return x if self.training else (torch.cat(z, 1), x) From 0529b77232d72c41557fb03753caa356f583e5fc Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 18:03:37 +0100 Subject: [PATCH 109/661] Update common.py lists for tuples (#7063) Improved profiling. --- models/common.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/models/common.py b/models/common.py index 5561d92ecb73..066f8774d3c3 100644 --- a/models/common.py +++ b/models/common.py @@ -31,7 +31,7 @@ def autopad(k, p=None): # kernel, padding # Pad to 'same' if p is None: - p = k // 2 if isinstance(k, int) else [x // 2 for x in k] # auto-pad + p = k // 2 if isinstance(k, int) else (x // 2 for x in k) # auto-pad return p @@ -133,7 +133,7 @@ def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, nu self.cv2 = Conv(c1, c_, 1, 1) self.cv3 = Conv(2 * c_, c2, 1) # act=FReLU(c2) self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n))) - # self.m = nn.Sequential(*[CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n)]) + # self.m = nn.Sequential(*(CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n))) def forward(self, x): return self.cv3(torch.cat((self.m(self.cv1(x)), self.cv2(x)), dim=1)) @@ -194,7 +194,7 @@ def forward(self, x): warnings.simplefilter('ignore') # suppress torch 1.9.0 max_pool2d() warning y1 = self.m(x) y2 = self.m(y1) - return self.cv2(torch.cat([x, y1, y2, self.m(y2)], 1)) + return self.cv2(torch.cat((x, y1, y2, self.m(y2)), 1)) class Focus(nn.Module): @@ -205,7 +205,7 @@ def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True): # ch_in, ch_out, k # self.contract = Contract(gain=2) def forward(self, x): # x(b,c,w,h) -> y(b,4c,w/2,h/2) - return self.conv(torch.cat([x[..., ::2, ::2], x[..., 1::2, ::2], x[..., ::2, 1::2], x[..., 1::2, 1::2]], 1)) + return self.conv(torch.cat((x[..., ::2, ::2], x[..., 1::2, ::2], x[..., ::2, 1::2], x[..., 1::2, 1::2]), 1)) # return self.conv(self.contract(x)) @@ -219,7 +219,7 @@ def __init__(self, c1, c2, k=1, s=1, g=1, act=True): # ch_in, ch_out, kernel, s def forward(self, x): y = self.cv1(x) - return torch.cat([y, self.cv2(y)], 1) + return torch.cat((y, self.cv2(y)), 1) class GhostBottleneck(nn.Module): From e278fd63ec6c09d264c2bc983ad91717c577e97c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 18:15:22 +0100 Subject: [PATCH 110/661] Update W&B message to `LOGGER.info()` (#7064) --- utils/loggers/__init__.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 86ccf38443a9..ce0bea00e1af 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -56,7 +56,7 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, if not wandb: prefix = colorstr('Weights & Biases: ') s = f"{prefix}run 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs (RECOMMENDED)" - print(emojis(s)) + self.logger.info(emojis(s)) # TensorBoard s = self.save_dir From 9e75cbf4c18457297cd7b28653ebeb5b1262e8c9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 18:17:04 +0100 Subject: [PATCH 111/661] Update __init__.py (#7065) --- utils/loggers/__init__.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index ce0bea00e1af..866bdc4be2f5 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -47,7 +47,7 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, 'metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95', # metrics 'val/box_loss', 'val/obj_loss', 'val/cls_loss', # val loss 'x/lr0', 'x/lr1', 'x/lr2'] # params - self.best_keys = ['best/epoch', 'best/precision', 'best/recall', 'best/mAP_0.5', 'best/mAP_0.5:0.95',] + self.best_keys = ['best/epoch', 'best/precision', 'best/recall', 'best/mAP_0.5', 'best/mAP_0.5:0.95'] for k in LOGGERS: setattr(self, k, None) # init empty logger dictionary self.csv = True # always log to csv From 178c1095768a81edefc4c4ae87984ab1962e0906 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 18:37:27 +0100 Subject: [PATCH 112/661] Add non-zero `da` `check_anchor_order()` condition (#7066) --- utils/autoanchor.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/autoanchor.py b/utils/autoanchor.py index 7eb46af91195..882712d45a38 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -20,7 +20,7 @@ def check_anchor_order(m): a = m.anchors.prod(-1).mean(-1).view(-1) # mean anchor area per output layer da = a[-1] - a[0] # delta a ds = m.stride[-1] - m.stride[0] # delta s - if da.sign() != ds.sign(): # same order + if da and (da.sign() != ds.sign()): # same order LOGGER.info(f'{PREFIX}Reversing anchor order') m.anchors[:] = m.anchors.flip(0) From 9cd89b75cca8bb165a3b19c9b8356f7b3bb22b31 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 20 Mar 2022 18:55:13 +0100 Subject: [PATCH 113/661] Fix2 `check_anchor_order()` in pixel-space not grid-space (#7067) Follows https://github.com/ultralytics/yolov5/pull/7060 which provided only a partial solution to this issue. #7060 resolved occurences in yolo.py, this applies the same fix in autoanchor.py. --- utils/autoanchor.py | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/utils/autoanchor.py b/utils/autoanchor.py index 882712d45a38..77518abe9889 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -40,7 +40,8 @@ def metric(k): # compute metric bpr = (best > 1 / thr).float().mean() # best possible recall return bpr, aat - anchors = m.anchors.clone() * m.stride.to(m.anchors.device).view(-1, 1, 1) # current anchors + stride = m.stride.to(m.anchors.device).view(-1, 1, 1) # model strides + anchors = m.anchors.clone() * stride # current anchors bpr, aat = metric(anchors.cpu().view(-1, 2)) s = f'\n{PREFIX}{aat:.2f} anchors/target, {bpr:.3f} Best Possible Recall (BPR). ' if bpr > 0.98: # threshold to recompute @@ -55,8 +56,9 @@ def metric(k): # compute metric new_bpr = metric(anchors)[0] if new_bpr > bpr: # replace anchors anchors = torch.tensor(anchors, device=m.anchors.device).type_as(m.anchors) - m.anchors[:] = anchors.clone().view_as(m.anchors) / m.stride.to(m.anchors.device).view(-1, 1, 1) # loss - check_anchor_order(m) + m.anchors[:] = anchors.clone().view_as(m.anchors) + check_anchor_order(m) # must be in pixel-space (not grid-space) + m.anchors /= stride s = f'{PREFIX}Done ✅ (optional: update model *.yaml to use these anchors in the future)' else: s = f'{PREFIX}Done ⚠️ (original anchors better than new anchors, proceeding with original anchors)' From 9b771a3e7112f864cb9c877733eca9240e8fb4a5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 21 Mar 2022 09:33:39 +0100 Subject: [PATCH 114/661] Revert "Update detect.py non-inplace with `y.tensor_split()` (#7062)" (#7074) This reverts commit d5e363f29d7619f2a186678eb6d61672f49b11f1. --- models/yolo.py | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index 09215101e8a0..2f4bbe0f71d1 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -62,10 +62,9 @@ def forward(self, x): y[..., 0:2] = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953 - xy, wh, conf = y.tensor_split((2, 4), 4) - xy = (xy * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy - wh = (wh * 2) ** 2 * self.anchor_grid[i] # wh - y = torch.cat((xy, wh, conf), 4) + xy = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy + wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh + y = torch.cat((xy, wh, y[..., 4:]), -1) z.append(y.view(bs, -1, self.no)) return x if self.training else (torch.cat(z, 1), x) From 6f128031d073754ee8ed6b6a85ecb6c0619cd0a7 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 21 Mar 2022 18:35:36 +0100 Subject: [PATCH 115/661] Update loss.py with `if self.gr < 1:` (#7087) * Update loss.py with `if self.gr < 1:` * Update loss.py --- utils/loss.py | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/utils/loss.py b/utils/loss.py index 0f0137817955..b49cc7f66e66 100644 --- a/utils/loss.py +++ b/utils/loss.py @@ -139,11 +139,13 @@ def __call__(self, p, targets): # predictions, targets lbox += (1.0 - iou).mean() # iou loss # Objectness - score_iou = iou.detach().clamp(0).type(tobj.dtype) + iou = iou.detach().clamp(0).type(tobj.dtype) if self.sort_obj_iou: - sort_id = torch.argsort(score_iou) - b, a, gj, gi, score_iou = b[sort_id], a[sort_id], gj[sort_id], gi[sort_id], score_iou[sort_id] - tobj[b, a, gj, gi] = (1.0 - self.gr) + self.gr * score_iou # iou ratio + j = iou.argsort() + b, a, gj, gi, iou = b[j], a[j], gj[j], gi[j], iou[j] + if self.gr < 1: + iou = (1.0 - self.gr) + self.gr * iou + tobj[b, a, gj, gi] = iou # iou ratio # Classification if self.nc > 1: # cls loss (only if multiple classes) From a2d617ece94dcd8c9bc205ea70f1223c84fdbe3a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 21 Mar 2022 19:18:34 +0100 Subject: [PATCH 116/661] Update loss for FP16 `tobj` (#7088) --- utils/loss.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/loss.py b/utils/loss.py index b49cc7f66e66..a06330e034bc 100644 --- a/utils/loss.py +++ b/utils/loss.py @@ -125,7 +125,7 @@ def __call__(self, p, targets): # predictions, targets # Losses for i, pi in enumerate(p): # layer index, layer predictions b, a, gj, gi = indices[i] # image, anchor, gridy, gridx - tobj = torch.zeros(pi.shape[:4], device=self.device) # target obj + tobj = torch.zeros(pi.shape[:4], dtype=pi.dtype, device=self.device) # target obj n = b.shape[0] # number of targets if n: From a600baed8efc6407ec4fb7a71fa1dbe3be23d441 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 22 Mar 2022 15:41:19 +0100 Subject: [PATCH 117/661] Update model summary to display model name (#7101) --- utils/torch_utils.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index efcacc9ca735..793c9c184a44 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -229,7 +229,8 @@ def model_info(model, verbose=False, img_size=640): except (ImportError, Exception): fs = '' - LOGGER.info(f"Model Summary: {len(list(model.modules()))} layers, {n_p} parameters, {n_g} gradients{fs}") + name = model.yaml_file.rstrip('.yaml').replace('yolov5', 'YOLOv5') if hasattr(model, 'yaml_file') else 'Model' + LOGGER.info(f"{name} summary: {len(list(model.modules()))} layers, {n_p} parameters, {n_g} gradients{fs}") def scale_img(img, ratio=1.0, same_shape=False, gs=32): # img(16,3,256,416) From 05aae1733352289e4c4dca031159df7f0354d049 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 22 Mar 2022 17:36:05 +0100 Subject: [PATCH 118/661] `torch.split()` 1.7.0 compatibility fix (#7102) * Update loss.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update loss.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/loss.py | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/utils/loss.py b/utils/loss.py index a06330e034bc..bf9b592d4ad2 100644 --- a/utils/loss.py +++ b/utils/loss.py @@ -108,13 +108,15 @@ def __init__(self, model, autobalance=False): if g > 0: BCEcls, BCEobj = FocalLoss(BCEcls, g), FocalLoss(BCEobj, g) - det = de_parallel(model).model[-1] # Detect() module - self.balance = {3: [4.0, 1.0, 0.4]}.get(det.nl, [4.0, 1.0, 0.25, 0.06, 0.02]) # P3-P7 - self.ssi = list(det.stride).index(16) if autobalance else 0 # stride 16 index + m = de_parallel(model).model[-1] # Detect() module + self.balance = {3: [4.0, 1.0, 0.4]}.get(m.nl, [4.0, 1.0, 0.25, 0.06, 0.02]) # P3-P7 + self.ssi = list(m.stride).index(16) if autobalance else 0 # stride 16 index self.BCEcls, self.BCEobj, self.gr, self.hyp, self.autobalance = BCEcls, BCEobj, 1.0, h, autobalance + self.na = m.na # number of anchors + self.nc = m.nc # number of classes + self.nl = m.nl # number of layers + self.anchors = m.anchors self.device = device - for k in 'na', 'nc', 'nl', 'anchors': - setattr(self, k, getattr(det, k)) def __call__(self, p, targets): # predictions, targets lcls = torch.zeros(1, device=self.device) # class loss @@ -129,7 +131,8 @@ def __call__(self, p, targets): # predictions, targets n = b.shape[0] # number of targets if n: - pxy, pwh, _, pcls = pi[b, a, gj, gi].tensor_split((2, 4, 5), dim=1) # target-subset of predictions + # pxy, pwh, _, pcls = pi[b, a, gj, gi].tensor_split((2, 4, 5), dim=1) # faster, requires torch 1.8.0 + pxy, pwh, _, pcls = pi[b, a, gj, gi].split((2, 2, 1, self.nc), 1) # target-subset of predictions # Regression pxy = pxy.sigmoid() * 2 - 0.5 From ee0b3b2a953bd50ba29b39119a09ef9521596416 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 22 Mar 2022 18:02:35 +0100 Subject: [PATCH 119/661] Update benchmarks significant digits (#7103) --- utils/benchmarks.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/benchmarks.py b/utils/benchmarks.py index bdbbdc43b639..446248c03f68 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -67,7 +67,7 @@ def run(weights=ROOT / 'yolov5s.pt', # weights path result = val.run(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) metrics = result[0] # metrics (mp, mr, map50, map, *losses(box, obj, cls)) speeds = result[2] # times (preprocess, inference, postprocess) - y.append([name, metrics[3], speeds[1]]) # mAP, t_inference + y.append([name, round(metrics[3], 4), round(speeds[1], 2)]) # mAP, t_inference except Exception as e: LOGGER.warning(f'WARNING: Benchmark failure for {name}: {e}') y.append([name, None, None]) # mAP, t_inference From 6134ec5d9484ac9ac743121b1c74709e93c68a88 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 22 Mar 2022 20:05:07 +0100 Subject: [PATCH 120/661] Model summary `pathlib` fix (#7104) Stems not working correctly for YOLOv5l with current .rstrip() implementation. After fix: ``` YOLOv5l summary: 468 layers, 46563709 parameters, 46563709 gradients, 109.3 GFLOPs ``` --- utils/torch_utils.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 793c9c184a44..72f8a0fd1659 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -11,6 +11,7 @@ import warnings from contextlib import contextmanager from copy import deepcopy +from pathlib import Path import torch import torch.distributed as dist @@ -229,7 +230,7 @@ def model_info(model, verbose=False, img_size=640): except (ImportError, Exception): fs = '' - name = model.yaml_file.rstrip('.yaml').replace('yolov5', 'YOLOv5') if hasattr(model, 'yaml_file') else 'Model' + name = Path(model.yaml_file).stem.replace('yolov5', 'YOLOv5') if hasattr(model, 'yaml_file') else 'Model' LOGGER.info(f"{name} summary: {len(list(model.modules()))} layers, {n_p} parameters, {n_g} gradients{fs}") From ecc2c7ba73e71211b192cba69e255afad92de67a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 22 Mar 2022 20:44:07 +0100 Subject: [PATCH 121/661] Remove named arguments where possible (#7105) * Remove named arguments where possible Speed improvements. * Update yolo.py * Update yolo.py * Update yolo.py --- models/common.py | 14 +++++++------- models/yolo.py | 10 +++++----- 2 files changed, 12 insertions(+), 12 deletions(-) diff --git a/models/common.py b/models/common.py index 066f8774d3c3..0286c74fe8cd 100644 --- a/models/common.py +++ b/models/common.py @@ -121,7 +121,7 @@ def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, nu def forward(self, x): y1 = self.cv3(self.m(self.cv1(x))) y2 = self.cv2(x) - return self.cv4(self.act(self.bn(torch.cat((y1, y2), dim=1)))) + return self.cv4(self.act(self.bn(torch.cat((y1, y2), 1)))) class C3(nn.Module): @@ -136,7 +136,7 @@ def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, nu # self.m = nn.Sequential(*(CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n))) def forward(self, x): - return self.cv3(torch.cat((self.m(self.cv1(x)), self.cv2(x)), dim=1)) + return self.cv3(torch.cat((self.m(self.cv1(x)), self.cv2(x)), 1)) class C3TR(C3): @@ -527,7 +527,7 @@ def forward(self, imgs, size=640, augment=False, profile=False): p = next(self.model.parameters()) if self.pt else torch.zeros(1) # for device and type autocast = self.amp and (p.device.type != 'cpu') # Automatic Mixed Precision (AMP) inference if isinstance(imgs, torch.Tensor): # torch - with amp.autocast(enabled=autocast): + with amp.autocast(autocast): return self.model(imgs.to(p.device).type_as(p), augment, profile) # inference # Pre-process @@ -550,19 +550,19 @@ def forward(self, imgs, size=640, augment=False, profile=False): shape1.append([y * g for y in s]) imgs[i] = im if im.data.contiguous else np.ascontiguousarray(im) # update shape1 = [make_divisible(x, self.stride) if self.pt else size for x in np.array(shape1).max(0)] # inf shape - x = [letterbox(im, new_shape=shape1, auto=False)[0] for im in imgs] # pad + x = [letterbox(im, shape1, auto=False)[0] for im in imgs] # pad x = np.ascontiguousarray(np.array(x).transpose((0, 3, 1, 2))) # stack and BHWC to BCHW x = torch.from_numpy(x).to(p.device).type_as(p) / 255 # uint8 to fp16/32 t.append(time_sync()) - with amp.autocast(enabled=autocast): + with amp.autocast(autocast): # Inference y = self.model(x, augment, profile) # forward t.append(time_sync()) # Post-process - y = non_max_suppression(y if self.dmb else y[0], self.conf, iou_thres=self.iou, classes=self.classes, - agnostic=self.agnostic, multi_label=self.multi_label, max_det=self.max_det) # NMS + y = non_max_suppression(y if self.dmb else y[0], self.conf, self.iou, self.classes, self.agnostic, + self.multi_label, max_det=self.max_det) # NMS for i in range(n): scale_coords(shape1, y[i][:, :4], shape0[i]) diff --git a/models/yolo.py b/models/yolo.py index 2f4bbe0f71d1..9f4701c49f9d 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -71,13 +71,13 @@ def forward(self, x): def _make_grid(self, nx=20, ny=20, i=0): d = self.anchors[i].device + shape = 1, self.na, ny, nx, 2 # grid shape if check_version(torch.__version__, '1.10.0'): # torch>=1.10.0 meshgrid workaround for torch>=0.7 compatibility - yv, xv = torch.meshgrid([torch.arange(ny, device=d), torch.arange(nx, device=d)], indexing='ij') + yv, xv = torch.meshgrid(torch.arange(ny, device=d), torch.arange(nx, device=d), indexing='ij') else: - yv, xv = torch.meshgrid([torch.arange(ny, device=d), torch.arange(nx, device=d)]) - grid = torch.stack((xv, yv), 2).expand((1, self.na, ny, nx, 2)).float() - anchor_grid = (self.anchors[i].clone() * self.stride[i]) \ - .view((1, self.na, 1, 1, 2)).expand((1, self.na, ny, nx, 2)).float() + yv, xv = torch.meshgrid(torch.arange(ny, device=d), torch.arange(nx, device=d)) + grid = torch.stack((xv, yv), 2).expand(shape).float() + anchor_grid = (self.anchors[i] * self.stride[i]).view((1, self.na, 1, 1, 2)).expand(shape).float() return grid, anchor_grid From c3ae4e4af6f75aff537b876adc11da3de441dd60 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 23 Mar 2022 01:19:37 +0100 Subject: [PATCH 122/661] Multi-threaded VisDrone and VOC downloads (#7108) * Multi-threaded VOC download * Update VOC.yaml * Update * Update general.py * Update general.py --- data/GlobalWheat2020.yaml | 1 + data/Objects365.yaml | 1 + data/SKU-110K.yaml | 1 + data/VOC.yaml | 2 +- data/VisDrone.yaml | 2 +- data/coco.yaml | 1 + utils/general.py | 11 +++++++---- 7 files changed, 13 insertions(+), 6 deletions(-) diff --git a/data/GlobalWheat2020.yaml b/data/GlobalWheat2020.yaml index 869dace0be2b..c1ba289f2833 100644 --- a/data/GlobalWheat2020.yaml +++ b/data/GlobalWheat2020.yaml @@ -34,6 +34,7 @@ names: ['wheat_head'] # class names download: | from utils.general import download, Path + # Download dir = Path(yaml['path']) # dataset root dir urls = ['https://zenodo.org/record/4298502/files/global-wheat-codalab-official.zip', diff --git a/data/Objects365.yaml b/data/Objects365.yaml index 4c7cf3fdb2c8..bd6e5d6e1144 100644 --- a/data/Objects365.yaml +++ b/data/Objects365.yaml @@ -65,6 +65,7 @@ download: | from utils.general import Path, download, np, xyxy2xywhn + # Make Directories dir = Path(yaml['path']) # dataset root dir for p in 'images', 'labels': diff --git a/data/SKU-110K.yaml b/data/SKU-110K.yaml index 9481b7a04aee..46459eab6bb7 100644 --- a/data/SKU-110K.yaml +++ b/data/SKU-110K.yaml @@ -24,6 +24,7 @@ download: | from tqdm import tqdm from utils.general import np, pd, Path, download, xyxy2xywh + # Download dir = Path(yaml['path']) # dataset root dir parent = Path(dir.parent) # download dir diff --git a/data/VOC.yaml b/data/VOC.yaml index 975d56466de1..be04fb1e2ecb 100644 --- a/data/VOC.yaml +++ b/data/VOC.yaml @@ -62,7 +62,7 @@ download: | urls = [url + 'VOCtrainval_06-Nov-2007.zip', # 446MB, 5012 images url + 'VOCtest_06-Nov-2007.zip', # 438MB, 4953 images url + 'VOCtrainval_11-May-2012.zip'] # 1.95GB, 17126 images - download(urls, dir=dir / 'images', delete=False) + download(urls, dir=dir / 'images', delete=False, threads=3) # Convert path = dir / f'images/VOCdevkit' diff --git a/data/VisDrone.yaml b/data/VisDrone.yaml index 83a5c7d55e06..2a3b2f03e674 100644 --- a/data/VisDrone.yaml +++ b/data/VisDrone.yaml @@ -54,7 +54,7 @@ download: | 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-val.zip', 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-test-dev.zip', 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-test-challenge.zip'] - download(urls, dir=dir) + download(urls, dir=dir, threads=4) # Convert for d in 'VisDrone2019-DET-train', 'VisDrone2019-DET-val', 'VisDrone2019-DET-test-dev': diff --git a/data/coco.yaml b/data/coco.yaml index 3ed7e48a2185..7494fc2f9cd1 100644 --- a/data/coco.yaml +++ b/data/coco.yaml @@ -30,6 +30,7 @@ names: ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 't download: | from utils.general import download, Path + # Download labels segments = False # segment or box labels dir = Path(yaml['path']) # dataset root dir diff --git a/utils/general.py b/utils/general.py index e8b3b05c5fe1..b0c5e9d69ab7 100755 --- a/utils/general.py +++ b/utils/general.py @@ -449,8 +449,9 @@ def check_dataset(data, autodownload=True): if val: val = [Path(x).resolve() for x in (val if isinstance(val, list) else [val])] # val path if not all(x.exists() for x in val): - LOGGER.info('\nDataset not found, missing paths: %s' % [str(x) for x in val if not x.exists()]) + LOGGER.info(emojis('\nDataset not found ⚠️, missing paths %s' % [str(x) for x in val if not x.exists()])) if s and autodownload: # download script + t = time.time() root = path.parent if 'path' in data else '..' # unzip directory i.e. '../' if s.startswith('http') and s.endswith('.zip'): # URL f = Path(s).name # filename @@ -465,9 +466,11 @@ def check_dataset(data, autodownload=True): r = os.system(s) else: # python script r = exec(s, {'yaml': data}) # return None - LOGGER.info(f"Dataset autodownload {f'success, saved to {root}' if r in (0, None) else 'failure'}\n") + dt = f'({round(time.time() - t, 1)}s)' + s = f"success ✅ {dt}, saved to {colorstr('bold', root)}" if r in (0, None) else f"failure {dt} ❌" + LOGGER.info(emojis(f"Dataset download {s}")) else: - raise Exception('Dataset not found.') + raise Exception(emojis('Dataset not found ❌')) return data # dictionary @@ -491,7 +494,7 @@ def download_one(url, dir): if curl: os.system(f"curl -L '{url}' -o '{f}' --retry 9 -C -") # curl download, retry and resume on fail else: - torch.hub.download_url_to_file(url, f, progress=True) # torch download + torch.hub.download_url_to_file(url, f, progress=threads == 1) # torch download if unzip and f.suffix in ('.zip', '.gz'): LOGGER.info(f'Unzipping {f}...') if f.suffix == '.zip': From bc3ed957ce0f0990a3cb408e462197b83b0d075f Mon Sep 17 00:00:00 2001 From: yeshanliu <41566254+yeshanliu@users.noreply.github.com> Date: Wed, 23 Mar 2022 22:35:15 +0800 Subject: [PATCH 123/661] `np.fromfile()` Chinese image paths fix (#6979) * :tada: :new: now can read Chinese image path. use "cv2.imdecode(np.fromfile(f, np.uint8), cv2.IMREAD_COLOR)" instead of "cv2.imread(f)" for Chinese image path. * Update datasets.py * Update __init__.py Co-authored-by: Glenn Jocher --- utils/datasets.py | 3 +++ utils/loggers/__init__.py | 3 +++ 2 files changed, 6 insertions(+) diff --git a/utils/datasets.py b/utils/datasets.py index 8627344af7b4..f212e54633be 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -32,6 +32,9 @@ segments2boxes, xyn2xy, xywh2xyxy, xywhn2xyxy, xyxy2xywhn) from utils.torch_utils import torch_distributed_zero_first +# Remap +cv2.imread = lambda x: cv2.imdecode(np.fromfile(x, np.uint8), cv2.IMREAD_COLOR) # for Chinese filenames + # Parameters HELP_URL = 'https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' IMG_FORMATS = 'bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp' # include image suffixes diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 866bdc4be2f5..ff6722ecd48a 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -148,6 +148,9 @@ def on_train_end(self, last, best, plots, epoch, results): if self.tb: import cv2 + import numpy as np + + cv2.imread = lambda x: cv2.imdecode(np.fromfile(x, np.uint8), cv2.IMREAD_COLOR) # remap for Chinese files for f in files: self.tb.add_image(f.stem, cv2.imread(str(f))[..., ::-1], epoch, dataformats='HWC') From a0a4adf6de4de3d9d5ac00c23796c844a8e57200 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 24 Mar 2022 11:31:22 +0100 Subject: [PATCH 124/661] Add PyTorch Hub `results.save(labels=False)` option (#7129) Resolves https://github.com/ultralytics/yolov5/issues/388#issuecomment-1077121821 --- models/common.py | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/models/common.py b/models/common.py index 0286c74fe8cd..115e3c3145ff 100644 --- a/models/common.py +++ b/models/common.py @@ -131,7 +131,7 @@ def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, nu c_ = int(c2 * e) # hidden channels self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c1, c_, 1, 1) - self.cv3 = Conv(2 * c_, c2, 1) # act=FReLU(c2) + self.cv3 = Conv(2 * c_, c2, 1) # optional act=FReLU(c2) self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n))) # self.m = nn.Sequential(*(CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n))) @@ -589,7 +589,7 @@ def __init__(self, imgs, pred, files, times=(0, 0, 0, 0), names=None, shape=None self.t = tuple((times[i + 1] - times[i]) * 1000 / self.n for i in range(3)) # timestamps (ms) self.s = shape # inference BCHW shape - def display(self, pprint=False, show=False, save=False, crop=False, render=False, save_dir=Path('')): + def display(self, pprint=False, show=False, save=False, crop=False, render=False, labels=True, save_dir=Path('')): crops = [] for i, (im, pred) in enumerate(zip(self.imgs, self.pred)): s = f'image {i + 1}/{len(self.pred)}: {im.shape[0]}x{im.shape[1]} ' # string @@ -606,7 +606,7 @@ def display(self, pprint=False, show=False, save=False, crop=False, render=False crops.append({'box': box, 'conf': conf, 'cls': cls, 'label': label, 'im': save_one_box(box, im, file=file, save=save)}) else: # all others - annotator.box_label(box, label, color=colors(cls)) + annotator.box_label(box, label if labels else '', color=colors(cls)) im = annotator.im else: s += '(no detections)' @@ -633,19 +633,19 @@ def print(self): LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {tuple(self.s)}' % self.t) - def show(self): - self.display(show=True) # show results + def show(self, labels=True): + self.display(show=True, labels=labels) # show results - def save(self, save_dir='runs/detect/exp'): + def save(self, labels=True, save_dir='runs/detect/exp'): save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) # increment save_dir - self.display(save=True, save_dir=save_dir) # save results + self.display(save=True, labels=labels, save_dir=save_dir) # save results def crop(self, save=True, save_dir='runs/detect/exp'): save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) if save else None return self.display(crop=True, save=save, save_dir=save_dir) # crop results - def render(self): - self.display(render=True) # render results + def render(self, labels=True): + self.display(render=True, labels=labels) # render results return self.imgs def pandas(self): From d115bbf509aa86ed553d1dc57c078c842393cca8 Mon Sep 17 00:00:00 2001 From: RcINS Date: Fri, 25 Mar 2022 20:25:30 +0800 Subject: [PATCH 125/661] Fix `cv2.imwrite` on non-ASCII paths (#7139) * Fix imwrite on non-ASCII paths * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update general.py * Update __init__.py * Update __init__.py * Update datasets.py * Update hubconf.py * Update detect.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update general.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- detect.py | 3 +-- hubconf.py | 3 ++- utils/datasets.py | 6 +----- utils/general.py | 17 ++++++++++++++++- utils/loggers/__init__.py | 6 +----- 5 files changed, 21 insertions(+), 14 deletions(-) diff --git a/detect.py b/detect.py index ccb9fbf5103f..046f7ae57b5c 100644 --- a/detect.py +++ b/detect.py @@ -29,7 +29,6 @@ import sys from pathlib import Path -import cv2 import torch import torch.backends.cudnn as cudnn @@ -41,7 +40,7 @@ from models.common import DetectMultiBackend from utils.datasets import IMG_FORMATS, VID_FORMATS, LoadImages, LoadStreams -from utils.general import (LOGGER, check_file, check_img_size, check_imshow, check_requirements, colorstr, +from utils.general import (LOGGER, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, increment_path, non_max_suppression, print_args, scale_coords, strip_optimizer, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box from utils.torch_utils import select_device, time_sync diff --git a/hubconf.py b/hubconf.py index 39fa614b2e34..d719b80034af 100644 --- a/hubconf.py +++ b/hubconf.py @@ -127,10 +127,11 @@ def yolov5x6(pretrained=True, channels=3, classes=80, autoshape=True, verbose=Tr # Verify inference from pathlib import Path - import cv2 import numpy as np from PIL import Image + from utils.general import cv2 + imgs = ['data/images/zidane.jpg', # filename Path('data/images/zidane.jpg'), # Path 'https://ultralytics.com/images/zidane.jpg', # URI diff --git a/utils/datasets.py b/utils/datasets.py index f212e54633be..d0b35e808000 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -18,7 +18,6 @@ from urllib.parse import urlparse from zipfile import ZipFile -import cv2 import numpy as np import torch import torch.nn.functional as F @@ -29,12 +28,9 @@ from utils.augmentations import Albumentations, augment_hsv, copy_paste, letterbox, mixup, random_perspective from utils.general import (DATASETS_DIR, LOGGER, NUM_THREADS, check_dataset, check_requirements, check_yaml, clean_str, - segments2boxes, xyn2xy, xywh2xyxy, xywhn2xyxy, xyxy2xywhn) + cv2, segments2boxes, xyn2xy, xywh2xyxy, xywhn2xyxy, xyxy2xywhn) from utils.torch_utils import torch_distributed_zero_first -# Remap -cv2.imread = lambda x: cv2.imdecode(np.fromfile(x, np.uint8), cv2.IMREAD_COLOR) # for Chinese filenames - # Parameters HELP_URL = 'https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' IMG_FORMATS = 'bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp' # include image suffixes diff --git a/utils/general.py b/utils/general.py index b0c5e9d69ab7..dcdb58cb0f51 100755 --- a/utils/general.py +++ b/utils/general.py @@ -904,5 +904,20 @@ def increment_path(path, exist_ok=False, sep='', mkdir=False): return path -# Variables +# OpenCV Chinese-friendly functions ------------------------------------------------------------------------------------ +def imread(path): + return cv2.imdecode(np.fromfile(path, np.uint8), cv2.IMREAD_COLOR) + + +def imwrite(path, im): + try: + cv2.imencode(Path(path).suffix, im)[1].tofile(path) + return True + except Exception: + return False + + +cv2.imread, cv2.imwrite = imread, imwrite # redefine + +# Variables ------------------------------------------------------------------------------------------------------------ NCOLS = 0 if is_docker() else shutil.get_terminal_size().columns # terminal window size for tqdm diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index ff6722ecd48a..bb8523c0219e 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -11,7 +11,7 @@ import torch from torch.utils.tensorboard import SummaryWriter -from utils.general import colorstr, emojis +from utils.general import colorstr, cv2, emojis from utils.loggers.wandb.wandb_utils import WandbLogger from utils.plots import plot_images, plot_results from utils.torch_utils import de_parallel @@ -147,10 +147,6 @@ def on_train_end(self, last, best, plots, epoch, results): files = [(self.save_dir / f) for f in files if (self.save_dir / f).exists()] # filter if self.tb: - import cv2 - import numpy as np - - cv2.imread = lambda x: cv2.imdecode(np.fromfile(x, np.uint8), cv2.IMREAD_COLOR) # remap for Chinese files for f in files: self.tb.add_image(f.stem, cv2.imread(str(f))[..., ::-1], epoch, dataformats='HWC') From a4c661873f1edfe3d687bd01c4477e56739c7db3 Mon Sep 17 00:00:00 2001 From: Zengyf-CVer <41098760+Zengyf-CVer@users.noreply.github.com> Date: Fri, 25 Mar 2022 20:40:55 +0800 Subject: [PATCH 126/661] Fix `detect.py --view-img` for non-ASCII paths (#7093) * Update detect.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update general.py * Update detect.py * Update general.py * Update general.py * Update general.py * Update general.py * Update general.py * Update general.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update general.py * Update general.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- utils/general.py | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index dcdb58cb0f51..45e23edff062 100755 --- a/utils/general.py +++ b/utils/general.py @@ -905,6 +905,9 @@ def increment_path(path, exist_ok=False, sep='', mkdir=False): # OpenCV Chinese-friendly functions ------------------------------------------------------------------------------------ +imshow_ = cv2.imshow # copy to avoid recursion errors + + def imread(path): return cv2.imdecode(np.fromfile(path, np.uint8), cv2.IMREAD_COLOR) @@ -917,7 +920,11 @@ def imwrite(path, im): return False -cv2.imread, cv2.imwrite = imread, imwrite # redefine +def imshow(path, im): + imshow_(path.encode('unicode_escape').decode(), im) + + +cv2.imread, cv2.imwrite, cv2.imshow = imread, imwrite, imshow # redefine # Variables ------------------------------------------------------------------------------------------------------------ NCOLS = 0 if is_docker() else shutil.get_terminal_size().columns # terminal window size for tqdm From 7a2a11893b56c67903f0dc4e313235e544189601 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 25 Mar 2022 14:45:23 +0100 Subject: [PATCH 127/661] Add Architecture Summary to README Tutorials (#7146) * Add Architecture Summary to README Tutorials Per https://github.com/ultralytics/yolov5/issues/6998#issuecomment-1073517405 * Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 3ebc085b6c33..54c5cbd83f5b 100644 --- a/README.md +++ b/README.md @@ -162,7 +162,7 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 * [Model Pruning/Sparsity](https://github.com/ultralytics/yolov5/issues/304) * [Hyperparameter Evolution](https://github.com/ultralytics/yolov5/issues/607) * [Transfer Learning with Frozen Layers](https://github.com/ultralytics/yolov5/issues/1314)  ⭐ NEW -* [TensorRT Deployment](https://github.com/wang-xinyu/tensorrtx) +* [Architecture Summary](https://github.com/ultralytics/yolov5/issues/6998)  ⭐ NEW From 26bfd4446559814ab5b1a2fa34584dcb3a49da6c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 26 Mar 2022 11:45:28 +0100 Subject: [PATCH 128/661] Adjust NMS time limit warning to batch size (#7156) --- utils/general.py | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/utils/general.py b/utils/general.py index 45e23edff062..e1751c4bb62d 100755 --- a/utils/general.py +++ b/utils/general.py @@ -709,6 +709,7 @@ def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=Non list of detections, on (n,6) tensor per image [xyxy, conf, cls] """ + bs = prediction.shape[0] # batch size nc = prediction.shape[2] - 5 # number of classes xc = prediction[..., 4] > conf_thres # candidates @@ -719,13 +720,13 @@ def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=Non # Settings min_wh, max_wh = 2, 7680 # (pixels) minimum and maximum box width and height max_nms = 30000 # maximum number of boxes into torchvision.ops.nms() - time_limit = 10.0 # seconds to quit after + time_limit = 0.030 * bs # seconds to quit after redundant = True # require redundant detections multi_label &= nc > 1 # multiple labels per box (adds 0.5ms/img) merge = False # use merge-NMS - t = time.time() - output = [torch.zeros((0, 6), device=prediction.device)] * prediction.shape[0] + t, warn_time = time.time(), True + output = [torch.zeros((0, 6), device=prediction.device)] * bs for xi, x in enumerate(prediction): # image index, image inference # Apply constraints x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0 # width-height @@ -789,7 +790,9 @@ def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=Non output[xi] = x[i] if (time.time() - t) > time_limit: - LOGGER.warning(f'WARNING: NMS time limit {time_limit}s exceeded') + if warn_time: + LOGGER.warning(f'WARNING: NMS time limit {time_limit:3f}s exceeded') + warn_time = False break # time limit exceeded return output From e19f87eb4bcdc01ee0570cf283fb3d031dbe5451 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 26 Mar 2022 14:18:53 +0100 Subject: [PATCH 129/661] Sidestep `os.path.relpath()` Windows bug (#7158) * Sidestep os.path.relpath() Windows bug os.path.relpath() seems to have a major bug on Windows due to Windows horrible path handling. This fix attempts to sidestep the issue. ``` File "C:\Users\mkokg/.cache\torch\hub\ultralytics_yolov5_master\export.py", line 64, in ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative File "C:\Users\mkokg\AppData\Local\Programs\Python\Python310\lib\ntpath.py", line 718, in relpath raise ValueError("path is on mount %r, start on mount %r" % ( ValueError: path is on mount 'C:', start on mount 'D:' ``` * Update yolo.py * Update yolo.py * Update yolo.py * Update export.py --- export.py | 3 ++- models/yolo.py | 5 ++++- 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/export.py b/export.py index 2d4a68e62f89..7517dc4678da 100644 --- a/export.py +++ b/export.py @@ -61,7 +61,8 @@ ROOT = FILE.parents[0] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH -ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative +if platform.system() != 'Windows': + ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.common import Conv from models.experimental import attempt_load diff --git a/models/yolo.py b/models/yolo.py index 9f4701c49f9d..11e17d28fa47 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -7,6 +7,8 @@ """ import argparse +import os +import platform import sys from copy import deepcopy from pathlib import Path @@ -15,7 +17,8 @@ ROOT = FILE.parents[1] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH -# ROOT = ROOT.relative_to(Path.cwd()) # relative +if platform.system() != 'Windows': + ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.common import * from models.experimental import * From 3373aab56c28ce2160d6e8f09035db49061a2619 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 26 Mar 2022 16:52:58 +0100 Subject: [PATCH 130/661] NMS unused variable fix (#7161) * NMS unused variable fix * Update general.py --- utils/general.py | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) diff --git a/utils/general.py b/utils/general.py index e1751c4bb62d..5905211cfa59 100755 --- a/utils/general.py +++ b/utils/general.py @@ -703,7 +703,7 @@ def clip_coords(boxes, shape): def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=None, agnostic=False, multi_label=False, labels=(), max_det=300): - """Runs Non-Maximum Suppression (NMS) on inference results + """Non-Maximum Suppression (NMS) on inference results to reject overlapping bounding boxes Returns: list of detections, on (n,6) tensor per image [xyxy, conf, cls] @@ -718,18 +718,19 @@ def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=Non assert 0 <= iou_thres <= 1, f'Invalid IoU {iou_thres}, valid values are between 0.0 and 1.0' # Settings - min_wh, max_wh = 2, 7680 # (pixels) minimum and maximum box width and height + # min_wh = 2 # (pixels) minimum box width and height + max_wh = 7680 # (pixels) maximum box width and height max_nms = 30000 # maximum number of boxes into torchvision.ops.nms() time_limit = 0.030 * bs # seconds to quit after redundant = True # require redundant detections multi_label &= nc > 1 # multiple labels per box (adds 0.5ms/img) merge = False # use merge-NMS - t, warn_time = time.time(), True + t = time.time() output = [torch.zeros((0, 6), device=prediction.device)] * bs for xi, x in enumerate(prediction): # image index, image inference # Apply constraints - x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0 # width-height + # x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0 # width-height x = x[xc[xi]] # confidence # Cat apriori labels if autolabelling @@ -790,9 +791,7 @@ def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=Non output[xi] = x[i] if (time.time() - t) > time_limit: - if warn_time: - LOGGER.warning(f'WARNING: NMS time limit {time_limit:3f}s exceeded') - warn_time = False + LOGGER.warning(f'WARNING: NMS time limit {time_limit:.3f}s exceeded') break # time limit exceeded return output From 7830e91b9aec29180de9505316f8c8de607a6014 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 26 Mar 2022 16:53:42 +0100 Subject: [PATCH 131/661] `yolo.py --profile` default GPU batch size 16 --- models/yolo.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/yolo.py b/models/yolo.py index 11e17d28fa47..fb01aaafedcf 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -314,7 +314,7 @@ def parse_model(d, ch): # model_dict, input_channels(3) # Profile if opt.profile: - img = torch.rand(8 if torch.cuda.is_available() else 1, 3, 640, 640).to(device) + img = torch.rand(16 if torch.cuda.is_available() else 1, 3, 640, 640).to(device) y = model(img, profile=True) # Test all models From b2194b90156e74e5a1480cd2457d1b41ec2dc181 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 27 Mar 2022 20:24:42 +0200 Subject: [PATCH 132/661] `yolo.py --profile` updates (#7170) --- models/yolo.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index fb01aaafedcf..e88db79ca8c7 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -310,11 +310,11 @@ def parse_model(d, ch): # model_dict, input_channels(3) # Create model model = Model(opt.cfg).to(device) - model.train() # Profile if opt.profile: - img = torch.rand(16 if torch.cuda.is_available() else 1, 3, 640, 640).to(device) + model.eval().fuse() + img = torch.rand(8 if torch.cuda.is_available() else 1, 3, 640, 640).to(device) y = model(img, profile=True) # Test all models From 1832264dd684256715384dd12e6c40696c89d903 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 28 Mar 2022 02:26:44 +0200 Subject: [PATCH 133/661] Update --- models/common.py | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/models/common.py b/models/common.py index 115e3c3145ff..5dd4843ed66d 100644 --- a/models/common.py +++ b/models/common.py @@ -124,6 +124,9 @@ def forward(self, x): return self.cv4(self.act(self.bn(torch.cat((y1, y2), 1)))) +from models.experimental import CrossConv + + class C3(nn.Module): # CSP Bottleneck with 3 convolutions def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, number, shortcut, groups, expansion @@ -132,8 +135,8 @@ def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, nu self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c1, c_, 1, 1) self.cv3 = Conv(2 * c_, c2, 1) # optional act=FReLU(c2) - self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n))) - # self.m = nn.Sequential(*(CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n))) + # self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n))) + self.m = nn.Sequential(*(CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n))) def forward(self, x): return self.cv3(torch.cat((self.m(self.cv1(x)), self.cv2(x)), 1)) From ee77632393b5f0989e92f39d2c3aeef9d4ebf0e6 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 28 Mar 2022 02:31:00 +0200 Subject: [PATCH 134/661] Revert `C3()` change (#7172) --- models/common.py | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/models/common.py b/models/common.py index 5dd4843ed66d..115e3c3145ff 100644 --- a/models/common.py +++ b/models/common.py @@ -124,9 +124,6 @@ def forward(self, x): return self.cv4(self.act(self.bn(torch.cat((y1, y2), 1)))) -from models.experimental import CrossConv - - class C3(nn.Module): # CSP Bottleneck with 3 convolutions def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, number, shortcut, groups, expansion @@ -135,8 +132,8 @@ def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5): # ch_in, ch_out, nu self.cv1 = Conv(c1, c_, 1, 1) self.cv2 = Conv(c1, c_, 1, 1) self.cv3 = Conv(2 * c_, c2, 1) # optional act=FReLU(c2) - # self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n))) - self.m = nn.Sequential(*(CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n))) + self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n))) + # self.m = nn.Sequential(*(CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n))) def forward(self, x): return self.cv3(torch.cat((self.m(self.cv1(x)), self.cv2(x)), 1)) From d51f9b2ff6e60b7eaaafc7e8d991f0d6dbb786cd Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Mon, 28 Mar 2022 10:42:19 +0200 Subject: [PATCH 135/661] Bump actions/cache from 2.1.7 to 3 (#7175) Bumps [actions/cache](https://github.com/actions/cache) from 2.1.7 to 3. - [Release notes](https://github.com/actions/cache/releases) - [Commits](https://github.com/actions/cache/compare/v2.1.7...v3) --- updated-dependencies: - dependency-name: actions/cache dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> --- .github/workflows/ci-testing.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index f2096ce17a17..59193e05e08c 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -39,7 +39,7 @@ jobs: python -c "from pip._internal.locations import USER_CACHE_DIR; print('::set-output name=dir::' + USER_CACHE_DIR)" - name: Cache pip - uses: actions/cache@v2.1.7 + uses: actions/cache@v3 with: path: ${{ steps.pip-cache.outputs.dir }} key: ${{ runner.os }}-${{ matrix.python-version }}-pip-${{ hashFiles('requirements.txt') }} From cf4f3c3455d14c62e11dcce9f1d30211509da72f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 29 Mar 2022 10:15:53 +0200 Subject: [PATCH 136/661] yolo.py profiling updates (#7178) * yolo.py profiling updates * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- models/yolo.py | 26 ++++++++++++-------------- 1 file changed, 12 insertions(+), 14 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index e88db79ca8c7..81ab539deffa 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -25,7 +25,8 @@ from utils.autoanchor import check_anchor_order from utils.general import LOGGER, check_version, check_yaml, make_divisible, print_args from utils.plots import feature_visualization -from utils.torch_utils import fuse_conv_and_bn, initialize_weights, model_info, scale_img, select_device, time_sync +from utils.torch_utils import (fuse_conv_and_bn, initialize_weights, model_info, profile, scale_img, select_device, + time_sync) try: import thop # for FLOPs computation @@ -300,8 +301,10 @@ def parse_model(d, ch): # model_dict, input_channels(3) if __name__ == '__main__': parser = argparse.ArgumentParser() parser.add_argument('--cfg', type=str, default='yolov5s.yaml', help='model.yaml') + parser.add_argument('--batch-size', type=int, default=1, help='total batch size for all GPUs') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--profile', action='store_true', help='profile model speed') + parser.add_argument('--line-profile', action='store_true', help='profile model speed layer by layer') parser.add_argument('--test', action='store_true', help='test all yolo*.yaml') opt = parser.parse_args() opt.cfg = check_yaml(opt.cfg) # check YAML @@ -309,24 +312,19 @@ def parse_model(d, ch): # model_dict, input_channels(3) device = select_device(opt.device) # Create model + im = torch.rand(opt.batch_size, 3, 640, 640).to(device) model = Model(opt.cfg).to(device) - # Profile - if opt.profile: - model.eval().fuse() - img = torch.rand(8 if torch.cuda.is_available() else 1, 3, 640, 640).to(device) - y = model(img, profile=True) + # Options + if opt.line_profile: # profile layer by layer + _ = model(im, profile=True) - # Test all models - if opt.test: + elif opt.profile: # profile forward-backward + results = profile(input=im, ops=[model], n=3) + + elif opt.test: # test all models for cfg in Path(ROOT / 'models').rglob('yolo*.yaml'): try: _ = Model(cfg) except Exception as e: print(f'Error in {cfg}: {e}') - - # Tensorboard (not working https://github.com/ultralytics/yolov5/issues/2898) - # from torch.utils.tensorboard import SummaryWriter - # tb_writer = SummaryWriter('.') - # LOGGER.info("Run 'tensorboard --logdir=models' to view tensorboard at http://localhost:6006/") - # tb_writer.add_graph(torch.jit.trace(model, img, strict=False), []) # add model graph From 9c1e91aea2437f56f1729ad8f92ce7a7d54f1268 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 30 Mar 2022 12:53:49 +0200 Subject: [PATCH 137/661] Update tutorial.ipynb (#7212) --- tutorial.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 1479a164cd8e..0379fb1a3c57 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -1077,7 +1077,7 @@ }, "source": [ "# VOC\n", - "for b, m in zip([64, 64, 32, 16], ['yolov5s', 'yolov5m', 'yolov5l', 'yolov5x']): # zip(batch_size, model)\n", + "for b, m in zip([64, 64, 64, 32, 16], ['yolov5n', 'yolov5s', 'yolov5m', 'yolov5l', 'yolov5x']): # zip(batch_size, model)\n", " !python train.py --batch {b} --weights {m}.pt --data VOC.yaml --epochs 50 --cache --img 512 --nosave --hyp hyp.VOC.yaml --project VOC --name {m}" ], "execution_count": null, From c94736acece384ed2d5a7299ee82af2969abb48b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 30 Mar 2022 16:01:55 +0200 Subject: [PATCH 138/661] `ENV OMP_NUM_THREADS=8` (#7215) --- Dockerfile | 1 + 1 file changed, 1 insertion(+) diff --git a/Dockerfile b/Dockerfile index 304e8b2801a9..59aa99faa1d6 100644 --- a/Dockerfile +++ b/Dockerfile @@ -26,6 +26,7 @@ RUN git clone https://github.com/ultralytics/yolov5 /usr/src/app ADD https://ultralytics.com/assets/Arial.ttf /root/.config/Ultralytics/ # Set environment variables +ENV OMP_NUM_THREADS=8 # ENV HOME=/usr/src/app From df9008ee69cac78524cc84500c7fc282a1a1d4bd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 31 Mar 2022 13:17:22 +0200 Subject: [PATCH 139/661] Add train.py `--name cfg` option (#7202) Automatically names run as --cfg argument --- train.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/train.py b/train.py index 60be962d447f..36a0e7a7ba66 100644 --- a/train.py +++ b/train.py @@ -519,6 +519,8 @@ def main(opt, callbacks=Callbacks()): if opt.project == str(ROOT / 'runs/train'): # if default project name, rename to runs/evolve opt.project = str(ROOT / 'runs/evolve') opt.exist_ok, opt.resume = opt.resume, False # pass resume to exist_ok and disable resume + if opt.name == 'cfg': + opt.name = Path(opt.cfg).stem # use model.yaml as name opt.save_dir = str(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok)) # DDP mode From c3d5ac151eaedb61495e5866f13a9746d3706abc Mon Sep 17 00:00:00 2001 From: Jirka Borovec Date: Thu, 31 Mar 2022 23:52:34 +0900 Subject: [PATCH 140/661] precommit: yapf (#5494) * precommit: yapf * align isort * fix # Conflicts: # utils/plots.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update setup.cfg * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update setup.cfg * Update setup.cfg * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update wandb_utils.py * Update augmentations.py * Update setup.cfg * Update yolo.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update val.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * simplify colorstr * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * val run fix * export.py last comma * Update export.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update hubconf.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * PyTorch Hub tuple fix * PyTorch Hub tuple fix2 * PyTorch Hub tuple fix3 * Update setup Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- .pre-commit-config.yaml | 11 +-- detect.py | 5 +- export.py | 110 ++++++++++++--------- hubconf.py | 13 +-- models/common.py | 37 +++++--- models/experimental.py | 4 +- models/tf.py | 67 +++++++++---- models/yolo.py | 4 +- setup.cfg | 14 +++ train.py | 147 ++++++++++++++++------------- utils/activations.py | 2 - utils/augmentations.py | 15 ++- utils/benchmarks.py | 5 +- utils/callbacks.py | 7 +- utils/datasets.py | 112 ++++++++++++++-------- utils/downloads.py | 17 ++-- utils/general.py | 74 ++++++++------- utils/loggers/__init__.py | 21 ++++- utils/loggers/wandb/wandb_utils.py | 112 ++++++++++++---------- utils/loss.py | 14 ++- utils/metrics.py | 11 ++- utils/plots.py | 30 ++++-- utils/torch_utils.py | 1 - val.py | 25 +++-- 24 files changed, 527 insertions(+), 331 deletions(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 526a5609fdd7..0b4fedcd2d43 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -36,12 +36,11 @@ repos: - id: isort name: Sort imports - # TODO - #- repo: https://github.com/pre-commit/mirrors-yapf - # rev: v0.31.0 - # hooks: - # - id: yapf - # name: formatting + - repo: https://github.com/pre-commit/mirrors-yapf + rev: v0.31.0 + hooks: + - id: yapf + name: formatting # TODO #- repo: https://github.com/executablebooks/mdformat diff --git a/detect.py b/detect.py index 046f7ae57b5c..2875285ee314 100644 --- a/detect.py +++ b/detect.py @@ -47,7 +47,8 @@ @torch.no_grad() -def run(weights=ROOT / 'yolov5s.pt', # model.pt path(s) +def run( + weights=ROOT / 'yolov5s.pt', # model.pt path(s) source=ROOT / 'data/images', # file/dir/URL/glob, 0 for webcam data=ROOT / 'data/coco128.yaml', # dataset.yaml path imgsz=(640, 640), # inference size (height, width) @@ -73,7 +74,7 @@ def run(weights=ROOT / 'yolov5s.pt', # model.pt path(s) hide_conf=False, # hide confidences half=False, # use FP16 half-precision inference dnn=False, # use OpenCV DNN for ONNX inference - ): +): source = str(source) save_img = not nosave and not source.endswith('.txt') # save inference images is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS) diff --git a/export.py b/export.py index 7517dc4678da..78b886fa3a6b 100644 --- a/export.py +++ b/export.py @@ -76,16 +76,11 @@ def export_formats(): # YOLOv5 export formats - x = [['PyTorch', '-', '.pt', True], - ['TorchScript', 'torchscript', '.torchscript', True], - ['ONNX', 'onnx', '.onnx', True], - ['OpenVINO', 'openvino', '_openvino_model', False], - ['TensorRT', 'engine', '.engine', True], - ['CoreML', 'coreml', '.mlmodel', False], - ['TensorFlow SavedModel', 'saved_model', '_saved_model', True], - ['TensorFlow GraphDef', 'pb', '.pb', True], - ['TensorFlow Lite', 'tflite', '.tflite', False], - ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False], + x = [['PyTorch', '-', '.pt', True], ['TorchScript', 'torchscript', '.torchscript', True], + ['ONNX', 'onnx', '.onnx', True], ['OpenVINO', 'openvino', '_openvino_model', False], + ['TensorRT', 'engine', '.engine', True], ['CoreML', 'coreml', '.mlmodel', False], + ['TensorFlow SavedModel', 'saved_model', '_saved_model', True], ['TensorFlow GraphDef', 'pb', '.pb', True], + ['TensorFlow Lite', 'tflite', '.tflite', False], ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False], ['TensorFlow.js', 'tfjs', '_web_model', False]] return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix', 'GPU']) @@ -119,14 +114,25 @@ def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorst LOGGER.info(f'\n{prefix} starting export with onnx {onnx.__version__}...') f = file.with_suffix('.onnx') - torch.onnx.export(model, im, f, verbose=False, opset_version=opset, - training=torch.onnx.TrainingMode.TRAINING if train else torch.onnx.TrainingMode.EVAL, - do_constant_folding=not train, - input_names=['images'], - output_names=['output'], - dynamic_axes={'images': {0: 'batch', 2: 'height', 3: 'width'}, # shape(1,3,640,640) - 'output': {0: 'batch', 1: 'anchors'} # shape(1,25200,85) - } if dynamic else None) + torch.onnx.export( + model, + im, + f, + verbose=False, + opset_version=opset, + training=torch.onnx.TrainingMode.TRAINING if train else torch.onnx.TrainingMode.EVAL, + do_constant_folding=not train, + input_names=['images'], + output_names=['output'], + dynamic_axes={ + 'images': { + 0: 'batch', + 2: 'height', + 3: 'width'}, # shape(1,3,640,640) + 'output': { + 0: 'batch', + 1: 'anchors'} # shape(1,25200,85) + } if dynamic else None) # Checks model_onnx = onnx.load(f) # load onnx model @@ -140,10 +146,9 @@ def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorst import onnxsim LOGGER.info(f'{prefix} simplifying with onnx-simplifier {onnxsim.__version__}...') - model_onnx, check = onnxsim.simplify( - model_onnx, - dynamic_input_shape=dynamic, - input_shapes={'images': list(im.shape)} if dynamic else None) + model_onnx, check = onnxsim.simplify(model_onnx, + dynamic_input_shape=dynamic, + input_shapes={'images': list(im.shape)} if dynamic else None) assert check, 'assert check failed' onnx.save(model_onnx, f) except Exception as e: @@ -246,9 +251,18 @@ def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=F LOGGER.info(f'\n{prefix} export failure: {e}') -def export_saved_model(model, im, file, dynamic, - tf_nms=False, agnostic_nms=False, topk_per_class=100, topk_all=100, iou_thres=0.45, - conf_thres=0.25, keras=False, prefix=colorstr('TensorFlow SavedModel:')): +def export_saved_model(model, + im, + file, + dynamic, + tf_nms=False, + agnostic_nms=False, + topk_per_class=100, + topk_all=100, + iou_thres=0.45, + conf_thres=0.25, + keras=False, + prefix=colorstr('TensorFlow SavedModel:')): # YOLOv5 TensorFlow SavedModel export try: import tensorflow as tf @@ -278,11 +292,10 @@ def export_saved_model(model, im, file, dynamic, tfm = tf.Module() tfm.__call__ = tf.function(lambda x: frozen_func(x)[0], [spec]) tfm.__call__(im) - tf.saved_model.save( - tfm, - f, - options=tf.saved_model.SaveOptions(experimental_custom_gradients=False) if - check_version(tf.__version__, '2.6') else tf.saved_model.SaveOptions()) + tf.saved_model.save(tfm, + f, + options=tf.saved_model.SaveOptions(experimental_custom_gradients=False) + if check_version(tf.__version__, '2.6') else tf.saved_model.SaveOptions()) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') return keras_model, f except Exception as e: @@ -352,10 +365,10 @@ def export_edgetpu(keras_model, im, file, prefix=colorstr('Edge TPU:')): if subprocess.run(cmd + ' >/dev/null', shell=True).returncode != 0: LOGGER.info(f'\n{prefix} export requires Edge TPU compiler. Attempting install from {help_url}') sudo = subprocess.run('sudo --version >/dev/null', shell=True).returncode == 0 # sudo installed on system - for c in ['curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -', - 'echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list', - 'sudo apt-get update', - 'sudo apt-get install edgetpu-compiler']: + for c in ( + 'curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -', + 'echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list', + 'sudo apt-get update', 'sudo apt-get install edgetpu-compiler'): subprocess.run(c if sudo else c.replace('sudo ', ''), shell=True, check=True) ver = subprocess.run(cmd, shell=True, capture_output=True, check=True).stdout.decode().split()[-1] @@ -395,12 +408,10 @@ def export_tfjs(keras_model, im, file, prefix=colorstr('TensorFlow.js:')): r'{"outputs": {"Identity.?.?": {"name": "Identity.?.?"}, ' r'"Identity.?.?": {"name": "Identity.?.?"}, ' r'"Identity.?.?": {"name": "Identity.?.?"}, ' - r'"Identity.?.?": {"name": "Identity.?.?"}}}', - r'{"outputs": {"Identity": {"name": "Identity"}, ' + r'"Identity.?.?": {"name": "Identity.?.?"}}}', r'{"outputs": {"Identity": {"name": "Identity"}, ' r'"Identity_1": {"name": "Identity_1"}, ' r'"Identity_2": {"name": "Identity_2"}, ' - r'"Identity_3": {"name": "Identity_3"}}}', - json) + r'"Identity_3": {"name": "Identity_3"}}}', json) j.write(subst) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') @@ -410,7 +421,8 @@ def export_tfjs(keras_model, im, file, prefix=colorstr('TensorFlow.js:')): @torch.no_grad() -def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path' +def run( + data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path' weights=ROOT / 'yolov5s.pt', # weights path imgsz=(640, 640), # image (height, width) batch_size=1, # batch size @@ -431,8 +443,8 @@ def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path' topk_per_class=100, # TF.js NMS: topk per class to keep topk_all=100, # TF.js NMS: topk for all classes to keep iou_thres=0.45, # TF.js NMS: IoU threshold - conf_thres=0.25 # TF.js NMS: confidence threshold - ): + conf_thres=0.25, # TF.js NMS: confidence threshold +): t = time.time() include = [x.lower() for x in include] # to lowercase formats = tuple(export_formats()['Argument'][1:]) # --include arguments @@ -495,9 +507,16 @@ def run(data=ROOT / 'data/coco128.yaml', # 'dataset.yaml path' if int8 or edgetpu: # TFLite --int8 bug https://github.com/ultralytics/yolov5/issues/5707 check_requirements(('flatbuffers==1.12',)) # required before `import tensorflow` assert not (tflite and tfjs), 'TFLite and TF.js models must be exported separately, please pass only one type.' - model, f[5] = export_saved_model(model.cpu(), im, file, dynamic, tf_nms=nms or agnostic_nms or tfjs, - agnostic_nms=agnostic_nms or tfjs, topk_per_class=topk_per_class, - topk_all=topk_all, conf_thres=conf_thres, iou_thres=iou_thres) # keras model + model, f[5] = export_saved_model(model.cpu(), + im, + file, + dynamic, + tf_nms=nms or agnostic_nms or tfjs, + agnostic_nms=agnostic_nms or tfjs, + topk_per_class=topk_per_class, + topk_all=topk_all, + conf_thres=conf_thres, + iou_thres=iou_thres) # keras model if pb or tfjs: # pb prerequisite to tfjs f[6] = export_pb(model, im, file) if tflite or edgetpu: @@ -542,7 +561,8 @@ def parse_opt(): parser.add_argument('--topk-all', type=int, default=100, help='TF.js NMS: topk for all classes to keep') parser.add_argument('--iou-thres', type=float, default=0.45, help='TF.js NMS: IoU threshold') parser.add_argument('--conf-thres', type=float, default=0.25, help='TF.js NMS: confidence threshold') - parser.add_argument('--include', nargs='+', + parser.add_argument('--include', + nargs='+', default=['torchscript', 'onnx'], help='torchscript, onnx, openvino, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs') opt = parser.parse_args() diff --git a/hubconf.py b/hubconf.py index d719b80034af..86aa07b9466f 100644 --- a/hubconf.py +++ b/hubconf.py @@ -132,12 +132,13 @@ def yolov5x6(pretrained=True, channels=3, classes=80, autoshape=True, verbose=Tr from utils.general import cv2 - imgs = ['data/images/zidane.jpg', # filename - Path('data/images/zidane.jpg'), # Path - 'https://ultralytics.com/images/zidane.jpg', # URI - cv2.imread('data/images/bus.jpg')[:, :, ::-1], # OpenCV - Image.open('data/images/bus.jpg'), # PIL - np.zeros((320, 640, 3))] # numpy + imgs = [ + 'data/images/zidane.jpg', # filename + Path('data/images/zidane.jpg'), # Path + 'https://ultralytics.com/images/zidane.jpg', # URI + cv2.imread('data/images/bus.jpg')[:, :, ::-1], # OpenCV + Image.open('data/images/bus.jpg'), # PIL + np.zeros((320, 640, 3))] # numpy results = model(imgs, size=320) # batched inference results.print() diff --git a/models/common.py b/models/common.py index 115e3c3145ff..8396caa1af5c 100644 --- a/models/common.py +++ b/models/common.py @@ -227,11 +227,12 @@ class GhostBottleneck(nn.Module): def __init__(self, c1, c2, k=3, s=1): # ch_in, ch_out, kernel, stride super().__init__() c_ = c2 // 2 - self.conv = nn.Sequential(GhostConv(c1, c_, 1, 1), # pw - DWConv(c_, c_, k, s, act=False) if s == 2 else nn.Identity(), # dw - GhostConv(c_, c2, 1, 1, act=False)) # pw-linear - self.shortcut = nn.Sequential(DWConv(c1, c1, k, s, act=False), - Conv(c1, c2, 1, 1, act=False)) if s == 2 else nn.Identity() + self.conv = nn.Sequential( + GhostConv(c1, c_, 1, 1), # pw + DWConv(c_, c_, k, s, act=False) if s == 2 else nn.Identity(), # dw + GhostConv(c_, c2, 1, 1, act=False)) # pw-linear + self.shortcut = nn.Sequential(DWConv(c1, c1, k, s, act=False), Conv(c1, c2, 1, 1, + act=False)) if s == 2 else nn.Identity() def forward(self, x): return self.conv(x) + self.shortcut(x) @@ -387,9 +388,10 @@ def wrap_frozen_graph(gd, inputs, outputs): Interpreter, load_delegate = tf.lite.Interpreter, tf.lite.experimental.load_delegate, if edgetpu: # Edge TPU https://coral.ai/software/#edgetpu-runtime LOGGER.info(f'Loading {w} for TensorFlow Lite Edge TPU inference...') - delegate = {'Linux': 'libedgetpu.so.1', - 'Darwin': 'libedgetpu.1.dylib', - 'Windows': 'edgetpu.dll'}[platform.system()] + delegate = { + 'Linux': 'libedgetpu.so.1', + 'Darwin': 'libedgetpu.1.dylib', + 'Windows': 'edgetpu.dll'}[platform.system()] interpreter = Interpreter(model_path=w, experimental_delegates=[load_delegate(delegate)]) else: # Lite LOGGER.info(f'Loading {w} for TensorFlow Lite inference...') @@ -531,7 +533,7 @@ def forward(self, imgs, size=640, augment=False, profile=False): return self.model(imgs.to(p.device).type_as(p), augment, profile) # inference # Pre-process - n, imgs = (len(imgs), imgs) if isinstance(imgs, list) else (1, [imgs]) # number of images, list of images + n, imgs = (len(imgs), list(imgs)) if isinstance(imgs, (list, tuple)) else (1, [imgs]) # number, list of images shape0, shape1, files = [], [], [] # image and inference shapes, filenames for i, im in enumerate(imgs): f = f'image{i}' # filename @@ -561,8 +563,13 @@ def forward(self, imgs, size=640, augment=False, profile=False): t.append(time_sync()) # Post-process - y = non_max_suppression(y if self.dmb else y[0], self.conf, self.iou, self.classes, self.agnostic, - self.multi_label, max_det=self.max_det) # NMS + y = non_max_suppression(y if self.dmb else y[0], + self.conf, + self.iou, + self.classes, + self.agnostic, + self.multi_label, + max_det=self.max_det) # NMS for i in range(n): scale_coords(shape1, y[i][:, :4], shape0[i]) @@ -603,8 +610,12 @@ def display(self, pprint=False, show=False, save=False, crop=False, render=False label = f'{self.names[int(cls)]} {conf:.2f}' if crop: file = save_dir / 'crops' / self.names[int(cls)] / self.files[i] if save else None - crops.append({'box': box, 'conf': conf, 'cls': cls, 'label': label, - 'im': save_one_box(box, im, file=file, save=save)}) + crops.append({ + 'box': box, + 'conf': conf, + 'cls': cls, + 'label': label, + 'im': save_one_box(box, im, file=file, save=save)}) else: # all others annotator.box_label(box, label if labels else '', color=colors(cls)) im = annotator.im diff --git a/models/experimental.py b/models/experimental.py index 1230f4656c8f..e166722cbfca 100644 --- a/models/experimental.py +++ b/models/experimental.py @@ -63,8 +63,8 @@ def __init__(self, c1, c2, k=(1, 3), s=1, equal_ch=True): # ch_in, ch_out, kern a[0] = 1 c_ = np.linalg.lstsq(a, b, rcond=None)[0].round() # solve for equal weight indices, ax = b - self.m = nn.ModuleList( - [nn.Conv2d(c1, int(c_), k, s, k // 2, groups=math.gcd(c1, int(c_)), bias=False) for k, c_ in zip(k, c_)]) + self.m = nn.ModuleList([ + nn.Conv2d(c1, int(c_), k, s, k // 2, groups=math.gcd(c1, int(c_)), bias=False) for k, c_ in zip(k, c_)]) self.bn = nn.BatchNorm2d(c2) self.act = nn.SiLU() diff --git a/models/tf.py b/models/tf.py index 728907f8fb47..c6fb6b82a72e 100644 --- a/models/tf.py +++ b/models/tf.py @@ -69,7 +69,11 @@ def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True, w=None): # see https://stackoverflow.com/questions/52975843/comparing-conv2d-with-padding-between-tensorflow-and-pytorch conv = keras.layers.Conv2D( - c2, k, s, 'SAME' if s == 1 else 'VALID', use_bias=False if hasattr(w, 'bn') else True, + c2, + k, + s, + 'SAME' if s == 1 else 'VALID', + use_bias=False if hasattr(w, 'bn') else True, kernel_initializer=keras.initializers.Constant(w.conv.weight.permute(2, 3, 1, 0).numpy()), bias_initializer='zeros' if hasattr(w, 'bn') else keras.initializers.Constant(w.conv.bias.numpy())) self.conv = conv if s == 1 else keras.Sequential([TFPad(autopad(k, p)), conv]) @@ -98,10 +102,10 @@ def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True, w=None): def call(self, inputs): # x(b,w,h,c) -> y(b,w/2,h/2,4c) # inputs = inputs / 255 # normalize 0-255 to 0-1 - return self.conv(tf.concat([inputs[:, ::2, ::2, :], - inputs[:, 1::2, ::2, :], - inputs[:, ::2, 1::2, :], - inputs[:, 1::2, 1::2, :]], 3)) + return self.conv( + tf.concat( + [inputs[:, ::2, ::2, :], inputs[:, 1::2, ::2, :], inputs[:, ::2, 1::2, :], inputs[:, 1::2, 1::2, :]], + 3)) class TFBottleneck(keras.layers.Layer): @@ -123,9 +127,14 @@ def __init__(self, c1, c2, k, s=1, g=1, bias=True, w=None): super().__init__() assert g == 1, "TF v2.2 Conv2D does not support 'groups' argument" self.conv = keras.layers.Conv2D( - c2, k, s, 'VALID', use_bias=bias, + c2, + k, + s, + 'VALID', + use_bias=bias, kernel_initializer=keras.initializers.Constant(w.weight.permute(2, 3, 1, 0).numpy()), - bias_initializer=keras.initializers.Constant(w.bias.numpy()) if bias else None, ) + bias_initializer=keras.initializers.Constant(w.bias.numpy()) if bias else None, + ) def call(self, inputs): return self.conv(inputs) @@ -206,8 +215,7 @@ def __init__(self, nc=80, anchors=(), ch=(), imgsz=(640, 640), w=None): # detec self.na = len(anchors[0]) // 2 # number of anchors self.grid = [tf.zeros(1)] * self.nl # init grid self.anchors = tf.convert_to_tensor(w.anchors.numpy(), dtype=tf.float32) - self.anchor_grid = tf.reshape(self.anchors * tf.reshape(self.stride, [self.nl, 1, 1]), - [self.nl, 1, -1, 1, 2]) + self.anchor_grid = tf.reshape(self.anchors * tf.reshape(self.stride, [self.nl, 1, 1]), [self.nl, 1, -1, 1, 2]) self.m = [TFConv2d(x, self.no * self.na, 1, w=w.m[i]) for i, x in enumerate(ch)] self.training = False # set to False after building model self.imgsz = imgsz @@ -339,7 +347,13 @@ def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, model=None, imgsz=(640, 64 self.yaml['nc'] = nc # override yaml value self.model, self.savelist = parse_model(deepcopy(self.yaml), ch=[ch], model=model, imgsz=imgsz) - def predict(self, inputs, tf_nms=False, agnostic_nms=False, topk_per_class=100, topk_all=100, iou_thres=0.45, + def predict(self, + inputs, + tf_nms=False, + agnostic_nms=False, + topk_per_class=100, + topk_all=100, + iou_thres=0.45, conf_thres=0.25): y = [] # outputs x = inputs @@ -361,8 +375,13 @@ def predict(self, inputs, tf_nms=False, agnostic_nms=False, topk_per_class=100, return nms, x[1] else: boxes = tf.expand_dims(boxes, 2) - nms = tf.image.combined_non_max_suppression( - boxes, scores, topk_per_class, topk_all, iou_thres, conf_thres, clip_boxes=False) + nms = tf.image.combined_non_max_suppression(boxes, + scores, + topk_per_class, + topk_all, + iou_thres, + conf_thres, + clip_boxes=False) return nms, x[1] return x[0] # output only first tensor [1,6300,85] = [xywh, conf, class0, class1, ...] @@ -383,7 +402,8 @@ class AgnosticNMS(keras.layers.Layer): # TF Agnostic NMS def call(self, input, topk_all, iou_thres, conf_thres): # wrap map_fn to avoid TypeSpec related error https://stackoverflow.com/a/65809989/3036450 - return tf.map_fn(lambda x: self._nms(x, topk_all, iou_thres, conf_thres), input, + return tf.map_fn(lambda x: self._nms(x, topk_all, iou_thres, conf_thres), + input, fn_output_signature=(tf.float32, tf.float32, tf.float32, tf.int32), name='agnostic_nms') @@ -392,20 +412,26 @@ def _nms(x, topk_all=100, iou_thres=0.45, conf_thres=0.25): # agnostic NMS boxes, classes, scores = x class_inds = tf.cast(tf.argmax(classes, axis=-1), tf.float32) scores_inp = tf.reduce_max(scores, -1) - selected_inds = tf.image.non_max_suppression( - boxes, scores_inp, max_output_size=topk_all, iou_threshold=iou_thres, score_threshold=conf_thres) + selected_inds = tf.image.non_max_suppression(boxes, + scores_inp, + max_output_size=topk_all, + iou_threshold=iou_thres, + score_threshold=conf_thres) selected_boxes = tf.gather(boxes, selected_inds) padded_boxes = tf.pad(selected_boxes, paddings=[[0, topk_all - tf.shape(selected_boxes)[0]], [0, 0]], - mode="CONSTANT", constant_values=0.0) + mode="CONSTANT", + constant_values=0.0) selected_scores = tf.gather(scores_inp, selected_inds) padded_scores = tf.pad(selected_scores, paddings=[[0, topk_all - tf.shape(selected_boxes)[0]]], - mode="CONSTANT", constant_values=-1.0) + mode="CONSTANT", + constant_values=-1.0) selected_classes = tf.gather(class_inds, selected_inds) padded_classes = tf.pad(selected_classes, paddings=[[0, topk_all - tf.shape(selected_boxes)[0]]], - mode="CONSTANT", constant_values=-1.0) + mode="CONSTANT", + constant_values=-1.0) valid_detections = tf.shape(selected_inds)[0] return padded_boxes, padded_scores, padded_classes, valid_detections @@ -421,11 +447,12 @@ def representative_dataset_gen(dataset, ncalib=100): break -def run(weights=ROOT / 'yolov5s.pt', # weights path +def run( + weights=ROOT / 'yolov5s.pt', # weights path imgsz=(640, 640), # inference size h,w batch_size=1, # batch size dynamic=False, # dynamic batch size - ): +): # PyTorch model im = torch.zeros((batch_size, 3, *imgsz)) # BCHW image model = attempt_load(weights, map_location=torch.device('cpu'), inplace=True, fuse=False) diff --git a/models/yolo.py b/models/yolo.py index 81ab539deffa..4cdfea34d63e 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -260,8 +260,8 @@ def parse_model(d, ch): # model_dict, input_channels(3) pass n = n_ = max(round(n * gd), 1) if n > 1 else n # depth gain - if m in [Conv, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv, - BottleneckCSP, C3, C3TR, C3SPP, C3Ghost]: + if m in (Conv, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv, + BottleneckCSP, C3, C3TR, C3SPP, C3Ghost): c1, c2 = ch[f], args[0] if c2 != no: # if not output c2 = make_divisible(c2 * gw, 8) diff --git a/setup.cfg b/setup.cfg index 20ea49a8b4d6..c387d84a33e2 100644 --- a/setup.cfg +++ b/setup.cfg @@ -1,5 +1,6 @@ # Project-wide configuration file, can be used for package metadata and other toll configurations # Example usage: global configuration for PEP8 (via flake8) setting or default pytest arguments +# Local usage: pip install pre-commit, pre-commit run --all-files [metadata] license_file = LICENSE @@ -42,4 +43,17 @@ ignore = [isort] # https://pycqa.github.io/isort/docs/configuration/options.html line_length = 120 +# see: https://pycqa.github.io/isort/docs/configuration/multi_line_output_modes.html multi_line_output = 0 + + +[yapf] +based_on_style = pep8 +spaces_before_comment = 2 +COLUMN_LIMIT = 120 +COALESCE_BRACKETS = True +SPACES_AROUND_POWER_OPERATOR = True +SPACE_BETWEEN_ENDING_COMMA_AND_CLOSING_BRACKET = False +SPLIT_BEFORE_CLOSING_BRACKET = False +SPLIT_BEFORE_FIRST_ARGUMENT = False +# EACH_DICT_ENTRY_ON_SEPARATE_LINE = False diff --git a/train.py b/train.py index 36a0e7a7ba66..fbaaeb8ef930 100644 --- a/train.py +++ b/train.py @@ -62,11 +62,7 @@ WORLD_SIZE = int(os.getenv('WORLD_SIZE', 1)) -def train(hyp, # path/to/hyp.yaml or hyp dictionary - opt, - device, - callbacks - ): +def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictionary save_dir, epochs, batch_size, weights, single_cls, evolve, data, cfg, resume, noval, nosave, workers, freeze = \ Path(opt.save_dir), opt.epochs, opt.batch_size, opt.weights, opt.single_cls, opt.evolve, opt.data, opt.cfg, \ opt.resume, opt.noval, opt.nosave, opt.workers, opt.freeze @@ -220,20 +216,38 @@ def train(hyp, # path/to/hyp.yaml or hyp dictionary LOGGER.info('Using SyncBatchNorm()') # Trainloader - train_loader, dataset = create_dataloader(train_path, imgsz, batch_size // WORLD_SIZE, gs, single_cls, - hyp=hyp, augment=True, cache=None if opt.cache == 'val' else opt.cache, - rect=opt.rect, rank=LOCAL_RANK, workers=workers, - image_weights=opt.image_weights, quad=opt.quad, - prefix=colorstr('train: '), shuffle=True) + train_loader, dataset = create_dataloader(train_path, + imgsz, + batch_size // WORLD_SIZE, + gs, + single_cls, + hyp=hyp, + augment=True, + cache=None if opt.cache == 'val' else opt.cache, + rect=opt.rect, + rank=LOCAL_RANK, + workers=workers, + image_weights=opt.image_weights, + quad=opt.quad, + prefix=colorstr('train: '), + shuffle=True) mlc = int(np.concatenate(dataset.labels, 0)[:, 0].max()) # max label class nb = len(train_loader) # number of batches assert mlc < nc, f'Label class {mlc} exceeds nc={nc} in {data}. Possible class labels are 0-{nc - 1}' # Process 0 if RANK in [-1, 0]: - val_loader = create_dataloader(val_path, imgsz, batch_size // WORLD_SIZE * 2, gs, single_cls, - hyp=hyp, cache=None if noval else opt.cache, - rect=True, rank=-1, workers=workers * 2, pad=0.5, + val_loader = create_dataloader(val_path, + imgsz, + batch_size // WORLD_SIZE * 2, + gs, + single_cls, + hyp=hyp, + cache=None if noval else opt.cache, + rect=True, + rank=-1, + workers=workers * 2, + pad=0.5, prefix=colorstr('val: '))[0] if not resume: @@ -350,8 +364,8 @@ def train(hyp, # path/to/hyp.yaml or hyp dictionary if RANK in [-1, 0]: mloss = (mloss * i + loss_items) / (i + 1) # update mean losses mem = f'{torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0:.3g}G' # (GB) - pbar.set_description(('%10s' * 2 + '%10.4g' * 5) % ( - f'{epoch}/{epochs - 1}', mem, *mloss, targets.shape[0], imgs.shape[-1])) + pbar.set_description(('%10s' * 2 + '%10.4g' * 5) % + (f'{epoch}/{epochs - 1}', mem, *mloss, targets.shape[0], imgs.shape[-1])) callbacks.run('on_train_batch_end', ni, model, imgs, targets, paths, plots, opt.sync_bn) if callbacks.stop_training: return @@ -387,14 +401,15 @@ def train(hyp, # path/to/hyp.yaml or hyp dictionary # Save model if (not nosave) or (final_epoch and not evolve): # if save - ckpt = {'epoch': epoch, - 'best_fitness': best_fitness, - 'model': deepcopy(de_parallel(model)).half(), - 'ema': deepcopy(ema.ema).half(), - 'updates': ema.updates, - 'optimizer': optimizer.state_dict(), - 'wandb_id': loggers.wandb.wandb_run.id if loggers.wandb else None, - 'date': datetime.now().isoformat()} + ckpt = { + 'epoch': epoch, + 'best_fitness': best_fitness, + 'model': deepcopy(de_parallel(model)).half(), + 'ema': deepcopy(ema.ema).half(), + 'updates': ema.updates, + 'optimizer': optimizer.state_dict(), + 'wandb_id': loggers.wandb.wandb_run.id if loggers.wandb else None, + 'date': datetime.now().isoformat()} # Save last, best and delete torch.save(ckpt, last) @@ -428,19 +443,20 @@ def train(hyp, # path/to/hyp.yaml or hyp dictionary strip_optimizer(f) # strip optimizers if f is best: LOGGER.info(f'\nValidating {f}...') - results, _, _ = val.run(data_dict, - batch_size=batch_size // WORLD_SIZE * 2, - imgsz=imgsz, - model=attempt_load(f, device).half(), - iou_thres=0.65 if is_coco else 0.60, # best pycocotools results at 0.65 - single_cls=single_cls, - dataloader=val_loader, - save_dir=save_dir, - save_json=is_coco, - verbose=True, - plots=True, - callbacks=callbacks, - compute_loss=compute_loss) # val best model with plots + results, _, _ = val.run( + data_dict, + batch_size=batch_size // WORLD_SIZE * 2, + imgsz=imgsz, + model=attempt_load(f, device).half(), + iou_thres=0.65 if is_coco else 0.60, # best pycocotools results at 0.65 + single_cls=single_cls, + dataloader=val_loader, + save_dir=save_dir, + save_json=is_coco, + verbose=True, + plots=True, + callbacks=callbacks, + compute_loss=compute_loss) # val best model with plots if is_coco: callbacks.run('on_fit_epoch_end', list(mloss) + list(results) + lr, epoch, best_fitness, fi) @@ -546,35 +562,36 @@ def main(opt, callbacks=Callbacks()): # Evolve hyperparameters (optional) else: # Hyperparameter evolution metadata (mutation scale 0-1, lower_limit, upper_limit) - meta = {'lr0': (1, 1e-5, 1e-1), # initial learning rate (SGD=1E-2, Adam=1E-3) - 'lrf': (1, 0.01, 1.0), # final OneCycleLR learning rate (lr0 * lrf) - 'momentum': (0.3, 0.6, 0.98), # SGD momentum/Adam beta1 - 'weight_decay': (1, 0.0, 0.001), # optimizer weight decay - 'warmup_epochs': (1, 0.0, 5.0), # warmup epochs (fractions ok) - 'warmup_momentum': (1, 0.0, 0.95), # warmup initial momentum - 'warmup_bias_lr': (1, 0.0, 0.2), # warmup initial bias lr - 'box': (1, 0.02, 0.2), # box loss gain - 'cls': (1, 0.2, 4.0), # cls loss gain - 'cls_pw': (1, 0.5, 2.0), # cls BCELoss positive_weight - 'obj': (1, 0.2, 4.0), # obj loss gain (scale with pixels) - 'obj_pw': (1, 0.5, 2.0), # obj BCELoss positive_weight - 'iou_t': (0, 0.1, 0.7), # IoU training threshold - 'anchor_t': (1, 2.0, 8.0), # anchor-multiple threshold - 'anchors': (2, 2.0, 10.0), # anchors per output grid (0 to ignore) - 'fl_gamma': (0, 0.0, 2.0), # focal loss gamma (efficientDet default gamma=1.5) - 'hsv_h': (1, 0.0, 0.1), # image HSV-Hue augmentation (fraction) - 'hsv_s': (1, 0.0, 0.9), # image HSV-Saturation augmentation (fraction) - 'hsv_v': (1, 0.0, 0.9), # image HSV-Value augmentation (fraction) - 'degrees': (1, 0.0, 45.0), # image rotation (+/- deg) - 'translate': (1, 0.0, 0.9), # image translation (+/- fraction) - 'scale': (1, 0.0, 0.9), # image scale (+/- gain) - 'shear': (1, 0.0, 10.0), # image shear (+/- deg) - 'perspective': (0, 0.0, 0.001), # image perspective (+/- fraction), range 0-0.001 - 'flipud': (1, 0.0, 1.0), # image flip up-down (probability) - 'fliplr': (0, 0.0, 1.0), # image flip left-right (probability) - 'mosaic': (1, 0.0, 1.0), # image mixup (probability) - 'mixup': (1, 0.0, 1.0), # image mixup (probability) - 'copy_paste': (1, 0.0, 1.0)} # segment copy-paste (probability) + meta = { + 'lr0': (1, 1e-5, 1e-1), # initial learning rate (SGD=1E-2, Adam=1E-3) + 'lrf': (1, 0.01, 1.0), # final OneCycleLR learning rate (lr0 * lrf) + 'momentum': (0.3, 0.6, 0.98), # SGD momentum/Adam beta1 + 'weight_decay': (1, 0.0, 0.001), # optimizer weight decay + 'warmup_epochs': (1, 0.0, 5.0), # warmup epochs (fractions ok) + 'warmup_momentum': (1, 0.0, 0.95), # warmup initial momentum + 'warmup_bias_lr': (1, 0.0, 0.2), # warmup initial bias lr + 'box': (1, 0.02, 0.2), # box loss gain + 'cls': (1, 0.2, 4.0), # cls loss gain + 'cls_pw': (1, 0.5, 2.0), # cls BCELoss positive_weight + 'obj': (1, 0.2, 4.0), # obj loss gain (scale with pixels) + 'obj_pw': (1, 0.5, 2.0), # obj BCELoss positive_weight + 'iou_t': (0, 0.1, 0.7), # IoU training threshold + 'anchor_t': (1, 2.0, 8.0), # anchor-multiple threshold + 'anchors': (2, 2.0, 10.0), # anchors per output grid (0 to ignore) + 'fl_gamma': (0, 0.0, 2.0), # focal loss gamma (efficientDet default gamma=1.5) + 'hsv_h': (1, 0.0, 0.1), # image HSV-Hue augmentation (fraction) + 'hsv_s': (1, 0.0, 0.9), # image HSV-Saturation augmentation (fraction) + 'hsv_v': (1, 0.0, 0.9), # image HSV-Value augmentation (fraction) + 'degrees': (1, 0.0, 45.0), # image rotation (+/- deg) + 'translate': (1, 0.0, 0.9), # image translation (+/- fraction) + 'scale': (1, 0.0, 0.9), # image scale (+/- gain) + 'shear': (1, 0.0, 10.0), # image shear (+/- deg) + 'perspective': (0, 0.0, 0.001), # image perspective (+/- fraction), range 0-0.001 + 'flipud': (1, 0.0, 1.0), # image flip up-down (probability) + 'fliplr': (0, 0.0, 1.0), # image flip left-right (probability) + 'mosaic': (1, 0.0, 1.0), # image mixup (probability) + 'mixup': (1, 0.0, 1.0), # image mixup (probability) + 'copy_paste': (1, 0.0, 1.0)} # segment copy-paste (probability) with open(opt.hyp, errors='ignore') as f: hyp = yaml.safe_load(f) # load hyps dict diff --git a/utils/activations.py b/utils/activations.py index a4ff789cf336..b104ac18b03b 100644 --- a/utils/activations.py +++ b/utils/activations.py @@ -64,7 +64,6 @@ class AconC(nn.Module): AconC: (p1*x-p2*x) * sigmoid(beta*(p1*x-p2*x)) + p2*x, beta is a learnable parameter according to "Activate or Not: Learning Customized Activation" . """ - def __init__(self, c1): super().__init__() self.p1 = nn.Parameter(torch.randn(1, c1, 1, 1)) @@ -81,7 +80,6 @@ class MetaAconC(nn.Module): MetaAconC: (p1*x-p2*x) * sigmoid(beta*(p1*x-p2*x)) + p2*x, beta is generated by a small network according to "Activate or Not: Learning Customized Activation" . """ - def __init__(self, c1, k=1, s=1, r=16): # ch_in, kernel, stride, r super().__init__() c2 = max(r, c1 // r) diff --git a/utils/augmentations.py b/utils/augmentations.py index 0311b97b63db..3f764c06ae3b 100644 --- a/utils/augmentations.py +++ b/utils/augmentations.py @@ -21,15 +21,15 @@ def __init__(self): import albumentations as A check_version(A.__version__, '1.0.3', hard=True) # version requirement - self.transform = A.Compose([ + T = [ A.Blur(p=0.01), A.MedianBlur(p=0.01), A.ToGray(p=0.01), A.CLAHE(p=0.01), A.RandomBrightnessContrast(p=0.0), A.RandomGamma(p=0.0), - A.ImageCompression(quality_lower=75, p=0.0)], - bbox_params=A.BboxParams(format='yolo', label_fields=['class_labels'])) + A.ImageCompression(quality_lower=75, p=0.0)] # transforms + self.transform = A.Compose(T, bbox_params=A.BboxParams(format='yolo', label_fields=['class_labels'])) LOGGER.info(colorstr('albumentations: ') + ', '.join(f'{x}' for x in self.transform.transforms if x.p)) except ImportError: # package not installed, skip @@ -121,7 +121,14 @@ def letterbox(im, new_shape=(640, 640), color=(114, 114, 114), auto=True, scaleF return im, ratio, (dw, dh) -def random_perspective(im, targets=(), segments=(), degrees=10, translate=.1, scale=.1, shear=10, perspective=0.0, +def random_perspective(im, + targets=(), + segments=(), + degrees=10, + translate=.1, + scale=.1, + shear=10, + perspective=0.0, border=(0, 0)): # torchvision.transforms.RandomAffine(degrees=(-10, 10), translate=(0.1, 0.1), scale=(0.9, 1.1), shear=(-10, 10)) # targets = [cls, xyxy] diff --git a/utils/benchmarks.py b/utils/benchmarks.py index 446248c03f68..5bfa872cc3fb 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -45,13 +45,14 @@ from utils.torch_utils import select_device -def run(weights=ROOT / 'yolov5s.pt', # weights path +def run( + weights=ROOT / 'yolov5s.pt', # weights path imgsz=640, # inference size (pixels) batch_size=1, # batch size data=ROOT / 'data/coco128.yaml', # dataset.yaml path device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu half=False, # use FP16 half-precision inference - ): +): y, t = [], time.time() formats = export.export_formats() device = select_device(device) diff --git a/utils/callbacks.py b/utils/callbacks.py index c51c268f20d6..6323985b8098 100644 --- a/utils/callbacks.py +++ b/utils/callbacks.py @@ -8,13 +8,11 @@ class Callbacks: """" Handles all registered callbacks for YOLOv5 Hooks """ - def __init__(self): # Define the available callbacks self._callbacks = { 'on_pretrain_routine_start': [], 'on_pretrain_routine_end': [], - 'on_train_start': [], 'on_train_epoch_start': [], 'on_train_batch_start': [], @@ -22,19 +20,16 @@ def __init__(self): 'on_before_zero_grad': [], 'on_train_batch_end': [], 'on_train_epoch_end': [], - 'on_val_start': [], 'on_val_batch_start': [], 'on_val_image_end': [], 'on_val_batch_end': [], 'on_val_end': [], - 'on_fit_epoch_end': [], # fit = train + val 'on_model_save': [], 'on_train_end': [], 'on_params_update': [], - 'teardown': [], - } + 'teardown': [],} self.stop_training = False # set True to interrupt training def register_action(self, hook, name='', callback=None): diff --git a/utils/datasets.py b/utils/datasets.py index d0b35e808000..7e8b423c3174 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -77,14 +77,14 @@ def exif_transpose(image): exif = image.getexif() orientation = exif.get(0x0112, 1) # default 1 if orientation > 1: - method = {2: Image.FLIP_LEFT_RIGHT, - 3: Image.ROTATE_180, - 4: Image.FLIP_TOP_BOTTOM, - 5: Image.TRANSPOSE, - 6: Image.ROTATE_270, - 7: Image.TRANSVERSE, - 8: Image.ROTATE_90, - }.get(orientation) + method = { + 2: Image.FLIP_LEFT_RIGHT, + 3: Image.ROTATE_180, + 4: Image.FLIP_TOP_BOTTOM, + 5: Image.TRANSPOSE, + 6: Image.ROTATE_270, + 7: Image.TRANSVERSE, + 8: Image.ROTATE_90,}.get(orientation) if method is not None: image = image.transpose(method) del exif[0x0112] @@ -92,22 +92,39 @@ def exif_transpose(image): return image -def create_dataloader(path, imgsz, batch_size, stride, single_cls=False, hyp=None, augment=False, cache=False, pad=0.0, - rect=False, rank=-1, workers=8, image_weights=False, quad=False, prefix='', shuffle=False): +def create_dataloader(path, + imgsz, + batch_size, + stride, + single_cls=False, + hyp=None, + augment=False, + cache=False, + pad=0.0, + rect=False, + rank=-1, + workers=8, + image_weights=False, + quad=False, + prefix='', + shuffle=False): if rect and shuffle: LOGGER.warning('WARNING: --rect is incompatible with DataLoader shuffle, setting shuffle=False') shuffle = False with torch_distributed_zero_first(rank): # init dataset *.cache only once if DDP - dataset = LoadImagesAndLabels(path, imgsz, batch_size, - augment=augment, # augmentation - hyp=hyp, # hyperparameters - rect=rect, # rectangular batches - cache_images=cache, - single_cls=single_cls, - stride=int(stride), - pad=pad, - image_weights=image_weights, - prefix=prefix) + dataset = LoadImagesAndLabels( + path, + imgsz, + batch_size, + augment=augment, # augmentation + hyp=hyp, # hyperparameters + rect=rect, # rectangular batches + cache_images=cache, + single_cls=single_cls, + stride=int(stride), + pad=pad, + image_weights=image_weights, + prefix=prefix) batch_size = min(batch_size, len(dataset)) nd = torch.cuda.device_count() # number of CUDA devices @@ -128,7 +145,6 @@ class InfiniteDataLoader(dataloader.DataLoader): Uses same syntax as vanilla DataLoader """ - def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) object.__setattr__(self, 'batch_sampler', _RepeatSampler(self.batch_sampler)) @@ -148,7 +164,6 @@ class _RepeatSampler: Args: sampler (Sampler) """ - def __init__(self, sampler): self.sampler = sampler @@ -380,8 +395,19 @@ class LoadImagesAndLabels(Dataset): # YOLOv5 train_loader/val_loader, loads images and labels for training and validation cache_version = 0.6 # dataset labels *.cache version - def __init__(self, path, img_size=640, batch_size=16, augment=False, hyp=None, rect=False, image_weights=False, - cache_images=False, single_cls=False, stride=32, pad=0.0, prefix=''): + def __init__(self, + path, + img_size=640, + batch_size=16, + augment=False, + hyp=None, + rect=False, + image_weights=False, + cache_images=False, + single_cls=False, + stride=32, + pad=0.0, + prefix=''): self.img_size = img_size self.augment = augment self.hyp = hyp @@ -510,7 +536,9 @@ def cache_labels(self, path=Path('./labels.cache'), prefix=''): desc = f"{prefix}Scanning '{path.parent / path.stem}' images and labels..." with Pool(NUM_THREADS) as pool: pbar = tqdm(pool.imap(verify_image_label, zip(self.im_files, self.label_files, repeat(prefix))), - desc=desc, total=len(self.im_files), bar_format=BAR_FORMAT) + desc=desc, + total=len(self.im_files), + bar_format=BAR_FORMAT) for im_file, lb, shape, segments, nm_f, nf_f, ne_f, nc_f, msg in pbar: nm += nm_f nf += nf_f @@ -576,7 +604,8 @@ def __getitem__(self, index): labels[:, 1:] = xywhn2xyxy(labels[:, 1:], ratio[0] * w, ratio[1] * h, padw=pad[0], padh=pad[1]) if self.augment: - img, labels = random_perspective(img, labels, + img, labels = random_perspective(img, + labels, degrees=hyp['degrees'], translate=hyp['translate'], scale=hyp['scale'], @@ -633,8 +662,7 @@ def load_image(self, i): h0, w0 = im.shape[:2] # orig hw r = self.img_size / max(h0, w0) # ratio if r != 1: # if sizes are not equal - im = cv2.resize(im, - (int(w0 * r), int(h0 * r)), + im = cv2.resize(im, (int(w0 * r), int(h0 * r)), interpolation=cv2.INTER_LINEAR if (self.augment or r > 1) else cv2.INTER_AREA) return im, (h0, w0), im.shape[:2] # im, hw_original, hw_resized else: @@ -692,7 +720,9 @@ def load_mosaic(self, index): # Augment img4, labels4, segments4 = copy_paste(img4, labels4, segments4, p=self.hyp['copy_paste']) - img4, labels4 = random_perspective(img4, labels4, segments4, + img4, labels4 = random_perspective(img4, + labels4, + segments4, degrees=self.hyp['degrees'], translate=self.hyp['translate'], scale=self.hyp['scale'], @@ -766,7 +796,9 @@ def load_mosaic9(self, index): # img9, labels9 = replicate(img9, labels9) # replicate # Augment - img9, labels9 = random_perspective(img9, labels9, segments9, + img9, labels9 = random_perspective(img9, + labels9, + segments9, degrees=self.hyp['degrees'], translate=self.hyp['translate'], scale=self.hyp['scale'], @@ -795,8 +827,8 @@ def collate_fn4(batch): for i in range(n): # zidane torch.zeros(16,3,720,1280) # BCHW i *= 4 if random.random() < 0.5: - im = F.interpolate(img[i].unsqueeze(0).float(), scale_factor=2.0, mode='bilinear', align_corners=False)[ - 0].type(img[i].type()) + im = F.interpolate(img[i].unsqueeze(0).float(), scale_factor=2.0, mode='bilinear', + align_corners=False)[0].type(img[i].type()) lb = label[i] else: im = torch.cat((torch.cat((img[i], img[i + 1]), 1), torch.cat((img[i + 2], img[i + 3]), 1)), 2) @@ -946,7 +978,6 @@ def dataset_stats(path='coco128.yaml', autodownload=False, verbose=False, profil autodownload: Attempt to download dataset if not found locally verbose: Print stats dictionary """ - def round_labels(labels): # Update labels to integer class and 6 decimal place floats return [[int(c), *(round(x, 4) for x in points)] for c, *points in labels] @@ -996,11 +1027,16 @@ def hub_ops(f, max_dim=1920): for label in tqdm(dataset.labels, total=dataset.n, desc='Statistics'): x.append(np.bincount(label[:, 0].astype(int), minlength=data['nc'])) x = np.array(x) # shape(128x80) - stats[split] = {'instance_stats': {'total': int(x.sum()), 'per_class': x.sum(0).tolist()}, - 'image_stats': {'total': dataset.n, 'unlabelled': int(np.all(x == 0, 1).sum()), - 'per_class': (x > 0).sum(0).tolist()}, - 'labels': [{str(Path(k).name): round_labels(v.tolist())} for k, v in - zip(dataset.im_files, dataset.labels)]} + stats[split] = { + 'instance_stats': { + 'total': int(x.sum()), + 'per_class': x.sum(0).tolist()}, + 'image_stats': { + 'total': dataset.n, + 'unlabelled': int(np.all(x == 0, 1).sum()), + 'per_class': (x > 0).sum(0).tolist()}, + 'labels': [{ + str(Path(k).name): round_labels(v.tolist())} for k, v in zip(dataset.im_files, dataset.labels)]} if hub: im_dir = hub_dir / 'images' diff --git a/utils/downloads.py b/utils/downloads.py index d7b87cb2cadd..4a012cc05849 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -63,19 +63,21 @@ def attempt_download(file, repo='ultralytics/yolov5'): # from utils.downloads i assets = [x['name'] for x in response['assets']] # release assets, i.e. ['yolov5s.pt', 'yolov5m.pt', ...] tag = response['tag_name'] # i.e. 'v1.0' except Exception: # fallback plan - assets = ['yolov5n.pt', 'yolov5s.pt', 'yolov5m.pt', 'yolov5l.pt', 'yolov5x.pt', - 'yolov5n6.pt', 'yolov5s6.pt', 'yolov5m6.pt', 'yolov5l6.pt', 'yolov5x6.pt'] + assets = [ + 'yolov5n.pt', 'yolov5s.pt', 'yolov5m.pt', 'yolov5l.pt', 'yolov5x.pt', 'yolov5n6.pt', 'yolov5s6.pt', + 'yolov5m6.pt', 'yolov5l6.pt', 'yolov5x6.pt'] try: tag = subprocess.check_output('git tag', shell=True, stderr=subprocess.STDOUT).decode().split()[-1] except Exception: tag = 'v6.0' # current release if name in assets: - safe_download(file, - url=f'https://github.com/{repo}/releases/download/{tag}/{name}', - # url2=f'https://storage.googleapis.com/{repo}/ckpt/{name}', # backup url (optional) - min_bytes=1E5, - error_msg=f'{file} missing, try downloading from https://github.com/{repo}/releases/') + safe_download( + file, + url=f'https://github.com/{repo}/releases/download/{tag}/{name}', + # url2=f'https://storage.googleapis.com/{repo}/ckpt/{name}', # backup url (optional) + min_bytes=1E5, + error_msg=f'{file} missing, try downloading from https://github.com/{repo}/releases/') return str(file) @@ -122,6 +124,7 @@ def get_token(cookie="./cookie"): return line.split()[-1] return "" + # Google utils: https://cloud.google.com/storage/docs/reference/libraries ---------------------------------------------- # # diff --git a/utils/general.py b/utils/general.py index 5905211cfa59..a64680bc06e5 100755 --- a/utils/general.py +++ b/utils/general.py @@ -536,25 +536,26 @@ def one_cycle(y1=0.0, y2=1.0, steps=100): def colorstr(*input): # Colors a string https://en.wikipedia.org/wiki/ANSI_escape_code, i.e. colorstr('blue', 'hello world') *args, string = input if len(input) > 1 else ('blue', 'bold', input[0]) # color arguments, string - colors = {'black': '\033[30m', # basic colors - 'red': '\033[31m', - 'green': '\033[32m', - 'yellow': '\033[33m', - 'blue': '\033[34m', - 'magenta': '\033[35m', - 'cyan': '\033[36m', - 'white': '\033[37m', - 'bright_black': '\033[90m', # bright colors - 'bright_red': '\033[91m', - 'bright_green': '\033[92m', - 'bright_yellow': '\033[93m', - 'bright_blue': '\033[94m', - 'bright_magenta': '\033[95m', - 'bright_cyan': '\033[96m', - 'bright_white': '\033[97m', - 'end': '\033[0m', # misc - 'bold': '\033[1m', - 'underline': '\033[4m'} + colors = { + 'black': '\033[30m', # basic colors + 'red': '\033[31m', + 'green': '\033[32m', + 'yellow': '\033[33m', + 'blue': '\033[34m', + 'magenta': '\033[35m', + 'cyan': '\033[36m', + 'white': '\033[37m', + 'bright_black': '\033[90m', # bright colors + 'bright_red': '\033[91m', + 'bright_green': '\033[92m', + 'bright_yellow': '\033[93m', + 'bright_blue': '\033[94m', + 'bright_magenta': '\033[95m', + 'bright_cyan': '\033[96m', + 'bright_white': '\033[97m', + 'end': '\033[0m', # misc + 'bold': '\033[1m', + 'underline': '\033[4m'} return ''.join(colors[x] for x in args) + f'{string}' + colors['end'] @@ -591,9 +592,10 @@ def coco80_to_coco91_class(): # converts 80-index (val2014) to 91-index (paper) # b = np.loadtxt('data/coco_paper.names', dtype='str', delimiter='\n') # x1 = [list(a[i] == b).index(True) + 1 for i in range(80)] # darknet to coco # x2 = [list(b[i] == a).index(True) if any(b[i] == a) else None for i in range(91)] # coco to darknet - x = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 27, 28, 31, 32, 33, 34, - 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, - 64, 65, 67, 70, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 84, 85, 86, 87, 88, 89, 90] + x = [ + 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 27, 28, 31, 32, 33, 34, + 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, + 64, 65, 67, 70, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 84, 85, 86, 87, 88, 89, 90] return x @@ -701,8 +703,14 @@ def clip_coords(boxes, shape): boxes[:, [1, 3]] = boxes[:, [1, 3]].clip(0, shape[0]) # y1, y2 -def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=None, agnostic=False, multi_label=False, - labels=(), max_det=300): +def non_max_suppression(prediction, + conf_thres=0.25, + iou_thres=0.45, + classes=None, + agnostic=False, + multi_label=False, + labels=(), + max_det=300): """Non-Maximum Suppression (NMS) on inference results to reject overlapping bounding boxes Returns: @@ -816,8 +824,8 @@ def strip_optimizer(f='best.pt', s=''): # from utils.general import *; strip_op def print_mutation(results, hyp, save_dir, bucket, prefix=colorstr('evolve: ')): evolve_csv = save_dir / 'evolve.csv' evolve_yaml = save_dir / 'hyp_evolve.yaml' - keys = ('metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95', - 'val/box_loss', 'val/obj_loss', 'val/cls_loss') + tuple(hyp.keys()) # [results + hyps] + keys = ('metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95', 'val/box_loss', + 'val/obj_loss', 'val/cls_loss') + tuple(hyp.keys()) # [results + hyps] keys = tuple(x.strip() for x in keys) vals = results + tuple(hyp.values()) n = len(keys) @@ -839,17 +847,15 @@ def print_mutation(results, hyp, save_dir, bucket, prefix=colorstr('evolve: ')): data = data.rename(columns=lambda x: x.strip()) # strip keys i = np.argmax(fitness(data.values[:, :4])) # generations = len(data) - f.write('# YOLOv5 Hyperparameter Evolution Results\n' + - f'# Best generation: {i}\n' + - f'# Last generation: {generations - 1}\n' + - '# ' + ', '.join(f'{x.strip():>20s}' for x in keys[:7]) + '\n' + - '# ' + ', '.join(f'{x:>20.5g}' for x in data.values[i, :7]) + '\n\n') + f.write('# YOLOv5 Hyperparameter Evolution Results\n' + f'# Best generation: {i}\n' + + f'# Last generation: {generations - 1}\n' + '# ' + ', '.join(f'{x.strip():>20s}' for x in keys[:7]) + + '\n' + '# ' + ', '.join(f'{x:>20.5g}' for x in data.values[i, :7]) + '\n\n') yaml.safe_dump(data.loc[i][7:].to_dict(), f, sort_keys=False) # Print to screen - LOGGER.info(prefix + f'{generations} generations finished, current result:\n' + - prefix + ', '.join(f'{x.strip():>20s}' for x in keys) + '\n' + - prefix + ', '.join(f'{x:20.5g}' for x in vals) + '\n\n') + LOGGER.info(prefix + f'{generations} generations finished, current result:\n' + prefix + + ', '.join(f'{x.strip():>20s}' for x in keys) + '\n' + prefix + ', '.join(f'{x:20.5g}' + for x in vals) + '\n\n') if bucket: os.system(f'gsutil cp {evolve_csv} {evolve_yaml} gs://{bucket}') # upload diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index bb8523c0219e..2e639dfb9b53 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -43,10 +43,20 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, self.hyp = hyp self.logger = logger # for printing results to console self.include = include - self.keys = ['train/box_loss', 'train/obj_loss', 'train/cls_loss', # train loss - 'metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95', # metrics - 'val/box_loss', 'val/obj_loss', 'val/cls_loss', # val loss - 'x/lr0', 'x/lr1', 'x/lr2'] # params + self.keys = [ + 'train/box_loss', + 'train/obj_loss', + 'train/cls_loss', # train loss + 'metrics/precision', + 'metrics/recall', + 'metrics/mAP_0.5', + 'metrics/mAP_0.5:0.95', # metrics + 'val/box_loss', + 'val/obj_loss', + 'val/cls_loss', # val loss + 'x/lr0', + 'x/lr1', + 'x/lr2'] # params self.best_keys = ['best/epoch', 'best/precision', 'best/recall', 'best/mAP_0.5', 'best/mAP_0.5:0.95'] for k in LOGGERS: setattr(self, k, None) # init empty logger dictionary @@ -155,7 +165,8 @@ def on_train_end(self, last, best, plots, epoch, results): self.wandb.log({"Results": [wandb.Image(str(f), caption=f.name) for f in files]}) # Calling wandb.log. TODO: Refactor this into WandbLogger.log_model if not self.opt.evolve: - wandb.log_artifact(str(best if best.exists() else last), type='model', + wandb.log_artifact(str(best if best.exists() else last), + type='model', name='run_' + self.wandb.wandb_run.id + '_model', aliases=['latest', 'best', 'stripped']) self.wandb.finish_run() diff --git a/utils/loggers/wandb/wandb_utils.py b/utils/loggers/wandb/wandb_utils.py index 786e58a19972..6ec2559e29ac 100644 --- a/utils/loggers/wandb/wandb_utils.py +++ b/utils/loggers/wandb/wandb_utils.py @@ -46,10 +46,10 @@ def check_wandb_dataset(data_file): if check_file(data_file) and data_file.endswith('.yaml'): with open(data_file, errors='ignore') as f: data_dict = yaml.safe_load(f) - is_trainset_wandb_artifact = (isinstance(data_dict['train'], str) and - data_dict['train'].startswith(WANDB_ARTIFACT_PREFIX)) - is_valset_wandb_artifact = (isinstance(data_dict['val'], str) and - data_dict['val'].startswith(WANDB_ARTIFACT_PREFIX)) + is_trainset_wandb_artifact = isinstance(data_dict['train'], + str) and data_dict['train'].startswith(WANDB_ARTIFACT_PREFIX) + is_valset_wandb_artifact = isinstance(data_dict['val'], + str) and data_dict['val'].startswith(WANDB_ARTIFACT_PREFIX) if is_trainset_wandb_artifact or is_valset_wandb_artifact: return data_dict else: @@ -116,7 +116,6 @@ class WandbLogger(): For more on how this logger is used, see the Weights & Biases documentation: https://docs.wandb.com/guides/integrations/yolov5 """ - def __init__(self, opt, run_id=None, job_type='Training'): """ - Initialize WandbLogger instance @@ -181,8 +180,7 @@ def __init__(self, opt, run_id=None, job_type='Training'): self.wandb_artifact_data_dict = self.wandb_artifact_data_dict or self.data_dict # write data_dict to config. useful for resuming from artifacts. Do this only when not resuming. - self.wandb_run.config.update({'data_dict': self.wandb_artifact_data_dict}, - allow_val_change=True) + self.wandb_run.config.update({'data_dict': self.wandb_artifact_data_dict}, allow_val_change=True) self.setup_training(opt) if self.job_type == 'Dataset Creation': @@ -200,8 +198,7 @@ def check_and_upload_dataset(self, opt): Updated dataset info dictionary where local dataset paths are replaced by WAND_ARFACT_PREFIX links. """ assert wandb, 'Install wandb to upload dataset' - config_path = self.log_dataset_artifact(opt.data, - opt.single_cls, + config_path = self.log_dataset_artifact(opt.data, opt.single_cls, 'YOLOv5' if opt.project == 'runs/train' else Path(opt.project).stem) with open(config_path, errors='ignore') as f: wandb_data_dict = yaml.safe_load(f) @@ -230,10 +227,10 @@ def setup_training(self, opt): config.hyp, config.imgsz data_dict = self.data_dict if self.val_artifact is None: # If --upload_dataset is set, use the existing artifact, don't download - self.train_artifact_path, self.train_artifact = self.download_dataset_artifact(data_dict.get('train'), - opt.artifact_alias) - self.val_artifact_path, self.val_artifact = self.download_dataset_artifact(data_dict.get('val'), - opt.artifact_alias) + self.train_artifact_path, self.train_artifact = self.download_dataset_artifact( + data_dict.get('train'), opt.artifact_alias) + self.val_artifact_path, self.val_artifact = self.download_dataset_artifact( + data_dict.get('val'), opt.artifact_alias) if self.train_artifact_path is not None: train_path = Path(self.train_artifact_path) / 'data/images/' @@ -308,14 +305,15 @@ def log_model(self, path, opt, epoch, fitness_score, best_model=False): fitness_score (float) -- fitness score for current epoch best_model (boolean) -- Boolean representing if the current checkpoint is the best yet. """ - model_artifact = wandb.Artifact('run_' + wandb.run.id + '_model', type='model', metadata={ - 'original_url': str(path), - 'epochs_trained': epoch + 1, - 'save period': opt.save_period, - 'project': opt.project, - 'total_epochs': opt.epochs, - 'fitness_score': fitness_score - }) + model_artifact = wandb.Artifact('run_' + wandb.run.id + '_model', + type='model', + metadata={ + 'original_url': str(path), + 'epochs_trained': epoch + 1, + 'save period': opt.save_period, + 'project': opt.project, + 'total_epochs': opt.epochs, + 'fitness_score': fitness_score}) model_artifact.add_file(str(path / 'last.pt'), name='last.pt') wandb.log_artifact(model_artifact, aliases=['latest', 'last', 'epoch ' + str(self.current_epoch), 'best' if best_model else '']) @@ -344,13 +342,14 @@ def log_dataset_artifact(self, data_file, single_cls, project, overwrite_config= # log train set if not log_val_only: - self.train_artifact = self.create_dataset_table(LoadImagesAndLabels( - data['train'], rect=True, batch_size=1), names, name='train') if data.get('train') else None + self.train_artifact = self.create_dataset_table(LoadImagesAndLabels(data['train'], rect=True, batch_size=1), + names, + name='train') if data.get('train') else None if data.get('train'): data['train'] = WANDB_ARTIFACT_PREFIX + str(Path(project) / 'train') - self.val_artifact = self.create_dataset_table(LoadImagesAndLabels( - data['val'], rect=True, batch_size=1), names, name='val') if data.get('val') else None + self.val_artifact = self.create_dataset_table( + LoadImagesAndLabels(data['val'], rect=True, batch_size=1), names, name='val') if data.get('val') else None if data.get('val'): data['val'] = WANDB_ARTIFACT_PREFIX + str(Path(project) / 'val') @@ -412,17 +411,21 @@ def create_dataset_table(self, dataset: LoadImagesAndLabels, class_to_id: Dict[i else: artifact.add_file(img_file, name='data/images/' + Path(img_file).name) label_file = Path(img2label_paths([img_file])[0]) - artifact.add_file(str(label_file), - name='data/labels/' + label_file.name) if label_file.exists() else None + artifact.add_file(str(label_file), name='data/labels/' + + label_file.name) if label_file.exists() else None table = wandb.Table(columns=["id", "train_image", "Classes", "name"]) class_set = wandb.Classes([{'id': id, 'name': name} for id, name in class_to_id.items()]) for si, (img, labels, paths, shapes) in enumerate(tqdm(dataset)): box_data, img_classes = [], {} for cls, *xywh in labels[:, 1:].tolist(): cls = int(cls) - box_data.append({"position": {"middle": [xywh[0], xywh[1]], "width": xywh[2], "height": xywh[3]}, - "class_id": cls, - "box_caption": "%s" % (class_to_id[cls])}) + box_data.append({ + "position": { + "middle": [xywh[0], xywh[1]], + "width": xywh[2], + "height": xywh[3]}, + "class_id": cls, + "box_caption": "%s" % (class_to_id[cls])}) img_classes[cls] = class_to_id[cls] boxes = {"ground_truth": {"box_data": box_data, "class_labels": class_to_id}} # inference-space table.add_data(si, wandb.Image(paths, classes=class_set, boxes=boxes), list(img_classes.values()), @@ -446,12 +449,17 @@ def log_training_progress(self, predn, path, names): for *xyxy, conf, cls in predn.tolist(): if conf >= 0.25: cls = int(cls) - box_data.append( - {"position": {"minX": xyxy[0], "minY": xyxy[1], "maxX": xyxy[2], "maxY": xyxy[3]}, - "class_id": cls, - "box_caption": f"{names[cls]} {conf:.3f}", - "scores": {"class_score": conf}, - "domain": "pixel"}) + box_data.append({ + "position": { + "minX": xyxy[0], + "minY": xyxy[1], + "maxX": xyxy[2], + "maxY": xyxy[3]}, + "class_id": cls, + "box_caption": f"{names[cls]} {conf:.3f}", + "scores": { + "class_score": conf}, + "domain": "pixel"}) avg_conf_per_class[cls] += conf if cls in pred_class_count: @@ -464,12 +472,9 @@ def log_training_progress(self, predn, path, names): boxes = {"predictions": {"box_data": box_data, "class_labels": names}} # inference-space id = self.val_table_path_map[Path(path).name] - self.result_table.add_data(self.current_epoch, - id, - self.val_table.data[id][1], + self.result_table.add_data(self.current_epoch, id, self.val_table.data[id][1], wandb.Image(self.val_table.data[id][1], boxes=boxes, classes=class_set), - *avg_conf_per_class - ) + *avg_conf_per_class) def val_one_image(self, pred, predn, path, names, im): """ @@ -485,11 +490,17 @@ def val_one_image(self, pred, predn, path, names, im): if len(self.bbox_media_panel_images) < self.max_imgs_to_log and self.current_epoch > 0: if self.current_epoch % self.bbox_interval == 0: - box_data = [{"position": {"minX": xyxy[0], "minY": xyxy[1], "maxX": xyxy[2], "maxY": xyxy[3]}, - "class_id": int(cls), - "box_caption": f"{names[int(cls)]} {conf:.3f}", - "scores": {"class_score": conf}, - "domain": "pixel"} for *xyxy, conf, cls in pred.tolist()] + box_data = [{ + "position": { + "minX": xyxy[0], + "minY": xyxy[1], + "maxX": xyxy[2], + "maxY": xyxy[3]}, + "class_id": int(cls), + "box_caption": f"{names[int(cls)]} {conf:.3f}", + "scores": { + "class_score": conf}, + "domain": "pixel"} for *xyxy, conf, cls in pred.tolist()] boxes = {"predictions": {"box_data": box_data, "class_labels": names}} # inference-space self.bbox_media_panel_images.append(wandb.Image(im, boxes=boxes, caption=path.name)) @@ -519,7 +530,8 @@ def end_epoch(self, best_result=False): wandb.log(self.log_dict) except BaseException as e: LOGGER.info( - f"An error occurred in wandb logger. The training will proceed without interruption. More info\n{e}") + f"An error occurred in wandb logger. The training will proceed without interruption. More info\n{e}" + ) self.wandb_run.finish() self.wandb_run = None @@ -527,8 +539,10 @@ def end_epoch(self, best_result=False): self.bbox_media_panel_images = [] if self.result_artifact: self.result_artifact.add(self.result_table, 'result') - wandb.log_artifact(self.result_artifact, aliases=['latest', 'last', 'epoch ' + str(self.current_epoch), - ('best' if best_result else '')]) + wandb.log_artifact(self.result_artifact, + aliases=[ + 'latest', 'last', 'epoch ' + str(self.current_epoch), + ('best' if best_result else '')]) wandb.log({"evaluation": self.result_table}) columns = ["epoch", "id", "ground truth", "prediction"] diff --git a/utils/loss.py b/utils/loss.py index bf9b592d4ad2..fa8095515477 100644 --- a/utils/loss.py +++ b/utils/loss.py @@ -183,10 +183,16 @@ def build_targets(self, p, targets): targets = torch.cat((targets.repeat(na, 1, 1), ai[:, :, None]), 2) # append anchor indices g = 0.5 # bias - off = torch.tensor([[0, 0], - [1, 0], [0, 1], [-1, 0], [0, -1], # j,k,l,m - # [1, 1], [1, -1], [-1, 1], [-1, -1], # jk,jm,lk,lm - ], device=self.device).float() * g # offsets + off = torch.tensor( + [ + [0, 0], + [1, 0], + [0, 1], + [-1, 0], + [0, -1], # j,k,l,m + # [1, 1], [1, -1], [-1, 1], [-1, -1], # jk,jm,lk,lm + ], + device=self.device).float() * g # offsets for i in range(self.nl): anchors = self.anchors[i] diff --git a/utils/metrics.py b/utils/metrics.py index 857fa5d81f91..216956e90ecc 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -184,7 +184,14 @@ def plot(self, normalize=True, save_dir='', names=()): labels = (0 < nn < 99) and (nn == nc) # apply names to ticklabels with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress empty matrix RuntimeWarning: All-NaN slice encountered - sn.heatmap(array, annot=nc < 30, annot_kws={"size": 8}, cmap='Blues', fmt='.2f', square=True, vmin=0.0, + sn.heatmap(array, + annot=nc < 30, + annot_kws={ + "size": 8}, + cmap='Blues', + fmt='.2f', + square=True, + vmin=0.0, xticklabels=names + ['background FP'] if labels else "auto", yticklabels=names + ['background FN'] if labels else "auto").set_facecolor((1, 1, 1)) fig.axes[0].set_xlabel('True') @@ -253,7 +260,6 @@ def box_iou(box1, box2): iou (Tensor[N, M]): the NxM matrix containing the pairwise IoU values for every element in boxes1 and boxes2 """ - def box_area(box): # box = 4xn return (box[2] - box[0]) * (box[3] - box[1]) @@ -300,6 +306,7 @@ def wh_iou(wh1, wh2): # Plots ---------------------------------------------------------------------------------------------------------------- + def plot_pr_curve(px, py, ap, save_dir='pr_curve.png', names=()): # Precision-recall curve fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True) diff --git a/utils/plots.py b/utils/plots.py index a30c0faf962a..51e9cfdf6e04 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -89,10 +89,11 @@ def box_label(self, box, label='', color=(128, 128, 128), txt_color=(255, 255, 2 if label: w, h = self.font.getsize(label) # text width, height outside = box[1] - h >= 0 # label fits outside box - self.draw.rectangle((box[0], - box[1] - h if outside else box[1], - box[0] + w + 1, - box[1] + 1 if outside else box[1] + h + 1), fill=color) + self.draw.rectangle( + (box[0], box[1] - h if outside else box[1], box[0] + w + 1, + box[1] + 1 if outside else box[1] + h + 1), + fill=color, + ) # self.draw.text((box[0], box[1]), label, fill=txt_color, font=self.font, anchor='ls') # for PIL>8.0 self.draw.text((box[0], box[1] - h if outside else box[1]), label, fill=txt_color, font=self.font) else: # cv2 @@ -104,8 +105,13 @@ def box_label(self, box, label='', color=(128, 128, 128), txt_color=(255, 255, 2 outside = p1[1] - h - 3 >= 0 # label fits outside box p2 = p1[0] + w, p1[1] - h - 3 if outside else p1[1] + h + 3 cv2.rectangle(self.im, p1, p2, color, -1, cv2.LINE_AA) # filled - cv2.putText(self.im, label, (p1[0], p1[1] - 2 if outside else p1[1] + h + 2), 0, self.lw / 3, txt_color, - thickness=tf, lineType=cv2.LINE_AA) + cv2.putText(self.im, + label, (p1[0], p1[1] - 2 if outside else p1[1] + h + 2), + 0, + self.lw / 3, + txt_color, + thickness=tf, + lineType=cv2.LINE_AA) def rectangle(self, xy, fill=None, outline=None, width=1): # Add rectangle to image (PIL-only) @@ -307,11 +313,19 @@ def plot_val_study(file='', dir='', x=None): # from utils.plots import *; plot_ ax[i].set_title(s[i]) j = y[3].argmax() + 1 - ax2.plot(y[5, 1:j], y[3, 1:j] * 1E2, '.-', linewidth=2, markersize=8, + ax2.plot(y[5, 1:j], + y[3, 1:j] * 1E2, + '.-', + linewidth=2, + markersize=8, label=f.stem.replace('study_coco_', '').replace('yolo', 'YOLO')) ax2.plot(1E3 / np.array([209, 140, 97, 58, 35, 18]), [34.6, 40.5, 43.0, 47.5, 49.7, 51.5], - 'k.-', linewidth=2, markersize=8, alpha=.25, label='EfficientDet') + 'k.-', + linewidth=2, + markersize=8, + alpha=.25, + label='EfficientDet') ax2.grid(alpha=0.2) ax2.set_yticks(np.arange(20, 60, 5)) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 72f8a0fd1659..bc96ec75be7c 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -284,7 +284,6 @@ class ModelEMA: Keeps a moving average of everything in the model state_dict (parameters and buffers) For EMA details see https://www.tensorflow.org/api_docs/python/tf/train/ExponentialMovingAverage """ - def __init__(self, model, decay=0.9999, tau=2000, updates=0): # Create EMA self.ema = deepcopy(de_parallel(model)).eval() # FP32 EMA diff --git a/val.py b/val.py index 2dd2aec679f9..bc4abc248dc8 100644 --- a/val.py +++ b/val.py @@ -62,10 +62,11 @@ def save_one_json(predn, jdict, path, class_map): box = xyxy2xywh(predn[:, :4]) # xywh box[:, :2] -= box[:, 2:] / 2 # xy center to top-left corner for p, b in zip(predn.tolist(), box.tolist()): - jdict.append({'image_id': image_id, - 'category_id': class_map[int(p[5])], - 'bbox': [round(x, 3) for x in b], - 'score': round(p[4], 5)}) + jdict.append({ + 'image_id': image_id, + 'category_id': class_map[int(p[5])], + 'bbox': [round(x, 3) for x in b], + 'score': round(p[4], 5)}) def process_batch(detections, labels, iouv): @@ -93,7 +94,8 @@ def process_batch(detections, labels, iouv): @torch.no_grad() -def run(data, +def run( + data, weights=None, # model.pt path(s) batch_size=32, # batch size imgsz=640, # inference size (pixels) @@ -120,7 +122,7 @@ def run(data, plots=True, callbacks=Callbacks(), compute_loss=None, - ): +): # Initialize/load model and set device training = model is not None if training: # called by train.py @@ -164,8 +166,15 @@ def run(data, pad = 0.0 if task in ('speed', 'benchmark') else 0.5 rect = False if task == 'benchmark' else pt # square inference for benchmarks task = task if task in ('train', 'val', 'test') else 'val' # path to train/val/test images - dataloader = create_dataloader(data[task], imgsz, batch_size, stride, single_cls, pad=pad, rect=rect, - workers=workers, prefix=colorstr(f'{task}: '))[0] + dataloader = create_dataloader(data[task], + imgsz, + batch_size, + stride, + single_cls, + pad=pad, + rect=rect, + workers=workers, + prefix=colorstr(f'{task}: '))[0] seen = 0 confusion_matrix = ConfusionMatrix(nc=nc) From 2c3221844b604c7e3f26c1f26d0c5ed78f700fd5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 31 Mar 2022 17:11:43 +0200 Subject: [PATCH 141/661] CLI `fire` prep updates (#7229) * CLI fire prep updates * revert unintentional TF export change --- detect.py | 2 +- export.py | 2 +- models/tf.py | 2 +- models/yolo.py | 2 +- train.py | 2 +- utils/benchmarks.py | 2 +- utils/general.py | 15 ++++++++++++--- val.py | 2 +- 8 files changed, 19 insertions(+), 10 deletions(-) diff --git a/detect.py b/detect.py index 2875285ee314..14ff9a6ab421 100644 --- a/detect.py +++ b/detect.py @@ -238,7 +238,7 @@ def parse_opt(): parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') opt = parser.parse_args() opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand - print_args(FILE.stem, opt) + print_args(vars(opt)) return opt diff --git a/export.py b/export.py index 78b886fa3a6b..e146dad42980 100644 --- a/export.py +++ b/export.py @@ -566,7 +566,7 @@ def parse_opt(): default=['torchscript', 'onnx'], help='torchscript, onnx, openvino, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs') opt = parser.parse_args() - print_args(FILE.stem, opt) + print_args(vars(opt)) return opt diff --git a/models/tf.py b/models/tf.py index c6fb6b82a72e..1b7653bce8f6 100644 --- a/models/tf.py +++ b/models/tf.py @@ -480,7 +480,7 @@ def parse_opt(): parser.add_argument('--dynamic', action='store_true', help='dynamic batch size') opt = parser.parse_args() opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand - print_args(FILE.stem, opt) + print_args(vars(opt)) return opt diff --git a/models/yolo.py b/models/yolo.py index 4cdfea34d63e..e18614cb37bd 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -308,7 +308,7 @@ def parse_model(d, ch): # model_dict, input_channels(3) parser.add_argument('--test', action='store_true', help='test all yolo*.yaml') opt = parser.parse_args() opt.cfg = check_yaml(opt.cfg) # check YAML - print_args(FILE.stem, opt) + print_args(vars(opt)) device = select_device(opt.device) # Create model diff --git a/train.py b/train.py index fbaaeb8ef930..38c25c053e26 100644 --- a/train.py +++ b/train.py @@ -515,7 +515,7 @@ def parse_opt(known=False): def main(opt, callbacks=Callbacks()): # Checks if RANK in [-1, 0]: - print_args(FILE.stem, opt) + print_args(vars(opt)) check_git_status() check_requirements(exclude=['thop']) diff --git a/utils/benchmarks.py b/utils/benchmarks.py index 5bfa872cc3fb..69243725b48a 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -92,7 +92,7 @@ def parse_opt(): parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') opt = parser.parse_args() - print_args(FILE.stem, opt) + print_args(vars(opt)) return opt diff --git a/utils/general.py b/utils/general.py index a64680bc06e5..9622a32c5c70 100755 --- a/utils/general.py +++ b/utils/general.py @@ -5,6 +5,7 @@ import contextlib import glob +import inspect import logging import math import os @@ -20,6 +21,7 @@ from multiprocessing.pool import ThreadPool from pathlib import Path from subprocess import check_output +from typing import Optional from zipfile import ZipFile import cv2 @@ -163,9 +165,15 @@ def methods(instance): return [f for f in dir(instance) if callable(getattr(instance, f)) and not f.startswith("__")] -def print_args(name, opt): - # Print argparser arguments - LOGGER.info(colorstr(f'{name}: ') + ', '.join(f'{k}={v}' for k, v in vars(opt).items())) +def print_args(args: Optional[dict] = None, show_file=True, show_fcn=False): + # Print function arguments (optional args dict) + x = inspect.currentframe().f_back # previous frame + file, _, fcn, _, _ = inspect.getframeinfo(x) + if args is None: # get args automatically + args, _, _, frm = inspect.getargvalues(x) + args = {k: v for k, v in frm.items() if k in args} + s = (f'{Path(file).stem}: ' if show_file else '') + (f'{fcn}: ' if show_fcn else '') + LOGGER.info(colorstr(s) + ', '.join(f'{k}={v}' for k, v in args.items())) def init_seeds(seed=0): @@ -346,6 +354,7 @@ def check_img_size(imgsz, s=32, floor=0): if isinstance(imgsz, int): # integer i.e. img_size=640 new_size = max(make_divisible(imgsz, int(s)), floor) else: # list i.e. img_size=[640, 480] + imgsz = list(imgsz) # convert to list if tuple new_size = [max(make_divisible(x, int(s)), floor) for x in imgsz] if new_size != imgsz: LOGGER.warning(f'WARNING: --img-size {imgsz} must be multiple of max stride {s}, updating to {new_size}') diff --git a/val.py b/val.py index bc4abc248dc8..019beedea61a 100644 --- a/val.py +++ b/val.py @@ -350,7 +350,7 @@ def parse_opt(): opt.data = check_yaml(opt.data) # check YAML opt.save_json |= opt.data.endswith('coco.yaml') opt.save_txt |= opt.save_hybrid - print_args(FILE.stem, opt) + print_args(vars(opt)) return opt From 4d157f578a7bbff08d1e17a4e6e47aece4d91207 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 31 Mar 2022 17:26:34 +0200 Subject: [PATCH 142/661] Update .pre-commit-config.yaml (#7230) --- .pre-commit-config.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 0b4fedcd2d43..208cb072c8aa 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -40,7 +40,7 @@ repos: rev: v0.31.0 hooks: - id: yapf - name: formatting + name: YAPF formatting # TODO #- repo: https://github.com/executablebooks/mdformat From 734ab033fdd7542bde14cab6c040415eb51dc9ac Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 1 Apr 2022 00:07:23 +0200 Subject: [PATCH 143/661] SavedModel TF Serve Fix (#7228) * SavedModel TF Serve Fix Fix for https://github.com/ultralytics/yolov5/issues/7205 proposed by @tylertroy * Update export.py --- export.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/export.py b/export.py index e146dad42980..cc7a74db9af2 100644 --- a/export.py +++ b/export.py @@ -285,12 +285,12 @@ def export_saved_model(model, if keras: keras_model.save(f, save_format='tf') else: - m = tf.function(lambda x: keras_model(x)) # full model spec = tf.TensorSpec(keras_model.inputs[0].shape, keras_model.inputs[0].dtype) + m = tf.function(lambda x: keras_model(x)) # full model m = m.get_concrete_function(spec) frozen_func = convert_variables_to_constants_v2(m) tfm = tf.Module() - tfm.__call__ = tf.function(lambda x: frozen_func(x)[0], [spec]) + tfm.__call__ = tf.function(lambda x: frozen_func(x)[:4] if tf_nms else frozen_func(x)[0], [spec]) tfm.__call__(im) tf.saved_model.save(tfm, f, From 71621df87589faea19ba4c4098bb68e73201f30c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 1 Apr 2022 00:24:37 +0200 Subject: [PATCH 144/661] Create CODE_OF_CONDUCT.md (#7233) --- .github/CODE_OF_CONDUCT.md | 128 +++++++++++++++++++++++++++++++++++++ 1 file changed, 128 insertions(+) create mode 100644 .github/CODE_OF_CONDUCT.md diff --git a/.github/CODE_OF_CONDUCT.md b/.github/CODE_OF_CONDUCT.md new file mode 100644 index 000000000000..ef10b05fc88e --- /dev/null +++ b/.github/CODE_OF_CONDUCT.md @@ -0,0 +1,128 @@ +# YOLOv5 🚀 Contributor Covenant Code of Conduct + +## Our Pledge + +We as members, contributors, and leaders pledge to make participation in our +community a harassment-free experience for everyone, regardless of age, body +size, visible or invisible disability, ethnicity, sex characteristics, gender +identity and expression, level of experience, education, socio-economic status, +nationality, personal appearance, race, religion, or sexual identity +and orientation. + +We pledge to act and interact in ways that contribute to an open, welcoming, +diverse, inclusive, and healthy community. + +## Our Standards + +Examples of behavior that contributes to a positive environment for our +community include: + +* Demonstrating empathy and kindness toward other people +* Being respectful of differing opinions, viewpoints, and experiences +* Giving and gracefully accepting constructive feedback +* Accepting responsibility and apologizing to those affected by our mistakes, + and learning from the experience +* Focusing on what is best not just for us as individuals, but for the + overall community + +Examples of unacceptable behavior include: + +* The use of sexualized language or imagery, and sexual attention or + advances of any kind +* Trolling, insulting or derogatory comments, and personal or political attacks +* Public or private harassment +* Publishing others' private information, such as a physical or email + address, without their explicit permission +* Other conduct which could reasonably be considered inappropriate in a + professional setting + +## Enforcement Responsibilities + +Community leaders are responsible for clarifying and enforcing our standards of +acceptable behavior and will take appropriate and fair corrective action in +response to any behavior that they deem inappropriate, threatening, offensive, +or harmful. + +Community leaders have the right and responsibility to remove, edit, or reject +comments, commits, code, wiki edits, issues, and other contributions that are +not aligned to this Code of Conduct, and will communicate reasons for moderation +decisions when appropriate. + +## Scope + +This Code of Conduct applies within all community spaces, and also applies when +an individual is officially representing the community in public spaces. +Examples of representing our community include using an official e-mail address, +posting via an official social media account, or acting as an appointed +representative at an online or offline event. + +## Enforcement + +Instances of abusive, harassing, or otherwise unacceptable behavior may be +reported to the community leaders responsible for enforcement at +hello@ultralytics.com. +All complaints will be reviewed and investigated promptly and fairly. + +All community leaders are obligated to respect the privacy and security of the +reporter of any incident. + +## Enforcement Guidelines + +Community leaders will follow these Community Impact Guidelines in determining +the consequences for any action they deem in violation of this Code of Conduct: + +### 1. Correction + +**Community Impact**: Use of inappropriate language or other behavior deemed +unprofessional or unwelcome in the community. + +**Consequence**: A private, written warning from community leaders, providing +clarity around the nature of the violation and an explanation of why the +behavior was inappropriate. A public apology may be requested. + +### 2. Warning + +**Community Impact**: A violation through a single incident or series +of actions. + +**Consequence**: A warning with consequences for continued behavior. No +interaction with the people involved, including unsolicited interaction with +those enforcing the Code of Conduct, for a specified period of time. This +includes avoiding interactions in community spaces as well as external channels +like social media. Violating these terms may lead to a temporary or +permanent ban. + +### 3. Temporary Ban + +**Community Impact**: A serious violation of community standards, including +sustained inappropriate behavior. + +**Consequence**: A temporary ban from any sort of interaction or public +communication with the community for a specified period of time. No public or +private interaction with the people involved, including unsolicited interaction +with those enforcing the Code of Conduct, is allowed during this period. +Violating these terms may lead to a permanent ban. + +### 4. Permanent Ban + +**Community Impact**: Demonstrating a pattern of violation of community +standards, including sustained inappropriate behavior, harassment of an +individual, or aggression toward or disparagement of classes of individuals. + +**Consequence**: A permanent ban from any sort of public interaction within +the community. + +## Attribution + +This Code of Conduct is adapted from the [Contributor Covenant][homepage], +version 2.0, available at +https://www.contributor-covenant.org/version/2/0/code_of_conduct.html. + +Community Impact Guidelines were inspired by [Mozilla's code of conduct +enforcement ladder](https://github.com/mozilla/diversity). + +[homepage]: https://www.contributor-covenant.org + +For answers to common questions about this code of conduct, see the FAQ at +https://www.contributor-covenant.org/faq. Translations are available at +https://www.contributor-covenant.org/translations. From 37675e110f3d2635dbc3acc8794e782c452e4ad5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 1 Apr 2022 21:38:49 +0200 Subject: [PATCH 145/661] Fix `www.youtube.com` hostname (#7242) * Fix `www.youtube.com` hostname * Update datasets.py --- utils/datasets.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/datasets.py b/utils/datasets.py index 7e8b423c3174..b2d4fa54ae0d 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -316,7 +316,7 @@ def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True): for i, s in enumerate(sources): # index, source # Start thread to read frames from video stream st = f'{i + 1}/{n}: {s}... ' - if urlparse(s).hostname in ('youtube.com', 'youtu.be'): # if source is YouTube video + if urlparse(s).hostname in ('www.youtube.com', 'youtube.com', 'youtu.be'): # if source is YouTube video check_requirements(('pafy', 'youtube_dl==2020.12.2')) import pafy s = pafy.new(s).getbest(preftype="mp4").url # YouTube URL From a19406b39dbc45db0bbae8d0b7da9d6281f9af1e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 2 Apr 2022 15:05:00 +0200 Subject: [PATCH 146/661] Update minimum Python>=3.7.0 (#7247) --- utils/general.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index 9622a32c5c70..379e9e0f47a4 100755 --- a/utils/general.py +++ b/utils/general.py @@ -295,7 +295,7 @@ def check_git_status(): LOGGER.info(emojis(s)) # emoji-safe -def check_python(minimum='3.6.2'): +def check_python(minimum='3.7.0'): # Check current python version vs. required python version check_version(platform.python_version(), minimum, name='Python ', hard=True) From 6f4eb95af72589c0f751111978631db8d38da7f0 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 00:18:18 +0200 Subject: [PATCH 147/661] Update setup.cfg to `description_file` field (#7248) Resolve `UserWarning: Usage of dash-separated 'description-file' will not be supported in future versions. Please use the underscore name 'description_file' instead` --- setup.cfg | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/setup.cfg b/setup.cfg index c387d84a33e2..020a75740e97 100644 --- a/setup.cfg +++ b/setup.cfg @@ -4,7 +4,7 @@ [metadata] license_file = LICENSE -description-file = README.md +description_file = README.md [tool:pytest] From 3d3483cf0c085977d66684c0e2439ea31f38ab60 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 12:14:12 +0200 Subject: [PATCH 148/661] Update tutorial.ipynb (#7254) --- tutorial.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 0379fb1a3c57..1a6d41526140 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -1078,7 +1078,7 @@ "source": [ "# VOC\n", "for b, m in zip([64, 64, 64, 32, 16], ['yolov5n', 'yolov5s', 'yolov5m', 'yolov5l', 'yolov5x']): # zip(batch_size, model)\n", - " !python train.py --batch {b} --weights {m}.pt --data VOC.yaml --epochs 50 --cache --img 512 --nosave --hyp hyp.VOC.yaml --project VOC --name {m}" + " !python train.py --batch {b} --weights {m}.pt --data VOC.yaml --epochs 50 --img 512 --hyp hyp.VOC.yaml --project VOC --name {m} --cache" ], "execution_count": null, "outputs": [] From 035b5548e47541767565a1934054bf47404757df Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 12:18:24 +0200 Subject: [PATCH 149/661] Update tutorial.ipynb (#7255) --- tutorial.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 1a6d41526140..d5a10dfd5952 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -1077,7 +1077,7 @@ }, "source": [ "# VOC\n", - "for b, m in zip([64, 64, 64, 32, 16], ['yolov5n', 'yolov5s', 'yolov5m', 'yolov5l', 'yolov5x']): # zip(batch_size, model)\n", + "for b, m in zip([64, 64, 64, 32, 16], ['yolov5n', 'yolov5s', 'yolov5m', 'yolov5l', 'yolov5x']): # batch, model\n", " !python train.py --batch {b} --weights {m}.pt --data VOC.yaml --epochs 50 --img 512 --hyp hyp.VOC.yaml --project VOC --name {m} --cache" ], "execution_count": null, From dda669a12c4df7b282a1378e251f8314e6179bcb Mon Sep 17 00:00:00 2001 From: Zengyf-CVer <41098760+Zengyf-CVer@users.noreply.github.com> Date: Sun, 3 Apr 2022 19:19:26 +0800 Subject: [PATCH 150/661] Fix Flask REST API (#7210) * Update restapi.py * Update restapi.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Cleanup * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/flask_rest_api/restapi.py | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/utils/flask_rest_api/restapi.py b/utils/flask_rest_api/restapi.py index b93ad16a0f58..38868cc98d84 100644 --- a/utils/flask_rest_api/restapi.py +++ b/utils/flask_rest_api/restapi.py @@ -1,5 +1,5 @@ """ -Run a rest API exposing the yolov5s object detection model +Run a Flask REST API exposing a YOLOv5s model """ import argparse import io @@ -31,7 +31,10 @@ def predict(): if __name__ == "__main__": parser = argparse.ArgumentParser(description="Flask API exposing YOLOv5 model") parser.add_argument("--port", default=5000, type=int, help="port number") - args = parser.parse_args() + opt = parser.parse_args() + + # Fix known issue urllib.error.HTTPError 403: rate limit exceeded https://github.com/ultralytics/yolov5/pull/7210 + torch.hub._validate_not_a_forked_repo = lambda a, b, c: True model = torch.hub.load("ultralytics/yolov5", "yolov5s", force_reload=True) # force_reload to recache - app.run(host="0.0.0.0", port=args.port) # debug=True causes Restarting with stat + app.run(host="0.0.0.0", port=opt.port) # debug=True causes Restarting with stat From ffcbd8ca97f037a83c5e0bc30a691e745b1c3cc9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 18:45:05 +0200 Subject: [PATCH 151/661] Export with official `nn.SiLU()` (#7256) * Update * Update time_limit --- export.py | 11 ++++------- utils/general.py | 2 +- 2 files changed, 5 insertions(+), 8 deletions(-) diff --git a/export.py b/export.py index cc7a74db9af2..e73715ea13e9 100644 --- a/export.py +++ b/export.py @@ -54,7 +54,6 @@ import pandas as pd import torch -import torch.nn as nn from torch.utils.mobile_optimizer import optimize_for_mobile FILE = Path(__file__).resolve() @@ -64,10 +63,8 @@ if platform.system() != 'Windows': ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative -from models.common import Conv from models.experimental import attempt_load from models.yolo import Detect -from utils.activations import SiLU from utils.datasets import LoadImages from utils.general import (LOGGER, check_dataset, check_img_size, check_requirements, check_version, colorstr, file_size, print_args, url2file) @@ -474,10 +471,10 @@ def run( im, model = im.half(), model.half() # to FP16 model.train() if train else model.eval() # training mode = no Detect() layer grid construction for k, m in model.named_modules(): - if isinstance(m, Conv): # assign export-friendly activations - if isinstance(m.act, nn.SiLU): - m.act = SiLU() - elif isinstance(m, Detect): + # if isinstance(m, Conv): # assign export-friendly activations + # if isinstance(m.act, nn.SiLU): + # m.act = SiLU() + if isinstance(m, Detect): m.inplace = inplace m.onnx_dynamic = dynamic if hasattr(m, 'forward_export'): diff --git a/utils/general.py b/utils/general.py index 379e9e0f47a4..da7dbb6d3e55 100755 --- a/utils/general.py +++ b/utils/general.py @@ -738,7 +738,7 @@ def non_max_suppression(prediction, # min_wh = 2 # (pixels) minimum box width and height max_wh = 7680 # (pixels) maximum box width and height max_nms = 30000 # maximum number of boxes into torchvision.ops.nms() - time_limit = 0.030 * bs # seconds to quit after + time_limit = 0.1 + 0.03 * bs # seconds to quit after redundant = True # require redundant detections multi_label &= nc > 1 # multiple labels per box (adds 0.5ms/img) merge = False # use merge-NMS From 4f839b7970555f100c4380fa7a6e0e089a93ac2a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 19:26:23 +0200 Subject: [PATCH 152/661] Refactor out-of-place `Detect()` for reduced ops (#7257) --- models/yolo.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index e18614cb37bd..f255a812b11a 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -66,9 +66,9 @@ def forward(self, x): y[..., 0:2] = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953 - xy = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy + xy = (y[..., 0:2] * 2 + (self.grid[i] - 0.5)) * self.stride[i] # xy wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh - y = torch.cat((xy, wh, y[..., 4:]), -1) + y = torch.cat((xy, wh, y[..., 4:]), 4) z.append(y.view(bs, -1, self.no)) return x if self.training else (torch.cat(z, 1), x) From ad0e4d5d199dc2da92d2058b57b0970fe2924bca Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 20:05:50 +0200 Subject: [PATCH 153/661] `torch.split()` replace slicing on out-of-place inference (#7258) --- models/yolo.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index f255a812b11a..3dd5fe9dcd25 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -66,9 +66,10 @@ def forward(self, x): y[..., 0:2] = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953 - xy = (y[..., 0:2] * 2 + (self.grid[i] - 0.5)) * self.stride[i] # xy - wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh - y = torch.cat((xy, wh, y[..., 4:]), 4) + xy, wh, conf = y.split((2, 2, self.nc + 1), 4) # y.tensor_split((2, 4, 5), 4) # torch 1.8.0 + xy = (xy * 2 + (self.grid[i] - 0.5)) * self.stride[i] # xy + wh = (wh * 2) ** 2 * self.anchor_grid[i] # wh + y = torch.cat((xy, wh, conf), 4) z.append(y.view(bs, -1, self.no)) return x if self.training else (torch.cat(z, 1), x) From 779efbb9ca26b9ed4177a59936ec1d0dfdc9365e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 21:21:55 +0200 Subject: [PATCH 154/661] Update --- utils/benchmarks.py | 1 + 1 file changed, 1 insertion(+) diff --git a/utils/benchmarks.py b/utils/benchmarks.py index 69243725b48a..36e827848584 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -58,6 +58,7 @@ def run( device = select_device(device) for i, (name, f, suffix, gpu) in formats.iterrows(): # index, (name, file, suffix, gpu-capable) try: + assert i < 9, 'Edge TPU and TF.js not supported' if device.type != 'cpu': assert gpu, f'{name} inference not supported on GPU' if f == '-': From 05cf0d1a44430230e75339ff7cfdd26bdf554502 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 21:29:20 +0200 Subject: [PATCH 155/661] Export single output only (#7259) * Update * Update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 1 + models/yolo.py | 3 ++- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/export.py b/export.py index e73715ea13e9..574bf8d9ed61 100644 --- a/export.py +++ b/export.py @@ -477,6 +477,7 @@ def run( if isinstance(m, Detect): m.inplace = inplace m.onnx_dynamic = dynamic + m.export = True if hasattr(m, 'forward_export'): m.forward = m.forward_export # assign custom forward (optional) diff --git a/models/yolo.py b/models/yolo.py index 3dd5fe9dcd25..fee5e932fd4d 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -37,6 +37,7 @@ class Detect(nn.Module): stride = None # strides computed during build onnx_dynamic = False # ONNX export parameter + export = False # export mode def __init__(self, nc=80, anchors=(), ch=(), inplace=True): # detection layer super().__init__() @@ -72,7 +73,7 @@ def forward(self, x): y = torch.cat((xy, wh, conf), 4) z.append(y.view(bs, -1, self.no)) - return x if self.training else (torch.cat(z, 1), x) + return x if self.training else (torch.cat(z, 1),) if self.export else (torch.cat(z, 1), x) def _make_grid(self, nx=20, ny=20, i=0): d = self.anchors[i].device From 8bc839ed8e423c7baeb778e60e4d6f67eb0d5f3d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 22:51:11 +0200 Subject: [PATCH 156/661] TorchScript single-output fix (#7261) --- export.py | 18 ++++++++++++------ models/common.py | 7 ++++--- 2 files changed, 16 insertions(+), 9 deletions(-) diff --git a/export.py b/export.py index 574bf8d9ed61..87be00376778 100644 --- a/export.py +++ b/export.py @@ -73,12 +73,18 @@ def export_formats(): # YOLOv5 export formats - x = [['PyTorch', '-', '.pt', True], ['TorchScript', 'torchscript', '.torchscript', True], - ['ONNX', 'onnx', '.onnx', True], ['OpenVINO', 'openvino', '_openvino_model', False], - ['TensorRT', 'engine', '.engine', True], ['CoreML', 'coreml', '.mlmodel', False], - ['TensorFlow SavedModel', 'saved_model', '_saved_model', True], ['TensorFlow GraphDef', 'pb', '.pb', True], - ['TensorFlow Lite', 'tflite', '.tflite', False], ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False], - ['TensorFlow.js', 'tfjs', '_web_model', False]] + x = [ + ['PyTorch', '-', '.pt', True], + ['TorchScript', 'torchscript', '.torchscript', True], + ['ONNX', 'onnx', '.onnx', True], + ['OpenVINO', 'openvino', '_openvino_model', False], + ['TensorRT', 'engine', '.engine', True], + ['CoreML', 'coreml', '.mlmodel', False], + ['TensorFlow SavedModel', 'saved_model', '_saved_model', True], + ['TensorFlow GraphDef', 'pb', '.pb', True], + ['TensorFlow Lite', 'tflite', '.tflite', False], + ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False], + ['TensorFlow.js', 'tfjs', '_web_model', False],] return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix', 'GPU']) diff --git a/models/common.py b/models/common.py index 8396caa1af5c..dcd3e5f408dd 100644 --- a/models/common.py +++ b/models/common.py @@ -406,9 +406,10 @@ def wrap_frozen_graph(gd, inputs, outputs): def forward(self, im, augment=False, visualize=False, val=False): # YOLOv5 MultiBackend inference b, ch, h, w = im.shape # batch, channel, height, width - if self.pt or self.jit: # PyTorch - y = self.model(im) if self.jit else self.model(im, augment=augment, visualize=visualize) - return y if val else y[0] + if self.pt: # PyTorch + y = self.model(im, augment=augment, visualize=visualize)[0] + elif self.jit: # TorchScript + y = self.model(im)[0] elif self.dnn: # ONNX OpenCV DNN im = im.cpu().numpy() # torch to numpy self.net.setInput(im) From ea72b84f5e690cb516642ce2d9ae200145b0af34 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 3 Apr 2022 23:40:23 +0200 Subject: [PATCH 157/661] Integrate offset into grid (#7262) Eliminate 1 op during training and inference. --- models/yolo.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index fee5e932fd4d..d6f5c0961e0d 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -64,11 +64,11 @@ def forward(self, x): y = x[i].sigmoid() if self.inplace: - y[..., 0:2] = (y[..., 0:2] * 2 - 0.5 + self.grid[i]) * self.stride[i] # xy + y[..., 0:2] = (y[..., 0:2] * 2 + self.grid[i]) * self.stride[i] # xy y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953 xy, wh, conf = y.split((2, 2, self.nc + 1), 4) # y.tensor_split((2, 4, 5), 4) # torch 1.8.0 - xy = (xy * 2 + (self.grid[i] - 0.5)) * self.stride[i] # xy + xy = (xy * 2 + self.grid[i]) * self.stride[i] # xy wh = (wh * 2) ** 2 * self.anchor_grid[i] # wh y = torch.cat((xy, wh, conf), 4) z.append(y.view(bs, -1, self.no)) @@ -82,7 +82,7 @@ def _make_grid(self, nx=20, ny=20, i=0): yv, xv = torch.meshgrid(torch.arange(ny, device=d), torch.arange(nx, device=d), indexing='ij') else: yv, xv = torch.meshgrid(torch.arange(ny, device=d), torch.arange(nx, device=d)) - grid = torch.stack((xv, yv), 2).expand(shape).float() + grid = torch.stack((xv, yv), 2).expand(shape).float() - 0.5 # add grid offset, i.e. y = 2.0 * x - 0.5 anchor_grid = (self.anchors[i] * self.stride[i]).view((1, self.na, 1, 1, 2)).expand(shape).float() return grid, anchor_grid From 7882950577116eff9085b96abd8036522f2de7ca Mon Sep 17 00:00:00 2001 From: "pre-commit-ci[bot]" <66853113+pre-commit-ci[bot]@users.noreply.github.com> Date: Mon, 4 Apr 2022 22:47:00 +0200 Subject: [PATCH 158/661] [pre-commit.ci] pre-commit suggestions (#7279) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * [pre-commit.ci] pre-commit suggestions updates: - [github.com/asottile/pyupgrade: v2.31.0 → v2.31.1](https://github.com/asottile/pyupgrade/compare/v2.31.0...v2.31.1) - [github.com/pre-commit/mirrors-yapf: v0.31.0 → v0.32.0](https://github.com/pre-commit/mirrors-yapf/compare/v0.31.0...v0.32.0) * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update yolo.py * Update activations.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update activations.py * Update tf.py * Update tf.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- .pre-commit-config.yaml | 4 ++-- models/tf.py | 5 +++++ models/yolo.py | 1 + utils/activations.py | 20 ++++++++++++-------- utils/callbacks.py | 1 + utils/datasets.py | 3 +++ utils/loggers/wandb/wandb_utils.py | 1 + utils/metrics.py | 1 + utils/torch_utils.py | 1 + 9 files changed, 27 insertions(+), 10 deletions(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 208cb072c8aa..ae61892b68b2 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -24,7 +24,7 @@ repos: - id: check-docstring-first - repo: https://github.com/asottile/pyupgrade - rev: v2.31.0 + rev: v2.31.1 hooks: - id: pyupgrade args: [--py36-plus] @@ -37,7 +37,7 @@ repos: name: Sort imports - repo: https://github.com/pre-commit/mirrors-yapf - rev: v0.31.0 + rev: v0.32.0 hooks: - id: yapf name: YAPF formatting diff --git a/models/tf.py b/models/tf.py index 1b7653bce8f6..a15569e3b465 100644 --- a/models/tf.py +++ b/models/tf.py @@ -50,6 +50,7 @@ def call(self, inputs): class TFPad(keras.layers.Layer): + def __init__(self, pad): super().__init__() self.pad = tf.constant([[0, 0], [pad, pad], [pad, pad], [0, 0]]) @@ -206,6 +207,7 @@ def call(self, inputs): class TFDetect(keras.layers.Layer): + # TF YOLOv5 Detect layer def __init__(self, nc=80, anchors=(), ch=(), imgsz=(640, 640), w=None): # detection layer super().__init__() self.stride = tf.convert_to_tensor(w.stride.numpy(), dtype=tf.float32) @@ -255,6 +257,7 @@ def _make_grid(nx=20, ny=20): class TFUpsample(keras.layers.Layer): + # TF version of torch.nn.Upsample() def __init__(self, size, scale_factor, mode, w=None): # warning: all arguments needed including 'w' super().__init__() assert scale_factor == 2, "scale_factor must be 2" @@ -269,6 +272,7 @@ def call(self, inputs): class TFConcat(keras.layers.Layer): + # TF version of torch.concat() def __init__(self, dimension=1, w=None): super().__init__() assert dimension == 1, "convert only NCHW to NHWC concat" @@ -331,6 +335,7 @@ def parse_model(d, ch, model, imgsz): # model_dict, input_channels(3) class TFModel: + # TF YOLOv5 model def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, model=None, imgsz=(640, 640)): # model, channels, classes super().__init__() if isinstance(cfg, dict): diff --git a/models/yolo.py b/models/yolo.py index d6f5c0961e0d..85c5a96997f2 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -88,6 +88,7 @@ def _make_grid(self, nx=20, ny=20, i=0): class Model(nn.Module): + # YOLOv5 model def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, anchors=None): # model, input channels, number of classes super().__init__() if isinstance(cfg, dict): diff --git a/utils/activations.py b/utils/activations.py index b104ac18b03b..084ce8c41230 100644 --- a/utils/activations.py +++ b/utils/activations.py @@ -8,29 +8,32 @@ import torch.nn.functional as F -# SiLU https://arxiv.org/pdf/1606.08415.pdf ---------------------------------------------------------------------------- -class SiLU(nn.Module): # export-friendly version of nn.SiLU() +class SiLU(nn.Module): + # SiLU activation https://arxiv.org/pdf/1606.08415.pdf @staticmethod def forward(x): return x * torch.sigmoid(x) -class Hardswish(nn.Module): # export-friendly version of nn.Hardswish() +class Hardswish(nn.Module): + # Hard-SiLU activation @staticmethod def forward(x): # return x * F.hardsigmoid(x) # for TorchScript and CoreML return x * F.hardtanh(x + 3, 0.0, 6.0) / 6.0 # for TorchScript, CoreML and ONNX -# Mish https://github.com/digantamisra98/Mish -------------------------------------------------------------------------- class Mish(nn.Module): + # Mish activation https://github.com/digantamisra98/Mish @staticmethod def forward(x): return x * F.softplus(x).tanh() class MemoryEfficientMish(nn.Module): + # Mish activation memory-efficient class F(torch.autograd.Function): + @staticmethod def forward(ctx, x): ctx.save_for_backward(x) @@ -47,8 +50,8 @@ def forward(self, x): return self.F.apply(x) -# FReLU https://arxiv.org/abs/2007.11824 ------------------------------------------------------------------------------- class FReLU(nn.Module): + # FReLU activation https://arxiv.org/abs/2007.11824 def __init__(self, c1, k=3): # ch_in, kernel super().__init__() self.conv = nn.Conv2d(c1, c1, k, 1, 1, groups=c1, bias=False) @@ -58,12 +61,12 @@ def forward(self, x): return torch.max(x, self.bn(self.conv(x))) -# ACON https://arxiv.org/pdf/2009.04759.pdf ---------------------------------------------------------------------------- class AconC(nn.Module): - r""" ACON activation (activate or not). + r""" ACON activation (activate or not) AconC: (p1*x-p2*x) * sigmoid(beta*(p1*x-p2*x)) + p2*x, beta is a learnable parameter according to "Activate or Not: Learning Customized Activation" . """ + def __init__(self, c1): super().__init__() self.p1 = nn.Parameter(torch.randn(1, c1, 1, 1)) @@ -76,10 +79,11 @@ def forward(self, x): class MetaAconC(nn.Module): - r""" ACON activation (activate or not). + r""" ACON activation (activate or not) MetaAconC: (p1*x-p2*x) * sigmoid(beta*(p1*x-p2*x)) + p2*x, beta is generated by a small network according to "Activate or Not: Learning Customized Activation" . """ + def __init__(self, c1, k=1, s=1, r=16): # ch_in, kernel, stride, r super().__init__() c2 = max(r, c1 // r) diff --git a/utils/callbacks.py b/utils/callbacks.py index 6323985b8098..c6b3be1cbd69 100644 --- a/utils/callbacks.py +++ b/utils/callbacks.py @@ -8,6 +8,7 @@ class Callbacks: """" Handles all registered callbacks for YOLOv5 Hooks """ + def __init__(self): # Define the available callbacks self._callbacks = { diff --git a/utils/datasets.py b/utils/datasets.py index b2d4fa54ae0d..c12d3d9b9649 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -145,6 +145,7 @@ class InfiniteDataLoader(dataloader.DataLoader): Uses same syntax as vanilla DataLoader """ + def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) object.__setattr__(self, 'batch_sampler', _RepeatSampler(self.batch_sampler)) @@ -164,6 +165,7 @@ class _RepeatSampler: Args: sampler (Sampler) """ + def __init__(self, sampler): self.sampler = sampler @@ -978,6 +980,7 @@ def dataset_stats(path='coco128.yaml', autodownload=False, verbose=False, profil autodownload: Attempt to download dataset if not found locally verbose: Print stats dictionary """ + def round_labels(labels): # Update labels to integer class and 6 decimal place floats return [[int(c), *(round(x, 4) for x in points)] for c, *points in labels] diff --git a/utils/loggers/wandb/wandb_utils.py b/utils/loggers/wandb/wandb_utils.py index 6ec2559e29ac..08b568d074a2 100644 --- a/utils/loggers/wandb/wandb_utils.py +++ b/utils/loggers/wandb/wandb_utils.py @@ -116,6 +116,7 @@ class WandbLogger(): For more on how this logger is used, see the Weights & Biases documentation: https://docs.wandb.com/guides/integrations/yolov5 """ + def __init__(self, opt, run_id=None, job_type='Training'): """ - Initialize WandbLogger instance diff --git a/utils/metrics.py b/utils/metrics.py index 216956e90ecc..0674beddc0fb 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -260,6 +260,7 @@ def box_iou(box1, box2): iou (Tensor[N, M]): the NxM matrix containing the pairwise IoU values for every element in boxes1 and boxes2 """ + def box_area(box): # box = 4xn return (box[2] - box[0]) * (box[3] - box[1]) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index bc96ec75be7c..72f8a0fd1659 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -284,6 +284,7 @@ class ModelEMA: Keeps a moving average of everything in the model state_dict (parameters and buffers) For EMA details see https://www.tensorflow.org/api_docs/python/tf/train/ExponentialMovingAverage """ + def __init__(self, model, decay=0.9999, tau=2000, updates=0): # Create EMA self.ema = deepcopy(de_parallel(model)).eval() # FP32 EMA From 2da68664b51b847ff73d007e1eba6364ec452764 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 5 Apr 2022 00:52:37 +0200 Subject: [PATCH 159/661] Update Dockerfile (#7282) --- Dockerfile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Dockerfile b/Dockerfile index 59aa99faa1d6..7df6c1854156 100644 --- a/Dockerfile +++ b/Dockerfile @@ -19,8 +19,8 @@ RUN mkdir -p /usr/src/app WORKDIR /usr/src/app # Copy contents -RUN git clone https://github.com/ultralytics/yolov5 /usr/src/app -# COPY . /usr/src/app +COPY . /usr/src/app +RUN git clone https://github.com/ultralytics/yolov5 /usr/src/yolov5 # Downloads to user config dir ADD https://ultralytics.com/assets/Arial.ttf /root/.config/Ultralytics/ From 8d0291f3af881c315d8a6c1d39d1af2b1ff45359 Mon Sep 17 00:00:00 2001 From: leeflix <41200990+leeflix@users.noreply.github.com> Date: Tue, 5 Apr 2022 11:33:08 +0200 Subject: [PATCH 160/661] Enable TensorFlow ops for `--nms` and `--agnostic-nms` (#7281) * enable TensorFlow ops if flag --nms or --agnostic-nms is used * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update export.py * Update export.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- export.py | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/export.py b/export.py index 87be00376778..c0b98ce40fd5 100644 --- a/export.py +++ b/export.py @@ -327,7 +327,7 @@ def export_pb(keras_model, im, file, prefix=colorstr('TensorFlow GraphDef:')): LOGGER.info(f'\n{prefix} export failure: {e}') -def export_tflite(keras_model, im, file, int8, data, ncalib, prefix=colorstr('TensorFlow Lite:')): +def export_tflite(keras_model, im, file, int8, data, nms, agnostic_nms, prefix=colorstr('TensorFlow Lite:')): # YOLOv5 TensorFlow Lite export try: import tensorflow as tf @@ -343,13 +343,15 @@ def export_tflite(keras_model, im, file, int8, data, ncalib, prefix=colorstr('Te if int8: from models.tf import representative_dataset_gen dataset = LoadImages(check_dataset(data)['train'], img_size=imgsz, auto=False) # representative data - converter.representative_dataset = lambda: representative_dataset_gen(dataset, ncalib) + converter.representative_dataset = lambda: representative_dataset_gen(dataset, ncalib=100) converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8] converter.target_spec.supported_types = [] converter.inference_input_type = tf.uint8 # or tf.int8 converter.inference_output_type = tf.uint8 # or tf.int8 converter.experimental_new_quantizer = True f = str(file).replace('.pt', '-int8.tflite') + if nms or agnostic_nms: + converter.target_spec.supported_ops.append(tf.lite.OpsSet.SELECT_TF_OPS) tflite_model = converter.convert() open(f, "wb").write(tflite_model) @@ -524,7 +526,7 @@ def run( if pb or tfjs: # pb prerequisite to tfjs f[6] = export_pb(model, im, file) if tflite or edgetpu: - f[7] = export_tflite(model, im, file, int8=int8 or edgetpu, data=data, ncalib=100) + f[7] = export_tflite(model, im, file, int8=int8 or edgetpu, data=data, nms=nms, agnostic_nms=agnostic_nms) if edgetpu: f[8] = export_edgetpu(model, im, file) if tfjs: From 2181ef371e5493eb3cddcfa50b59804cbabce73d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 5 Apr 2022 11:49:32 +0200 Subject: [PATCH 161/661] Update `cv2.imread()` patch with flags argument (#7287) --- utils/general.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/general.py b/utils/general.py index da7dbb6d3e55..65dd9326797e 100755 --- a/utils/general.py +++ b/utils/general.py @@ -925,8 +925,8 @@ def increment_path(path, exist_ok=False, sep='', mkdir=False): imshow_ = cv2.imshow # copy to avoid recursion errors -def imread(path): - return cv2.imdecode(np.fromfile(path, np.uint8), cv2.IMREAD_COLOR) +def imread(path, flags=cv2.IMREAD_COLOR): + return cv2.imdecode(np.fromfile(path, np.uint8), flags) def imwrite(path, im): From 5f97001ed4e5deb5c92eb200a79b5cb9da861130 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 5 Apr 2022 12:54:25 +0200 Subject: [PATCH 162/661] Context manager `open(file) as f` fixes (#7289) * Flask context manager `open()` fix * Additional read context manager fixes --- data/VOC.yaml | 3 ++- export.py | 3 ++- models/common.py | 3 ++- utils/flask_rest_api/example_request.py | 12 +++++++++--- utils/flask_rest_api/restapi.py | 2 ++ 5 files changed, 17 insertions(+), 6 deletions(-) diff --git a/data/VOC.yaml b/data/VOC.yaml index be04fb1e2ecb..9865967dd028 100644 --- a/data/VOC.yaml +++ b/data/VOC.yaml @@ -72,7 +72,8 @@ download: | imgs_path.mkdir(exist_ok=True, parents=True) lbs_path.mkdir(exist_ok=True, parents=True) - image_ids = open(path / f'VOC{year}/ImageSets/Main/{image_set}.txt').read().strip().split() + with open(path / f'VOC{year}/ImageSets/Main/{image_set}.txt') as f: + image_ids = f.read().strip().split() for id in tqdm(image_ids, desc=f'{image_set}{year}'): f = path / f'VOC{year}/JPEGImages/{id}.jpg' # old img path lb_path = (lbs_path / f.name).with_suffix('.txt') # new label path diff --git a/export.py b/export.py index c0b98ce40fd5..df4f3b6e05ef 100644 --- a/export.py +++ b/export.py @@ -407,7 +407,8 @@ def export_tfjs(keras_model, im, file, prefix=colorstr('TensorFlow.js:')): f'--output_node_names="Identity,Identity_1,Identity_2,Identity_3" {f_pb} {f}' subprocess.run(cmd, shell=True) - json = open(f_json).read() + with open(f_json) as j: + json = j.read() with open(f_json, 'w') as j: # sort JSON Identity_* in ascending order subst = re.sub( r'{"outputs": {"Identity.?.?": {"name": "Identity.?.?"}, ' diff --git a/models/common.py b/models/common.py index dcd3e5f408dd..5a83bce33fc8 100644 --- a/models/common.py +++ b/models/common.py @@ -378,7 +378,8 @@ def wrap_frozen_graph(gd, inputs, outputs): return x.prune(tf.nest.map_structure(ge, inputs), tf.nest.map_structure(ge, outputs)) gd = tf.Graph().as_graph_def() # graph_def - gd.ParseFromString(open(w, 'rb').read()) + with open(w, 'rb') as f: + gd.ParseFromString(f.read()) frozen_func = wrap_frozen_graph(gd, inputs="x:0", outputs="Identity:0") elif tflite or edgetpu: # https://www.tensorflow.org/lite/guide/python#install_tensorflow_lite_for_python try: # https://coral.ai/docs/edgetpu/tflite-python/#update-existing-tf-lite-code-for-the-edge-tpu diff --git a/utils/flask_rest_api/example_request.py b/utils/flask_rest_api/example_request.py index ff21f30f93ca..773ad8932967 100644 --- a/utils/flask_rest_api/example_request.py +++ b/utils/flask_rest_api/example_request.py @@ -1,12 +1,18 @@ -"""Perform test request""" +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Perform test request +""" + import pprint import requests DETECTION_URL = "http://localhost:5000/v1/object-detection/yolov5s" -TEST_IMAGE = "zidane.jpg" +IMAGE = "zidane.jpg" -image_data = open(TEST_IMAGE, "rb").read() +# Read image +with open(IMAGE, "rb") as f: + image_data = f.read() response = requests.post(DETECTION_URL, files={"image": image_data}).json() diff --git a/utils/flask_rest_api/restapi.py b/utils/flask_rest_api/restapi.py index 38868cc98d84..62adb4bbf716 100644 --- a/utils/flask_rest_api/restapi.py +++ b/utils/flask_rest_api/restapi.py @@ -1,6 +1,8 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Run a Flask REST API exposing a YOLOv5s model """ + import argparse import io From d2e7ba2a3af8f6f17fa5240422b964a1ecf717d5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 5 Apr 2022 14:23:15 +0200 Subject: [PATCH 163/661] val.py `--weights` and `--data` compatibility check (#7292) Improved error messages for understanding of user error with val.py. May help https://github.com/ultralytics/yolov5/issues/7291 --- val.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/val.py b/val.py index 019beedea61a..50a6d91edfff 100644 --- a/val.py +++ b/val.py @@ -162,6 +162,10 @@ def run( # Dataloader if not training: + if pt and not single_cls: # check --weights are trained on --data + ncm = model.model.yaml['nc'] + assert ncm == nc, f'{weights[0]} ({ncm} classes) trained on different --data than what you passed ({nc} ' \ + f'classes). Pass correct combination of --weights and --data that are trained together.' model.warmup(imgsz=(1 if pt else batch_size, 3, imgsz, imgsz)) # warmup pad = 0.0 if task in ('speed', 'benchmark') else 0.5 rect = False if task == 'benchmark' else pt # square inference for benchmarks From b1300f3e0b7f1f5971b1d3abc6b7a0c0bd92b389 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 5 Apr 2022 15:14:54 +0200 Subject: [PATCH 164/661] Add dataset sizes (zipped) (#7293) --- data/Argoverse.yaml | 2 +- data/GlobalWheat2020.yaml | 2 +- data/Objects365.yaml | 2 +- data/SKU-110K.yaml | 2 +- data/VOC.yaml | 2 +- data/VisDrone.yaml | 2 +- data/coco.yaml | 2 +- data/coco128.yaml | 2 +- data/xView.yaml | 2 +- 9 files changed, 9 insertions(+), 9 deletions(-) diff --git a/data/Argoverse.yaml b/data/Argoverse.yaml index 312791b33a2d..43426f5ebe15 100644 --- a/data/Argoverse.yaml +++ b/data/Argoverse.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── Argoverse ← downloads here +# └── Argoverse ← downloads here (31.3 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] diff --git a/data/GlobalWheat2020.yaml b/data/GlobalWheat2020.yaml index c1ba289f2833..4c43693f1d82 100644 --- a/data/GlobalWheat2020.yaml +++ b/data/GlobalWheat2020.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── GlobalWheat2020 ← downloads here +# └── GlobalWheat2020 ← downloads here (7.0 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] diff --git a/data/Objects365.yaml b/data/Objects365.yaml index bd6e5d6e1144..1e09fd718479 100644 --- a/data/Objects365.yaml +++ b/data/Objects365.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── Objects365 ← downloads here +# └── Objects365 ← downloads here (750 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] diff --git a/data/SKU-110K.yaml b/data/SKU-110K.yaml index 46459eab6bb7..2acf34d155bd 100644 --- a/data/SKU-110K.yaml +++ b/data/SKU-110K.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── SKU-110K ← downloads here +# └── SKU-110K ← downloads here (13.6 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] diff --git a/data/VOC.yaml b/data/VOC.yaml index 9865967dd028..4fec304133be 100644 --- a/data/VOC.yaml +++ b/data/VOC.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── VOC ← downloads here +# └── VOC ← downloads here (2.8 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] diff --git a/data/VisDrone.yaml b/data/VisDrone.yaml index 2a3b2f03e674..fe87588ee870 100644 --- a/data/VisDrone.yaml +++ b/data/VisDrone.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── VisDrone ← downloads here +# └── VisDrone ← downloads here (2.3 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] diff --git a/data/coco.yaml b/data/coco.yaml index 7494fc2f9cd1..0c0c4adab05d 100644 --- a/data/coco.yaml +++ b/data/coco.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── coco ← downloads here +# └── coco ← downloads here (20.1 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] diff --git a/data/coco128.yaml b/data/coco128.yaml index d07c704407a1..2517d2079257 100644 --- a/data/coco128.yaml +++ b/data/coco128.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── coco128 ← downloads here +# └── coco128 ← downloads here (7 MB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] diff --git a/data/xView.yaml b/data/xView.yaml index fd82828dcb8c..3b38f1ff4439 100644 --- a/data/xView.yaml +++ b/data/xView.yaml @@ -5,7 +5,7 @@ # parent # ├── yolov5 # └── datasets -# └── xView ← downloads here +# └── xView ← downloads here (20.7 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] From c759bbdf19f3c430e778a84a76849145ebf58d25 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 5 Apr 2022 15:55:16 +0200 Subject: [PATCH 165/661] Add `check_requirements(('pycocotools>=2.0',))` (#7295) Add `check_requirements(('pycocotools>=2.0',))` --- data/Objects365.yaml | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/data/Objects365.yaml b/data/Objects365.yaml index 1e09fd718479..82b42a120d40 100644 --- a/data/Objects365.yaml +++ b/data/Objects365.yaml @@ -60,11 +60,12 @@ names: ['Person', 'Sneakers', 'Chair', 'Other Shoes', 'Hat', 'Car', 'Lamp', 'Gla # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | - from pycocotools.coco import COCO from tqdm import tqdm - - from utils.general import Path, download, np, xyxy2xywhn - + + from utils.general import Path, check_requirements, download, np, xyxy2xywhn + + check_requirements(('pycocotools>=2.0',)) + from pycocotools.coco import COCO # Make Directories dir = Path(yaml['path']) # dataset root dir From 741fac815e366d74eed020efb8c68a23828ee3e9 Mon Sep 17 00:00:00 2001 From: Max Strobel Date: Tue, 5 Apr 2022 17:38:13 +0200 Subject: [PATCH 166/661] fix: disable usage of root logger (#7296) * fix: disable usage of root logger `logging.basicConfig` configures Python's root logger. This prohibits fine control of logging, overwrites logging configuration done outside the package, and is not best practice. Instead, the used logger is now configured directly, and the root logger is untouched. Example: If yolov5 is used as part of another project with some sophisticated logging, the internal `logging.basicConfig` call overwrites all the external configuration. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update general.py * Update general.py * Comment kaggle * Uncomment kaggle Co-authored-by: Maximilian Strobel Co-authored-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- data/Objects365.yaml | 4 ++-- utils/general.py | 12 +++++++++--- 2 files changed, 11 insertions(+), 5 deletions(-) diff --git a/data/Objects365.yaml b/data/Objects365.yaml index 82b42a120d40..114bee2b159c 100644 --- a/data/Objects365.yaml +++ b/data/Objects365.yaml @@ -61,9 +61,9 @@ names: ['Person', 'Sneakers', 'Chair', 'Other Shoes', 'Hat', 'Car', 'Lamp', 'Gla # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | from tqdm import tqdm - + from utils.general import Path, check_requirements, download, np, xyxy2xywhn - + check_requirements(('pycocotools>=2.0',)) from pycocotools.coco import COCO diff --git a/utils/general.py b/utils/general.py index 65dd9326797e..5316f504871a 100755 --- a/utils/general.py +++ b/utils/general.py @@ -82,11 +82,17 @@ def set_logging(name=None, verbose=VERBOSE): for h in logging.root.handlers: logging.root.removeHandler(h) # remove all handlers associated with the root logger object rank = int(os.getenv('RANK', -1)) # rank in world for Multi-GPU trainings - logging.basicConfig(format="%(message)s", level=logging.INFO if (verbose and rank in (-1, 0)) else logging.WARNING) - return logging.getLogger(name) + level = logging.INFO if (verbose and rank in (-1, 0)) else logging.WARNING + log = logging.getLogger(name) + log.setLevel(level) + handler = logging.StreamHandler() + handler.setFormatter(logging.Formatter("%(message)s")) + handler.setLevel(level) + log.addHandler(handler) -LOGGER = set_logging('yolov5') # define globally (used in train.py, val.py, detect.py, etc.) +set_logging() # run before defining LOGGER +LOGGER = logging.getLogger("yolov5") # define globally (used in train.py, val.py, detect.py, etc.) def user_config_dir(dir='Ultralytics', env_var='YOLOV5_CONFIG_DIR'): From d257c75c848ccab4d9195300a61195cf0dfef1bf Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 5 Apr 2022 21:21:57 +0200 Subject: [PATCH 167/661] Update export.py (#7301) * Update export.py Simplify code. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 5 ----- 1 file changed, 5 deletions(-) diff --git a/export.py b/export.py index df4f3b6e05ef..16ba2ffce3ec 100644 --- a/export.py +++ b/export.py @@ -480,15 +480,10 @@ def run( im, model = im.half(), model.half() # to FP16 model.train() if train else model.eval() # training mode = no Detect() layer grid construction for k, m in model.named_modules(): - # if isinstance(m, Conv): # assign export-friendly activations - # if isinstance(m.act, nn.SiLU): - # m.act = SiLU() if isinstance(m, Detect): m.inplace = inplace m.onnx_dynamic = dynamic m.export = True - if hasattr(m, 'forward_export'): - m.forward = m.forward_export # assign custom forward (optional) for _ in range(2): y = model(im) # dry runs From f735458987f7e80c32739bfe0440cbcad36aeae3 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 6 Apr 2022 12:20:24 +0200 Subject: [PATCH 168/661] Use `tqdm.auto` (#7311) --- data/Argoverse.yaml | 2 +- data/Objects365.yaml | 2 +- data/SKU-110K.yaml | 2 +- data/VOC.yaml | 2 +- data/VisDrone.yaml | 2 +- data/xView.yaml | 2 +- train.py | 2 +- utils/autoanchor.py | 2 +- utils/datasets.py | 2 +- utils/loggers/wandb/wandb_utils.py | 2 +- val.py | 2 +- 11 files changed, 11 insertions(+), 11 deletions(-) diff --git a/data/Argoverse.yaml b/data/Argoverse.yaml index 43426f5ebe15..9d114f55dce8 100644 --- a/data/Argoverse.yaml +++ b/data/Argoverse.yaml @@ -22,7 +22,7 @@ names: ['person', 'bicycle', 'car', 'motorcycle', 'bus', 'truck', 'traffic download: | import json - from tqdm import tqdm + from tqdm.auto import tqdm from utils.general import download, Path diff --git a/data/Objects365.yaml b/data/Objects365.yaml index 114bee2b159c..ab8207d200f5 100644 --- a/data/Objects365.yaml +++ b/data/Objects365.yaml @@ -60,7 +60,7 @@ names: ['Person', 'Sneakers', 'Chair', 'Other Shoes', 'Hat', 'Car', 'Lamp', 'Gla # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | - from tqdm import tqdm + from tqdm.auto import tqdm from utils.general import Path, check_requirements, download, np, xyxy2xywhn diff --git a/data/SKU-110K.yaml b/data/SKU-110K.yaml index 2acf34d155bd..2fd689b1bcac 100644 --- a/data/SKU-110K.yaml +++ b/data/SKU-110K.yaml @@ -21,7 +21,7 @@ names: ['object'] # class names # Download script/URL (optional) --------------------------------------------------------------------------------------- download: | import shutil - from tqdm import tqdm + from tqdm.auto import tqdm from utils.general import np, pd, Path, download, xyxy2xywh diff --git a/data/VOC.yaml b/data/VOC.yaml index 4fec304133be..fbe3b193bf2e 100644 --- a/data/VOC.yaml +++ b/data/VOC.yaml @@ -29,7 +29,7 @@ names: ['aeroplane', 'bicycle', 'bird', 'boat', 'bottle', 'bus', 'car', 'cat', ' download: | import xml.etree.ElementTree as ET - from tqdm import tqdm + from tqdm.auto import tqdm from utils.general import download, Path diff --git a/data/VisDrone.yaml b/data/VisDrone.yaml index fe87588ee870..ef7e6c4fed35 100644 --- a/data/VisDrone.yaml +++ b/data/VisDrone.yaml @@ -24,7 +24,7 @@ download: | def visdrone2yolo(dir): from PIL import Image - from tqdm import tqdm + from tqdm.auto import tqdm def convert_box(size, box): # Convert VisDrone box to YOLO xywh box diff --git a/data/xView.yaml b/data/xView.yaml index 3b38f1ff4439..aac2d026e424 100644 --- a/data/xView.yaml +++ b/data/xView.yaml @@ -34,7 +34,7 @@ download: | import numpy as np from PIL import Image - from tqdm import tqdm + from tqdm.auto import tqdm from utils.datasets import autosplit from utils.general import download, xyxy2xywhn diff --git a/train.py b/train.py index 38c25c053e26..dfce5a195660 100644 --- a/train.py +++ b/train.py @@ -30,7 +30,7 @@ from torch.cuda import amp from torch.nn.parallel import DistributedDataParallel as DDP from torch.optim import SGD, Adam, AdamW, lr_scheduler -from tqdm import tqdm +from tqdm.auto import tqdm FILE = Path(__file__).resolve() ROOT = FILE.parents[0] # YOLOv5 root directory diff --git a/utils/autoanchor.py b/utils/autoanchor.py index 77518abe9889..cdcecd855a51 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -8,7 +8,7 @@ import numpy as np import torch import yaml -from tqdm import tqdm +from tqdm.auto import tqdm from utils.general import LOGGER, colorstr, emojis diff --git a/utils/datasets.py b/utils/datasets.py index c12d3d9b9649..578e5b829dc0 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -24,7 +24,7 @@ import yaml from PIL import ExifTags, Image, ImageOps from torch.utils.data import DataLoader, Dataset, dataloader, distributed -from tqdm import tqdm +from tqdm.auto import tqdm from utils.augmentations import Albumentations, augment_hsv, copy_paste, letterbox, mixup, random_perspective from utils.general import (DATASETS_DIR, LOGGER, NUM_THREADS, check_dataset, check_requirements, check_yaml, clean_str, diff --git a/utils/loggers/wandb/wandb_utils.py b/utils/loggers/wandb/wandb_utils.py index 08b568d074a2..e65d028f28db 100644 --- a/utils/loggers/wandb/wandb_utils.py +++ b/utils/loggers/wandb/wandb_utils.py @@ -8,7 +8,7 @@ from typing import Dict import yaml -from tqdm import tqdm +from tqdm.auto import tqdm FILE = Path(__file__).resolve() ROOT = FILE.parents[3] # YOLOv5 root directory diff --git a/val.py b/val.py index 50a6d91edfff..58a12ceae254 100644 --- a/val.py +++ b/val.py @@ -27,7 +27,7 @@ import numpy as np import torch -from tqdm import tqdm +from tqdm.auto import tqdm FILE = Path(__file__).resolve() ROOT = FILE.parents[0] # YOLOv5 root directory From 32661f75ac6eaa8c5dfd0ad36abfaa8d4e4fe700 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 6 Apr 2022 13:12:41 +0200 Subject: [PATCH 169/661] Add `retry=3` to `download()` (#7313) * Add `retry=3` to `download()` * Update general.py * Update general.py * Update general.py * Update VOC.yaml * Update VisDrone.yaml --- data/VOC.yaml | 2 +- data/VisDrone.yaml | 2 +- utils/general.py | 24 ++++++++++++++++++------ 3 files changed, 20 insertions(+), 8 deletions(-) diff --git a/data/VOC.yaml b/data/VOC.yaml index fbe3b193bf2e..93a1f181ce8c 100644 --- a/data/VOC.yaml +++ b/data/VOC.yaml @@ -62,7 +62,7 @@ download: | urls = [url + 'VOCtrainval_06-Nov-2007.zip', # 446MB, 5012 images url + 'VOCtest_06-Nov-2007.zip', # 438MB, 4953 images url + 'VOCtrainval_11-May-2012.zip'] # 1.95GB, 17126 images - download(urls, dir=dir / 'images', delete=False, threads=3) + download(urls, dir=dir / 'images', delete=False, curl=True, threads=3) # Convert path = dir / f'images/VOCdevkit' diff --git a/data/VisDrone.yaml b/data/VisDrone.yaml index ef7e6c4fed35..c38fb2ab769e 100644 --- a/data/VisDrone.yaml +++ b/data/VisDrone.yaml @@ -54,7 +54,7 @@ download: | 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-val.zip', 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-test-dev.zip', 'https://github.com/ultralytics/yolov5/releases/download/v1.0/VisDrone2019-DET-test-challenge.zip'] - download(urls, dir=dir, threads=4) + download(urls, dir=dir, curl=True, threads=4) # Convert for d in 'VisDrone2019-DET-train', 'VisDrone2019-DET-val', 'VisDrone2019-DET-test-dev': diff --git a/utils/general.py b/utils/general.py index 5316f504871a..6c2558db74c4 100755 --- a/utils/general.py +++ b/utils/general.py @@ -497,20 +497,32 @@ def url2file(url): return file -def download(url, dir='.', unzip=True, delete=True, curl=False, threads=1): +def download(url, dir='.', unzip=True, delete=True, curl=False, threads=1, retry=3): # Multi-threaded file download and unzip function, used in data.yaml for autodownload def download_one(url, dir): # Download 1 file + success = True f = dir / Path(url).name # filename if Path(url).is_file(): # exists in current path Path(url).rename(f) # move to dir elif not f.exists(): LOGGER.info(f'Downloading {url} to {f}...') - if curl: - os.system(f"curl -L '{url}' -o '{f}' --retry 9 -C -") # curl download, retry and resume on fail - else: - torch.hub.download_url_to_file(url, f, progress=threads == 1) # torch download - if unzip and f.suffix in ('.zip', '.gz'): + for i in range(retry + 1): + if curl: + s = 'sS' if threads > 1 else '' # silent + r = os.system(f"curl -{s}L '{url}' -o '{f}' --retry 9 -C -") # curl download + success = r == 0 + else: + torch.hub.download_url_to_file(url, f, progress=threads == 1) # torch download + success = f.is_file() + if success: + break + elif i < retry: + LOGGER.warning(f'Download failure, retrying {i + 1}/{retry} {url}...') + else: + LOGGER.warning(f'Failed to download {url}...') + + if unzip and success and f.suffix in ('.zip', '.gz'): LOGGER.info(f'Unzipping {f}...') if f.suffix == '.zip': ZipFile(f).extractall(path=dir) # unzip From 245d6459a93bb707d9624027bf9ebf40bd925ca8 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 6 Apr 2022 17:23:34 +0200 Subject: [PATCH 170/661] Add callbacks (#7315) * Add `on_train_start()` callback * Update * Update --- train.py | 4 ++++ utils/loggers/__init__.py | 4 ++++ val.py | 4 ++++ 3 files changed, 12 insertions(+) diff --git a/train.py b/train.py index dfce5a195660..b7f70ab5bea4 100644 --- a/train.py +++ b/train.py @@ -66,6 +66,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio save_dir, epochs, batch_size, weights, single_cls, evolve, data, cfg, resume, noval, nosave, workers, freeze = \ Path(opt.save_dir), opt.epochs, opt.batch_size, opt.weights, opt.single_cls, opt.evolve, opt.data, opt.cfg, \ opt.resume, opt.noval, opt.nosave, opt.workers, opt.freeze + callbacks.run('on_pretrain_routine_start') # Directories w = save_dir / 'weights' # weights dir @@ -291,11 +292,13 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio scaler = amp.GradScaler(enabled=cuda) stopper = EarlyStopping(patience=opt.patience) compute_loss = ComputeLoss(model) # init loss class + callbacks.run('on_train_start') LOGGER.info(f'Image sizes {imgsz} train, {imgsz} val\n' f'Using {train_loader.num_workers * WORLD_SIZE} dataloader workers\n' f"Logging results to {colorstr('bold', save_dir)}\n" f'Starting training for {epochs} epochs...') for epoch in range(start_epoch, epochs): # epoch ------------------------------------------------------------------ + callbacks.run('on_train_epoch_start') model.train() # Update image weights (optional, single-GPU only) @@ -317,6 +320,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio pbar = tqdm(pbar, total=nb, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar optimizer.zero_grad() for i, (imgs, targets, paths, _) in pbar: # batch ------------------------------------------------------------- + callbacks.run('on_train_batch_start') ni = i + nb * epoch # number integrated batches (since train start) imgs = imgs.to(device, non_blocking=True).float() / 255 # uint8 to float32, 0-255 to 0.0-1.0 diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 2e639dfb9b53..bab133cc35a9 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -84,6 +84,10 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, else: self.wandb = None + def on_train_start(self): + # Callback runs on train start + pass + def on_pretrain_routine_end(self): # Callback runs on pre-train routine end paths = self.save_dir.glob('*labels*.jpg') # training labels diff --git a/val.py b/val.py index 58a12ceae254..48f396626b54 100644 --- a/val.py +++ b/val.py @@ -188,8 +188,10 @@ def run( dt, p, r, f1, mp, mr, map50, map = [0.0, 0.0, 0.0], 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 loss = torch.zeros(3, device=device) jdict, stats, ap, ap_class = [], [], [], [] + callbacks.run('on_val_start') pbar = tqdm(dataloader, desc=s, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar for batch_i, (im, targets, paths, shapes) in enumerate(pbar): + callbacks.run('on_val_batch_start') t1 = time_sync() if cuda: im = im.to(device, non_blocking=True) @@ -260,6 +262,8 @@ def run( f = save_dir / f'val_batch{batch_i}_pred.jpg' # predictions Thread(target=plot_images, args=(im, output_to_target(out), paths, f, names), daemon=True).start() + callbacks.run('on_val_batch_end') + # Compute metrics stats = [np.concatenate(x, 0) for x in zip(*stats)] # to numpy if len(stats) and stats[0].any(): From a88a81469a54838abfbba0885e1c47c9e87ce3e2 Mon Sep 17 00:00:00 2001 From: Nick Martin Date: Wed, 6 Apr 2022 09:35:33 -0700 Subject: [PATCH 171/661] Copy wandb param dict before training to avoid overwrites (#7317) * Copy wandb param dict before training to avoid overwrites. Copy the hyperparameter dict retrieved from wandb configuration before passing it to `train()`. Training overwrites parameters in the dictionary (eg scaling obj/box/cls gains), which causes the values reported in wandb to not match the input values. This is confusing as it makes it hard to reproduce a run, and also throws off wandb's Bayesian sweep algorithm. * Cleanup Co-authored-by: Glenn Jocher --- utils/loggers/wandb/sweep.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/loggers/wandb/sweep.py b/utils/loggers/wandb/sweep.py index 206059bc30bf..d49ea6f2778b 100644 --- a/utils/loggers/wandb/sweep.py +++ b/utils/loggers/wandb/sweep.py @@ -16,8 +16,8 @@ def sweep(): wandb.init() - # Get hyp dict from sweep agent - hyp_dict = vars(wandb.config).get("_items") + # Get hyp dict from sweep agent. Copy because train() modifies parameters which confused wandb. + hyp_dict = vars(wandb.config).get("_items").copy() # Workaround: get necessary opt args opt = parse_opt(known=True) From 0ca85ed65f124871fa7686dcf0efbd8dc9699856 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 6 Apr 2022 23:52:19 +0200 Subject: [PATCH 172/661] Update Objects365.yaml (#7323) Updated dataset size to 712GB (includes undeleted zips). --- data/Objects365.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/data/Objects365.yaml b/data/Objects365.yaml index ab8207d200f5..8e6326b38595 100644 --- a/data/Objects365.yaml +++ b/data/Objects365.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── Objects365 ← downloads here (750 GB) +# └── Objects365 ← downloads here (712 GB) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] From b7faeda0f225f909ce87ffe504e829062ac44ca4 Mon Sep 17 00:00:00 2001 From: Nrupatunga Date: Thu, 7 Apr 2022 17:22:44 +0530 Subject: [PATCH 173/661] Fix Tf export for BottleneckCSP (#7330) --- models/tf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/tf.py b/models/tf.py index a15569e3b465..04b1cd378f18 100644 --- a/models/tf.py +++ b/models/tf.py @@ -152,7 +152,7 @@ def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5, w=None): self.cv3 = TFConv2d(c_, c_, 1, 1, bias=False, w=w.cv3) self.cv4 = TFConv(2 * c_, c2, 1, 1, w=w.cv4) self.bn = TFBN(w.bn) - self.act = lambda x: keras.activations.relu(x, alpha=0.1) + self.act = lambda x: keras.activations.swish(x) self.m = keras.Sequential([TFBottleneck(c_, c_, shortcut, g, e=1.0, w=w.m[j]) for j in range(n)]) def call(self, inputs): From 5783de26fe14d8a890090329d6ce17c468f47dfa Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 7 Apr 2022 16:12:44 +0200 Subject: [PATCH 174/661] Objects365 dataset breakdown images vs zips (#7335) --- data/Objects365.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/data/Objects365.yaml b/data/Objects365.yaml index 8e6326b38595..334c23c359cf 100644 --- a/data/Objects365.yaml +++ b/data/Objects365.yaml @@ -4,7 +4,7 @@ # parent # ├── yolov5 # └── datasets -# └── Objects365 ← downloads here (712 GB) +# └── Objects365 ← downloads here (712 GB = 367G data + 345G zips) # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] From 676e10cf1abc03360b56d8030adea2cd0d0af353 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 7 Apr 2022 16:15:01 +0200 Subject: [PATCH 175/661] Simplify callbacks.py return (#7333) * Simplify callbacks.py return * Indent args (pytorch convention) * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/callbacks.py | 19 ++++++++----------- 1 file changed, 8 insertions(+), 11 deletions(-) diff --git a/utils/callbacks.py b/utils/callbacks.py index c6b3be1cbd69..2b32df0bf1c1 100644 --- a/utils/callbacks.py +++ b/utils/callbacks.py @@ -38,9 +38,9 @@ def register_action(self, hook, name='', callback=None): Register a new action to a callback hook Args: - hook The callback hook name to register the action to - name The name of the action for later reference - callback The callback to fire + hook: The callback hook name to register the action to + name: The name of the action for later reference + callback: The callback to fire """ assert hook in self._callbacks, f"hook '{hook}' not found in callbacks {self._callbacks}" assert callable(callback), f"callback '{callback}' is not callable" @@ -51,21 +51,18 @@ def get_registered_actions(self, hook=None): Returns all the registered actions by callback hook Args: - hook The name of the hook to check, defaults to all + hook: The name of the hook to check, defaults to all """ - if hook: - return self._callbacks[hook] - else: - return self._callbacks + return self._callbacks[hook] if hook else self._callbacks def run(self, hook, *args, **kwargs): """ Loop through the registered actions and fire all callbacks Args: - hook The name of the hook to check, defaults to all - args Arguments to receive from YOLOv5 - kwargs Keyword Arguments to receive from YOLOv5 + hook: The name of the hook to check, defaults to all + args: Arguments to receive from YOLOv5 + kwargs: Keyword Arguments to receive from YOLOv5 """ assert hook in self._callbacks, f"hook '{hook}' not found in callbacks {self._callbacks}" From 5f941a84efdd45c986cd1c3764ced99e7c8e8294 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 7 Apr 2022 16:44:08 +0200 Subject: [PATCH 176/661] Print dataset scan only `if RANK in (-1, 0)` (#7337) * Print dataset scan only `if RANK in (-1, 0)` * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- train.py | 10 +++++----- utils/datasets.py | 3 ++- 2 files changed, 7 insertions(+), 6 deletions(-) diff --git a/train.py b/train.py index b7f70ab5bea4..d6764116b27c 100644 --- a/train.py +++ b/train.py @@ -316,7 +316,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio train_loader.sampler.set_epoch(epoch) pbar = enumerate(train_loader) LOGGER.info(('\n' + '%10s' * 7) % ('Epoch', 'gpu_mem', 'box', 'obj', 'cls', 'labels', 'img_size')) - if RANK in [-1, 0]: + if RANK in (-1, 0): pbar = tqdm(pbar, total=nb, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar optimizer.zero_grad() for i, (imgs, targets, paths, _) in pbar: # batch ------------------------------------------------------------- @@ -365,7 +365,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio last_opt_step = ni # Log - if RANK in [-1, 0]: + if RANK in (-1, 0): mloss = (mloss * i + loss_items) / (i + 1) # update mean losses mem = f'{torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0:.3g}G' # (GB) pbar.set_description(('%10s' * 2 + '%10.4g' * 5) % @@ -379,7 +379,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio lr = [x['lr'] for x in optimizer.param_groups] # for loggers scheduler.step() - if RANK in [-1, 0]: + if RANK in (-1, 0): # mAP callbacks.run('on_train_epoch_end', epoch=epoch) ema.update_attr(model, include=['yaml', 'nc', 'hyp', 'names', 'stride', 'class_weights']) @@ -440,7 +440,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # end epoch ---------------------------------------------------------------------------------------------------- # end training ----------------------------------------------------------------------------------------------------- - if RANK in [-1, 0]: + if RANK in (-1, 0): LOGGER.info(f'\n{epoch - start_epoch + 1} epochs completed in {(time.time() - t0) / 3600:.3f} hours.') for f in last, best: if f.exists(): @@ -518,7 +518,7 @@ def parse_opt(known=False): def main(opt, callbacks=Callbacks()): # Checks - if RANK in [-1, 0]: + if RANK in (-1, 0): print_args(vars(opt)) check_git_status() check_requirements(exclude=['thop']) diff --git a/utils/datasets.py b/utils/datasets.py index 578e5b829dc0..3fa9aa4c6ca1 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -36,6 +36,7 @@ IMG_FORMATS = 'bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp' # include image suffixes VID_FORMATS = 'asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'ts', 'wmv' # include video suffixes BAR_FORMAT = '{l_bar}{bar:10}{r_bar}{bar:-10b}' # tqdm bar format +LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html # Get orientation exif tag for orientation in ExifTags.TAGS.keys(): @@ -454,7 +455,7 @@ def __init__(self, # Display cache nf, nm, ne, nc, n = cache.pop('results') # found, missing, empty, corrupt, total - if exists: + if exists and LOCAL_RANK in (-1, 0): d = f"Scanning '{cache_path}' images and labels... {nf} found, {nm} missing, {ne} empty, {nc} corrupt" tqdm(None, desc=prefix + d, total=n, initial=n, bar_format=BAR_FORMAT) # display cache results if cache['msgs']: From 302b00b5f4b93bb6cdb3c651dc9f06b66d06016d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 8 Apr 2022 12:55:16 +0200 Subject: [PATCH 177/661] Update `_make_grid()` (#7346) --- models/yolo.py | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index 85c5a96997f2..f072aeeb8eac 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -77,13 +77,15 @@ def forward(self, x): def _make_grid(self, nx=20, ny=20, i=0): d = self.anchors[i].device + t = self.anchors[i].dtype shape = 1, self.na, ny, nx, 2 # grid shape + y, x = torch.arange(ny, device=d, dtype=t), torch.arange(nx, device=d, dtype=t) if check_version(torch.__version__, '1.10.0'): # torch>=1.10.0 meshgrid workaround for torch>=0.7 compatibility - yv, xv = torch.meshgrid(torch.arange(ny, device=d), torch.arange(nx, device=d), indexing='ij') + yv, xv = torch.meshgrid(y, x, indexing='ij') else: - yv, xv = torch.meshgrid(torch.arange(ny, device=d), torch.arange(nx, device=d)) - grid = torch.stack((xv, yv), 2).expand(shape).float() - 0.5 # add grid offset, i.e. y = 2.0 * x - 0.5 - anchor_grid = (self.anchors[i] * self.stride[i]).view((1, self.na, 1, 1, 2)).expand(shape).float() + yv, xv = torch.meshgrid(y, x) + grid = torch.stack((xv, yv), 2).expand(shape) - 0.5 # add grid offset, i.e. y = 2.0 * x - 0.5 + anchor_grid = (self.anchors[i] * self.stride[i]).view((1, self.na, 1, 1, 2)).expand(shape) return grid, anchor_grid From 446e6f563af1e92358603dda07c7462134c02b14 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 8 Apr 2022 23:05:15 +0200 Subject: [PATCH 178/661] Rename 'MacOS' to 'macOS' (#7349) --- .github/workflows/greetings.yml | 2 +- detect.py | 2 +- export.py | 2 +- tutorial.ipynb | 2 +- utils/loggers/wandb/README.md | 2 +- val.py | 2 +- 6 files changed, 6 insertions(+), 6 deletions(-) diff --git a/.github/workflows/greetings.yml b/.github/workflows/greetings.yml index 58fbcbfa90af..0b749f438dd2 100644 --- a/.github/workflows/greetings.yml +++ b/.github/workflows/greetings.yml @@ -56,4 +56,4 @@ jobs: CI CPU testing - If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), validation ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on MacOS, Windows, and Ubuntu every 24 hours and on every commit. + If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), validation ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on macOS, Windows, and Ubuntu every 24 hours and on every commit. diff --git a/detect.py b/detect.py index 14ff9a6ab421..bc93631caa4e 100644 --- a/detect.py +++ b/detect.py @@ -17,7 +17,7 @@ yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn yolov5s.xml # OpenVINO yolov5s.engine # TensorRT - yolov5s.mlmodel # CoreML (MacOS-only) + yolov5s.mlmodel # CoreML (macOS-only) yolov5s_saved_model # TensorFlow SavedModel yolov5s.pb # TensorFlow GraphDef yolov5s.tflite # TensorFlow Lite diff --git a/export.py b/export.py index 16ba2ffce3ec..ceb7862a49be 100644 --- a/export.py +++ b/export.py @@ -29,7 +29,7 @@ yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn yolov5s.xml # OpenVINO yolov5s.engine # TensorRT - yolov5s.mlmodel # CoreML (MacOS-only) + yolov5s.mlmodel # CoreML (macOS-only) yolov5s_saved_model # TensorFlow SavedModel yolov5s.pb # TensorFlow GraphDef yolov5s.tflite # TensorFlow Lite diff --git a/tutorial.ipynb b/tutorial.ipynb index d5a10dfd5952..dd6f520334b0 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -962,7 +962,7 @@ "\n", "![CI CPU testing](https://github.com/ultralytics/yolov5/workflows/CI%20CPU%20testing/badge.svg)\n", "\n", - "If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), testing ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on MacOS, Windows, and Ubuntu every 24 hours and on every commit.\n" + "If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), testing ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on macOS, Windows, and Ubuntu every 24 hours and on every commit.\n" ] }, { diff --git a/utils/loggers/wandb/README.md b/utils/loggers/wandb/README.md index 63d999859e6d..3e9c9fd38433 100644 --- a/utils/loggers/wandb/README.md +++ b/utils/loggers/wandb/README.md @@ -149,4 +149,4 @@ YOLOv5 may be run in any of the following up-to-date verified environments (with ![CI CPU testing](https://github.com/ultralytics/yolov5/workflows/CI%20CPU%20testing/badge.svg) -If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), validation ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on MacOS, Windows, and Ubuntu every 24 hours and on every commit. +If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), validation ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on macOS, Windows, and Ubuntu every 24 hours and on every commit. diff --git a/val.py b/val.py index 48f396626b54..5841437051c2 100644 --- a/val.py +++ b/val.py @@ -11,7 +11,7 @@ yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn yolov5s.xml # OpenVINO yolov5s.engine # TensorRT - yolov5s.mlmodel # CoreML (MacOS-only) + yolov5s.mlmodel # CoreML (macOS-only) yolov5s_saved_model # TensorFlow SavedModel yolov5s.pb # TensorFlow GraphDef yolov5s.tflite # TensorFlow Lite From 698a5d7f26002e7b0b0d535d981c2b92f25bc76e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 9 Apr 2022 01:32:16 +0200 Subject: [PATCH 179/661] Add `python benchmarks.py --test` for export-only (#7350) * Test exports * Fix precommit --- utils/benchmarks.py | 44 +++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 41 insertions(+), 3 deletions(-) diff --git a/utils/benchmarks.py b/utils/benchmarks.py index 36e827848584..1c1bb7a8f2ed 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -52,20 +52,26 @@ def run( data=ROOT / 'data/coco128.yaml', # dataset.yaml path device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu half=False, # use FP16 half-precision inference + test=False, # test exports only ): y, t = [], time.time() formats = export.export_formats() device = select_device(device) for i, (name, f, suffix, gpu) in formats.iterrows(): # index, (name, file, suffix, gpu-capable) try: - assert i < 9, 'Edge TPU and TF.js not supported' + assert i != 9, 'Edge TPU not supported' + assert i != 10, 'TF.js not supported' if device.type != 'cpu': assert gpu, f'{name} inference not supported on GPU' + + # Export if f == '-': w = weights # PyTorch format else: w = export.run(weights=weights, imgsz=[imgsz], include=[f], device=device, half=half)[-1] # all others assert suffix in str(w), 'export failed' + + # Validate result = val.run(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) metrics = result[0] # metrics (mp, mr, map50, map, *losses(box, obj, cls)) speeds = result[2] # times (preprocess, inference, postprocess) @@ -78,8 +84,39 @@ def run( LOGGER.info('\n') parse_opt() notebook_init() # print system info - py = pd.DataFrame(y, columns=['Format', 'mAP@0.5:0.95', 'Inference time (ms)']) + py = pd.DataFrame(y, columns=['Format', 'mAP@0.5:0.95', 'Inference time (ms)'] if map else ['Format', 'Export', '']) LOGGER.info(f'\nBenchmarks complete ({time.time() - t:.2f}s)') + LOGGER.info(str(py if map else py.iloc[:, :2])) + return py + + +def test( + weights=ROOT / 'yolov5s.pt', # weights path + imgsz=640, # inference size (pixels) + batch_size=1, # batch size + data=ROOT / 'data/coco128.yaml', # dataset.yaml path + device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu + half=False, # use FP16 half-precision inference + test=False, # test exports only +): + y, t = [], time.time() + formats = export.export_formats() + device = select_device(device) + for i, (name, f, suffix, gpu) in formats.iterrows(): # index, (name, file, suffix, gpu-capable) + try: + w = weights if f == '-' else \ + export.run(weights=weights, imgsz=[imgsz], include=[f], device=device, half=half)[-1] # weights + assert suffix in str(w), 'export failed' + y.append([name, True]) + except Exception: + y.append([name, False]) # mAP, t_inference + + # Print results + LOGGER.info('\n') + parse_opt() + notebook_init() # print system info + py = pd.DataFrame(y, columns=['Format', 'Export']) + LOGGER.info(f'\nExports complete ({time.time() - t:.2f}s)') LOGGER.info(str(py)) return py @@ -92,13 +129,14 @@ def parse_opt(): parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') + parser.add_argument('--test', action='store_true', help='test exports only') opt = parser.parse_args() print_args(vars(opt)) return opt def main(opt): - run(**vars(opt)) + test(**vars(opt)) if opt.test else run(**vars(opt)) if __name__ == "__main__": From 3bb233a7fb5b23e8128855eba1aaf347b1e86f49 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 9 Apr 2022 13:27:49 +0200 Subject: [PATCH 180/661] Add ONNX export metadata (#7353) --- export.py | 8 +++++++- models/common.py | 3 +++ 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/export.py b/export.py index ceb7862a49be..ecead3ef5a90 100644 --- a/export.py +++ b/export.py @@ -140,7 +140,13 @@ def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorst # Checks model_onnx = onnx.load(f) # load onnx model onnx.checker.check_model(model_onnx) # check onnx model - # LOGGER.info(onnx.helper.printable_graph(model_onnx.graph)) # print + + # Metadata + d = {'stride': int(max(model.stride)), 'names': model.names} + for k, v in d.items(): + meta = model_onnx.metadata_props.add() + meta.key, meta.value = k, str(v) + onnx.save(model_onnx, f) # Simplify if simplify: diff --git a/models/common.py b/models/common.py index 5a83bce33fc8..49175f76a53a 100644 --- a/models/common.py +++ b/models/common.py @@ -328,6 +328,9 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, import onnxruntime providers = ['CUDAExecutionProvider', 'CPUExecutionProvider'] if cuda else ['CPUExecutionProvider'] session = onnxruntime.InferenceSession(w, providers=providers) + meta = session.get_modelmeta().custom_metadata_map # metadata + if 'stride' in meta: + stride, names = int(meta['stride']), eval(meta['names']) elif xml: # OpenVINO LOGGER.info(f'Loading {w} for OpenVINO inference...') check_requirements(('openvino-dev',)) # requires openvino-dev: https://pypi.org/project/openvino-dev/ From aa542ce6a65658ff931fee9bbab77c0145c152f0 Mon Sep 17 00:00:00 2001 From: rglkt <50093021+rglkt@users.noreply.github.com> Date: Sun, 10 Apr 2022 01:11:55 +0800 Subject: [PATCH 181/661] DetectMultiBackend() default `stride=32` (#7342) * set common default stride as 32 * restore default stride, and set it on argument optional * fix wrong use of opt * fix missing parameter of stride * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix format of parameters * Update val.py * Update common.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- models/common.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/common.py b/models/common.py index 49175f76a53a..6ab82ab51ff4 100644 --- a/models/common.py +++ b/models/common.py @@ -296,7 +296,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, super().__init__() w = str(weights[0] if isinstance(weights, list) else weights) pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs = self.model_type(w) # get backend - stride, names = 64, [f'class{i}' for i in range(1000)] # assign defaults + stride, names = 32, [f'class{i}' for i in range(1000)] # assign defaults w = attempt_download(w) # download if not local fp16 &= (pt or jit or onnx or engine) and device.type != 'cpu' # FP16 if data: # data.yaml path (optional) From 406ee528f0fb78e6f814b9a53765bc54183f0a0b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 10 Apr 2022 13:46:07 +0200 Subject: [PATCH 182/661] Loss and IoU speed improvements (#7361) * Loss speed improvements * bbox_iou speed improvements * bbox_ioa speed improvements * box_iou speed improvements * box_iou speed improvements --- utils/loss.py | 8 +++---- utils/metrics.py | 54 +++++++++++++++++++++++------------------------- val.py | 4 ++-- 3 files changed, 32 insertions(+), 34 deletions(-) diff --git a/utils/loss.py b/utils/loss.py index fa8095515477..b5d050e46047 100644 --- a/utils/loss.py +++ b/utils/loss.py @@ -138,7 +138,7 @@ def __call__(self, p, targets): # predictions, targets pxy = pxy.sigmoid() * 2 - 0.5 pwh = (pwh.sigmoid() * 2) ** 2 * anchors[i] pbox = torch.cat((pxy, pwh), 1) # predicted box - iou = bbox_iou(pbox.T, tbox[i], x1y1x2y2=False, CIoU=True) # iou(prediction, target) + iou = bbox_iou(pbox, tbox[i], CIoU=True).squeeze() # iou(prediction, target) lbox += (1.0 - iou).mean() # iou loss # Objectness @@ -180,7 +180,7 @@ def build_targets(self, p, targets): tcls, tbox, indices, anch = [], [], [], [] gain = torch.ones(7, device=self.device) # normalized to gridspace gain ai = torch.arange(na, device=self.device).float().view(na, 1).repeat(1, nt) # same as .repeat_interleave(nt) - targets = torch.cat((targets.repeat(na, 1, 1), ai[:, :, None]), 2) # append anchor indices + targets = torch.cat((targets.repeat(na, 1, 1), ai[..., None]), 2) # append anchor indices g = 0.5 # bias off = torch.tensor( @@ -199,10 +199,10 @@ def build_targets(self, p, targets): gain[2:6] = torch.tensor(p[i].shape)[[3, 2, 3, 2]] # xyxy gain # Match targets to anchors - t = targets * gain + t = targets * gain # shape(3,n,7) if nt: # Matches - r = t[:, :, 4:6] / anchors[:, None] # wh ratio + r = t[..., 4:6] / anchors[:, None] # wh ratio j = torch.max(r, 1 / r).max(2)[0] < self.hyp['anchor_t'] # compare # j = wh_iou(anchors, t[:, 4:6]) > model.hyp['iou_t'] # iou(3,n)=wh_iou(anchors(3,2), gwh(n,2)) t = t[j] # filter diff --git a/utils/metrics.py b/utils/metrics.py index 0674beddc0fb..ff43a3073062 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -206,37 +206,36 @@ def print(self): print(' '.join(map(str, self.matrix[i]))) -def bbox_iou(box1, box2, x1y1x2y2=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7): - # Returns the IoU of box1 to box2. box1 is 4, box2 is nx4 - box2 = box2.T +def bbox_iou(box1, box2, xywh=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7): + # Returns Intersection over Union (IoU) of box1(1,4) to box2(n,4) # Get the coordinates of bounding boxes - if x1y1x2y2: # x1, y1, x2, y2 = box1 - b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3] - b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3] - else: # transform from xywh to xyxy - b1_x1, b1_x2 = box1[0] - box1[2] / 2, box1[0] + box1[2] / 2 - b1_y1, b1_y2 = box1[1] - box1[3] / 2, box1[1] + box1[3] / 2 - b2_x1, b2_x2 = box2[0] - box2[2] / 2, box2[0] + box2[2] / 2 - b2_y1, b2_y2 = box2[1] - box2[3] / 2, box2[1] + box2[3] / 2 + if xywh: # transform from xywh to xyxy + (x1, y1, w1, h1), (x2, y2, w2, h2) = box1.chunk(4, 1), box2.chunk(4, 1) + w1_, h1_, w2_, h2_ = w1 / 2, h1 / 2, w2 / 2, h2 / 2 + b1_x1, b1_x2, b1_y1, b1_y2 = x1 - w1_, x1 + w1_, y1 - h1_, y1 + h1_ + b2_x1, b2_x2, b2_y1, b2_y2 = x2 - w2_, x2 + w2_, y2 - h2_, y2 + h2_ + else: # x1, y1, x2, y2 = box1 + b1_x1, b1_y1, b1_x2, b1_y2 = box1.chunk(4, 1) + b2_x1, b2_y1, b2_x2, b2_y2 = box2.chunk(4, 1) + w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps + w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps # Intersection area inter = (torch.min(b1_x2, b2_x2) - torch.max(b1_x1, b2_x1)).clamp(0) * \ (torch.min(b1_y2, b2_y2) - torch.max(b1_y1, b2_y1)).clamp(0) # Union Area - w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps - w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps union = w1 * h1 + w2 * h2 - inter + eps + # IoU iou = inter / union if CIoU or DIoU or GIoU: cw = torch.max(b1_x2, b2_x2) - torch.min(b1_x1, b2_x1) # convex (smallest enclosing box) width ch = torch.max(b1_y2, b2_y2) - torch.min(b1_y1, b2_y1) # convex height if CIoU or DIoU: # Distance or Complete IoU https://arxiv.org/abs/1911.08287v1 c2 = cw ** 2 + ch ** 2 + eps # convex diagonal squared - rho2 = ((b2_x1 + b2_x2 - b1_x1 - b1_x2) ** 2 + - (b2_y1 + b2_y2 - b1_y1 - b1_y2) ** 2) / 4 # center distance squared + rho2 = ((b2_x1 + b2_x2 - b1_x1 - b1_x2) ** 2 + (b2_y1 + b2_y2 - b1_y1 - b1_y2) ** 2) / 4 # center dist ** 2 if CIoU: # https://github.com/Zzh-tju/DIoU-SSD-pytorch/blob/master/utils/box/box_utils.py#L47 v = (4 / math.pi ** 2) * torch.pow(torch.atan(w2 / h2) - torch.atan(w1 / h1), 2) with torch.no_grad(): @@ -248,6 +247,11 @@ def bbox_iou(box1, box2, x1y1x2y2=True, GIoU=False, DIoU=False, CIoU=False, eps= return iou # IoU +def box_area(box): + # box = xyxy(4,n) + return (box[2] - box[0]) * (box[3] - box[1]) + + def box_iou(box1, box2): # https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py """ @@ -261,16 +265,12 @@ def box_iou(box1, box2): IoU values for every element in boxes1 and boxes2 """ - def box_area(box): - # box = 4xn - return (box[2] - box[0]) * (box[3] - box[1]) - - area1 = box_area(box1.T) - area2 = box_area(box2.T) - # inter(N,M) = (rb(N,M,2) - lt(N,M,2)).clamp(0).prod(2) - inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) - torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2) - return inter / (area1[:, None] + area2 - inter) # iou = inter / (area1 + area2 - inter) + (a1, a2), (b1, b2) = box1[:, None].chunk(2, 2), box2.chunk(2, 1) + inter = (torch.min(a2, b2) - torch.max(a1, b1)).clamp(0).prod(2) + + # IoU = inter / (area1 + area2 - inter) + return inter / (box_area(box1.T)[:, None] + box_area(box2.T) - inter) def bbox_ioa(box1, box2, eps=1E-7): @@ -280,11 +280,9 @@ def bbox_ioa(box1, box2, eps=1E-7): returns: np.array of shape(n) """ - box2 = box2.transpose() - # Get the coordinates of bounding boxes - b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3] - b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3] + b1_x1, b1_y1, b1_x2, b1_y2 = box1 + b2_x1, b2_y1, b2_x2, b2_y2 = box2.T # Intersection area inter_area = (np.minimum(b1_x2, b2_x2) - np.maximum(b1_x1, b2_x1)).clip(0) * \ diff --git a/val.py b/val.py index 5841437051c2..36f2a6c0284b 100644 --- a/val.py +++ b/val.py @@ -38,10 +38,10 @@ from models.common import DetectMultiBackend from utils.callbacks import Callbacks from utils.datasets import create_dataloader -from utils.general import (LOGGER, box_iou, check_dataset, check_img_size, check_requirements, check_yaml, +from utils.general import (LOGGER, check_dataset, check_img_size, check_requirements, check_yaml, coco80_to_coco91_class, colorstr, increment_path, non_max_suppression, print_args, scale_coords, xywh2xyxy, xyxy2xywh) -from utils.metrics import ConfusionMatrix, ap_per_class +from utils.metrics import ConfusionMatrix, ap_per_class, box_iou from utils.plots import output_to_target, plot_images, plot_val_study from utils.torch_utils import select_device, time_sync From 1993efd59e54e990add1b562ac147e57722987f9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 10 Apr 2022 13:53:28 +0200 Subject: [PATCH 183/661] Swap `unsafe_chunk()` for `chunk()` (#7362) Eliminates all unsafe function in YOLOv5 out of an abundance of caution. --- utils/loss.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/loss.py b/utils/loss.py index b5d050e46047..a1b0ff6c1244 100644 --- a/utils/loss.py +++ b/utils/loss.py @@ -220,7 +220,7 @@ def build_targets(self, p, targets): offsets = 0 # Define - bc, gxy, gwh, a = t.unsafe_chunk(4, dim=1) # (image, class), grid xy, grid wh, anchors + bc, gxy, gwh, a = t.chunk(4, 1) # (image, class), grid xy, grid wh, anchors a, (b, c) = a.long().view(-1), bc.long().T # anchors, image, class gij = (gxy - offsets).long() gi, gj = gij.T # grid indices From db36f13c7afa1d0b2a77d3437e46f6f5fe58c020 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 10 Apr 2022 14:40:33 +0200 Subject: [PATCH 184/661] Delete FUNDING.yml (#7363) Deleting as redundant with FUNDING.yml present in organization repo at https://github.com/ultralytics/.github --- .github/FUNDING.yml | 5 ----- 1 file changed, 5 deletions(-) delete mode 100644 .github/FUNDING.yml diff --git a/.github/FUNDING.yml b/.github/FUNDING.yml deleted file mode 100644 index 3da386f7e724..000000000000 --- a/.github/FUNDING.yml +++ /dev/null @@ -1,5 +0,0 @@ -# These are supported funding model platforms - -github: glenn-jocher -patreon: ultralytics -open_collective: ultralytics From b8d4f2bf74812fc299d6d363b441a99feb14af27 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 10 Apr 2022 14:50:01 +0200 Subject: [PATCH 185/661] Replace Slack with Community Forum in issues (#7364) --- .github/ISSUE_TEMPLATE/config.yml | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/.github/ISSUE_TEMPLATE/config.yml b/.github/ISSUE_TEMPLATE/config.yml index f388d7bacf66..4db7cefb2707 100644 --- a/.github/ISSUE_TEMPLATE/config.yml +++ b/.github/ISSUE_TEMPLATE/config.yml @@ -1,8 +1,8 @@ blank_issues_enabled: true contact_links: - - name: Slack - url: https://join.slack.com/t/ultralytics/shared_invite/zt-w29ei8bp-jczz7QYUmDtgo6r6KcMIAg - about: Ask on Ultralytics Slack Forum + - name: 💬 Forum + url: https://community.ultralytics.com/ + about: Ask on Ultralytics Community Forum - name: Stack Overflow url: https://stackoverflow.com/search?q=YOLOv5 about: Ask on Stack Overflow with 'YOLOv5' tag From 8c420c4c1fb3b83ef0e60749d46bcc2ec9967fc5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 10 Apr 2022 15:17:25 +0200 Subject: [PATCH 186/661] Update ci-testing.yml (#7365) Remove keras==2.6.0 patch --- .github/workflows/ci-testing.yml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 59193e05e08c..e5d5fc434f06 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -50,8 +50,8 @@ jobs: - name: Install dependencies run: | python -m pip install --upgrade pip - pip install -qr requirements.txt -f https://download.pytorch.org/whl/cpu/torch_stable.html - pip install -q onnx tensorflow-cpu keras==2.6.0 # wandb # extras + pip install -qr requirements.txt -f https://download.pytorch.org/whl/cpu/torch_stable.html \ + onnx tensorflow-cpu # wandb python --version pip --version pip list From 71685cbf91a9f60eb2f9c46ced8fa7becf6813d9 Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Mon, 11 Apr 2022 10:26:13 +0200 Subject: [PATCH 187/661] Bump actions/stale from 4 to 5 (#7371) Bumps [actions/stale](https://github.com/actions/stale) from 4 to 5. - [Release notes](https://github.com/actions/stale/releases) - [Changelog](https://github.com/actions/stale/blob/main/CHANGELOG.md) - [Commits](https://github.com/actions/stale/compare/v4...v5) --- updated-dependencies: - dependency-name: actions/stale dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> --- .github/workflows/stale.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/stale.yml b/.github/workflows/stale.yml index 7a83950c17b7..78b2161f73b0 100644 --- a/.github/workflows/stale.yml +++ b/.github/workflows/stale.yml @@ -9,7 +9,7 @@ jobs: stale: runs-on: ubuntu-latest steps: - - uses: actions/stale@v4 + - uses: actions/stale@v5 with: repo-token: ${{ secrets.GITHUB_TOKEN }} stale-issue-message: | From bd2dda8e64b384acd34f54a1aacfa7fc8997be13 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 11 Apr 2022 12:34:35 +0200 Subject: [PATCH 188/661] Update optimizer param group strategy (#7376) * Update optimizer param group strategy Avoid empty lists on missing BathNorm2d models as in https://github.com/ultralytics/yolov5/issues/7375 * fix init --- train.py | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/train.py b/train.py index d6764116b27c..e023a3418454 100644 --- a/train.py +++ b/train.py @@ -150,27 +150,27 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio hyp['weight_decay'] *= batch_size * accumulate / nbs # scale weight_decay LOGGER.info(f"Scaled weight_decay = {hyp['weight_decay']}") - g0, g1, g2 = [], [], [] # optimizer parameter groups + g = [], [], [] # optimizer parameter groups for v in model.modules(): if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter): # bias - g2.append(v.bias) + g[2].append(v.bias) if isinstance(v, nn.BatchNorm2d): # weight (no decay) - g0.append(v.weight) + g[1].append(v.weight) elif hasattr(v, 'weight') and isinstance(v.weight, nn.Parameter): # weight (with decay) - g1.append(v.weight) + g[0].append(v.weight) if opt.optimizer == 'Adam': - optimizer = Adam(g0, lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum + optimizer = Adam(g[2], lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum elif opt.optimizer == 'AdamW': - optimizer = AdamW(g0, lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum + optimizer = AdamW(g[2], lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum else: - optimizer = SGD(g0, lr=hyp['lr0'], momentum=hyp['momentum'], nesterov=True) + optimizer = SGD(g[2], lr=hyp['lr0'], momentum=hyp['momentum'], nesterov=True) - optimizer.add_param_group({'params': g1, 'weight_decay': hyp['weight_decay']}) # add g1 with weight_decay - optimizer.add_param_group({'params': g2}) # add g2 (biases) + optimizer.add_param_group({'params': g[0], 'weight_decay': hyp['weight_decay']}) # add g0 with weight_decay + optimizer.add_param_group({'params': g[1]}) # add g1 (BatchNorm2d weights) LOGGER.info(f"{colorstr('optimizer:')} {type(optimizer).__name__} with parameter groups " - f"{len(g0)} weight (no decay), {len(g1)} weight, {len(g2)} bias") - del g0, g1, g2 + f"{len(g[1])} weight (no decay), {len(g[0])} weight, {len(g[2])} bias") + del g # Scheduler if opt.cos_lr: From fa569cdae52dfd3074561129c3a5185bded60b16 Mon Sep 17 00:00:00 2001 From: Vardan Agarwal <35430842+vardanagarwal@users.noreply.github.com> Date: Mon, 11 Apr 2022 17:34:22 +0530 Subject: [PATCH 189/661] Add support for different normalization layers (#7377) * Add support for different normalization layers. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Cleanup Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- train.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/train.py b/train.py index e023a3418454..80bff18fd653 100644 --- a/train.py +++ b/train.py @@ -151,10 +151,11 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio LOGGER.info(f"Scaled weight_decay = {hyp['weight_decay']}") g = [], [], [] # optimizer parameter groups + bn = nn.BatchNorm2d, nn.LazyBatchNorm2d, nn.GroupNorm, nn.InstanceNorm2d, nn.LazyInstanceNorm2d, nn.LayerNorm for v in model.modules(): if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter): # bias g[2].append(v.bias) - if isinstance(v, nn.BatchNorm2d): # weight (no decay) + if isinstance(v, bn): # weight (no decay) g[1].append(v.weight) elif hasattr(v, 'weight') and isinstance(v.weight, nn.Parameter): # weight (with decay) g[0].append(v.weight) From 4bb7eb8b849fc8a90823a60e2b7a8ec9e38926bf Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 12 Apr 2022 11:02:11 +0200 Subject: [PATCH 190/661] Dynamic normalization layer selection (#7392) * Dynamic normalization layer selection Based on actual available layers. Torch 1.7 compatible, resolves https://github.com/ultralytics/yolov5/issues/7381 * Update train.py --- train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train.py b/train.py index 80bff18fd653..806e2cebe561 100644 --- a/train.py +++ b/train.py @@ -151,7 +151,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio LOGGER.info(f"Scaled weight_decay = {hyp['weight_decay']}") g = [], [], [] # optimizer parameter groups - bn = nn.BatchNorm2d, nn.LazyBatchNorm2d, nn.GroupNorm, nn.InstanceNorm2d, nn.LazyInstanceNorm2d, nn.LayerNorm + bn = tuple(v for k, v in nn.__dict__.items() if 'Norm' in k) # normalization layers, i.e. BatchNorm2d() for v in model.modules(): if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter): # bias g[2].append(v.bias) From 74aaab33129724e0f9f663cff268f7bb296c386b Mon Sep 17 00:00:00 2001 From: Ayush Chaurasia Date: Tue, 12 Apr 2022 15:16:56 +0530 Subject: [PATCH 191/661] Add version warning for wandb (#7385) * add version warning * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Cleanup * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update __init__.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- utils/loggers/__init__.py | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index bab133cc35a9..3a3ec1ee455b 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -81,6 +81,11 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, run_id = torch.load(self.weights).get('wandb_id') if self.opt.resume and not wandb_artifact_resume else None self.opt.hyp = self.hyp # add hyperparameters self.wandb = WandbLogger(self.opt, run_id) + # temp warn. because nested artifacts not supported after 0.12.10 + if pkg.parse_version(wandb.__version__) >= pkg.parse_version('0.12.11'): + self.logger.warning( + "YOLOv5 temporarily requires wandb version 0.12.10 or below. Some features may not work as expected." + ) else: self.wandb = None From 5333b55e7403f1f2db629eadf63b81200f8f8db2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 12 Apr 2022 14:57:50 +0200 Subject: [PATCH 192/661] Remove OpenVINO ONNX `opset<=12` check (#7398) No longer needed. --- export.py | 1 - 1 file changed, 1 deletion(-) diff --git a/export.py b/export.py index ecead3ef5a90..e1e7207058b5 100644 --- a/export.py +++ b/export.py @@ -473,7 +473,6 @@ def run( # Checks imgsz *= 2 if len(imgsz) == 1 else 1 # expand - opset = 12 if ('openvino' in include) else opset # OpenVINO requires opset <= 12 assert nc == len(names), f'Model class count {nc} != len(names) {len(names)}' # Input From 2da2466168116a9fa81f4acab744dc9fe8f90cac Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 12 Apr 2022 15:08:53 +0200 Subject: [PATCH 193/661] Fix EdgeTPU output directory (#7399) * Fix EdgeTPU output directory Outputs to same directory as --weights * Update export.py --- export.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/export.py b/export.py index e1e7207058b5..00b98517cdf6 100644 --- a/export.py +++ b/export.py @@ -387,7 +387,7 @@ def export_edgetpu(keras_model, im, file, prefix=colorstr('Edge TPU:')): f = str(file).replace('.pt', '-int8_edgetpu.tflite') # Edge TPU model f_tfl = str(file).replace('.pt', '-int8.tflite') # TFLite model - cmd = f"edgetpu_compiler -s {f_tfl}" + cmd = f"edgetpu_compiler -s -o {file.parent} {f_tfl}" subprocess.run(cmd, shell=True, check=True) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') From 014acde79daee83e1f1801412cc7a48293e6e1f2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 12 Apr 2022 17:26:53 +0200 Subject: [PATCH 194/661] Update `git_describe()` (#7402) * Update `git_describe()` Add .git path check to avoid `fatal: not a git repository (or any of the parent directories): .git` printout * Update general.py --- utils/general.py | 1 + 1 file changed, 1 insertion(+) diff --git a/utils/general.py b/utils/general.py index 6c2558db74c4..daef2a427111 100755 --- a/utils/general.py +++ b/utils/general.py @@ -275,6 +275,7 @@ def check_online(): def git_describe(path=ROOT): # path must be a directory # Return human-readable git description, i.e. v5.0-5-g3e25f1e https://git-scm.com/docs/git-describe try: + assert (Path(path) / '.git').is_dir() return check_output(f'git -C {path} describe --tags --long --always', shell=True).decode()[:-1] except Exception: return '' From 3eefab1bb109214a614485b6c5f80f22c122f2b2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 15 Apr 2022 21:48:52 +0200 Subject: [PATCH 195/661] Remove `tensorrt` pip install check (#7439) --- export.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/export.py b/export.py index 00b98517cdf6..f97df4710b6f 100644 --- a/export.py +++ b/export.py @@ -209,8 +209,7 @@ def export_coreml(model, im, file, prefix=colorstr('CoreML:')): def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=False, prefix=colorstr('TensorRT:')): # YOLOv5 TensorRT export https://developer.nvidia.com/tensorrt try: - check_requirements(('tensorrt',)) - import tensorrt as trt + import tensorrt as trt # pip install -U nvidia-tensorrt --index-url https://pypi.ngc.nvidia.com if trt.__version__[0] == '7': # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012 grid = model.model[-1].anchor_grid From c9a3b14a749edf77e2faf7ad41f5cd779bd106fd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 16 Apr 2022 15:12:38 +0200 Subject: [PATCH 196/661] Disable `pbar` for DDP ranks > 0 (#7440) --- utils/datasets.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/datasets.py b/utils/datasets.py index 3fa9aa4c6ca1..ef04f51dffef 100755 --- a/utils/datasets.py +++ b/utils/datasets.py @@ -522,7 +522,7 @@ def __init__(self, self.im_hw0, self.im_hw = [None] * n, [None] * n fcn = self.cache_images_to_disk if cache_images == 'disk' else self.load_image results = ThreadPool(NUM_THREADS).imap(fcn, range(n)) - pbar = tqdm(enumerate(results), total=n, bar_format=BAR_FORMAT) + pbar = tqdm(enumerate(results), total=n, bar_format=BAR_FORMAT, disable=LOCAL_RANK > 0) for i, x in pbar: if cache_images == 'disk': gb += self.npy_files[i].stat().st_size From 7926afccde1a95a4c8dbeb9d2b8a901d9f220ca7 Mon Sep 17 00:00:00 2001 From: Cedric Perauer <40869163+Cedric-Perauer@users.noreply.github.com> Date: Sat, 16 Apr 2022 18:00:50 +0200 Subject: [PATCH 197/661] Add `--half` support for FP16 CoreML exports with (#7446) * add fp16 for coreml using --half * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix * Cleanup * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update export.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- export.py | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) diff --git a/export.py b/export.py index f97df4710b6f..2a5eff23c1a6 100644 --- a/export.py +++ b/export.py @@ -186,7 +186,7 @@ def export_openvino(model, im, file, prefix=colorstr('OpenVINO:')): LOGGER.info(f'\n{prefix} export failure: {e}') -def export_coreml(model, im, file, prefix=colorstr('CoreML:')): +def export_coreml(model, im, file, int8, half, prefix=colorstr('CoreML:')): # YOLOv5 CoreML export try: check_requirements(('coremltools',)) @@ -197,6 +197,14 @@ def export_coreml(model, im, file, prefix=colorstr('CoreML:')): ts = torch.jit.trace(model, im, strict=False) # TorchScript model ct_model = ct.convert(ts, inputs=[ct.ImageType('image', shape=im.shape, scale=1 / 255, bias=[0, 0, 0])]) + bits, mode = (8, 'kmeans_lut') if int8 else (16, 'linear') if half else (32, None) + if bits < 32: + if platform.system() == 'Darwin': # quantization only supported on macOS + with warnings.catch_warnings(): + warnings.filterwarnings("ignore", category=DeprecationWarning) # suppress numpy==1.20 float warning + ct_model = ct.models.neural_network.quantization_utils.quantize_weights(ct_model, bits, mode) + else: + print(f'{prefix} quantization only supported on macOS, skipping...') ct_model.save(f) LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') @@ -466,7 +474,8 @@ def run( # Load PyTorch model device = select_device(device) - assert not (device.type == 'cpu' and half), '--half only compatible with GPU export, i.e. use --device 0' + if half: + assert device.type != 'cpu' or coreml, '--half only compatible with GPU export, i.e. use --device 0' model = attempt_load(weights, map_location=device, inplace=True, fuse=True) # load FP32 model nc, names = model.nc, model.names # number of classes, class names @@ -480,7 +489,7 @@ def run( im = torch.zeros(batch_size, 3, *imgsz).to(device) # image size(1,3,320,192) BCHW iDetection # Update model - if half: + if half and not coreml: im, model = im.half(), model.half() # to FP16 model.train() if train else model.eval() # training mode = no Detect() layer grid construction for k, m in model.named_modules(): @@ -506,7 +515,7 @@ def run( if xml: # OpenVINO f[3] = export_openvino(model, im, file) if coreml: - _, f[4] = export_coreml(model, im, file) + _, f[4] = export_coreml(model, im, file, int8, half) # TensorFlow Exports if any((saved_model, pb, tflite, edgetpu, tfjs)): From 3a25e81b303b0b80b79e1c99f4bc2a602e23ab65 Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Tue, 19 Apr 2022 15:07:05 -0700 Subject: [PATCH 198/661] Bump cirrus-actions/rebase from 1.5 to 1.6 (#7462) Bumps [cirrus-actions/rebase](https://github.com/cirrus-actions/rebase) from 1.5 to 1.6. - [Release notes](https://github.com/cirrus-actions/rebase/releases) - [Commits](https://github.com/cirrus-actions/rebase/compare/1.5...1.6) --- updated-dependencies: - dependency-name: cirrus-actions/rebase dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> --- .github/workflows/rebase.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/rebase.yml b/.github/workflows/rebase.yml index 75c57546166b..d79d5cfb20c4 100644 --- a/.github/workflows/rebase.yml +++ b/.github/workflows/rebase.yml @@ -16,6 +16,6 @@ jobs: token: ${{ secrets.ACTIONS_TOKEN }} fetch-depth: 0 # otherwise, you will fail to push refs to dest repo - name: Automatic Rebase - uses: cirrus-actions/rebase@1.5 + uses: cirrus-actions/rebase@1.6 env: GITHUB_TOKEN: ${{ secrets.ACTIONS_TOKEN }} From d876caab4d8f54d11988c277eb2a237bbe405841 Mon Sep 17 00:00:00 2001 From: HERIUN Date: Wed, 20 Apr 2022 07:40:06 +0900 Subject: [PATCH 199/661] Update val.py (#7478) * Update val.py is_coco doesn't work!! '/' -> os.sep!! * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Cleanup * fix Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- val.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/val.py b/val.py index 36f2a6c0284b..13971612ac78 100644 --- a/val.py +++ b/val.py @@ -155,7 +155,7 @@ def run( # Configure model.eval() cuda = device.type != 'cpu' - is_coco = isinstance(data.get('val'), str) and data['val'].endswith('coco/val2017.txt') # COCO dataset + is_coco = isinstance(data.get('val'), str) and data['val'].endswith(f'coco{os.sep}val2017.txt') # COCO dataset nc = 1 if single_cls else int(data['nc']) # number of classes iouv = torch.linspace(0.5, 0.95, 10, device=device) # iou vector for mAP@0.5:0.95 niou = iouv.numel() From c9042dc2adbb635aeca407c10cf492a6eb14d772 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 19 Apr 2022 17:32:15 -0700 Subject: [PATCH 200/661] Improved non-latin `Annotator()` plotting (#7488) * Improved non-latin labels Annotator plotting May resolve https://github.com/ultralytics/yolov5/issues/7460 * Update train.py * Update train.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * add progress arg Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- train.py | 8 +++++--- utils/general.py | 4 ++-- utils/plots.py | 7 ++++--- 3 files changed, 11 insertions(+), 8 deletions(-) diff --git a/train.py b/train.py index 806e2cebe561..c774430df293 100644 --- a/train.py +++ b/train.py @@ -48,13 +48,13 @@ from utils.downloads import attempt_download from utils.general import (LOGGER, check_dataset, check_file, check_git_status, check_img_size, check_requirements, check_suffix, check_yaml, colorstr, get_latest_run, increment_path, init_seeds, - intersect_dicts, labels_to_class_weights, labels_to_image_weights, methods, one_cycle, - print_args, print_mutation, strip_optimizer) + intersect_dicts, is_ascii, labels_to_class_weights, labels_to_image_weights, methods, + one_cycle, print_args, print_mutation, strip_optimizer) from utils.loggers import Loggers from utils.loggers.wandb.wandb_utils import check_wandb_resume from utils.loss import ComputeLoss from utils.metrics import fitness -from utils.plots import plot_evolve, plot_labels +from utils.plots import check_font, plot_evolve, plot_labels from utils.torch_utils import EarlyStopping, ModelEMA, de_parallel, select_device, torch_distributed_zero_first LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html @@ -105,6 +105,8 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio init_seeds(1 + RANK) with torch_distributed_zero_first(LOCAL_RANK): data_dict = data_dict or check_dataset(data) # check if None + if not is_ascii(data_dict['names']): # non-latin labels, i.e. asian, arabic, cyrillic + check_font('Arial.Unicode.ttf', progress=True) train_path, val_path = data_dict['train'], data_dict['val'] nc = 1 if single_cls else int(data_dict['nc']) # number of classes names = ['item'] if single_cls and len(data_dict['names']) != 1 else data_dict['names'] # class names diff --git a/utils/general.py b/utils/general.py index daef2a427111..a4bc3cae9315 100755 --- a/utils/general.py +++ b/utils/general.py @@ -424,13 +424,13 @@ def check_file(file, suffix=''): return files[0] # return file -def check_font(font=FONT): +def check_font(font=FONT, progress=False): # Download font to CONFIG_DIR if necessary font = Path(font) if not font.exists() and not (CONFIG_DIR / font.name).exists(): url = "https://ultralytics.com/assets/" + font.name LOGGER.info(f'Downloading {url} to {CONFIG_DIR / font.name}...') - torch.hub.download_url_to_file(url, str(font), progress=False) + torch.hub.download_url_to_file(url, str(font), progress=progress) def check_dataset(data, autodownload=True): diff --git a/utils/plots.py b/utils/plots.py index 51e9cfdf6e04..842894e745df 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -19,7 +19,7 @@ from PIL import Image, ImageDraw, ImageFont from utils.general import (CONFIG_DIR, FONT, LOGGER, Timeout, check_font, check_requirements, clip_coords, - increment_path, is_ascii, is_chinese, try_except, xywh2xyxy, xyxy2xywh) + increment_path, is_ascii, try_except, xywh2xyxy, xyxy2xywh) from utils.metrics import fitness # Settings @@ -72,11 +72,12 @@ class Annotator: # YOLOv5 Annotator for train/val mosaics and jpgs and detect/hub inference annotations def __init__(self, im, line_width=None, font_size=None, font='Arial.ttf', pil=False, example='abc'): assert im.data.contiguous, 'Image not contiguous. Apply np.ascontiguousarray(im) to Annotator() input images.' - self.pil = pil or not is_ascii(example) or is_chinese(example) + non_ascii = not is_ascii(example) # non-latin labels, i.e. asian, arabic, cyrillic + self.pil = pil or non_ascii if self.pil: # use PIL self.im = im if isinstance(im, Image.Image) else Image.fromarray(im) self.draw = ImageDraw.Draw(self.im) - self.font = check_pil_font(font='Arial.Unicode.ttf' if is_chinese(example) else font, + self.font = check_pil_font(font='Arial.Unicode.ttf' if non_ascii else font, size=font_size or max(round(sum(self.im.size) / 2 * 0.035), 12)) else: # use cv2 self.im = im From ab5b9174940f29a62374bddaf38cd5d2eeb68e25 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 19 Apr 2022 17:50:02 -0700 Subject: [PATCH 201/661] `check_fonts()` download to `CONFIG_DIR` fix (#7489) Follows https://github.com/ultralytics/yolov5/pull/7488. Correct bug where fonts were downloading to current working directory rather than global CONFIG_DIR --- utils/general.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/utils/general.py b/utils/general.py index a4bc3cae9315..cc37ad5fff62 100755 --- a/utils/general.py +++ b/utils/general.py @@ -427,10 +427,11 @@ def check_file(file, suffix=''): def check_font(font=FONT, progress=False): # Download font to CONFIG_DIR if necessary font = Path(font) - if not font.exists() and not (CONFIG_DIR / font.name).exists(): + file = CONFIG_DIR / font.name + if not font.exists() and not file.exists(): url = "https://ultralytics.com/assets/" + font.name - LOGGER.info(f'Downloading {url} to {CONFIG_DIR / font.name}...') - torch.hub.download_url_to_file(url, str(font), progress=progress) + LOGGER.info(f'Downloading {url} to {file}...') + torch.hub.download_url_to_file(url, str(file), progress=progress) def check_dataset(data, autodownload=True): From 3f3852e2ff755275098c07fe3bf4d2bde103ab30 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 19 Apr 2022 21:15:04 -0700 Subject: [PATCH 202/661] Fix val.py Ensemble() (#7490) --- models/experimental.py | 5 +++-- val.py | 2 +- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/models/experimental.py b/models/experimental.py index e166722cbfca..b8d4d70d26e8 100644 --- a/models/experimental.py +++ b/models/experimental.py @@ -115,7 +115,8 @@ def attempt_load(weights, map_location=None, inplace=True, fuse=True): return model[-1] # return model else: print(f'Ensemble created with {weights}\n') - for k in ['names']: - setattr(model, k, getattr(model[-1], k)) + for k in 'names', 'nc', 'yaml': + setattr(model, k, getattr(model[0], k)) model.stride = model[torch.argmax(torch.tensor([m.stride.max() for m in model])).int()].stride # max stride + assert all(model[0].nc == m.nc for m in model), f'Models have different class counts: {[m.nc for m in model]}' return model # return ensemble diff --git a/val.py b/val.py index 13971612ac78..a773ff3e4fa3 100644 --- a/val.py +++ b/val.py @@ -163,7 +163,7 @@ def run( # Dataloader if not training: if pt and not single_cls: # check --weights are trained on --data - ncm = model.model.yaml['nc'] + ncm = model.model.nc assert ncm == nc, f'{weights[0]} ({ncm} classes) trained on different --data than what you passed ({nc} ' \ f'classes). Pass correct combination of --weights and --data that are trained together.' model.warmup(imgsz=(1 if pt else batch_size, 3, imgsz, imgsz)) # warmup From b77c8d9d72031bbccdd2ed26febd70483b467d2e Mon Sep 17 00:00:00 2001 From: Joseph Kocherhans Date: Wed, 20 Apr 2022 12:08:22 -0700 Subject: [PATCH 203/661] Added `YOLOv5_AUTOINSTALL` environment variable (#7505) * Added a way to skip dependency auto-installation. Setting the environment variable `YOLOv5_AUTOINSTALL=False` will skip installing any missing dependencies as if the user had passed `install=False` to `check_requirements`. * Cleanup Co-authored-by: Glenn Jocher --- utils/general.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index cc37ad5fff62..92e3560de8c0 100755 --- a/utils/general.py +++ b/utils/general.py @@ -40,6 +40,7 @@ ROOT = FILE.parents[1] # YOLOv5 root directory DATASETS_DIR = ROOT.parent / 'datasets' # YOLOv5 datasets directory NUM_THREADS = min(8, max(1, os.cpu_count() - 1)) # number of YOLOv5 multiprocessing threads +AUTOINSTALL = str(os.getenv('YOLOv5_AUTOINSTALL', True)).lower() == 'true' # global auto-install mode VERBOSE = str(os.getenv('YOLOv5_VERBOSE', True)).lower() == 'true' # global verbose mode FONT = 'Arial.ttf' # https://ultralytics.com/assets/Arial.ttf @@ -338,7 +339,7 @@ def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), insta pkg.require(r) except Exception: # DistributionNotFound or VersionConflict if requirements not met s = f"{prefix} {r} not found and is required by YOLOv5" - if install: + if install and AUTOINSTALL: # check environment variable LOGGER.info(f"{s}, attempting auto-update...") try: assert check_online(), f"'pip install {r}' skipped (offline)" From 918d7b2b3f8433b80ff12b4407aa5ad524ddbf9d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 20 Apr 2022 14:23:55 -0700 Subject: [PATCH 204/661] Refactor Dockerfiles to `utils/docker` (#7510) * Refactor Docker files * Refactor Docker files * Update Dockerfile --- .dockerignore => utils/docker/.dockerignore | 0 Dockerfile => utils/docker/Dockerfile | 3 +- utils/docker/Dockerfile-cpu | 37 +++++++++++++++++++++ 3 files changed, 38 insertions(+), 2 deletions(-) rename .dockerignore => utils/docker/.dockerignore (100%) rename Dockerfile => utils/docker/Dockerfile (94%) create mode 100644 utils/docker/Dockerfile-cpu diff --git a/.dockerignore b/utils/docker/.dockerignore similarity index 100% rename from .dockerignore rename to utils/docker/.dockerignore diff --git a/Dockerfile b/utils/docker/Dockerfile similarity index 94% rename from Dockerfile rename to utils/docker/Dockerfile index 7df6c1854156..a2a0f0cd9c1a 100644 --- a/Dockerfile +++ b/utils/docker/Dockerfile @@ -23,11 +23,10 @@ COPY . /usr/src/app RUN git clone https://github.com/ultralytics/yolov5 /usr/src/yolov5 # Downloads to user config dir -ADD https://ultralytics.com/assets/Arial.ttf /root/.config/Ultralytics/ +ADD https://ultralytics.com/assets/Arial.ttf https://ultralytics.com/assets/Arial.Unicode.ttf /root/.config/Ultralytics/ # Set environment variables ENV OMP_NUM_THREADS=8 -# ENV HOME=/usr/src/app # Usage Examples ------------------------------------------------------------------------------------------------------- diff --git a/utils/docker/Dockerfile-cpu b/utils/docker/Dockerfile-cpu new file mode 100644 index 000000000000..6e757baa3ef1 --- /dev/null +++ b/utils/docker/Dockerfile-cpu @@ -0,0 +1,37 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license + +# Start FROM Ubuntu image https://hub.docker.com/_/ubuntu +FROM ubuntu:latest + +# Install linux packages +RUN apt update +RUN DEBIAN_FRONTEND=noninteractive TZ=Etc/UTC apt install -y tzdata +RUN apt install -y python3-pip git zip curl htop screen libgl1-mesa-glx libglib2.0-0 +RUN alias python=python3 + +# Install python dependencies +COPY requirements.txt . +RUN python3 -m pip install --upgrade pip +RUN pip install --no-cache -r requirements.txt albumentations gsutil notebook \ + coremltools onnx onnx-simplifier onnxruntime openvino-dev tensorflow-cpu tensorflowjs \ + torch==1.11.0+cpu torchvision==0.12.0+cpu -f https://download.pytorch.org/whl/cpu/torch_stable.html + +# Create working directory +RUN mkdir -p /usr/src/app +WORKDIR /usr/src/app + +# Copy contents +COPY . /usr/src/app +RUN git clone https://github.com/ultralytics/yolov5 /usr/src/yolov5 + +# Downloads to user config dir +ADD https://ultralytics.com/assets/Arial.ttf https://ultralytics.com/assets/Arial.Unicode.ttf /root/.config/Ultralytics/ + + +# Usage Examples ------------------------------------------------------------------------------------------------------- + +# Build and Push +# t=ultralytics/yolov5:latest-cpu && sudo docker build -t $t . && sudo docker push $t + +# Pull and Run +# t=ultralytics/yolov5:latest-cpu && sudo docker pull $t && sudo docker run -it --ipc=host -v "$(pwd)"/datasets:/usr/src/datasets $t From 6ea81bb3a9bb1701bc0aa9ccca546368ce1fa400 Mon Sep 17 00:00:00 2001 From: Zengyf-CVer <41098760+Zengyf-CVer@users.noreply.github.com> Date: Thu, 21 Apr 2022 09:44:52 +0800 Subject: [PATCH 205/661] Add yesqa to precommit checks (#7511) * Update .pre-commit-config.yaml * Update .pre-commit-config.yaml --- .pre-commit-config.yaml | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index ae61892b68b2..bff7f8a40093 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -52,11 +52,10 @@ repos: # - mdformat-black # - mdformat_frontmatter - # TODO - #- repo: https://github.com/asottile/yesqa - # rev: v1.2.3 - # hooks: - # - id: yesqa + - repo: https://github.com/asottile/yesqa + rev: v1.3.0 + hooks: + - id: yesqa - repo: https://github.com/PyCQA/flake8 rev: 4.0.1 From 23718df1c6b546e525d06a6e2f6a4ebc9737bb4b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 21 Apr 2022 18:21:01 -0700 Subject: [PATCH 206/661] Fix val `plots=plots` (#7524) --- train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train.py b/train.py index c774430df293..f6e66cb0ef09 100644 --- a/train.py +++ b/train.py @@ -461,7 +461,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio save_dir=save_dir, save_json=is_coco, verbose=True, - plots=True, + plots=plots, callbacks=callbacks, compute_loss=compute_loss) # val best model with plots if is_coco: From d2e698c75c4845757d31af4c9116f004624151e2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 21 Apr 2022 20:06:57 -0700 Subject: [PATCH 207/661] Reduce val device transfers (#7525) --- val.py | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/val.py b/val.py index a773ff3e4fa3..b2b3bc75911e 100644 --- a/val.py +++ b/val.py @@ -220,14 +220,14 @@ def run( # Metrics for si, pred in enumerate(out): labels = targets[targets[:, 0] == si, 1:] - nl = len(labels) - tcls = labels[:, 0].tolist() if nl else [] # target class + nl, npr = labels.shape[0], pred.shape[0] # number of labels, predictions path, shape = Path(paths[si]), shapes[si][0] + correct = torch.zeros(npr, niou, dtype=torch.bool, device=device) # init seen += 1 - if len(pred) == 0: + if npr == 0: if nl: - stats.append((torch.zeros(0, niou, dtype=torch.bool), torch.Tensor(), torch.Tensor(), tcls)) + stats.append((correct, *torch.zeros((3, 0)))) continue # Predictions @@ -244,9 +244,7 @@ def run( correct = process_batch(predn, labelsn, iouv) if plots: confusion_matrix.process_batch(predn, labelsn) - else: - correct = torch.zeros(pred.shape[0], niou, dtype=torch.bool) - stats.append((correct.cpu(), pred[:, 4].cpu(), pred[:, 5].cpu(), tcls)) # (correct, conf, pcls, tcls) + stats.append((correct, pred[:, 4], pred[:, 5], labels[:, 0])) # (correct, conf, pcls, tcls) # Save/log if save_txt: @@ -265,7 +263,7 @@ def run( callbacks.run('on_val_batch_end') # Compute metrics - stats = [np.concatenate(x, 0) for x in zip(*stats)] # to numpy + stats = [torch.cat(x, 0).cpu().numpy() for x in zip(*stats)] # to numpy if len(stats) and stats[0].any(): tp, fp, p, r, f1, ap, ap_class = ap_per_class(*stats, plot=plots, save_dir=save_dir, names=names) ap50, ap = ap[:, 0], ap.mean(1) # AP@0.5, AP@0.5:0.95 From b804b36bc4ea856ecec250add8ab39d4b5127eda Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 21 Apr 2022 20:31:26 -0700 Subject: [PATCH 208/661] Add Docker `--file` argument to build (#7527) --- utils/docker/Dockerfile | 2 +- utils/docker/Dockerfile-cpu | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index a2a0f0cd9c1a..9bb24bb6bf3e 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -32,7 +32,7 @@ ENV OMP_NUM_THREADS=8 # Usage Examples ------------------------------------------------------------------------------------------------------- # Build and Push -# t=ultralytics/yolov5:latest && sudo docker build -t $t . && sudo docker push $t +# t=ultralytics/yolov5:latest && sudo docker build -f utils/docker/Dockerfile -t $t . && sudo docker push $t # Pull and Run # t=ultralytics/yolov5:latest && sudo docker pull $t && sudo docker run -it --ipc=host --gpus all $t diff --git a/utils/docker/Dockerfile-cpu b/utils/docker/Dockerfile-cpu index 6e757baa3ef1..d30c07e81172 100644 --- a/utils/docker/Dockerfile-cpu +++ b/utils/docker/Dockerfile-cpu @@ -31,7 +31,7 @@ ADD https://ultralytics.com/assets/Arial.ttf https://ultralytics.com/assets/Aria # Usage Examples ------------------------------------------------------------------------------------------------------- # Build and Push -# t=ultralytics/yolov5:latest-cpu && sudo docker build -t $t . && sudo docker push $t +# t=ultralytics/yolov5:latest-cpu && sudo docker build -f utils/docker/Dockerfile-cpu -t $t . && sudo docker push $t # Pull and Run # t=ultralytics/yolov5:latest-cpu && sudo docker pull $t && sudo docker run -it --ipc=host -v "$(pwd)"/datasets:/usr/src/datasets $t From 813eba85b266fe46b0ac02a62fce8b25e3eeabac Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 22 Apr 2022 12:01:14 -0700 Subject: [PATCH 209/661] Empty val batch CUDA device fix (#7539) Verified fix for https://github.com/ultralytics/yolov5/pull/7525#issuecomment-1106081123 --- val.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/val.py b/val.py index b2b3bc75911e..58113f016a58 100644 --- a/val.py +++ b/val.py @@ -227,7 +227,7 @@ def run( if npr == 0: if nl: - stats.append((correct, *torch.zeros((3, 0)))) + stats.append((correct, *torch.zeros((3, 0), device=device))) continue # Predictions From cc1d7df03c7c3c37367e76b237ac4b087ea040d4 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 22 Apr 2022 12:31:33 -0700 Subject: [PATCH 210/661] Autoinstall TensorRT if missing (#7537) * Autoinstall TensorRT if missing May resolve https://github.com/ultralytics/yolov5/issues/7464 * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update export.py * Update export.py * Update export.py * Update export.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/export.py b/export.py index 2a5eff23c1a6..93d98c801d02 100644 --- a/export.py +++ b/export.py @@ -217,7 +217,15 @@ def export_coreml(model, im, file, int8, half, prefix=colorstr('CoreML:')): def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=False, prefix=colorstr('TensorRT:')): # YOLOv5 TensorRT export https://developer.nvidia.com/tensorrt try: - import tensorrt as trt # pip install -U nvidia-tensorrt --index-url https://pypi.ngc.nvidia.com + assert im.device.type != 'cpu', 'export running on CPU but must be on GPU, i.e. `python export.py --device 0`' + try: + import tensorrt as trt + except Exception: + s = f"\n{prefix} tensorrt not found and is required by YOLOv5" + LOGGER.info(f"{s}, attempting auto-update...") + r = '-U nvidia-tensorrt --index-url https://pypi.ngc.nvidia.com' + LOGGER.info(subprocess.check_output(f"pip install {r}", shell=True).decode()) + import tensorrt as trt if trt.__version__[0] == '7': # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012 grid = model.model[-1].anchor_grid @@ -230,7 +238,6 @@ def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=F onnx = file.with_suffix('.onnx') LOGGER.info(f'\n{prefix} starting export with TensorRT {trt.__version__}...') - assert im.device.type != 'cpu', 'export running on CPU but must be on GPU, i.e. `python export.py --device 0`' assert onnx.exists(), f'failed to export ONNX file: {onnx}' f = file.with_suffix('.engine') # TensorRT engine file logger = trt.Logger(trt.Logger.INFO) From c264795f50b685a8bef7f0d740482b0265ae4898 Mon Sep 17 00:00:00 2001 From: Zengyf-CVer <41098760+Zengyf-CVer@users.noreply.github.com> Date: Sat, 23 Apr 2022 04:36:27 +0800 Subject: [PATCH 211/661] Add mdformat to precommit checks and update other version (#7529) * Update .pre-commit-config.yaml * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update .pre-commit-config.yaml * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update CONTRIBUTING.md * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update README.md * Update README.md * Update README.md Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- .github/CODE_OF_CONDUCT.md | 24 ++++---- .pre-commit-config.yaml | 24 ++++---- CONTRIBUTING.md | 18 +++--- README.md | 53 ++++++++--------- utils/loggers/wandb/README.md | 106 +++++++++++++++++++--------------- 5 files changed, 119 insertions(+), 106 deletions(-) diff --git a/.github/CODE_OF_CONDUCT.md b/.github/CODE_OF_CONDUCT.md index ef10b05fc88e..27e59e9aab38 100644 --- a/.github/CODE_OF_CONDUCT.md +++ b/.github/CODE_OF_CONDUCT.md @@ -17,23 +17,23 @@ diverse, inclusive, and healthy community. Examples of behavior that contributes to a positive environment for our community include: -* Demonstrating empathy and kindness toward other people -* Being respectful of differing opinions, viewpoints, and experiences -* Giving and gracefully accepting constructive feedback -* Accepting responsibility and apologizing to those affected by our mistakes, +- Demonstrating empathy and kindness toward other people +- Being respectful of differing opinions, viewpoints, and experiences +- Giving and gracefully accepting constructive feedback +- Accepting responsibility and apologizing to those affected by our mistakes, and learning from the experience -* Focusing on what is best not just for us as individuals, but for the +- Focusing on what is best not just for us as individuals, but for the overall community Examples of unacceptable behavior include: -* The use of sexualized language or imagery, and sexual attention or +- The use of sexualized language or imagery, and sexual attention or advances of any kind -* Trolling, insulting or derogatory comments, and personal or political attacks -* Public or private harassment -* Publishing others' private information, such as a physical or email +- Trolling, insulting or derogatory comments, and personal or political attacks +- Public or private harassment +- Publishing others' private information, such as a physical or email address, without their explicit permission -* Other conduct which could reasonably be considered inappropriate in a +- Other conduct which could reasonably be considered inappropriate in a professional setting ## Enforcement Responsibilities @@ -121,8 +121,8 @@ https://www.contributor-covenant.org/version/2/0/code_of_conduct.html. Community Impact Guidelines were inspired by [Mozilla's code of conduct enforcement ladder](https://github.com/mozilla/diversity). -[homepage]: https://www.contributor-covenant.org - For answers to common questions about this code of conduct, see the FAQ at https://www.contributor-covenant.org/faq. Translations are available at https://www.contributor-covenant.org/translations. + +[homepage]: https://www.contributor-covenant.org diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index bff7f8a40093..924c940f2c1a 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -13,7 +13,7 @@ ci: repos: - repo: https://github.com/pre-commit/pre-commit-hooks - rev: v4.1.0 + rev: v4.2.0 hooks: - id: end-of-file-fixer - id: trailing-whitespace @@ -24,7 +24,7 @@ repos: - id: check-docstring-first - repo: https://github.com/asottile/pyupgrade - rev: v2.31.1 + rev: v2.32.0 hooks: - id: pyupgrade args: [--py36-plus] @@ -42,15 +42,17 @@ repos: - id: yapf name: YAPF formatting - # TODO - #- repo: https://github.com/executablebooks/mdformat - # rev: 0.7.7 - # hooks: - # - id: mdformat - # additional_dependencies: - # - mdformat-gfm - # - mdformat-black - # - mdformat_frontmatter + - repo: https://github.com/executablebooks/mdformat + rev: 0.7.14 + hooks: + - id: mdformat + additional_dependencies: + - mdformat-gfm + - mdformat-black + exclude: | + (?x)^( + README.md + )$ - repo: https://github.com/asottile/yesqa rev: v1.3.0 diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index ebde03a562a0..13b9b73b50cc 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -18,16 +18,19 @@ Submitting a PR is easy! This example shows how to submit a PR for updating `req ### 1. Select File to Update Select `requirements.txt` to update by clicking on it in GitHub. +

PR_step1

### 2. Click 'Edit this file' Button is in top-right corner. +

PR_step2

### 3. Make Changes Change `matplotlib` version from `3.2.2` to `3.3`. +

PR_step3

### 4. Preview Changes and Submit PR @@ -35,6 +38,7 @@ Change `matplotlib` version from `3.2.2` to `3.3`. Click on the **Preview changes** tab to verify your updates. At the bottom of the screen select 'Create a **new branch** for this commit', assign your branch a descriptive name such as `fix/matplotlib_version` and click the green **Propose changes** button. All done, your PR is now submitted to YOLOv5 for review and approval 😃! +

PR_step4

### PR recommendations @@ -70,21 +74,21 @@ understand and use to **reproduce** the problem. This is referred to by communit a [minimum reproducible example](https://stackoverflow.com/help/minimal-reproducible-example). Your code that reproduces the problem should be: -* ✅ **Minimal** – Use as little code as possible that still produces the same problem -* ✅ **Complete** – Provide **all** parts someone else needs to reproduce your problem in the question itself -* ✅ **Reproducible** – Test the code you're about to provide to make sure it reproduces the problem +- ✅ **Minimal** – Use as little code as possible that still produces the same problem +- ✅ **Complete** – Provide **all** parts someone else needs to reproduce your problem in the question itself +- ✅ **Reproducible** – Test the code you're about to provide to make sure it reproduces the problem In addition to the above requirements, for [Ultralytics](https://ultralytics.com/) to provide assistance your code should be: -* ✅ **Current** – Verify that your code is up-to-date with current +- ✅ **Current** – Verify that your code is up-to-date with current GitHub [master](https://github.com/ultralytics/yolov5/tree/master), and if necessary `git pull` or `git clone` a new copy to ensure your problem has not already been resolved by previous commits. -* ✅ **Unmodified** – Your problem must be reproducible without any modifications to the codebase in this +- ✅ **Unmodified** – Your problem must be reproducible without any modifications to the codebase in this repository. [Ultralytics](https://ultralytics.com/) does not provide support for custom code ⚠️. -If you believe your problem meets all of the above criteria, please close this issue and raise a new one using the 🐛 ** -Bug Report** [template](https://github.com/ultralytics/yolov5/issues/new/choose) and providing +If you believe your problem meets all of the above criteria, please close this issue and raise a new one using the 🐛 +**Bug Report** [template](https://github.com/ultralytics/yolov5/issues/new/choose) and providing a [minimum reproducible example](https://stackoverflow.com/help/minimal-reproducible-example) to help us better understand and diagnose your problem. diff --git a/README.md b/README.md index 54c5cbd83f5b..f1dd65b0a4b1 100644 --- a/README.md +++ b/README.md @@ -103,8 +103,6 @@ results.print() # or .show(), .save(), .crop(), .pandas(), etc. - -
Inference with detect.py @@ -149,20 +147,20 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12
Tutorials -* [Train Custom Data](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data)  🚀 RECOMMENDED -* [Tips for Best Training Results](https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results)  ☘️ +- [Train Custom Data](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data)  🚀 RECOMMENDED +- [Tips for Best Training Results](https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results)  ☘️ RECOMMENDED -* [Weights & Biases Logging](https://github.com/ultralytics/yolov5/issues/1289)  🌟 NEW -* [Roboflow for Datasets, Labeling, and Active Learning](https://github.com/ultralytics/yolov5/issues/4975)  🌟 NEW -* [Multi-GPU Training](https://github.com/ultralytics/yolov5/issues/475) -* [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36)  ⭐ NEW -* [TFLite, ONNX, CoreML, TensorRT Export](https://github.com/ultralytics/yolov5/issues/251) 🚀 -* [Test-Time Augmentation (TTA)](https://github.com/ultralytics/yolov5/issues/303) -* [Model Ensembling](https://github.com/ultralytics/yolov5/issues/318) -* [Model Pruning/Sparsity](https://github.com/ultralytics/yolov5/issues/304) -* [Hyperparameter Evolution](https://github.com/ultralytics/yolov5/issues/607) -* [Transfer Learning with Frozen Layers](https://github.com/ultralytics/yolov5/issues/1314)  ⭐ NEW -* [Architecture Summary](https://github.com/ultralytics/yolov5/issues/6998)  ⭐ NEW +- [Weights & Biases Logging](https://github.com/ultralytics/yolov5/issues/1289)  🌟 NEW +- [Roboflow for Datasets, Labeling, and Active Learning](https://github.com/ultralytics/yolov5/issues/4975)  🌟 NEW +- [Multi-GPU Training](https://github.com/ultralytics/yolov5/issues/475) +- [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36)  ⭐ NEW +- [TFLite, ONNX, CoreML, TensorRT Export](https://github.com/ultralytics/yolov5/issues/251) 🚀 +- [Test-Time Augmentation (TTA)](https://github.com/ultralytics/yolov5/issues/303) +- [Model Ensembling](https://github.com/ultralytics/yolov5/issues/318) +- [Model Pruning/Sparsity](https://github.com/ultralytics/yolov5/issues/304) +- [Hyperparameter Evolution](https://github.com/ultralytics/yolov5/issues/607) +- [Transfer Learning with Frozen Layers](https://github.com/ultralytics/yolov5/issues/1314)  ⭐ NEW +- [Architecture Summary](https://github.com/ultralytics/yolov5/issues/6998)  ⭐ NEW
@@ -203,7 +201,6 @@ Get started in seconds with our verified environments. Click each icon below for |:-:|:-:| |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) | - + + + +##
文件
+ +请参阅[YOLOv5 Docs](https://docs.ultralytics.com)，了解有关培训、测试和部署的完整文件。 + +##
快速开始案例
+ +
+安装 + +在[**Python>=3.7.0**](https://www.python.org/) 的环境中克隆版本仓并安装 [requirements.txt](https://github.com/ultralytics/yolov5/blob/master/requirements.txt)，包括[**PyTorch>=1.7**](https://pytorch.org/get-started/locally/)。 +```bash +git clone https://github.com/ultralytics/yolov5 # 克隆 +cd yolov5 +pip install -r requirements.txt # 安装 +``` + +
+ +
+推断 + +YOLOv5 [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36) 推断. [模型](https://github.com/ultralytics/yolov5/tree/master/models) 自动从最新YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases)下载。 + +```python +import torch + +# 模型 +model = torch.hub.load('ultralytics/yolov5', 'yolov5s') # or yolov5n - yolov5x6, custom + +# 图像 +img = 'https://ultralytics.com/images/zidane.jpg' # or file, Path, PIL, OpenCV, numpy, list + +# 推论 +results = model(img) + +# 结果 +results.print() # or .show(), .save(), .crop(), .pandas(), etc. +``` + +
+ +
+用 detect.py 进行推断 + +`detect.py` 在各种资源上运行推理, 从最新的YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases) 中自动下载 [模型](https://github.com/ultralytics/yolov5/tree/master/models) 并保存结果来运行/检测。 + +```bash +python detect.py --source 0 # 网络摄像头 + img.jpg # 图像 + vid.mp4 # 视频 + path/ # 文件夹 + path/*.jpg # glob + 'https://youtu.be/Zgi9g1ksQHc' # YouTube + 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP 流 +``` + +
+ +
+训练 + +以下指令再现了YOLOv5 [COCO](https://github.com/ultralytics/yolov5/blob/master/data/scripts/get_coco.sh) +数据集结果. [模型](https://github.com/ultralytics/yolov5/tree/master/models) 和 [数据集](https://github.com/ultralytics/yolov5/tree/master/data) 自动从最新的YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases)中下载。YOLOv5n/s/m/l/x的训练时间在V100 GPU上是1/2/4/6/8天（多GPU倍速）. 尽可能使用最大的 `--batch-size`, 或通过 `--batch-size -1` 来实现 YOLOv5 [自动批处理](https://github.com/ultralytics/yolov5/pull/5092). 批量大小显示为V100-16GB。 + +```bash +python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 128 + yolov5s 64 + yolov5m 40 + yolov5l 24 + yolov5x 16 +``` + + + +
+ +
+教程 + +- [训练自定义数据](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data) 🚀 推荐 +- [获得最佳训练效果的技巧](https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results) ☘️ 推荐 +- [Weights & Biases 登陆](https://github.com/ultralytics/yolov5/issues/1289) 🌟 新 +- [Roboflow：数据集、标签和主动学习](https://github.com/ultralytics/yolov5/issues/4975) 🌟 新 +- [多GPU训练](https://github.com/ultralytics/yolov5/issues/475) +- [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36) ⭐ 新 +- [TFLite, ONNX, CoreML, TensorRT 导出](https://github.com/ultralytics/yolov5/issues/251) 🚀 +- [测试时数据增强 (TTA)](https://github.com/ultralytics/yolov5/issues/303) +- [模型组合](https://github.com/ultralytics/yolov5/issues/318) +- [模型剪枝/稀疏性](https://github.com/ultralytics/yolov5/issues/304) +- [超参数进化](https://github.com/ultralytics/yolov5/issues/607) +- [带有冻结层的迁移学习](https://github.com/ultralytics/yolov5/issues/1314) ⭐ 新 +- [架构概要](https://github.com/ultralytics/yolov5/issues/6998) ⭐ 新 + +
+ +##
环境
+ +使用经过我们验证的环境，几秒钟就可以开始。点击下面的每个图标了解详情。 + + + +##
一体化
+ + + +|Weights and Biases|Roboflow ⭐ 新| +|:-:|:-:| +|通过 [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) 自动跟踪和可视化你在云端的所有YOLOv5训练运行状态。|标记并将您的自定义数据集直接导出到YOLOv5，以便用 [Roboflow](https://roboflow.com/?ref=ultralytics) 进行训练。 | + + + +##
为什么是 YOLOv5
+ +

+
+ YOLOv5-P5 640 图像 (点击扩展) + +

+
+
+ 图片注释 (点击扩展) + +- **COCO AP val** 表示 mAP@0.5:0.95 在5000张图像的[COCO val2017](http://cocodataset.org)数据集上，在256到1536的不同推理大小上测量的指标。 +- **GPU Speed** 衡量的是在 [COCO val2017](http://cocodataset.org) 数据集上使用 [AWS p3.2xlarge](https://aws.amazon.com/ec2/instance-types/p3/) V100实例在批量大小为32时每张图像的平均推理时间。 +- **EfficientDet** 数据来自 [google/automl](https://github.com/google/automl) ，批量大小为 8。 +- **重制** 于 `python val.py --task study --data coco.yaml --iou 0.7 --weights yolov5n6.pt yolov5s6.pt yolov5m6.pt yolov5l6.pt yolov5x6.pt` + +
+ +### 预训练检查点 + +|Model |size
(pixels) |mAPval
0.5:0.95 |mAPval
0.5 |Speed
CPU b1
(ms) |Speed
V100 b1
(ms) |Speed
V100 b32
(ms) |params
(M) |FLOPs
@640 (B) +|--- |--- |--- |--- |--- |--- |--- |--- |--- +|[YOLOv5n][assets] |640 |28.0 |45.7 |**45** |**6.3**|**0.6**|**1.9**|**4.5** +|[YOLOv5s][assets] |640 |37.4 |56.8 |98 |6.4 |0.9 |7.2 |16.5 +|[YOLOv5m][assets] |640 |45.4 |64.1 |224 |8.2 |1.7 |21.2 |49.0 +|[YOLOv5l][assets] |640 |49.0 |67.3 |430 |10.1 |2.7 |46.5 |109.1 +|[YOLOv5x][assets] |640 |50.7 |68.9 |766 |12.1 |4.8 |86.7 |205.7 +| | | | | | | | | +|[YOLOv5n6][assets] |1280 |36.0 |54.4 |153 |8.1 |2.1 |3.2 |4.6 +|[YOLOv5s6][assets] |1280 |44.8 |63.7 |385 |8.2 |3.6 |12.6 |16.8 +|[YOLOv5m6][assets] |1280 |51.3 |69.3 |887 |11.1 |6.8 |35.7 |50.0 +|[YOLOv5l6][assets] |1280 |53.7 |71.3 |1784 |15.8 |10.5 |76.8 |111.4 +|[YOLOv5x6][assets]
+ [TTA][TTA]|1280
1536 |55.0
**55.8** |72.7
**72.7** |3136
- |26.2
- |19.4
- |140.7
- |209.8
- + +
+ 表格注释 (点击扩展) + +- 所有检查点都以默认设置训练到300个时期. Nano和Small模型用 [hyp.scratch-low.yaml](https://github.com/ultralytics/yolov5/blob/master/data/hyps/hyp.scratch-low.yaml) hyps, 其他模型使用 [hyp.scratch-high.yaml](https://github.com/ultralytics/yolov5/blob/master/data/hyps/hyp.scratch-high.yaml). +- **mAPval** 值是 [COCO val2017](http://cocodataset.org) 数据集上的单模型单尺度的值。 +
重制于 `python val.py --data coco.yaml --img 640 --conf 0.001 --iou 0.65` +- 使用 [AWS p3.2xlarge](https://aws.amazon.com/ec2/instance-types/p3/) 实例对COCO val图像的平均速度。不包括NMS时间（~1 ms/img) +
重制于`python val.py --data coco.yaml --img 640 --task speed --batch 1` +- **TTA** [测试时数据增强](https://github.com/ultralytics/yolov5/issues/303) 包括反射和比例增强. +
重制于 `python val.py --data coco.yaml --img 1536 --iou 0.7 --augment` + +
+ +##
贡献
+ +我们重视您的意见! 我们希望大家对YOLOv5的贡献尽可能的简单和透明。开始之前请先点击并查看我们的 [贡献指南](CONTRIBUTING.md)，填写[YOLOv5调查问卷](https://ultralytics.com/survey?utm_source=github&utm_medium=social&utm_campaign=Survey) 来向我们发送您的经验反馈。真诚感谢我们所有的贡献者! + + +##
联系
+ +关于YOLOv5的漏洞和功能问题，请访问 [GitHub Issues](https://github.com/ultralytics/yolov5/issues)。业务咨询或技术支持服务请访问[https://ultralytics.com/contact](https://ultralytics.com/contact)。 + +
+ + + +[assets]: https://github.com/ultralytics/yolov5/releases +[tta]: https://github.com/ultralytics/yolov5/issues/303 diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 850527491859..0c24b1ee2a06 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -50,10 +50,7 @@ repos: additional_dependencies: - mdformat-gfm - mdformat-black - exclude: | - (?x)^( - README.md - )$ + exclude: "README.md|README_cn.md" - repo: https://github.com/asottile/yesqa rev: v1.3.0 diff --git a/README.md b/README.md index 953761229f77..b0ea0a5d814c 100644 --- a/README.md +++ b/README.md @@ -3,6 +3,8 @@

+ +English | [简体中文](.github/README_cn.md)
CI CPU testing From 0537e8dd13859c4b44db3bf6f39b9ff20eaf163b Mon Sep 17 00:00:00 2001 From: Nicholas Zolton <78943323+NicholasZolton@users.noreply.github.com> Date: Sun, 26 Jun 2022 17:04:11 -0500 Subject: [PATCH 325/661] Allow detect.py to use video size for initial window size (#8330) * fixed initial window size of detect.py being tiny * cleanup Co-authored-by: Glenn Jocher --- detect.py | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/detect.py b/detect.py index 9d92e4c169e4..bb09ce171a96 100644 --- a/detect.py +++ b/detect.py @@ -106,7 +106,7 @@ def run( # Run inference model.warmup(imgsz=(1 if pt else bs, 3, *imgsz)) # warmup - dt, seen = [0.0, 0.0, 0.0], 0 + seen, windows, dt = 0, [], [0.0, 0.0, 0.0] for path, im, im0s, vid_cap, s in dataset: t1 = time_sync() im = torch.from_numpy(im).to(device) @@ -173,7 +173,10 @@ def run( # Stream results im0 = annotator.result() if view_img: - cv2.namedWindow(str(p), cv2.WINDOW_NORMAL | cv2.WINDOW_KEEPRATIO) # allow window resize (Linux) + if p not in windows: + windows.append(p) + cv2.namedWindow(str(p), cv2.WINDOW_NORMAL | cv2.WINDOW_KEEPRATIO) # allow window resize (Linux) + cv2.resizeWindow(str(p), im0.shape[1], im0.shape[0]) cv2.imshow(str(p), im0) cv2.waitKey(1) # 1 millisecond From b0814c95214b7fd0464310b1cf151fd5c1337c6d Mon Sep 17 00:00:00 2001 From: Zhiqiang Wang Date: Mon, 27 Jun 2022 19:10:30 +0800 Subject: [PATCH 326/661] Revamp Chinese docs (#8350) Revamp Chines docs --- .github/README_cn.md | 38 +++++++++++++++++++------------------- 1 file changed, 19 insertions(+), 19 deletions(-) diff --git a/.github/README_cn.md b/.github/README_cn.md index 78719509ad85..7e90336d5157 100644 --- a/.github/README_cn.md +++ b/.github/README_cn.md @@ -60,7 +60,7 @@ YOLOv5🚀是一个在COCO数据集上预训练的物体检测架构和模型系 ##
文件
-请参阅[YOLOv5 Docs](https://docs.ultralytics.com)，了解有关培训、测试和部署的完整文件。 +请参阅[YOLOv5 Docs](https://docs.ultralytics.com)，了解有关训练、测试和部署的完整文件。 ##
快速开始案例
@@ -77,9 +77,9 @@ pip install -r requirements.txt # 安装
-推断 +推理 -YOLOv5 [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36) 推断. [模型](https://github.com/ultralytics/yolov5/tree/master/models) 自动从最新YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases)下载。 +YOLOv5 [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36) 推理. [模型](https://github.com/ultralytics/yolov5/tree/master/models) 自动从最新YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases)下载。 ```python import torch @@ -90,7 +90,7 @@ model = torch.hub.load('ultralytics/yolov5', 'yolov5s') # or yolov5n - yolov5x6 # 图像 img = 'https://ultralytics.com/images/zidane.jpg' # or file, Path, PIL, OpenCV, numpy, list -# 推论 +# 推理 results = model(img) # 结果 @@ -100,9 +100,9 @@ results.print() # or .show(), .save(), .crop(), .pandas(), etc.
-用 detect.py 进行推断 +用 detect.py 进行推理 -`detect.py` 在各种资源上运行推理, 从最新的YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases) 中自动下载 [模型](https://github.com/ultralytics/yolov5/tree/master/models) 并保存结果来运行/检测。 +`detect.py` 在各种数据源上运行推理, 其会从最新的 YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases) 中自动下载 [模型](https://github.com/ultralytics/yolov5/tree/master/models) 并将检测结果保存到 `runs/detect` 目录。 ```bash python detect.py --source 0 # 网络摄像头 @@ -119,8 +119,8 @@ python detect.py --source 0 # 网络摄像头
训练 -以下指令再现了YOLOv5 [COCO](https://github.com/ultralytics/yolov5/blob/master/data/scripts/get_coco.sh) -数据集结果. [模型](https://github.com/ultralytics/yolov5/tree/master/models) 和 [数据集](https://github.com/ultralytics/yolov5/tree/master/data) 自动从最新的YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases)中下载。YOLOv5n/s/m/l/x的训练时间在V100 GPU上是1/2/4/6/8天（多GPU倍速）. 尽可能使用最大的 `--batch-size`, 或通过 `--batch-size -1` 来实现 YOLOv5 [自动批处理](https://github.com/ultralytics/yolov5/pull/5092). 批量大小显示为V100-16GB。 +以下指令再现了 YOLOv5 [COCO](https://github.com/ultralytics/yolov5/blob/master/data/scripts/get_coco.sh) +数据集结果. [模型](https://github.com/ultralytics/yolov5/tree/master/models) 和 [数据集](https://github.com/ultralytics/yolov5/tree/master/data) 自动从最新的YOLOv5 [版本](https://github.com/ultralytics/yolov5/releases) 中下载。YOLOv5n/s/m/l/x的训练时间在V100 GPU上是 1/2/4/6/8天（多GPU倍速）. 尽可能使用最大的 `--batch-size`, 或通过 `--batch-size -1` 来实现 YOLOv5 [自动批处理](https://github.com/ultralytics/yolov5/pull/5092). 批量大小显示为 V100-16GB。 ```bash python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 128 @@ -139,13 +139,13 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 - [训练自定义数据](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data) 🚀 推荐 - [获得最佳训练效果的技巧](https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results) ☘️ 推荐 -- [Weights & Biases 登陆](https://github.com/ultralytics/yolov5/issues/1289) 🌟 新 +- [使用 Weights & Biases 记录实验](https://github.com/ultralytics/yolov5/issues/1289) 🌟 新 - [Roboflow：数据集、标签和主动学习](https://github.com/ultralytics/yolov5/issues/4975) 🌟 新 - [多GPU训练](https://github.com/ultralytics/yolov5/issues/475) - [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36) ⭐ 新 - [TFLite, ONNX, CoreML, TensorRT 导出](https://github.com/ultralytics/yolov5/issues/251) 🚀 - [测试时数据增强 (TTA)](https://github.com/ultralytics/yolov5/issues/303) -- [模型组合](https://github.com/ultralytics/yolov5/issues/318) +- [模型集成](https://github.com/ultralytics/yolov5/issues/318) - [模型剪枝/稀疏性](https://github.com/ultralytics/yolov5/issues/304) - [超参数进化](https://github.com/ultralytics/yolov5/issues/607) - [带有冻结层的迁移学习](https://github.com/ultralytics/yolov5/issues/1314) ⭐ 新 @@ -175,7 +175,7 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 -##
一体化
+##
如何与第三方集成
@@ -199,7 +199,7 @@ We are super excited about our first-ever Ultralytics YOLOv5 🚀 EXPORT Competi

--> -##
为什么是 YOLOv5
+##
为什么选择 YOLOv5

@@ -212,8 +212,8 @@ We are super excited about our first-ever Ultralytics YOLOv5 🚀 EXPORT Competi - **COCO AP val** 表示 mAP@0.5:0.95 在5000张图像的[COCO val2017](http://cocodataset.org)数据集上，在256到1536的不同推理大小上测量的指标。 - **GPU Speed** 衡量的是在 [COCO val2017](http://cocodataset.org) 数据集上使用 [AWS p3.2xlarge](https://aws.amazon.com/ec2/instance-types/p3/) V100实例在批量大小为32时每张图像的平均推理时间。 -- **EfficientDet** 数据来自 [google/automl](https://github.com/google/automl) ，批量大小为 8。 -- **重制** 于 `python val.py --task study --data coco.yaml --iou 0.7 --weights yolov5n6.pt yolov5s6.pt yolov5m6.pt yolov5l6.pt yolov5x6.pt` +- **EfficientDet** 数据来自 [google/automl](https://github.com/google/automl) ，批量大小设置为 8。 +- 复现 mAP 方法: `python val.py --task study --data coco.yaml --iou 0.7 --weights yolov5n6.pt yolov5s6.pt yolov5m6.pt yolov5l6.pt yolov5x6.pt`
@@ -238,22 +238,22 @@ We are super excited about our first-ever Ultralytics YOLOv5 🚀 EXPORT Competi - 所有检查点都以默认设置训练到300个时期. Nano和Small模型用 [hyp.scratch-low.yaml](https://github.com/ultralytics/yolov5/blob/master/data/hyps/hyp.scratch-low.yaml) hyps, 其他模型使用 [hyp.scratch-high.yaml](https://github.com/ultralytics/yolov5/blob/master/data/hyps/hyp.scratch-high.yaml). - **mAPval** 值是 [COCO val2017](http://cocodataset.org) 数据集上的单模型单尺度的值。 -
重制于 `python val.py --data coco.yaml --img 640 --conf 0.001 --iou 0.65` +
复现方法: `python val.py --data coco.yaml --img 640 --conf 0.001 --iou 0.65` - 使用 [AWS p3.2xlarge](https://aws.amazon.com/ec2/instance-types/p3/) 实例对COCO val图像的平均速度。不包括NMS时间（~1 ms/img) -
重制于`python val.py --data coco.yaml --img 640 --task speed --batch 1` +
复现方法: `python val.py --data coco.yaml --img 640 --task speed --batch 1` - **TTA** [测试时数据增强](https://github.com/ultralytics/yolov5/issues/303) 包括反射和比例增强. -
重制于 `python val.py --data coco.yaml --img 1536 --iou 0.7 --augment` +
复现方法: `python val.py --data coco.yaml --img 1536 --iou 0.7 --augment`
##
贡献
-我们重视您的意见! 我们希望大家对YOLOv5的贡献尽可能的简单和透明。开始之前请先点击并查看我们的 [贡献指南](CONTRIBUTING.md)，填写[YOLOv5调查问卷](https://ultralytics.com/survey?utm_source=github&utm_medium=social&utm_campaign=Survey) 来向我们发送您的经验反馈。真诚感谢我们所有的贡献者! +我们重视您的意见! 我们希望给大家提供尽可能的简单和透明的方式对 YOLOv5 做出贡献。开始之前请先点击并查看我们的 [贡献指南](CONTRIBUTING.md)，填写[YOLOv5调查问卷](https://ultralytics.com/survey?utm_source=github&utm_medium=social&utm_campaign=Survey) 来向我们发送您的经验反馈。真诚感谢我们所有的贡献者! ##
联系
-关于YOLOv5的漏洞和功能问题，请访问 [GitHub Issues](https://github.com/ultralytics/yolov5/issues)。业务咨询或技术支持服务请访问[https://ultralytics.com/contact](https://ultralytics.com/contact)。 +关于YOLOv5的漏洞和功能问题，请访问 [GitHub Issues](https://github.com/ultralytics/yolov5/issues)。商业咨询或技术支持服务请访问[https://ultralytics.com/contact](https://ultralytics.com/contact)。
From 8ebf569d14aca4f0e5b1f730501ac73644d71ae0 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 27 Jun 2022 16:11:24 +0200 Subject: [PATCH 327/661] Fix bias warmup LR init (#8356) Per https://github.com/ultralytics/yolov5/issues/8352 --- train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train.py b/train.py index a06ad5a418f8..e1393213bb4b 100644 --- a/train.py +++ b/train.py @@ -335,7 +335,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio accumulate = max(1, np.interp(ni, xi, [1, nbs / batch_size]).round()) for j, x in enumerate(optimizer.param_groups): # bias lr falls from 0.1 to lr0, all other lrs rise from 0.0 to lr0 - x['lr'] = np.interp(ni, xi, [hyp['warmup_bias_lr'] if j == 2 else 0.0, x['initial_lr'] * lf(epoch)]) + x['lr'] = np.interp(ni, xi, [hyp['warmup_bias_lr'] if j == 0 else 0.0, x['initial_lr'] * lf(epoch)]) if 'momentum' in x: x['momentum'] = np.interp(ni, xi, [hyp['warmup_momentum'], hyp['momentum']]) From 34df5032a7d2e83fe3d16770a03bd129b115d184 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 27 Jun 2022 17:46:49 +0200 Subject: [PATCH 328/661] Add File Size (MB) column to benchmarks (#8359) * Add filesize to benchmarks.py * Add filesize to benchmarks.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/benchmarks.py | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/utils/benchmarks.py b/utils/benchmarks.py index d0f2a2529c5d..69d653a20916 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -41,7 +41,7 @@ import export import val from utils import notebook_init -from utils.general import LOGGER, check_yaml, print_args +from utils.general import LOGGER, check_yaml, file_size, print_args from utils.torch_utils import select_device @@ -75,10 +75,10 @@ def run( result = val.run(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) metrics = result[0] # metrics (mp, mr, map50, map, *losses(box, obj, cls)) speeds = result[2] # times (preprocess, inference, postprocess) - y.append([name, round(metrics[3], 4), round(speeds[1], 2)]) # mAP, t_inference + y.append([name, round(file_size(w), 1), round(metrics[3], 4), round(speeds[1], 2)]) # MB, mAP, t_inference except Exception as e: LOGGER.warning(f'WARNING: Benchmark failure for {name}: {e}') - y.append([name, None, None]) # mAP, t_inference + y.append([name, None, None, None]) # mAP, t_inference if pt_only and i == 0: break # break after PyTorch @@ -86,7 +86,8 @@ def run( LOGGER.info('\n') parse_opt() notebook_init() # print system info - py = pd.DataFrame(y, columns=['Format', 'mAP@0.5:0.95', 'Inference time (ms)'] if map else ['Format', 'Export', '']) + c = ['Format', 'Size (MB)', 'mAP@0.5:0.95', 'Inference time (ms)'] if map else ['Format', 'Export', '', ''] + py = pd.DataFrame(y, columns=c) LOGGER.info(f'\nBenchmarks complete ({time.time() - t:.2f}s)') LOGGER.info(str(py if map else py.iloc[:, :2])) return py From 50ff6eee31c72fe88bdd35fc7299b201cce0e9a6 Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Tue, 28 Jun 2022 12:04:58 +0200 Subject: [PATCH 329/661] Update protobuf requirement from <=3.20.1 to <4.21.3 (#8346) Updates the requirements on [protobuf](https://github.com/protocolbuffers/protobuf) to permit the latest version. - [Release notes](https://github.com/protocolbuffers/protobuf/releases) - [Changelog](https://github.com/protocolbuffers/protobuf/blob/main/generate_changelog.py) - [Commits](https://github.com/protocolbuffers/protobuf/commits) --- updated-dependencies: - dependency-name: protobuf dependency-type: direct:production ... Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) mode change 100755 => 100644 requirements.txt diff --git a/requirements.txt b/requirements.txt old mode 100755 new mode 100644 index 1937b93b5dda..332a0b81c45b --- a/requirements.txt +++ b/requirements.txt @@ -12,7 +12,7 @@ scipy>=1.4.1 # Google Colab version torch>=1.7.0 torchvision>=0.8.1 tqdm>=4.41.0 -protobuf<=3.20.1 # https://github.com/ultralytics/yolov5/issues/8012 +protobuf<4.21.3 # https://github.com/ultralytics/yolov5/issues/8012 # Logging ------------------------------------- tensorboard>=2.4.1 From 0c1324067c348c985b0c689a1e71cd9ba01513e0 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 28 Jun 2022 15:22:15 +0200 Subject: [PATCH 330/661] Fix ONNX `--dynamic` export on GPU (#8378) * Fix ONNX `--dynamic` export on GPU Patch forces --dynamic export model and image to CPU. Resolves bug raised in https://github.com/ultralytics/yolov5/issues/8377 * Update export.py --- export.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/export.py b/export.py index 72e170a30bf2..9daf39f871c2 100644 --- a/export.py +++ b/export.py @@ -119,8 +119,8 @@ def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorst f = file.with_suffix('.onnx') torch.onnx.export( - model, - im, + model.cpu() if dynamic else model, # --dynamic only compatible with cpu + im.cpu() if dynamic else im, f, verbose=False, opset_version=opset, @@ -499,8 +499,6 @@ def run( im = torch.zeros(batch_size, 3, *imgsz).to(device) # image size(1,3,320,192) BCHW iDetection # Update model - if half and not coreml and not xml: - im, model = im.half(), model.half() # to FP16 model.train() if train else model.eval() # training mode = no Detect() layer grid construction for k, m in model.named_modules(): if isinstance(m, Detect): @@ -510,6 +508,8 @@ def run( for _ in range(2): y = model(im) # dry runs + if half and not coreml: + im, model = im.half(), model.half() # to FP16 shape = tuple(y[0].shape) # model output shape LOGGER.info(f"\n{colorstr('PyTorch:')} starting from {file} with output shape {shape} ({file_size(file):.1f} MB)") From f76a78e7078185ecdc67470d8658103cf2067c81 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 28 Jun 2022 17:34:24 +0200 Subject: [PATCH 331/661] Update tutorial.ipynb (#8380) --- tutorial.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 664cbc156082..7cd9a2d17e94 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -369,7 +369,7 @@ "colab_type": "text" }, "source": [ - "\"Open" + "\"Open" ] }, { From 6935a54e603d634f6b0a9026604dc5875d1ca990 Mon Sep 17 00:00:00 2001 From: Giacomo Guiduzzi <10937563+giacomoguiduzzi@users.noreply.github.com> Date: Wed, 29 Jun 2022 12:41:46 +0200 Subject: [PATCH 332/661] Implementation of Early Stopping for DDP training (#8345) * Implementation of Early Stopping for DDP training This edit correctly uses the broadcast_object_list() function to send slave processes a boolean so to end the training phase if the variable is True, thus allowing the master process to destroy the process group and terminate. * Update train.py * Update train.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update train.py * Update train.py * Update train.py * Further cleanup This cleans up the definition of broadcast_list and removes the requirement for clear() afterward. Co-authored-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- train.py | 24 ++++++++++-------------- 1 file changed, 10 insertions(+), 14 deletions(-) diff --git a/train.py b/train.py index e1393213bb4b..dd5eeb600a76 100644 --- a/train.py +++ b/train.py @@ -294,7 +294,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio results = (0, 0, 0, 0, 0, 0, 0) # P, R, mAP@.5, mAP@.5-.95, val_loss(box, obj, cls) scheduler.last_epoch = start_epoch - 1 # do not move scaler = torch.cuda.amp.GradScaler(enabled=amp) - stopper = EarlyStopping(patience=opt.patience) + stopper, stop = EarlyStopping(patience=opt.patience), False compute_loss = ComputeLoss(model) # init loss class callbacks.run('on_train_start') LOGGER.info(f'Image sizes {imgsz} train, {imgsz} val\n' @@ -402,6 +402,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # Update best mAP fi = fitness(np.array(results).reshape(1, -1)) # weighted combination of [P, R, mAP@.5, mAP@.5-.95] + stop = stopper(epoch=epoch, fitness=fi) # early stop check if fi > best_fitness: best_fitness = fi log_vals = list(mloss) + list(results) + lr @@ -428,19 +429,14 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio del ckpt callbacks.run('on_model_save', last, epoch, final_epoch, best_fitness, fi) - # Stop Single-GPU - if RANK == -1 and stopper(epoch=epoch, fitness=fi): - break - - # Stop DDP TODO: known issues shttps://github.com/ultralytics/yolov5/pull/4576 - # stop = stopper(epoch=epoch, fitness=fi) - # if RANK == 0: - # dist.broadcast_object_list([stop], 0) # broadcast 'stop' to all ranks - - # Stop DPP - # with torch_distributed_zero_first(RANK): - # if stop: - # break # must break all DDP ranks + # EarlyStopping + if RANK != -1: # if DDP training + broadcast_list = [stop if RANK == 0 else None] + dist.broadcast_object_list(broadcast_list, 0) # broadcast 'stop' to all ranks + if RANK != 0: + stop = broadcast_list[0] + if stop: + break # must break all DDP ranks # end epoch ---------------------------------------------------------------------------------------------------- # end training ----------------------------------------------------------------------------------------------------- From e50dc38d3687d18cd932aa342bca03ca7125bbe0 Mon Sep 17 00:00:00 2001 From: Amir Pourmand Date: Thu, 30 Jun 2022 17:31:31 +0430 Subject: [PATCH 333/661] Improve `--local_rank` arg comment (#8409) * add more docs * add more docs * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update train.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train.py b/train.py index dd5eeb600a76..3161159ba44d 100644 --- a/train.py +++ b/train.py @@ -504,7 +504,7 @@ def parse_opt(known=False): parser.add_argument('--patience', type=int, default=100, help='EarlyStopping patience (epochs without improvement)') parser.add_argument('--freeze', nargs='+', type=int, default=[0], help='Freeze layers: backbone=10, first3=0 1 2') parser.add_argument('--save-period', type=int, default=-1, help='Save checkpoint every x epochs (disabled if < 1)') - parser.add_argument('--local_rank', type=int, default=-1, help='DDP parameter, do not modify') + parser.add_argument('--local_rank', type=int, default=-1, help='Automatic DDP Multi-GPU argument, do not modify') # Weights & Biases arguments parser.add_argument('--entity', default=None, help='W&B: Entity') From 898332433a71b8846b15daa276a8ac45c9efa98b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 30 Jun 2022 16:19:22 +0200 Subject: [PATCH 334/661] Update cache comments (#8414) * Update cache comments For better readability * Update dataloaders.py --- utils/dataloaders.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 51d1612d3d5d..5d4dfc6e4d14 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -449,10 +449,10 @@ def __init__(self, cache_path = (p if p.is_file() else Path(self.label_files[0]).parent).with_suffix('.cache') try: cache, exists = np.load(cache_path, allow_pickle=True).item(), True # load dict - assert cache['version'] == self.cache_version # same version - assert cache['hash'] == get_hash(self.label_files + self.im_files) # same hash + assert cache['version'] == self.cache_version # matches current version + assert cache['hash'] == get_hash(self.label_files + self.im_files) # identical hash except Exception: - cache, exists = self.cache_labels(cache_path, prefix), False # cache + cache, exists = self.cache_labels(cache_path, prefix), False # run cache ops # Display cache nf, nm, ne, nc, n = cache.pop('results') # found, missing, empty, corrupt, total From d94b4705a65e751a8238696704a6300df4ac33db Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 1 Jul 2022 15:41:14 +0200 Subject: [PATCH 335/661] TRT `--half` fix autocast images to FP16 (#8435) * TRT `--half` fix autocast images to FP16 Resolves bug raised in https://github.com/ultralytics/yolov5/issues/7822 * Update common.py --- models/common.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/models/common.py b/models/common.py index 7690f714def8..a6488dd85648 100644 --- a/models/common.py +++ b/models/common.py @@ -441,6 +441,9 @@ def wrap_frozen_graph(gd, inputs, outputs): def forward(self, im, augment=False, visualize=False, val=False): # YOLOv5 MultiBackend inference b, ch, h, w = im.shape # batch, channel, height, width + if self.fp16 and im.dtype != torch.float16: + im = im.half() # to FP16 + if self.pt: # PyTorch y = self.model(im, augment=augment, visualize=visualize)[0] elif self.jit: # TorchScript From da2ee3934e2572d700000cc1e5fdac615ba4dd79 Mon Sep 17 00:00:00 2001 From: Colin Wong Date: Fri, 1 Jul 2022 15:15:09 -0500 Subject: [PATCH 336/661] Expose OpenVINO `batch_size` similarly to TensorRT (#8437) --- models/common.py | 1 + 1 file changed, 1 insertion(+) diff --git a/models/common.py b/models/common.py index a6488dd85648..a40207fd4d7b 100644 --- a/models/common.py +++ b/models/common.py @@ -366,6 +366,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, if not Path(w).is_file(): # if not *.xml w = next(Path(w).glob('*.xml')) # get *.xml file from *_openvino_model dir network = ie.read_model(model=w, weights=Path(w).with_suffix('.bin')) + batch_size = network.batch_size executable_network = ie.compile_model(network, device_name="CPU") # device_name="MYRIAD" for Intel NCS2 output_layer = next(iter(executable_network.outputs)) meta = Path(w).with_suffix('.yaml') From 29d79a6360d8c7da8875284246847db3312e270a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 2 Jul 2022 18:35:45 +0200 Subject: [PATCH 337/661] Do not prefer Apple MPS (#8446) Require explicit request for MPS, i.e. ```bash python detect.py --device mps ``` Reverts https://github.com/ultralytics/yolov5/pull/8210 for preferring MPS if available. Note that torch MPS is experiencing ongoing compatibility issues in https://github.com/pytorch/pytorch/issues/77886 --- utils/torch_utils.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index b1b107ee4f1b..c21dc6658c1e 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -62,7 +62,7 @@ def select_device(device='', batch_size=0, newline=True): assert torch.cuda.is_available() and torch.cuda.device_count() >= len(device.replace(',', '')), \ f"Invalid CUDA '--device {device}' requested, use '--device cpu' or pass valid CUDA device(s)" - if not cpu and torch.cuda.is_available(): # prefer GPU if available + if not (cpu or mps) and torch.cuda.is_available(): # prefer GPU if available devices = device.split(',') if device else '0' # range(torch.cuda.device_count()) # i.e. 0,1,6,7 n = len(devices) # device count if n > 1 and batch_size > 0: # check batch_size is divisible by device_count @@ -72,7 +72,7 @@ def select_device(device='', batch_size=0, newline=True): p = torch.cuda.get_device_properties(i) s += f"{'' if i == 0 else space}CUDA:{d} ({p.name}, {p.total_memory / (1 << 20):.0f}MiB)\n" # bytes to MB arg = 'cuda:0' - elif not cpu and getattr(torch, 'has_mps', False) and torch.backends.mps.is_available(): # prefer MPS if available + elif mps and getattr(torch, 'has_mps', False) and torch.backends.mps.is_available(): # prefer MPS if available s += 'MPS\n' arg = 'mps' else: # revert to CPU From c7689198bc66023378f71aa80c0829a763a928bd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 4 Jul 2022 15:01:11 +0200 Subject: [PATCH 338/661] Update stale.yml (#8465) --- .github/workflows/stale.yml | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/.github/workflows/stale.yml b/.github/workflows/stale.yml index ee08510b4a30..03d99790a4a7 100644 --- a/.github/workflows/stale.yml +++ b/.github/workflows/stale.yml @@ -32,7 +32,9 @@ jobs: Thank you for your contributions to YOLOv5 🚀 and Vision AI ⭐! stale-pr-message: 'This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions YOLOv5 🚀 and Vision AI ⭐.' - days-before-stale: 30 - days-before-close: 5 + days-before-issue-stale: 30 + days-before-issue-close: 10 + days-before-pr-stale: 90 + days-before-pr-close: 30 exempt-issue-labels: 'documentation,tutorial,TODO' operations-per-run: 300 # The maximum number of operations per run, used to control rate limiting. From fdc9d9198e0dea90d0536f63b6408b97b1399cc1 Mon Sep 17 00:00:00 2001 From: "pre-commit-ci[bot]" <66853113+pre-commit-ci[bot]@users.noreply.github.com> Date: Mon, 4 Jul 2022 22:09:24 +0200 Subject: [PATCH 339/661] [pre-commit.ci] pre-commit suggestions (#8470) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit updates: - [github.com/pre-commit/pre-commit-hooks: v4.2.0 → v4.3.0](https://github.com/pre-commit/pre-commit-hooks/compare/v4.2.0...v4.3.0) - [github.com/asottile/pyupgrade: v2.32.1 → v2.34.0](https://github.com/asottile/pyupgrade/compare/v2.32.1...v2.34.0) Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- .pre-commit-config.yaml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 0c24b1ee2a06..9b8f28c77506 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -13,7 +13,7 @@ ci: repos: - repo: https://github.com/pre-commit/pre-commit-hooks - rev: v4.2.0 + rev: v4.3.0 hooks: - id: end-of-file-fixer - id: trailing-whitespace @@ -24,7 +24,7 @@ repos: - id: check-docstring-first - repo: https://github.com/asottile/pyupgrade - rev: v2.32.1 + rev: v2.34.0 hooks: - id: pyupgrade name: Upgrade code From 1ab23fc67f52d44d5f8ce67a895e73c7cbd7aec5 Mon Sep 17 00:00:00 2001 From: Junya Morioka <77187490+mjun0812@users.noreply.github.com> Date: Thu, 7 Jul 2022 02:32:58 +0900 Subject: [PATCH 340/661] Exclude torch==1.12.0, torchvision==0.13.0 (Fix #8395) (#8497) Exclude torch==1.12.0, torchvision==0.13.0 --- requirements.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/requirements.txt b/requirements.txt index 332a0b81c45b..ad3fd49691d4 100644 --- a/requirements.txt +++ b/requirements.txt @@ -9,8 +9,8 @@ Pillow>=7.1.2 PyYAML>=5.3.1 requests>=2.23.0 scipy>=1.4.1 # Google Colab version -torch>=1.7.0 -torchvision>=0.8.1 +torch>=1.7.0,!=1.12.0 # https://github.com/ultralytics/yolov5/issues/8395 +torchvision>=0.8.1,!=0.13.0 # https://github.com/ultralytics/yolov5/issues/8395 tqdm>=4.41.0 protobuf<4.21.3 # https://github.com/ultralytics/yolov5/issues/8012 From 36f64a981d08c1fc34c50ae2ff8a15769ee6b49b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 7 Jul 2022 12:34:01 +0200 Subject: [PATCH 341/661] Update tutorial.ipynb (#8507) --- tutorial.ipynb | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 7cd9a2d17e94..bdfba399a883 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -1105,8 +1105,8 @@ "# TensorRT \n", "# https://docs.nvidia.com/deeplearning/tensorrt/install-guide/index.html#installing-pip\n", "!pip install -U nvidia-tensorrt --index-url https://pypi.ngc.nvidia.com # install\n", - "!python export.py --weights yolov5s.pt --include engine --imgsz 640 640 --device 0 # export\n", - "!python detect.py --weights yolov5s.engine --imgsz 640 640 --device 0 # inference" + "!python export.py --weights yolov5s.pt --include engine --imgsz 640 --device 0 # export\n", + "!python detect.py --weights yolov5s.engine --imgsz 640 --device 0 # inference" ], "execution_count": null, "outputs": [] From 27d831b6e4ae4b0286ba0159f5c8542e052cd3c9 Mon Sep 17 00:00:00 2001 From: Ayush Chaurasia Date: Thu, 7 Jul 2022 18:09:29 +0530 Subject: [PATCH 342/661] Training reproducibility improvements (#8213) * attempt at reproducibility * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * use deterministic algs * fix everything :) * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * revert dataloader changes * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * process_batch as np * remove newline * Remove dataloader init fcn * Update val.py * Update train.py * revert additional changes * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update train.py * Add --seed arg * Update general.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update train.py * Update train.py * Update val.py * Update train.py * Update general.py * Update general.py * Add deterministic argument to init_seeds() Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- train.py | 3 ++- utils/general.py | 10 +++++++++- 2 files changed, 11 insertions(+), 2 deletions(-) diff --git a/train.py b/train.py index 3161159ba44d..bf5b4c69d74c 100644 --- a/train.py +++ b/train.py @@ -101,7 +101,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # Config plots = not evolve and not opt.noplots # create plots cuda = device.type != 'cpu' - init_seeds(1 + RANK) + init_seeds(opt.seed + 1 + RANK, deterministic=True) with torch_distributed_zero_first(LOCAL_RANK): data_dict = data_dict or check_dataset(data) # check if None train_path, val_path = data_dict['train'], data_dict['val'] @@ -504,6 +504,7 @@ def parse_opt(known=False): parser.add_argument('--patience', type=int, default=100, help='EarlyStopping patience (epochs without improvement)') parser.add_argument('--freeze', nargs='+', type=int, default=[0], help='Freeze layers: backbone=10, first3=0 1 2') parser.add_argument('--save-period', type=int, default=-1, help='Save checkpoint every x epochs (disabled if < 1)') + parser.add_argument('--seed', type=int, default=0, help='Global training seed') parser.add_argument('--local_rank', type=int, default=-1, help='Automatic DDP Multi-GPU argument, do not modify') # Weights & Biases arguments diff --git a/utils/general.py b/utils/general.py index a3e242d78a17..17b689010b39 100755 --- a/utils/general.py +++ b/utils/general.py @@ -195,14 +195,22 @@ def print_args(args: Optional[dict] = None, show_file=True, show_fcn=False): LOGGER.info(colorstr(s) + ', '.join(f'{k}={v}' for k, v in args.items())) -def init_seeds(seed=0): +def init_seeds(seed=0, deterministic=False): # Initialize random number generator (RNG) seeds https://pytorch.org/docs/stable/notes/randomness.html # cudnn seed 0 settings are slower and more reproducible, else faster and less reproducible import torch.backends.cudnn as cudnn + + if deterministic and check_version(torch.__version__, '1.12.0'): # https://github.com/ultralytics/yolov5/pull/8213 + torch.use_deterministic_algorithms(True) + os.environ['CUBLAS_WORKSPACE_CONFIG'] = ':4096:8' + # os.environ['PYTHONHASHSEED'] = str(seed) + random.seed(seed) np.random.seed(seed) torch.manual_seed(seed) cudnn.benchmark, cudnn.deterministic = (False, True) if seed == 0 else (True, False) + # torch.cuda.manual_seed(seed) + # torch.cuda.manual_seed_all(seed) # for multi GPU, exception safe def intersect_dicts(da, db, exclude=()): From 9d7bc06ae7ea59eeb09be14a42cc4530cdb97a22 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 7 Jul 2022 20:13:42 +0200 Subject: [PATCH 343/661] Revert "Expose OpenVINO `batch_size` similarly to TensorRT" (#8510) Revert "Expose OpenVINO `batch_size` similarly to TensorRT (#8437)" This reverts commit da2ee3934e2572d700000cc1e5fdac615ba4dd79. --- models/common.py | 1 - 1 file changed, 1 deletion(-) diff --git a/models/common.py b/models/common.py index a40207fd4d7b..a6488dd85648 100644 --- a/models/common.py +++ b/models/common.py @@ -366,7 +366,6 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, if not Path(w).is_file(): # if not *.xml w = next(Path(w).glob('*.xml')) # get *.xml file from *_openvino_model dir network = ie.read_model(model=w, weights=Path(w).with_suffix('.bin')) - batch_size = network.batch_size executable_network = ie.compile_model(network, device_name="CPU") # device_name="MYRIAD" for Intel NCS2 output_layer = next(iter(executable_network.outputs)) meta = Path(w).with_suffix('.yaml') From dd28df98c2307abfe13f8857110bfcd6b5c4eb4b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 7 Jul 2022 20:36:23 +0200 Subject: [PATCH 344/661] Avoid FP64 ops for MPS support in train.py (#8511) Avoid FP64 ops for MPS support Resolves https://github.com/ultralytics/yolov5/pull/7878#issuecomment-1177952614 --- utils/general.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/utils/general.py b/utils/general.py index 17b689010b39..a85a2915a31a 100755 --- a/utils/general.py +++ b/utils/general.py @@ -644,7 +644,7 @@ def labels_to_class_weights(labels, nc=80): return torch.Tensor() labels = np.concatenate(labels, 0) # labels.shape = (866643, 5) for COCO - classes = labels[:, 0].astype(np.int) # labels = [class xywh] + classes = labels[:, 0].astype(int) # labels = [class xywh] weights = np.bincount(classes, minlength=nc) # occurrences per class # Prepend gridpoint count (for uCE training) @@ -654,13 +654,13 @@ def labels_to_class_weights(labels, nc=80): weights[weights == 0] = 1 # replace empty bins with 1 weights = 1 / weights # number of targets per class weights /= weights.sum() # normalize - return torch.from_numpy(weights) + return torch.from_numpy(weights).float() def labels_to_image_weights(labels, nc=80, class_weights=np.ones(80)): # Produces image weights based on class_weights and image contents # Usage: index = random.choices(range(n), weights=image_weights, k=1) # weighted image sample - class_counts = np.array([np.bincount(x[:, 0].astype(np.int), minlength=nc) for x in labels]) + class_counts = np.array([np.bincount(x[:, 0].astype(int), minlength=nc) for x in labels]) return (class_weights.reshape(1, nc) * class_counts).sum(1) From 39d7a93619083cb8e37f5ef7708cf50b34e20ee1 Mon Sep 17 00:00:00 2001 From: UnglvKitDe <100289696+UnglvKitDe@users.noreply.github.com> Date: Thu, 7 Jul 2022 20:42:09 +0200 Subject: [PATCH 345/661] Fix AP calculation bug #8464 (#8484) Co-authored-by: Glenn Jocher --- val.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/val.py b/val.py index f4f4bab7e92d..77f6bbf5b7c2 100644 --- a/val.py +++ b/val.py @@ -227,7 +227,7 @@ def run( if npr == 0: if nl: - stats.append((correct, *torch.zeros((3, 0), device=device))) + stats.append((correct, *torch.zeros((2, 0), device=device), labels[:, 0])) continue # Predictions From 3e54651fcaee59561a405b00458bf95df1c8b82e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 7 Jul 2022 23:41:34 +0200 Subject: [PATCH 346/661] Add `--hard-fail` argument to benchmarks for CI errors (#8513) * Add `--hard-fail` list argument to benchmarks for CI Will cause CI to fail on a benchmark failure for given indices. * Update ci-testing.yml * Attempt Failure (CI should fail) * Update benchmarks.py * Update export.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update benchmarks.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update ci-testing.yml * Update benchmarks.py * Update benchmarks.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- .github/workflows/ci-testing.yml | 2 +- export.py | 24 ++++++++++++------------ utils/benchmarks.py | 16 ++++++++++++---- 3 files changed, 25 insertions(+), 17 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 4083ac354c46..f3e36675f49d 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -39,7 +39,7 @@ jobs: pip list - name: Run benchmarks run: | - python utils/benchmarks.py --weights ${{ matrix.model }}.pt --img 320 + python utils/benchmarks.py --weights ${{ matrix.model }}.pt --img 320 --hard-fail Tests: timeout-minutes: 60 diff --git a/export.py b/export.py index 9daf39f871c2..1d8f07fc9e2f 100644 --- a/export.py +++ b/export.py @@ -75,18 +75,18 @@ def export_formats(): # YOLOv5 export formats x = [ - ['PyTorch', '-', '.pt', True], - ['TorchScript', 'torchscript', '.torchscript', True], - ['ONNX', 'onnx', '.onnx', True], - ['OpenVINO', 'openvino', '_openvino_model', False], - ['TensorRT', 'engine', '.engine', True], - ['CoreML', 'coreml', '.mlmodel', False], - ['TensorFlow SavedModel', 'saved_model', '_saved_model', True], - ['TensorFlow GraphDef', 'pb', '.pb', True], - ['TensorFlow Lite', 'tflite', '.tflite', False], - ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False], - ['TensorFlow.js', 'tfjs', '_web_model', False],] - return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix', 'GPU']) + ['PyTorch', '-', '.pt', True, True], + ['TorchScript', 'torchscript', '.torchscript', True, True], + ['ONNX', 'onnx', '.onnx', True, True], + ['OpenVINO', 'openvino', '_openvino_model', True, False], + ['TensorRT', 'engine', '.engine', False, True], + ['CoreML', 'coreml', '.mlmodel', True, False], + ['TensorFlow SavedModel', 'saved_model', '_saved_model', True, True], + ['TensorFlow GraphDef', 'pb', '.pb', True, True], + ['TensorFlow Lite', 'tflite', '.tflite', True, False], + ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False, False], + ['TensorFlow.js', 'tfjs', '_web_model', False, False],] + return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix', 'CPU', 'GPU']) def export_torchscript(model, im, file, optimize, prefix=colorstr('TorchScript:')): diff --git a/utils/benchmarks.py b/utils/benchmarks.py index 69d653a20916..03bab9b6ded2 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -26,6 +26,7 @@ """ import argparse +import platform import sys import time from pathlib import Path @@ -54,14 +55,17 @@ def run( half=False, # use FP16 half-precision inference test=False, # test exports only pt_only=False, # test PyTorch only + hard_fail=False, # throw error on benchmark failure ): y, t = [], time.time() device = select_device(device) - for i, (name, f, suffix, gpu) in export.export_formats().iterrows(): # index, (name, file, suffix, gpu-capable) + for i, (name, f, suffix, cpu, gpu) in export.export_formats().iterrows(): # index, (name, file, suffix, CPU, GPU) try: - assert i != 9, 'Edge TPU not supported' - assert i != 10, 'TF.js not supported' - if device.type != 'cpu': + assert i not in (9, 10), f'{name} inference not supported' # Edge TPU and TF.js are unsupported + assert i != 5 or platform.system() == 'Darwin', f'{name} inference only supported on macOS>=10.13' + if 'cpu' in device.type: + assert cpu, f'{name} inference not supported on CPU' + if 'cuda' in device.type: assert gpu, f'{name} inference not supported on GPU' # Export @@ -77,6 +81,8 @@ def run( speeds = result[2] # times (preprocess, inference, postprocess) y.append([name, round(file_size(w), 1), round(metrics[3], 4), round(speeds[1], 2)]) # MB, mAP, t_inference except Exception as e: + if hard_fail: + assert type(e) is AssertionError, f'Benchmark --hard-fail for {name}: {e}' LOGGER.warning(f'WARNING: Benchmark failure for {name}: {e}') y.append([name, None, None, None]) # mAP, t_inference if pt_only and i == 0: @@ -102,6 +108,7 @@ def test( half=False, # use FP16 half-precision inference test=False, # test exports only pt_only=False, # test PyTorch only + hard_fail=False, # throw error on benchmark failure ): y, t = [], time.time() device = select_device(device) @@ -134,6 +141,7 @@ def parse_opt(): parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') parser.add_argument('--test', action='store_true', help='test exports only') parser.add_argument('--pt-only', action='store_true', help='test PyTorch only') + parser.add_argument('--hard-fail', action='store_true', help='throw error on benchmark failure') opt = parser.parse_args() opt.data = check_yaml(opt.data) # check YAML print_args(vars(opt)) From f17444abcd647a299f23fe2bf6324b8947cdee22 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 7 Jul 2022 23:46:55 +0200 Subject: [PATCH 347/661] Simplify benchmarks.py assertions (#8515) --- utils/benchmarks.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/utils/benchmarks.py b/utils/benchmarks.py index 03bab9b6ded2..d412653c866f 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -61,12 +61,12 @@ def run( device = select_device(device) for i, (name, f, suffix, cpu, gpu) in export.export_formats().iterrows(): # index, (name, file, suffix, CPU, GPU) try: - assert i not in (9, 10), f'{name} inference not supported' # Edge TPU and TF.js are unsupported - assert i != 5 or platform.system() == 'Darwin', f'{name} inference only supported on macOS>=10.13' + assert i not in (9, 10), 'inference not supported' # Edge TPU and TF.js are unsupported + assert i != 5 or platform.system() == 'Darwin', 'inference only supported on macOS>=10.13' # CoreML if 'cpu' in device.type: - assert cpu, f'{name} inference not supported on CPU' + assert cpu, 'inference not supported on CPU' if 'cuda' in device.type: - assert gpu, f'{name} inference not supported on GPU' + assert gpu, 'inference not supported on GPU' # Export if f == '-': From be42a24d2376d997a98d10433373af84fa85917b Mon Sep 17 00:00:00 2001 From: Colin Wong Date: Thu, 7 Jul 2022 16:53:09 -0500 Subject: [PATCH 348/661] Properly expose `batch_size` from OpenVINO similarly to TensorRT (#8514) Properly expose `batch_size` from OpenVINO Co-authored-by: Glenn Jocher --- models/common.py | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/models/common.py b/models/common.py index a6488dd85648..61e94296b6d0 100644 --- a/models/common.py +++ b/models/common.py @@ -361,11 +361,16 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, elif xml: # OpenVINO LOGGER.info(f'Loading {w} for OpenVINO inference...') check_requirements(('openvino',)) # requires openvino-dev: https://pypi.org/project/openvino-dev/ - from openvino.runtime import Core + from openvino.runtime import Core, Layout, get_batch ie = Core() if not Path(w).is_file(): # if not *.xml w = next(Path(w).glob('*.xml')) # get *.xml file from *_openvino_model dir network = ie.read_model(model=w, weights=Path(w).with_suffix('.bin')) + if network.get_parameters()[0].get_layout().empty: + network.get_parameters()[0].set_layout(Layout("NCHW")) + batch_dim = get_batch(network) + if batch_dim.is_static: + batch_size = batch_dim.get_length() executable_network = ie.compile_model(network, device_name="CPU") # device_name="MYRIAD" for Intel NCS2 output_layer = next(iter(executable_network.outputs)) meta = Path(w).with_suffix('.yaml') From 63ba0cb18a59e882d7e50ba01b934178b0e4bc5a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 8 Jul 2022 00:46:56 +0200 Subject: [PATCH 349/661] Add `--half` arguments to export.py Usage examples (#8516) --- export.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/export.py b/export.py index 1d8f07fc9e2f..ec9024484a3d 100644 --- a/export.py +++ b/export.py @@ -555,11 +555,12 @@ def run( # Finish f = [str(x) for x in f if x] # filter out '' and None if any(f): + h = '--half' if half else '' # --half FP16 inference arg LOGGER.info(f'\nExport complete ({time.time() - t:.2f}s)' f"\nResults saved to {colorstr('bold', file.parent.resolve())}" - f"\nDetect: python detect.py --weights {f[-1]}" + f"\nDetect: python detect.py --weights {f[-1]} {h}" + f"\nValidate: python val.py --weights {f[-1]} {h}" f"\nPyTorch Hub: model = torch.hub.load('ultralytics/yolov5', 'custom', '{f[-1]}')" - f"\nValidate: python val.py --weights {f[-1]}" f"\nVisualize: https://netron.app") return f # return list of exported files/dirs From c215e87393977cc5dd5381a82c63fddb6a8d0428 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 8 Jul 2022 13:49:20 +0200 Subject: [PATCH 350/661] XML export `--half` fix (#8522) Improved error reporting for https://github.com/ultralytics/yolov5/issues/8519 --- export.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/export.py b/export.py index ec9024484a3d..623844ff3531 100644 --- a/export.py +++ b/export.py @@ -484,7 +484,7 @@ def run( # Load PyTorch model device = select_device(device) if half: - assert device.type != 'cpu' or coreml or xml, '--half only compatible with GPU export, i.e. use --device 0' + assert device.type != 'cpu' or coreml, '--half only compatible with GPU export, i.e. use --device 0' assert not dynamic, '--half not compatible with --dynamic, i.e. use either --half or --dynamic but not both' model = attempt_load(weights, device=device, inplace=True, fuse=True) # load FP32 model nc, names = model.nc, model.names # number of classes, class names From 526e650553819dbff67897b9c752c4072e989823 Mon Sep 17 00:00:00 2001 From: Colin Wong Date: Fri, 8 Jul 2022 07:32:40 -0500 Subject: [PATCH 351/661] Fix `LoadImages()` with dataset YAML lists (#8517) * Fix LoadImages with dataset yaml lists * Update dataloaders.py * Update dataloaders.py * Simplify/refactor PR * Update dataloaders.py Co-authored-by: Colin Wong Co-authored-by: Glenn Jocher --- utils/dataloaders.py | 22 ++++++++++++---------- 1 file changed, 12 insertions(+), 10 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 5d4dfc6e4d14..4f1c98fd880d 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -176,15 +176,17 @@ def __iter__(self): class LoadImages: # YOLOv5 image/video dataloader, i.e. `python detect.py --source image.jpg/vid.mp4` def __init__(self, path, img_size=640, stride=32, auto=True): - p = str(Path(path).resolve()) # os-agnostic absolute path - if '*' in p: - files = sorted(glob.glob(p, recursive=True)) # glob - elif os.path.isdir(p): - files = sorted(glob.glob(os.path.join(p, '*.*'))) # dir - elif os.path.isfile(p): - files = [p] # files - else: - raise Exception(f'ERROR: {p} does not exist') + files = [] + for p in sorted(path) if isinstance(path, (list, tuple)) else [path]: + p = str(Path(p).resolve()) + if '*' in p: + files.extend(sorted(glob.glob(p, recursive=True))) # glob + elif os.path.isdir(p): + files.extend(sorted(glob.glob(os.path.join(p, '*.*')))) # dir + elif os.path.isfile(p): + files.append(p) # files + else: + raise FileNotFoundError(f'{p} does not exist') images = [x for x in files if x.split('.')[-1].lower() in IMG_FORMATS] videos = [x for x in files if x.split('.')[-1].lower() in VID_FORMATS] @@ -437,7 +439,7 @@ def __init__(self, f += [x.replace('./', parent) if x.startswith('./') else x for x in t] # local to global path # f += [p.parent / x.lstrip(os.sep) for x in t] # local to global path (pathlib) else: - raise Exception(f'{prefix}{p} does not exist') + raise FileNotFoundError(f'{prefix}{p} does not exist') self.im_files = sorted(x.replace('/', os.sep) for x in f if x.split('.')[-1].lower() in IMG_FORMATS) # self.img_files = sorted([x for x in f if x.suffix[1:].lower() in IMG_FORMATS]) # pathlib assert self.im_files, f'{prefix}No images found' From 7dafd1cb297869032d98406afc9f3e74f68b5bcd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 11 Jul 2022 15:09:42 +0200 Subject: [PATCH 352/661] val.py `assert ncm == nc` fix (#8545) --- val.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/val.py b/val.py index 77f6bbf5b7c2..b0cc8e7f1577 100644 --- a/val.py +++ b/val.py @@ -164,7 +164,7 @@ def run( if not training: if pt and not single_cls: # check --weights are trained on --data ncm = model.model.nc - assert ncm == nc, f'{weights[0]} ({ncm} classes) trained on different --data than what you passed ({nc} ' \ + assert ncm == nc, f'{weights} ({ncm} classes) trained on different --data than what you passed ({nc} ' \ f'classes). Pass correct combination of --weights and --data that are trained together.' model.warmup(imgsz=(1 if pt else batch_size, 3, imgsz, imgsz)) # warmup pad = 0.0 if task in ('speed', 'benchmark') else 0.5 From a84cd02387d70fb5a6287682a221e8cd46dca87a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 11 Jul 2022 16:07:11 +0200 Subject: [PATCH 353/661] CIoU protected divides (#8546) Protected divides in IOU function to resolve https://github.com/ultralytics/yolov5/issues/8539 --- utils/metrics.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/utils/metrics.py b/utils/metrics.py index e17747b703fa..858af23efadb 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -225,8 +225,8 @@ def bbox_iou(box1, box2, xywh=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7 else: # x1, y1, x2, y2 = box1 b1_x1, b1_y1, b1_x2, b1_y2 = box1.chunk(4, 1) b2_x1, b2_y1, b2_x2, b2_y2 = box2.chunk(4, 1) - w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps - w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps + w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 # Intersection area inter = (torch.min(b1_x2, b2_x2) - torch.max(b1_x1, b2_x1)).clamp(0) * \ @@ -244,7 +244,7 @@ def bbox_iou(box1, box2, xywh=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7 c2 = cw ** 2 + ch ** 2 + eps # convex diagonal squared rho2 = ((b2_x1 + b2_x2 - b1_x1 - b1_x2) ** 2 + (b2_y1 + b2_y2 - b1_y1 - b1_y2) ** 2) / 4 # center dist ** 2 if CIoU: # https://github.com/Zzh-tju/DIoU-SSD-pytorch/blob/master/utils/box/box_utils.py#L47 - v = (4 / math.pi ** 2) * torch.pow(torch.atan(w2 / h2) - torch.atan(w1 / h1), 2) + v = (4 / math.pi ** 2) * torch.pow(torch.atan(w2 / (h2 + eps)) - torch.atan(w1 / (h1 + eps)), 2) with torch.no_grad(): alpha = v / (v - iou + (1 + eps)) return iou - (rho2 / c2 + v * alpha) # CIoU From 04146371b9940e144080430eb5e28b828d2f9c3a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 12 Jul 2022 01:58:25 +0200 Subject: [PATCH 354/661] Update metrics.py with IoU protected divides (#8550) --- utils/metrics.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/utils/metrics.py b/utils/metrics.py index 858af23efadb..6bba4cfe2a42 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -259,7 +259,7 @@ def box_area(box): return (box[2] - box[0]) * (box[3] - box[1]) -def box_iou(box1, box2): +def box_iou(box1, box2, eps=1e-7): # https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py """ Return intersection-over-union (Jaccard index) of boxes. @@ -277,10 +277,10 @@ def box_iou(box1, box2): inter = (torch.min(a2, b2) - torch.max(a1, b1)).clamp(0).prod(2) # IoU = inter / (area1 + area2 - inter) - return inter / (box_area(box1.T)[:, None] + box_area(box2.T) - inter) + return inter / (box_area(box1.T)[:, None] + box_area(box2.T) - inter + eps) -def bbox_ioa(box1, box2, eps=1E-7): +def bbox_ioa(box1, box2, eps=1e-7): """ Returns the intersection over box2 area given box1, box2. Boxes are x1y1x2y2 box1: np.array of shape(4) box2: np.array of shape(nx4) @@ -302,12 +302,12 @@ def bbox_ioa(box1, box2, eps=1E-7): return inter_area / box2_area -def wh_iou(wh1, wh2): +def wh_iou(wh1, wh2, eps=1e-7): # Returns the nxm IoU matrix. wh1 is nx2, wh2 is mx2 wh1 = wh1[:, None] # [N,1,2] wh2 = wh2[None] # [1,M,2] inter = torch.min(wh1, wh2).prod(2) # [N,M] - return inter / (wh1.prod(2) + wh2.prod(2) - inter) # iou = inter / (area1 + area2 - inter) + return inter / (wh1.prod(2) + wh2.prod(2) - inter + eps) # iou = inter / (area1 + area2 - inter) # Plots ---------------------------------------------------------------------------------------------------------------- From fbd30205257d956f6c9840e9e9863e4bb7e1f3aa Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E6=9B=BE=E9=80=B8=E5=A4=AB=EF=BC=88Zeng=20Yifu=EF=BC=89?= <41098760+Zengyf-CVer@users.noreply.github.com> Date: Tue, 12 Jul 2022 19:19:25 +0800 Subject: [PATCH 355/661] Add TensorRT dependencies (#8553) Update requirements.txt --- requirements.txt | 2 ++ 1 file changed, 2 insertions(+) diff --git a/requirements.txt b/requirements.txt index ad3fd49691d4..931f93646b73 100644 --- a/requirements.txt +++ b/requirements.txt @@ -26,6 +26,8 @@ seaborn>=0.11.0 # coremltools>=4.1 # CoreML export # onnx>=1.9.0 # ONNX export # onnx-simplifier>=0.3.6 # ONNX simplifier +# nvidia-pyindex # TensorRT export +# nvidia-tensorrt # TensorRT export # scikit-learn==0.19.2 # CoreML quantization # tensorflow>=2.4.1 # TFLite export # tensorflowjs>=3.9.0 # TF.js export From 574ceedfc5f171a89417175bfb14fda6a2646603 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 12 Jul 2022 14:49:54 +0200 Subject: [PATCH 356/661] Add `thop>=0.1.0` (#8558) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 931f93646b73..4a4f68539cad 100644 --- a/requirements.txt +++ b/requirements.txt @@ -36,7 +36,7 @@ seaborn>=0.11.0 # Extras -------------------------------------- ipython # interactive notebook psutil # system utilization -thop # FLOPs computation +thop>=0.1.0 # FLOPs computation # albumentations>=1.0.3 # pycocotools>=2.0 # COCO mAP # roboflow From f8722b4429e80f96be04b36e4efd84ce6583bfa1 Mon Sep 17 00:00:00 2001 From: Colin Wong Date: Wed, 13 Jul 2022 04:13:01 -0500 Subject: [PATCH 357/661] Raise error on suffix-less model path (#8561) Raise error on invalid model --- models/common.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/models/common.py b/models/common.py index 61e94296b6d0..fb5ac3a6f5a4 100644 --- a/models/common.py +++ b/models/common.py @@ -441,6 +441,8 @@ def wrap_frozen_graph(gd, inputs, outputs): output_details = interpreter.get_output_details() # outputs elif tfjs: raise Exception('ERROR: YOLOv5 TF.js inference is not supported') + else: + raise Exception(f'ERROR: {w} is not a supported format') self.__dict__.update(locals()) # assign all variables to self def forward(self, im, augment=False, visualize=False, val=False): From f4b05680f89795658e1c898a28ff51edbf22a63b Mon Sep 17 00:00:00 2001 From: Colin Wong Date: Fri, 15 Jul 2022 09:01:01 -0500 Subject: [PATCH 358/661] Assert `--optimize` not used with cuda device (#8569) --- export.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/export.py b/export.py index 623844ff3531..9868fcae95c3 100644 --- a/export.py +++ b/export.py @@ -492,6 +492,8 @@ def run( # Checks imgsz *= 2 if len(imgsz) == 1 else 1 # expand assert nc == len(names), f'Model class count {nc} != len(names) {len(names)}' + if optimize: + assert device.type != 'cuda', '--optimize not compatible with cuda devices, i.e. use --device cpu' # Input gs = int(max(model.stride)) # grid size (max stride) From 72a81e7a1c13cd3ae4675037f217d0ed3db9bc20 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 15 Jul 2022 16:01:29 +0200 Subject: [PATCH 359/661] Update requirements.txt comment spacing (#8562) --- requirements.txt | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/requirements.txt b/requirements.txt index 4a4f68539cad..c0f12ccdd018 100644 --- a/requirements.txt +++ b/requirements.txt @@ -10,7 +10,7 @@ PyYAML>=5.3.1 requests>=2.23.0 scipy>=1.4.1 # Google Colab version torch>=1.7.0,!=1.12.0 # https://github.com/ultralytics/yolov5/issues/8395 -torchvision>=0.8.1,!=0.13.0 # https://github.com/ultralytics/yolov5/issues/8395 +torchvision>=0.8.1,!=0.13.0 # https://github.com/ultralytics/yolov5/issues/8395 tqdm>=4.41.0 protobuf<4.21.3 # https://github.com/ultralytics/yolov5/issues/8012 @@ -26,8 +26,8 @@ seaborn>=0.11.0 # coremltools>=4.1 # CoreML export # onnx>=1.9.0 # ONNX export # onnx-simplifier>=0.3.6 # ONNX simplifier -# nvidia-pyindex # TensorRT export -# nvidia-tensorrt # TensorRT export +# nvidia-pyindex # TensorRT export +# nvidia-tensorrt # TensorRT export # scikit-learn==0.19.2 # CoreML quantization # tensorflow>=2.4.1 # TFLite export # tensorflowjs>=3.9.0 # TF.js export From 7204c1ca25fa69a911802edab719b4cc323103f4 Mon Sep 17 00:00:00 2001 From: Yonghye Kwon Date: Sat, 16 Jul 2022 22:51:48 +0900 Subject: [PATCH 360/661] Explicitly set `weight_decay` value (#8592) * explicitly set weight_decay value The default weight_decay value of AdamW is 1e-2, so we should set it to zero. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Cleanup Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- train.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/train.py b/train.py index bf5b4c69d74c..ff13f1e256ec 100644 --- a/train.py +++ b/train.py @@ -163,12 +163,12 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio if opt.optimizer == 'Adam': optimizer = Adam(g[2], lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum elif opt.optimizer == 'AdamW': - optimizer = AdamW(g[2], lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum + optimizer = AdamW(g[2], lr=hyp['lr0'], betas=(hyp['momentum'], 0.999), weight_decay=0.0) else: optimizer = SGD(g[2], lr=hyp['lr0'], momentum=hyp['momentum'], nesterov=True) optimizer.add_param_group({'params': g[0], 'weight_decay': hyp['weight_decay']}) # add g0 with weight_decay - optimizer.add_param_group({'params': g[1]}) # add g1 (BatchNorm2d weights) + optimizer.add_param_group({'params': g[1], 'weight_decay': 0.0}) # add g1 (BatchNorm2d weights) LOGGER.info(f"{colorstr('optimizer:')} {type(optimizer).__name__} with parameter groups " f"{len(g[1])} weight (no decay), {len(g[0])} weight, {len(g[2])} bias") del g From cf28dda3660fcda0bac56a9ca75ca3c8749d1baf Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 16 Jul 2022 15:54:34 +0200 Subject: [PATCH 361/661] Update `scipy>=1.7.3` (#8595) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index c0f12ccdd018..f5ae6175b6f1 100644 --- a/requirements.txt +++ b/requirements.txt @@ -8,7 +8,7 @@ opencv-python>=4.1.1 Pillow>=7.1.2 PyYAML>=5.3.1 requests>=2.23.0 -scipy>=1.4.1 # Google Colab version +scipy>=1.7.3 # Google Colab version torch>=1.7.0,!=1.12.0 # https://github.com/ultralytics/yolov5/issues/8395 torchvision>=0.8.1,!=0.13.0 # https://github.com/ultralytics/yolov5/issues/8395 tqdm>=4.41.0 From 5c45a4b13d1782a8ad9cb993a1d22430540bd197 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 16 Jul 2022 16:14:16 +0200 Subject: [PATCH 362/661] Update `tqdm>=4.64.0` and `thop>=0.1.1` (#8596) * Update `tqdm>=4.64.0` and `thop>=0.1.1` * Update requirements.txt --- requirements.txt | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/requirements.txt b/requirements.txt index f5ae6175b6f1..4550fc771b04 100644 --- a/requirements.txt +++ b/requirements.txt @@ -8,10 +8,10 @@ opencv-python>=4.1.1 Pillow>=7.1.2 PyYAML>=5.3.1 requests>=2.23.0 -scipy>=1.7.3 # Google Colab version +scipy>=1.4.1 torch>=1.7.0,!=1.12.0 # https://github.com/ultralytics/yolov5/issues/8395 torchvision>=0.8.1,!=0.13.0 # https://github.com/ultralytics/yolov5/issues/8395 -tqdm>=4.41.0 +tqdm>=4.64.0 protobuf<4.21.3 # https://github.com/ultralytics/yolov5/issues/8012 # Logging ------------------------------------- @@ -36,7 +36,7 @@ seaborn>=0.11.0 # Extras -------------------------------------- ipython # interactive notebook psutil # system utilization -thop>=0.1.0 # FLOPs computation +thop>=0.1.1 # FLOPs computation # albumentations>=1.0.3 # pycocotools>=2.0 # COCO mAP # roboflow From 6e86af3de85c449fa2574c2461d8919d86620e6c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 16 Jul 2022 16:41:43 +0200 Subject: [PATCH 363/661] Add `pip install wheel` to avoid legacy `setup.py install` (#8597) * Update ci-testing with `pip install wheel` * Update ci-testing.yml * Update dockerfiles --- .github/workflows/ci-testing.yml | 4 ++-- utils/docker/Dockerfile | 2 +- utils/docker/Dockerfile-arm64 | 2 +- utils/docker/Dockerfile-cpu | 2 +- 4 files changed, 5 insertions(+), 5 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index f3e36675f49d..e3359cd3a283 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -32,7 +32,7 @@ jobs: # restore-keys: ${{ runner.os }}-Benchmarks- - name: Install requirements run: | - python -m pip install --upgrade pip + python -m pip install --upgrade pip wheel pip install -r requirements.txt coremltools openvino-dev tensorflow-cpu --extra-index-url https://download.pytorch.org/whl/cpu python --version pip --version @@ -77,7 +77,7 @@ jobs: restore-keys: ${{ runner.os }}-${{ matrix.python-version }}-pip- - name: Install requirements run: | - python -m pip install --upgrade pip + python -m pip install --upgrade pip wheel pip install -r requirements.txt --extra-index-url https://download.pytorch.org/whl/cpu python --version pip --version diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index a5fc7cbd6c45..1a4b66b106b2 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -14,7 +14,7 @@ RUN apt update && apt install --no-install-recommends -y zip htop screen libgl1- # Install pip packages COPY requirements.txt . -RUN python -m pip install --upgrade pip +RUN python -m pip install --upgrade pip wheel RUN pip uninstall -y torch torchvision torchtext Pillow RUN pip install --no-cache -r requirements.txt albumentations wandb gsutil notebook Pillow>=9.1.0 \ 'opencv-python<4.6.0.66' \ diff --git a/utils/docker/Dockerfile-arm64 b/utils/docker/Dockerfile-arm64 index 2e261051dedd..bca161e67a37 100644 --- a/utils/docker/Dockerfile-arm64 +++ b/utils/docker/Dockerfile-arm64 @@ -17,7 +17,7 @@ RUN apt install --no-install-recommends -y python3-pip git zip curl htop gcc \ # Install pip packages COPY requirements.txt . -RUN python3 -m pip install --upgrade pip +RUN python3 -m pip install --upgrade pip wheel RUN pip install --no-cache -r requirements.txt gsutil notebook \ tensorflow-aarch64 # tensorflowjs \ diff --git a/utils/docker/Dockerfile-cpu b/utils/docker/Dockerfile-cpu index c8aa8c6a48c6..f05e920ad53f 100644 --- a/utils/docker/Dockerfile-cpu +++ b/utils/docker/Dockerfile-cpu @@ -16,7 +16,7 @@ RUN apt install --no-install-recommends -y python3-pip git zip curl htop libgl1- # Install pip packages COPY requirements.txt . -RUN python3 -m pip install --upgrade pip +RUN python3 -m pip install --upgrade pip wheel RUN pip install --no-cache -r requirements.txt albumentations gsutil notebook \ coremltools onnx onnx-simplifier onnxruntime openvino-dev tensorflow-cpu tensorflowjs \ --extra-index-url https://download.pytorch.org/whl/cpu From a34b376d0fb90076e698b1b95df55c9cafba899a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 16 Jul 2022 23:46:23 +0200 Subject: [PATCH 364/661] Link fuse() to AutoShape() for Hub models (#8599) --- hubconf.py | 3 +-- models/common.py | 4 ++-- 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/hubconf.py b/hubconf.py index df585f8cb411..6bb9484a856d 100644 --- a/hubconf.py +++ b/hubconf.py @@ -36,7 +36,6 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo if not verbose: LOGGER.setLevel(logging.WARNING) - check_requirements(exclude=('tensorboard', 'thop', 'opencv-python')) name = Path(name) path = name.with_suffix('.pt') if name.suffix == '' and not name.is_dir() else name # checkpoint path @@ -44,7 +43,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo device = select_device(device) if pretrained and channels == 3 and classes == 80: - model = DetectMultiBackend(path, device=device) # download/load FP32 model + model = DetectMultiBackend(path, device=device, fuse=autoshape) # download/load FP32 model # model = models.experimental.attempt_load(path, map_location=device) # download/load FP32 model else: cfg = list((Path(__file__).parent / 'models').rglob(f'{path.stem}.yaml'))[0] # model.yaml path diff --git a/models/common.py b/models/common.py index fb5ac3a6f5a4..5ea1c307f034 100644 --- a/models/common.py +++ b/models/common.py @@ -305,7 +305,7 @@ def forward(self, x): class DetectMultiBackend(nn.Module): # YOLOv5 MultiBackend class for python inference on various backends - def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, data=None, fp16=False): + def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, data=None, fp16=False, fuse=True): # Usage: # PyTorch: weights = *.pt # TorchScript: *.torchscript @@ -331,7 +331,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, names = yaml.safe_load(f)['names'] if pt: # PyTorch - model = attempt_load(weights if isinstance(weights, list) else w, device=device) + model = attempt_load(weights if isinstance(weights, list) else w, device=device, inplace=True, fuse=fuse) stride = max(int(model.stride.max()), 32) # model stride names = model.module.names if hasattr(model, 'module') else model.names # get class names model.half() if fp16 else model.float() From 24305787ae32b7e04f52a971a5865c461842662e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 17 Jul 2022 00:55:55 +0200 Subject: [PATCH 365/661] FROM nvcr.io/nvidia/pytorch:22.06-py3 (#8600) --- utils/docker/Dockerfile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index 1a4b66b106b2..312d169d1a76 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -3,7 +3,7 @@ # Image is CUDA-optimized for YOLOv5 single/multi-GPU training and inference # Start FROM NVIDIA PyTorch image https://ngc.nvidia.com/catalog/containers/nvidia:pytorch -FROM nvcr.io/nvidia/pytorch:22.05-py3 +FROM nvcr.io/nvidia/pytorch:22.06-py3 RUN rm -rf /opt/pytorch # remove 1.2GB dir # Downloads to user config dir @@ -15,7 +15,7 @@ RUN apt update && apt install --no-install-recommends -y zip htop screen libgl1- # Install pip packages COPY requirements.txt . RUN python -m pip install --upgrade pip wheel -RUN pip uninstall -y torch torchvision torchtext Pillow +RUN pip uninstall -y Pillow torchtext # torch torchvision RUN pip install --no-cache -r requirements.txt albumentations wandb gsutil notebook Pillow>=9.1.0 \ 'opencv-python<4.6.0.66' \ --extra-index-url https://download.pytorch.org/whl/cu113 From 51fb467b63191b5f0ff8391608bb96b5deb8c3ea Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 17 Jul 2022 11:43:52 +0200 Subject: [PATCH 366/661] Refactor optimizer initialization (#8607) * Refactor optimizer initialization * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update train.py * Update train.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- train.py | 29 ++++------------------------- utils/torch_utils.py | 32 +++++++++++++++++++++++++++++++- 2 files changed, 35 insertions(+), 26 deletions(-) diff --git a/train.py b/train.py index ff13f1e256ec..6b463bf56423 100644 --- a/train.py +++ b/train.py @@ -28,7 +28,7 @@ import torch.nn as nn import yaml from torch.nn.parallel import DistributedDataParallel as DDP -from torch.optim import SGD, Adam, AdamW, lr_scheduler +from torch.optim import lr_scheduler from tqdm import tqdm FILE = Path(__file__).resolve() @@ -54,7 +54,8 @@ from utils.loss import ComputeLoss from utils.metrics import fitness from utils.plots import plot_evolve, plot_labels -from utils.torch_utils import EarlyStopping, ModelEMA, de_parallel, select_device, torch_distributed_zero_first +from utils.torch_utils import (EarlyStopping, ModelEMA, de_parallel, select_device, smart_optimizer, + torch_distributed_zero_first) LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html RANK = int(os.getenv('RANK', -1)) @@ -149,29 +150,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio accumulate = max(round(nbs / batch_size), 1) # accumulate loss before optimizing hyp['weight_decay'] *= batch_size * accumulate / nbs # scale weight_decay LOGGER.info(f"Scaled weight_decay = {hyp['weight_decay']}") - - g = [], [], [] # optimizer parameter groups - bn = tuple(v for k, v in nn.__dict__.items() if 'Norm' in k) # normalization layers, i.e. BatchNorm2d() - for v in model.modules(): - if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter): # bias - g[2].append(v.bias) - if isinstance(v, bn): # weight (no decay) - g[1].append(v.weight) - elif hasattr(v, 'weight') and isinstance(v.weight, nn.Parameter): # weight (with decay) - g[0].append(v.weight) - - if opt.optimizer == 'Adam': - optimizer = Adam(g[2], lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum - elif opt.optimizer == 'AdamW': - optimizer = AdamW(g[2], lr=hyp['lr0'], betas=(hyp['momentum'], 0.999), weight_decay=0.0) - else: - optimizer = SGD(g[2], lr=hyp['lr0'], momentum=hyp['momentum'], nesterov=True) - - optimizer.add_param_group({'params': g[0], 'weight_decay': hyp['weight_decay']}) # add g0 with weight_decay - optimizer.add_param_group({'params': g[1], 'weight_decay': 0.0}) # add g1 (BatchNorm2d weights) - LOGGER.info(f"{colorstr('optimizer:')} {type(optimizer).__name__} with parameter groups " - f"{len(g[1])} weight (no decay), {len(g[0])} weight, {len(g[2])} bias") - del g + optimizer = smart_optimizer(model, opt.optimizer, hyp['lr0'], hyp['momentum'], hyp['weight_decay']) # Scheduler if opt.cos_lr: diff --git a/utils/torch_utils.py b/utils/torch_utils.py index c21dc6658c1e..d82368dc6271 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -18,7 +18,7 @@ import torch.nn as nn import torch.nn.functional as F -from utils.general import LOGGER, file_date, git_describe +from utils.general import LOGGER, colorstr, file_date, git_describe try: import thop # for FLOPs computation @@ -260,6 +260,36 @@ def copy_attr(a, b, include=(), exclude=()): setattr(a, k, v) +def smart_optimizer(model, name='Adam', lr=0.001, momentum=0.9, weight_decay=1e-5): + # YOLOv5 3-param group optimizer: 0) weights with decay, 1) weights no decay, 2) biases no decay + g = [], [], [] # optimizer parameter groups + bn = tuple(v for k, v in nn.__dict__.items() if 'Norm' in k) # normalization layers, i.e. BatchNorm2d() + for v in model.modules(): + if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter): # bias (no decay) + g[2].append(v.bias) + if isinstance(v, bn): # weight (no decay) + g[1].append(v.weight) + elif hasattr(v, 'weight') and isinstance(v.weight, nn.Parameter): # weight (with decay) + g[0].append(v.weight) + + if name == 'Adam': + optimizer = torch.optim.Adam(g[2], lr=lr, betas=(momentum, 0.999)) # adjust beta1 to momentum + elif name == 'AdamW': + optimizer = torch.optim.AdamW(g[2], lr=lr, betas=(momentum, 0.999), weight_decay=0.0) + elif name == 'RMSProp': + optimizer = torch.optim.RMSprop(g[2], lr=lr, momentum=momentum) + elif name == 'SGD': + optimizer = torch.optim.SGD(g[2], lr=lr, momentum=momentum, nesterov=True) + else: + raise NotImplementedError(f'Optimizer {name} not implemented.') + + optimizer.add_param_group({'params': g[0], 'weight_decay': weight_decay}) # add g0 with weight_decay + optimizer.add_param_group({'params': g[1], 'weight_decay': 0.0}) # add g1 (BatchNorm2d weights) + LOGGER.info(f"{colorstr('optimizer:')} {type(optimizer).__name__} with parameter groups " + f"{len(g[1])} weight (no decay), {len(g[0])} weight, {len(g[2])} bias") + return optimizer + + class EarlyStopping: # YOLOv5 simple early stopper def __init__(self, patience=30): From 9cf5fd5ac33c096ae06f60667dd6582ddb84aa4c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 18 Jul 2022 15:05:58 +0200 Subject: [PATCH 367/661] assert torch!=1.12.0 for DDP training (#8621) * assert torch!=1.12.0 for DDP training * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- requirements.txt | 4 ++-- train.py | 14 +++++--------- utils/torch_utils.py | 18 +++++++++++++++++- 3 files changed, 24 insertions(+), 12 deletions(-) diff --git a/requirements.txt b/requirements.txt index 4550fc771b04..a3284d6529eb 100644 --- a/requirements.txt +++ b/requirements.txt @@ -9,8 +9,8 @@ Pillow>=7.1.2 PyYAML>=5.3.1 requests>=2.23.0 scipy>=1.4.1 -torch>=1.7.0,!=1.12.0 # https://github.com/ultralytics/yolov5/issues/8395 -torchvision>=0.8.1,!=0.13.0 # https://github.com/ultralytics/yolov5/issues/8395 +torch>=1.7.0 +torchvision>=0.8.1 tqdm>=4.64.0 protobuf<4.21.3 # https://github.com/ultralytics/yolov5/issues/8012 diff --git a/train.py b/train.py index 6b463bf56423..c298692b7335 100644 --- a/train.py +++ b/train.py @@ -27,7 +27,6 @@ import torch.distributed as dist import torch.nn as nn import yaml -from torch.nn.parallel import DistributedDataParallel as DDP from torch.optim import lr_scheduler from tqdm import tqdm @@ -46,15 +45,15 @@ from utils.dataloaders import create_dataloader from utils.downloads import attempt_download from utils.general import (LOGGER, check_amp, check_dataset, check_file, check_git_status, check_img_size, - check_requirements, check_suffix, check_version, check_yaml, colorstr, get_latest_run, - increment_path, init_seeds, intersect_dicts, labels_to_class_weights, - labels_to_image_weights, methods, one_cycle, print_args, print_mutation, strip_optimizer) + check_requirements, check_suffix, check_yaml, colorstr, get_latest_run, increment_path, + init_seeds, intersect_dicts, labels_to_class_weights, labels_to_image_weights, methods, + one_cycle, print_args, print_mutation, strip_optimizer) from utils.loggers import Loggers from utils.loggers.wandb.wandb_utils import check_wandb_resume from utils.loss import ComputeLoss from utils.metrics import fitness from utils.plots import plot_evolve, plot_labels -from utils.torch_utils import (EarlyStopping, ModelEMA, de_parallel, select_device, smart_optimizer, +from utils.torch_utils import (EarlyStopping, ModelEMA, de_parallel, select_device, smart_DDP, smart_optimizer, torch_distributed_zero_first) LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html @@ -248,10 +247,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # DDP mode if cuda and RANK != -1: - if check_version(torch.__version__, '1.11.0'): - model = DDP(model, device_ids=[LOCAL_RANK], output_device=LOCAL_RANK, static_graph=True) - else: - model = DDP(model, device_ids=[LOCAL_RANK], output_device=LOCAL_RANK) + model = smart_DDP(model) # Model attributes nl = de_parallel(model).model[-1].nl # number of detection layers (to scale hyps) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index d82368dc6271..5f2a22c36f1a 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -17,8 +17,13 @@ import torch.distributed as dist import torch.nn as nn import torch.nn.functional as F +from torch.nn.parallel import DistributedDataParallel as DDP -from utils.general import LOGGER, colorstr, file_date, git_describe +from utils.general import LOGGER, check_version, colorstr, file_date, git_describe + +LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html +RANK = int(os.getenv('RANK', -1)) +WORLD_SIZE = int(os.getenv('WORLD_SIZE', 1)) try: import thop # for FLOPs computation @@ -29,6 +34,17 @@ warnings.filterwarnings('ignore', message='User provided device_type of \'cuda\', but CUDA is not available. Disabling') +def smart_DDP(model): + # Model DDP creation with checks + assert not check_version(torch.__version__, '1.12.0', pinned=True), \ + 'torch==1.12.0 torchvision==0.13.0 DDP training is not supported due to a known issue. ' \ + 'Please upgrade or downgrade torch to use DDP. See https://github.com/ultralytics/yolov5/issues/8395' + if check_version(torch.__version__, '1.11.0'): + return DDP(model, device_ids=[LOCAL_RANK], output_device=LOCAL_RANK, static_graph=True) + else: + return DDP(model, device_ids=[LOCAL_RANK], output_device=LOCAL_RANK) + + @contextmanager def torch_distributed_zero_first(local_rank: int): # Decorator to make all processes in distributed training wait for each local_master to do something From fbe67e465375231474a2ad80a4389efc77ecff99 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 18 Jul 2022 17:53:30 +0200 Subject: [PATCH 368/661] Fix `OMP_NUM_THREADS=1` for macOS (#8624) Resolves https://github.com/ultralytics/yolov5/issues/8623 --- utils/general.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index a85a2915a31a..cb5ca500b9f3 100755 --- a/utils/general.py +++ b/utils/general.py @@ -52,7 +52,7 @@ pd.options.display.max_columns = 10 cv2.setNumThreads(0) # prevent OpenCV from multithreading (incompatible with PyTorch DataLoader) os.environ['NUMEXPR_MAX_THREADS'] = str(NUM_THREADS) # NumExpr max threads -os.environ['OMP_NUM_THREADS'] = str(NUM_THREADS) # OpenMP max threads (PyTorch and SciPy) +os.environ['OMP_NUM_THREADS'] = '1' if platform.system() == 'darwin' else str(NUM_THREADS) # OpenMP (PyTorch and SciPy) def is_kaggle(): From 92e47b85d952274480c8c5efa5900e686241a96b Mon Sep 17 00:00:00 2001 From: daquexian Date: Wed, 20 Jul 2022 01:01:24 +0800 Subject: [PATCH 369/661] Upgrade onnxsim to v0.4.1 (#8632) * upgrade onnxsim to v0.4.1 Signed-off-by: daquexian * Update export.py * Update export.py * Update export.py * Update export.py * Update export.py Co-authored-by: Glenn Jocher --- export.py | 9 ++++----- requirements.txt | 2 +- 2 files changed, 5 insertions(+), 6 deletions(-) diff --git a/export.py b/export.py index 9868fcae95c3..3629915f028d 100644 --- a/export.py +++ b/export.py @@ -152,13 +152,12 @@ def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorst # Simplify if simplify: try: - check_requirements(('onnx-simplifier',)) + cuda = torch.cuda.is_available() + check_requirements(('onnxruntime-gpu' if cuda else 'onnxruntime', 'onnx-simplifier>=0.4.1')) import onnxsim LOGGER.info(f'{prefix} simplifying with onnx-simplifier {onnxsim.__version__}...') - model_onnx, check = onnxsim.simplify(model_onnx, - dynamic_input_shape=dynamic, - input_shapes={'images': list(im.shape)} if dynamic else None) + model_onnx, check = onnxsim.simplify(model_onnx) assert check, 'assert check failed' onnx.save(model_onnx, f) except Exception as e: @@ -493,7 +492,7 @@ def run( imgsz *= 2 if len(imgsz) == 1 else 1 # expand assert nc == len(names), f'Model class count {nc} != len(names) {len(names)}' if optimize: - assert device.type != 'cuda', '--optimize not compatible with cuda devices, i.e. use --device cpu' + assert device.type == 'cpu', '--optimize not compatible with cuda devices, i.e. use --device cpu' # Input gs = int(max(model.stride)) # grid size (max stride) diff --git a/requirements.txt b/requirements.txt index a3284d6529eb..8548f67b5a48 100644 --- a/requirements.txt +++ b/requirements.txt @@ -25,7 +25,7 @@ seaborn>=0.11.0 # Export -------------------------------------- # coremltools>=4.1 # CoreML export # onnx>=1.9.0 # ONNX export -# onnx-simplifier>=0.3.6 # ONNX simplifier +# onnx-simplifier>=0.4.1 # ONNX simplifier # nvidia-pyindex # TensorRT export # nvidia-tensorrt # TensorRT export # scikit-learn==0.19.2 # CoreML quantization From 602d7ffb0e8667c63bd0007ecf3cfd29a46f9cc4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?R=C3=BCdiger=20Busche?= Date: Thu, 21 Jul 2022 17:40:53 +0200 Subject: [PATCH 370/661] Check TensorBoard logger before adding graph (#8664) Otherwise, an error is thrown if the tensorboard logger is not included. --- utils/loggers/__init__.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 42b696ba644f..88bdb0521619 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -102,7 +102,7 @@ def on_train_batch_end(self, ni, model, imgs, targets, paths, plots): # Callback runs on train batch end if plots: if ni == 0: - if not self.opt.sync_bn: # --sync known issue https://github.com/ultralytics/yolov5/issues/3754 + if self.tb and not self.opt.sync_bn: # --sync known issue https://github.com/ultralytics/yolov5/issues/3754 with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress jit trace warning self.tb.add_graph(torch.jit.trace(de_parallel(model), imgs[0:1], strict=False), []) From 4c1784bd158d3215aa7170b33578e1032442a160 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 21 Jul 2022 23:12:49 +0200 Subject: [PATCH 371/661] Use contextlib's suppress method to silence an error (#8668) --- models/yolo.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index 02660e6c4130..56846815e08a 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -7,6 +7,7 @@ """ import argparse +import contextlib import os import platform import sys @@ -259,10 +260,8 @@ def parse_model(d, ch): # model_dict, input_channels(3) for i, (f, n, m, args) in enumerate(d['backbone'] + d['head']): # from, number, module, args m = eval(m) if isinstance(m, str) else m # eval strings for j, a in enumerate(args): - try: + with contextlib.suppress(NameError): args[j] = eval(a) if isinstance(a, str) else a # eval strings - except NameError: - pass n = n_ = max(round(n * gd), 1) if n > 1 else n # depth gain if m in (Conv, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv, From 38721de7ef6923f52c1ce1eb00a765a447c27d3c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 22 Jul 2022 11:54:31 +0200 Subject: [PATCH 372/661] Update hubconf.py to reset LOGGER.level after load (#8674) Resolves silent outputs after model load --- hubconf.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/hubconf.py b/hubconf.py index 6bb9484a856d..8748279e027a 100644 --- a/hubconf.py +++ b/hubconf.py @@ -34,6 +34,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo from utils.general import LOGGER, check_requirements, intersect_dicts, logging from utils.torch_utils import select_device + level = LOGGER.level if not verbose: LOGGER.setLevel(logging.WARNING) check_requirements(exclude=('tensorboard', 'thop', 'opencv-python')) @@ -57,6 +58,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo model.names = ckpt['model'].names # set class names attribute if autoshape: model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS + LOGGER.setLevel(level) return model.to(device) except Exception as e: From b17629e54f5a392c8e32219ba03b06b7eb11a48a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 22 Jul 2022 15:23:22 +0200 Subject: [PATCH 373/661] Update warning emojis (#8678) --- utils/general.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/general.py b/utils/general.py index cb5ca500b9f3..925f7fbf0ecb 100755 --- a/utils/general.py +++ b/utils/general.py @@ -474,7 +474,7 @@ def check_dataset(data, autodownload=True): for k in 'train', 'val', 'nc': assert k in data, emojis(f"data.yaml '{k}:' field missing ❌") if 'names' not in data: - LOGGER.warning(emojis("data.yaml 'names:' field missing ⚠, assigning default names 'class0', 'class1', etc.")) + LOGGER.warning(emojis("data.yaml 'names:' field missing ⚠️, assigning default names 'class0', 'class1', etc.")) data['names'] = [f'class{i}' for i in range(data['nc'])] # default names # Resolve paths @@ -490,7 +490,7 @@ def check_dataset(data, autodownload=True): if val: val = [Path(x).resolve() for x in (val if isinstance(val, list) else [val])] # val path if not all(x.exists() for x in val): - LOGGER.info(emojis('\nDataset not found ⚠, missing paths %s' % [str(x) for x in val if not x.exists()])) + LOGGER.info(emojis('\nDataset not found ⚠️, missing paths %s' % [str(x) for x in val if not x.exists()])) if not s or not autodownload: raise Exception(emojis('Dataset not found ❌')) t = time.time() From b92430a83bfe11dd3be74e486c37b836be46bc98 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 22 Jul 2022 19:01:16 +0200 Subject: [PATCH 374/661] Update hubconf.py to reset logging level to INFO (#8680) --- hubconf.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/hubconf.py b/hubconf.py index 8748279e027a..25f9d1b82c14 100644 --- a/hubconf.py +++ b/hubconf.py @@ -34,7 +34,6 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo from utils.general import LOGGER, check_requirements, intersect_dicts, logging from utils.torch_utils import select_device - level = LOGGER.level if not verbose: LOGGER.setLevel(logging.WARNING) check_requirements(exclude=('tensorboard', 'thop', 'opencv-python')) @@ -58,7 +57,8 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo model.names = ckpt['model'].names # set class names attribute if autoshape: model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS - LOGGER.setLevel(level) + if not verbose: + LOGGER.setLevel(logging.INFO) # reset to default return model.to(device) except Exception as e: From 1c5e92aba11f0dd007716821e7cd151d532342a8 Mon Sep 17 00:00:00 2001 From: UnglvKitDe <100289696+UnglvKitDe@users.noreply.github.com> Date: Sat, 23 Jul 2022 01:25:17 +0200 Subject: [PATCH 375/661] Add generator and worker seed (#8602) * Add generator and worker seed * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update dataloaders.py * Update dataloaders.py * Update dataloaders.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- utils/dataloaders.py | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 4f1c98fd880d..85a39ab52f82 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -91,6 +91,13 @@ def exif_transpose(image): return image +def seed_worker(worker_id): + # Set dataloader worker seed https://pytorch.org/docs/stable/notes/randomness.html#dataloader + worker_seed = torch.initial_seed() % 2 ** 32 + np.random.seed(worker_seed) + random.seed(worker_seed) + + def create_dataloader(path, imgsz, batch_size, @@ -130,13 +137,17 @@ def create_dataloader(path, nw = min([os.cpu_count() // max(nd, 1), batch_size if batch_size > 1 else 0, workers]) # number of workers sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle) loader = DataLoader if image_weights else InfiniteDataLoader # only DataLoader allows for attribute updates + generator = torch.Generator() + generator.manual_seed(0) return loader(dataset, batch_size=batch_size, shuffle=shuffle and sampler is None, num_workers=nw, sampler=sampler, pin_memory=True, - collate_fn=LoadImagesAndLabels.collate_fn4 if quad else LoadImagesAndLabels.collate_fn), dataset + collate_fn=LoadImagesAndLabels.collate_fn4 if quad else LoadImagesAndLabels.collate_fn, + worker_init_fn=seed_worker, + generator=generator), dataset class InfiniteDataLoader(dataloader.DataLoader): From 7f7bd6fbcd214886aa2a275500eb5e05933bea05 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 23 Jul 2022 18:24:14 +0200 Subject: [PATCH 376/661] Set `torch.cuda.manual_seed_all()` for DDP (#8688) * Set `torch.cuda.manual_seed_all()` for DDP * Update general.py * Update general.py --- utils/general.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/utils/general.py b/utils/general.py index 925f7fbf0ecb..b049ce469a71 100755 --- a/utils/general.py +++ b/utils/general.py @@ -203,14 +203,14 @@ def init_seeds(seed=0, deterministic=False): if deterministic and check_version(torch.__version__, '1.12.0'): # https://github.com/ultralytics/yolov5/pull/8213 torch.use_deterministic_algorithms(True) os.environ['CUBLAS_WORKSPACE_CONFIG'] = ':4096:8' - # os.environ['PYTHONHASHSEED'] = str(seed) + os.environ['PYTHONHASHSEED'] = str(seed) random.seed(seed) np.random.seed(seed) torch.manual_seed(seed) cudnn.benchmark, cudnn.deterministic = (False, True) if seed == 0 else (True, False) - # torch.cuda.manual_seed(seed) - # torch.cuda.manual_seed_all(seed) # for multi GPU, exception safe + torch.cuda.manual_seed(seed) + torch.cuda.manual_seed_all(seed) # for Multi-GPU, exception safe def intersect_dicts(da, db, exclude=()): From b510957650c890dee876146c43dcda1fdfc279d6 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 23 Jul 2022 18:50:19 +0200 Subject: [PATCH 377/661] Move .dockerignore to root (#8690) --- utils/docker/.dockerignore => .dockerignore | 0 1 file changed, 0 insertions(+), 0 deletions(-) rename utils/docker/.dockerignore => .dockerignore (100%) diff --git a/utils/docker/.dockerignore b/.dockerignore similarity index 100% rename from utils/docker/.dockerignore rename to .dockerignore From 916bdb1d61f23de92833bd491df54cda5c3ef0cc Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 23 Jul 2022 23:30:30 +0200 Subject: [PATCH 378/661] Faster crop saving (#8696) Faster crops Following https://github.com/ultralytics/yolov5/issues/8641#issuecomment-1193190325 --- utils/plots.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/plots.py b/utils/plots.py index 1bbb9c09c33a..53e326c23f6e 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -484,6 +484,6 @@ def save_one_box(xyxy, im, file=Path('im.jpg'), gain=1.02, pad=10, square=False, if save: file.parent.mkdir(parents=True, exist_ok=True) # make directory f = str(increment_path(file).with_suffix('.jpg')) - # cv2.imwrite(f, crop) # https://github.com/ultralytics/yolov5/issues/7007 chroma subsampling issue - Image.fromarray(cv2.cvtColor(crop, cv2.COLOR_BGR2RGB)).save(f, quality=95, subsampling=0) + # cv2.imwrite(f, crop) # save BGR, https://github.com/ultralytics/yolov5/issues/7007 chroma subsampling issue + Image.fromarray(crop[..., ::-1]).save(f, quality=95, subsampling=0) # save RGB return crop From 0ab303b04499b6b912d8212a4fa10fe3fcb78efa Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 25 Jul 2022 00:02:09 +0200 Subject: [PATCH 379/661] Remove `else:` from load_image() (#8692) --- utils/dataloaders.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 85a39ab52f82..36610c88980a 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -682,8 +682,7 @@ def load_image(self, i): interp = cv2.INTER_LINEAR if (self.augment or r > 1) else cv2.INTER_AREA im = cv2.resize(im, (int(w0 * r), int(h0 * r)), interpolation=interp) return im, (h0, w0), im.shape[:2] # im, hw_original, hw_resized - else: - return self.ims[i], self.im_hw0[i], self.im_hw[i] # im, hw_original, hw_resized + return self.ims[i], self.im_hw0[i], self.im_hw[i] # im, hw_original, hw_resized def cache_images_to_disk(self, i): # Saves an image as an *.npy file for faster loading From 7215a0fb41a90d8a0bf259fa708dff608a1f0262 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 25 Jul 2022 13:57:05 +0200 Subject: [PATCH 380/661] Avoid cv2 window init code on Windows (#8712) Resolves https://github.com/ultralytics/yolov5/issues/8642 --- detect.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/detect.py b/detect.py index bb09ce171a96..01ad797ae6f1 100644 --- a/detect.py +++ b/detect.py @@ -26,6 +26,7 @@ import argparse import os +import platform import sys from pathlib import Path @@ -173,7 +174,7 @@ def run( # Stream results im0 = annotator.result() if view_img: - if p not in windows: + if platform.system() == 'Linux' and p not in windows: windows.append(p) cv2.namedWindow(str(p), cv2.WINDOW_NORMAL | cv2.WINDOW_KEEPRATIO) # allow window resize (Linux) cv2.resizeWindow(str(p), im0.shape[1], im0.shape[0]) From a6f197ae79d546efd58e4a4f206621196ab5cacd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 25 Jul 2022 16:52:28 +0200 Subject: [PATCH 381/661] Update dataloaders.py (#8714) * Update dataloaders.py * Update dataloaders.py --- utils/dataloaders.py | 36 ++++++++++++++---------------------- 1 file changed, 14 insertions(+), 22 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 36610c88980a..c32f60fe4ec7 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -3,6 +3,7 @@ Dataloaders and dataset utils """ +import contextlib import glob import hashlib import json @@ -55,13 +56,10 @@ def get_hash(paths): def exif_size(img): # Returns exif-corrected PIL size s = img.size # (width, height) - try: + with contextlib.suppress(Exception): rotation = dict(img._getexif().items())[orientation] if rotation in [6, 8]: # rotation 270 or 90 s = (s[1], s[0]) - except Exception: - pass - return s @@ -859,18 +857,13 @@ def collate_fn4(batch): # Ancillary functions -------------------------------------------------------------------------------------------------- -def create_folder(path='./new'): - # Create folder - if os.path.exists(path): - shutil.rmtree(path) # delete output folder - os.makedirs(path) # make new output folder - - def flatten_recursive(path=DATASETS_DIR / 'coco128'): # Flatten a recursive directory by bringing all files to top level - new_path = Path(str(path) + '_flat') - create_folder(new_path) - for file in tqdm(glob.glob(str(Path(path)) + '/**/*.*', recursive=True)): + new_path = Path(f'{str(path)}_flat') + if os.path.exists(new_path): + shutil.rmtree(new_path) # delete output folder + os.makedirs(new_path) # make new output folder + for file in tqdm(glob.glob(f'{str(Path(path))}/**/*.*', recursive=True)): shutil.copyfile(file, new_path / Path(file).name) @@ -929,7 +922,7 @@ def autosplit(path=DATASETS_DIR / 'coco128/images', weights=(0.9, 0.1, 0.0), ann for i, img in tqdm(zip(indices, files), total=n): if not annotated_only or Path(img2label_paths([str(img)])[0]).exists(): # check label with open(path.parent / txt[i], 'a') as f: - f.write('./' + img.relative_to(path.parent).as_posix() + '\n') # add image to txt file + f.write(f'./{img.relative_to(path.parent).as_posix()}' + '\n') # add image to txt file def verify_image_label(args): @@ -1011,14 +1004,13 @@ def _find_yaml(dir): def _unzip(path): # Unzip data.zip - if str(path).endswith('.zip'): # path is data.zip - assert Path(path).is_file(), f'Error unzipping {path}, file not found' - ZipFile(path).extractall(path=path.parent) # unzip - dir = path.with_suffix('') # dataset directory == zip name - assert dir.is_dir(), f'Error unzipping {path}, {dir} not found. path/to/abc.zip MUST unzip to path/to/abc/' - return True, str(dir), _find_yaml(dir) # zipped, data_dir, yaml_path - else: # path is data.yaml + if not str(path).endswith('.zip'): # path is data.yaml return False, None, path + assert Path(path).is_file(), f'Error unzipping {path}, file not found' + ZipFile(path).extractall(path=path.parent) # unzip + dir = path.with_suffix('') # dataset directory == zip name + assert dir.is_dir(), f'Error unzipping {path}, {dir} not found. path/to/abc.zip MUST unzip to path/to/abc/' + return True, str(dir), _find_yaml(dir) # zipped, data_dir, yaml_path def _hub_ops(f, max_dim=1920): # HUB ops for 1 image 'f': resize and save at reduced quality in /dataset-hub for web/app viewing From b367860196a2590a5f44c9b18401dedfc0543077 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 25 Jul 2022 18:20:01 +0200 Subject: [PATCH 382/661] New `HUBDatasetStats()` class (#8716) * New `HUBDatasetStats()` class Usage examples: ``` from utils.dataloaders import * stats = HUBDatasetStats('coco128.yaml', autodownload=True) # method 1 stats = HUBDatasetStats('path/to/coco128_with_yaml.zip') # method 1 stats.get_json(save=False) stats.process_images() ``` @kalenmike * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update dataloaders.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update dataloaders.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update dataloaders.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update dataloaders.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/dataloaders.py | 146 +++++++++++++++++++++---------------------- 1 file changed, 70 insertions(+), 76 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index c32f60fe4ec7..9ccfe2545d75 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -977,21 +977,35 @@ def verify_image_label(args): return [None, None, None, None, nm, nf, ne, nc, msg] -def dataset_stats(path='coco128.yaml', autodownload=False, verbose=False, profile=False, hub=False): +class HUBDatasetStats(): """ Return dataset statistics dictionary with images and instances counts per split per class To run in parent directory: export PYTHONPATH="$PWD/yolov5" - Usage1: from utils.dataloaders import *; dataset_stats('coco128.yaml', autodownload=True) - Usage2: from utils.dataloaders import *; dataset_stats('path/to/coco128_with_yaml.zip') + Usage1: from utils.dataloaders import *; HUBDatasetStats('coco128.yaml', autodownload=True) + Usage2: from utils.dataloaders import *; HUBDatasetStats('path/to/coco128_with_yaml.zip') Arguments path: Path to data.yaml or data.zip (with data.yaml inside data.zip) autodownload: Attempt to download dataset if not found locally - verbose: Print stats dictionary """ - def _round_labels(labels): - # Update labels to integer class and 6 decimal place floats - return [[int(c), *(round(x, 4) for x in points)] for c, *points in labels] + def __init__(self, path='coco128.yaml', autodownload=False): + # Initialize class + zipped, data_dir, yaml_path = self._unzip(Path(path)) + try: + with open(check_yaml(yaml_path), errors='ignore') as f: + data = yaml.safe_load(f) # data dict + if zipped: + data['path'] = data_dir + except Exception as e: + raise Exception("error/HUB/dataset_stats/yaml_load") from e + + check_dataset(data, autodownload) # download dataset if missing + self.hub_dir = Path(data['path'] + '-hub') + self.im_dir = self.hub_dir / 'images' + self.im_dir.mkdir(parents=True, exist_ok=True) # makes /images + self.stats = {'nc': data['nc'], 'names': data['names']} # statistics dictionary + self.data = data + @staticmethod def _find_yaml(dir): # Return data.yaml file files = list(dir.glob('*.yaml')) or list(dir.rglob('*.yaml')) # try root level first and then recursive @@ -1002,7 +1016,7 @@ def _find_yaml(dir): assert len(files) == 1, f'Multiple *.yaml files found: {files}, only 1 *.yaml file allowed in {dir}' return files[0] - def _unzip(path): + def _unzip(self, path): # Unzip data.zip if not str(path).endswith('.zip'): # path is data.yaml return False, None, path @@ -1010,11 +1024,11 @@ def _unzip(path): ZipFile(path).extractall(path=path.parent) # unzip dir = path.with_suffix('') # dataset directory == zip name assert dir.is_dir(), f'Error unzipping {path}, {dir} not found. path/to/abc.zip MUST unzip to path/to/abc/' - return True, str(dir), _find_yaml(dir) # zipped, data_dir, yaml_path + return True, str(dir), self._find_yaml(dir) # zipped, data_dir, yaml_path - def _hub_ops(f, max_dim=1920): + def _hub_ops(self, f, max_dim=1920): # HUB ops for 1 image 'f': resize and save at reduced quality in /dataset-hub for web/app viewing - f_new = im_dir / Path(f).name # dataset-hub image filename + f_new = self.im_dir / Path(f).name # dataset-hub image filename try: # use PIL im = Image.open(f) r = max_dim / max(im.height, im.width) # ratio @@ -1030,69 +1044,49 @@ def _hub_ops(f, max_dim=1920): im = cv2.resize(im, (int(im_width * r), int(im_height * r)), interpolation=cv2.INTER_AREA) cv2.imwrite(str(f_new), im) - zipped, data_dir, yaml_path = _unzip(Path(path)) - try: - with open(check_yaml(yaml_path), errors='ignore') as f: - data = yaml.safe_load(f) # data dict - if zipped: - data['path'] = data_dir # TODO: should this be dir.resolve()?` - except Exception: - raise Exception("error/HUB/dataset_stats/yaml_load") - - check_dataset(data, autodownload) # download dataset if missing - hub_dir = Path(data['path'] + ('-hub' if hub else '')) - stats = {'nc': data['nc'], 'names': data['names']} # statistics dictionary - for split in 'train', 'val', 'test': - if data.get(split) is None: - stats[split] = None # i.e. no test set - continue - x = [] - dataset = LoadImagesAndLabels(data[split]) # load dataset - for label in tqdm(dataset.labels, total=dataset.n, desc='Statistics'): - x.append(np.bincount(label[:, 0].astype(int), minlength=data['nc'])) - x = np.array(x) # shape(128x80) - stats[split] = { - 'instance_stats': { - 'total': int(x.sum()), - 'per_class': x.sum(0).tolist()}, - 'image_stats': { - 'total': dataset.n, - 'unlabelled': int(np.all(x == 0, 1).sum()), - 'per_class': (x > 0).sum(0).tolist()}, - 'labels': [{ - str(Path(k).name): _round_labels(v.tolist())} for k, v in zip(dataset.im_files, dataset.labels)]} - - if hub: - im_dir = hub_dir / 'images' - im_dir.mkdir(parents=True, exist_ok=True) - for _ in tqdm(ThreadPool(NUM_THREADS).imap(_hub_ops, dataset.im_files), total=dataset.n, desc='HUB Ops'): + def get_json(self, save=False, verbose=False): + # Return dataset JSON for Ultralytics HUB + def _round(labels): + # Update labels to integer class and 6 decimal place floats + return [[int(c), *(round(x, 4) for x in points)] for c, *points in labels] + + for split in 'train', 'val', 'test': + if self.data.get(split) is None: + self.stats[split] = None # i.e. no test set + continue + dataset = LoadImagesAndLabels(self.data[split]) # load dataset + x = np.array([ + np.bincount(label[:, 0].astype(int), minlength=self.data['nc']) + for label in tqdm(dataset.labels, total=dataset.n, desc='Statistics')]) # shape(128x80) + self.stats[split] = { + 'instance_stats': { + 'total': int(x.sum()), + 'per_class': x.sum(0).tolist()}, + 'image_stats': { + 'total': dataset.n, + 'unlabelled': int(np.all(x == 0, 1).sum()), + 'per_class': (x > 0).sum(0).tolist()}, + 'labels': [{ + str(Path(k).name): _round(v.tolist())} for k, v in zip(dataset.im_files, dataset.labels)]} + + # Save, print and return + if save: + stats_path = self.hub_dir / 'stats.json' + print(f'Saving {stats_path.resolve()}...') + with open(stats_path, 'w') as f: + json.dump(self.stats, f) # save stats.json + if verbose: + print(json.dumps(self.stats, indent=2, sort_keys=False)) + return self.stats + + def process_images(self): + # Compress images for Ultralytics HUB + for split in 'train', 'val', 'test': + if self.data.get(split) is None: + continue + dataset = LoadImagesAndLabels(self.data[split]) # load dataset + desc = f'{split} images' + for _ in tqdm(ThreadPool(NUM_THREADS).imap(self._hub_ops, dataset.im_files), total=dataset.n, desc=desc): pass - - # Profile - stats_path = hub_dir / 'stats.json' - if profile: - for _ in range(1): - file = stats_path.with_suffix('.npy') - t1 = time.time() - np.save(file, stats) - t2 = time.time() - x = np.load(file, allow_pickle=True) - print(f'stats.npy times: {time.time() - t2:.3f}s read, {t2 - t1:.3f}s write') - - file = stats_path.with_suffix('.json') - t1 = time.time() - with open(file, 'w') as f: - json.dump(stats, f) # save stats *.json - t2 = time.time() - with open(file) as f: - x = json.load(f) # load hyps dict - print(f'stats.json times: {time.time() - t2:.3f}s read, {t2 - t1:.3f}s write') - - # Save, print and return - if hub: - print(f'Saving {stats_path.resolve()}...') - with open(stats_path, 'w') as f: - json.dump(stats, f) # save stats.json - if verbose: - print(json.dumps(stats, indent=2, sort_keys=False)) - return stats + print(f'Done. All images saved to {self.im_dir}') + return self.im_dir From 2e1291fdce26b3cff213e9e7ee8c196fa263b688 Mon Sep 17 00:00:00 2001 From: UnglvKitDe <100289696+UnglvKitDe@users.noreply.github.com> Date: Tue, 26 Jul 2022 13:52:56 +0200 Subject: [PATCH 383/661] Fix BGR->RGB Bug in albumentations #8641 (#8695) * Fix BGR->RGB Bug in albumentations https://github.com/ultralytics/yolov5/issues/8641 * Change transform methode from cv2 to numpy * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Simplify * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update augmentations.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- utils/augmentations.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/utils/augmentations.py b/utils/augmentations.py index 3f764c06ae3b..97506ae25123 100644 --- a/utils/augmentations.py +++ b/utils/augmentations.py @@ -39,8 +39,9 @@ def __init__(self): def __call__(self, im, labels, p=1.0): if self.transform and random.random() < p: - new = self.transform(image=im, bboxes=labels[:, 1:], class_labels=labels[:, 0]) # transformed - im, labels = new['image'], np.array([[c, *b] for c, b in zip(new['class_labels'], new['bboxes'])]) + new = self.transform(image=im[..., ::-1], bboxes=labels[:, 1:], class_labels=labels[:, 0]) # transformed + im = new['image'][..., ::-1] # RGB to BGR + labels = np.array([[c, *b] for c, b in zip(new['class_labels'], new['bboxes'])]) return im, labels From d5116bbe9c9411b7c0c969fce32b86abd74c6d4a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 26 Jul 2022 17:50:49 +0200 Subject: [PATCH 384/661] coremltools>=5.2 for CoreML export (#8725) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 8548f67b5a48..de3239cbdd42 100644 --- a/requirements.txt +++ b/requirements.txt @@ -23,7 +23,7 @@ pandas>=1.1.4 seaborn>=0.11.0 # Export -------------------------------------- -# coremltools>=4.1 # CoreML export +# coremltools>=5.2 # CoreML export # onnx>=1.9.0 # ONNX export # onnx-simplifier>=0.4.1 # ONNX simplifier # nvidia-pyindex # TensorRT export From c775a296a7db2e523a230b2a0900ecd12845ecde Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 26 Jul 2022 19:00:48 +0200 Subject: [PATCH 385/661] Revert "Fix BGR->RGB Bug in albumentations #8641" (#8727) Revert "Fix BGR->RGB Bug in albumentations #8641 (#8695)" This reverts commit 2e1291fdce26b3cff213e9e7ee8c196fa263b688. --- utils/augmentations.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/utils/augmentations.py b/utils/augmentations.py index 97506ae25123..3f764c06ae3b 100644 --- a/utils/augmentations.py +++ b/utils/augmentations.py @@ -39,9 +39,8 @@ def __init__(self): def __call__(self, im, labels, p=1.0): if self.transform and random.random() < p: - new = self.transform(image=im[..., ::-1], bboxes=labels[:, 1:], class_labels=labels[:, 0]) # transformed - im = new['image'][..., ::-1] # RGB to BGR - labels = np.array([[c, *b] for c, b in zip(new['class_labels'], new['bboxes'])]) + new = self.transform(image=im, bboxes=labels[:, 1:], class_labels=labels[:, 0]) # transformed + im, labels = new['image'], np.array([[c, *b] for c, b in zip(new['class_labels'], new['bboxes'])]) return im, labels From 0b5ac224aef287ac3ac9ebf70ade60159450a0b1 Mon Sep 17 00:00:00 2001 From: Max Strobel Date: Tue, 26 Jul 2022 18:02:44 +0100 Subject: [PATCH 386/661] fix: broken ``is_docker`` check (#8711) Checking if ``/workspace`` exists is not a reliable method to check if the process runs in a docker container. Reusing the logic from the npm "is-docker" package to check if the process runs in a container. References: https://github.com/sindresorhus/is-docker/blob/main/index.js Fixes #8710. Co-authored-by: Maximilian Strobel Co-authored-by: Glenn Jocher --- utils/general.py | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/utils/general.py b/utils/general.py index b049ce469a71..67078338d762 100755 --- a/utils/general.py +++ b/utils/general.py @@ -224,9 +224,15 @@ def get_latest_run(search_dir='.'): return max(last_list, key=os.path.getctime) if last_list else '' -def is_docker(): - # Is environment a Docker container? - return Path('/workspace').exists() # or Path('/.dockerenv').exists() +def is_docker() -> bool: + """Check if the process runs inside a docker container.""" + if Path("/.dockerenv").exists(): + return True + try: # check if docker is in control groups + with open("/proc/self/cgroup") as file: + return any("docker" in line for line in file) + except OSError: + return False def is_colab(): From 3e858633b283767f038b4cab910a95e40fe8577b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 27 Jul 2022 17:27:44 +0200 Subject: [PATCH 387/661] Revert protobuf<=3.20.1 (#8742) Resolve #8012 (again) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index de3239cbdd42..6313cecee578 100644 --- a/requirements.txt +++ b/requirements.txt @@ -12,7 +12,7 @@ scipy>=1.4.1 torch>=1.7.0 torchvision>=0.8.1 tqdm>=4.64.0 -protobuf<4.21.3 # https://github.com/ultralytics/yolov5/issues/8012 +protobuf<=3.20.1 # https://github.com/ultralytics/yolov5/issues/8012 # Logging ------------------------------------- tensorboard>=2.4.1 From 587a3a37c57661c3a0ef710d2b309199fad632d2 Mon Sep 17 00:00:00 2001 From: Colin Wong Date: Fri, 29 Jul 2022 06:51:16 -0500 Subject: [PATCH 388/661] Dynamic batch size support for TensorRT (#8526) * Dynamic batch size support for TensorRT * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update export.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix optimization profile when batch size is 1 * Warn users if they use batch-size=1 with dynamic * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * More descriptive assertion error * Fix syntax * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * pre-commit formatting sucked * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update export.py Co-authored-by: Colin Wong Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- export.py | 21 +++++++++++++++------ models/common.py | 22 ++++++++++++++++------ 2 files changed, 31 insertions(+), 12 deletions(-) diff --git a/export.py b/export.py index 3629915f028d..4846624541e4 100644 --- a/export.py +++ b/export.py @@ -216,8 +216,9 @@ def export_coreml(model, im, file, int8, half, prefix=colorstr('CoreML:')): return None, None -def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=False, prefix=colorstr('TensorRT:')): +def export_engine(model, im, file, train, half, dynamic, simplify, workspace=4, verbose=False): # YOLOv5 TensorRT export https://developer.nvidia.com/tensorrt + prefix = colorstr('TensorRT:') try: assert im.device.type != 'cpu', 'export running on CPU but must be on GPU, i.e. `python export.py --device 0`' try: @@ -230,11 +231,11 @@ def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=F if trt.__version__[0] == '7': # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012 grid = model.model[-1].anchor_grid model.model[-1].anchor_grid = [a[..., :1, :1, :] for a in grid] - export_onnx(model, im, file, 12, train, False, simplify) # opset 12 + export_onnx(model, im, file, 12, train, dynamic, simplify) # opset 12 model.model[-1].anchor_grid = grid else: # TensorRT >= 8 check_version(trt.__version__, '8.0.0', hard=True) # require tensorrt>=8.0.0 - export_onnx(model, im, file, 13, train, False, simplify) # opset 13 + export_onnx(model, im, file, 13, train, dynamic, simplify) # opset 13 onnx = file.with_suffix('.onnx') LOGGER.info(f'\n{prefix} starting export with TensorRT {trt.__version__}...') @@ -263,6 +264,14 @@ def export_engine(model, im, file, train, half, simplify, workspace=4, verbose=F for out in outputs: LOGGER.info(f'{prefix}\toutput "{out.name}" with shape {out.shape} and dtype {out.dtype}') + if dynamic: + if im.shape[0] <= 1: + LOGGER.warning(f"{prefix}WARNING: --dynamic model requires maximum --batch-size argument") + profile = builder.create_optimization_profile() + for inp in inputs: + profile.set_shape(inp.name, (1, *im.shape[1:]), (max(1, im.shape[0] // 2), *im.shape[1:]), im.shape) + config.add_optimization_profile(profile) + LOGGER.info(f'{prefix} building FP{16 if builder.platform_has_fast_fp16 and half else 32} engine in {f}') if builder.platform_has_fast_fp16 and half: config.set_flag(trt.BuilderFlag.FP16) @@ -460,7 +469,7 @@ def run( keras=False, # use Keras optimize=False, # TorchScript: optimize for mobile int8=False, # CoreML/TF INT8 quantization - dynamic=False, # ONNX/TF: dynamic axes + dynamic=False, # ONNX/TF/TensorRT: dynamic axes simplify=False, # ONNX: simplify model opset=12, # ONNX: opset version verbose=False, # TensorRT: verbose log @@ -520,7 +529,7 @@ def run( if jit: f[0] = export_torchscript(model, im, file, optimize) if engine: # TensorRT required before ONNX - f[1] = export_engine(model, im, file, train, half, simplify, workspace, verbose) + f[1] = export_engine(model, im, file, train, half, dynamic, simplify, workspace, verbose) if onnx or xml: # OpenVINO requires ONNX f[2] = export_onnx(model, im, file, opset, train, dynamic, simplify) if xml: # OpenVINO @@ -579,7 +588,7 @@ def parse_opt(): parser.add_argument('--keras', action='store_true', help='TF: use Keras') parser.add_argument('--optimize', action='store_true', help='TorchScript: optimize for mobile') parser.add_argument('--int8', action='store_true', help='CoreML/TF INT8 quantization') - parser.add_argument('--dynamic', action='store_true', help='ONNX/TF: dynamic axes') + parser.add_argument('--dynamic', action='store_true', help='ONNX/TF/TensorRT: dynamic axes') parser.add_argument('--simplify', action='store_true', help='ONNX: simplify model') parser.add_argument('--opset', type=int, default=12, help='ONNX: opset version') parser.add_argument('--verbose', action='store_true', help='TensorRT: verbose log') diff --git a/models/common.py b/models/common.py index 5ea1c307f034..959c965e6002 100644 --- a/models/common.py +++ b/models/common.py @@ -384,19 +384,24 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, logger = trt.Logger(trt.Logger.INFO) with open(w, 'rb') as f, trt.Runtime(logger) as runtime: model = runtime.deserialize_cuda_engine(f.read()) + context = model.create_execution_context() bindings = OrderedDict() fp16 = False # default updated below + dynamic_input = False for index in range(model.num_bindings): name = model.get_binding_name(index) dtype = trt.nptype(model.get_binding_dtype(index)) - shape = tuple(model.get_binding_shape(index)) + if model.binding_is_input(index): + if -1 in tuple(model.get_binding_shape(index)): # dynamic + dynamic_input = True + context.set_binding_shape(index, tuple(model.get_profile_shape(0, index)[2])) + if dtype == np.float16: + fp16 = True + shape = tuple(context.get_binding_shape(index)) data = torch.from_numpy(np.empty(shape, dtype=np.dtype(dtype))).to(device) bindings[name] = Binding(name, dtype, shape, data, int(data.data_ptr())) - if model.binding_is_input(index) and dtype == np.float16: - fp16 = True binding_addrs = OrderedDict((n, d.ptr) for n, d in bindings.items()) - context = model.create_execution_context() - batch_size = bindings['images'].shape[0] + batch_size = bindings['images'].shape[0] # if dynamic, this is instead max batch size elif coreml: # CoreML LOGGER.info(f'Loading {w} for CoreML inference...') import coremltools as ct @@ -466,7 +471,12 @@ def forward(self, im, augment=False, visualize=False, val=False): im = im.cpu().numpy() # FP32 y = self.executable_network([im])[self.output_layer] elif self.engine: # TensorRT - assert im.shape == self.bindings['images'].shape, (im.shape, self.bindings['images'].shape) + if im.shape != self.bindings['images'].shape and self.dynamic_input: + self.context.set_binding_shape(self.model.get_binding_index('images'), im.shape) # reshape if dynamic + self.bindings['images'] = self.bindings['images']._replace(shape=im.shape) + assert im.shape == self.bindings['images'].shape, ( + f"image shape {im.shape} exceeds model max shape {self.bindings['images'].shape}" if self.dynamic_input + else f"image shape {im.shape} does not match model shape {self.bindings['images'].shape}") self.binding_addrs['images'] = int(im.data_ptr()) self.context.execute_v2(list(self.binding_addrs.values())) y = self.bindings['output'].data From 567397d67ae173fb82e06672a763cc28c5cfeb2b Mon Sep 17 00:00:00 2001 From: jbutle55 Date: Fri, 29 Jul 2022 06:06:23 -0600 Subject: [PATCH 389/661] Fix confusion matrix update when no predictions are made (#8748) * Fix confusion matrix update when no predictions are made * Update metrics.py * Simply confusion matrix changes * Simply confusion matrix fix Co-authored-by: Glenn Jocher --- utils/metrics.py | 6 ++++++ val.py | 2 ++ 2 files changed, 8 insertions(+) diff --git a/utils/metrics.py b/utils/metrics.py index 6bba4cfe2a42..9bf084c78854 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -139,6 +139,12 @@ def process_batch(self, detections, labels): Returns: None, updates confusion matrix accordingly """ + if detections is None: + gt_classes = labels.int() + for i, gc in enumerate(gt_classes): + self.matrix[self.nc, gc] += 1 # background FN + return + detections = detections[detections[:, 4] > self.conf] gt_classes = labels[:, 0].int() detection_classes = detections[:, 5].int() diff --git a/val.py b/val.py index b0cc8e7f1577..48207a1130a6 100644 --- a/val.py +++ b/val.py @@ -228,6 +228,8 @@ def run( if npr == 0: if nl: stats.append((correct, *torch.zeros((2, 0), device=device), labels[:, 0])) + if plots: + confusion_matrix.process_batch(detections=None, labels=labels[:, 0]) continue # Predictions From e309a855860bc3f618c3541909c515a65ffc35b0 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 29 Jul 2022 14:45:29 +0200 Subject: [PATCH 390/661] Add val.py no label warning (#8782) Help resolve confusion around zero-metrics val.py results when no labels are found in https://github.com/ultralytics/yolov5/issues/8753 --- val.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/val.py b/val.py index 48207a1130a6..006ade37d03e 100644 --- a/val.py +++ b/val.py @@ -275,6 +275,8 @@ def run( # Print results pf = '%20s' + '%11i' * 2 + '%11.3g' * 4 # print format LOGGER.info(pf % ('all', seen, nt.sum(), mp, mr, map50, map)) + if nt.sum() == 0: + LOGGER.warning(emojis(f'WARNING: no labels found in {task} set, can not compute metrics without labels ⚠️')) # Print results per class if (verbose or (nc < 50 and not training)) and nc > 1 and len(stats): From 52d3a9aee1016604652898fed679e55783e264ed Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 29 Jul 2022 17:07:24 +0200 Subject: [PATCH 391/661] Fix `detect.py --update` list bug (#8783) Fix detect.py --update Resolves https://github.com/ultralytics/yolov5/issues/8776 --- detect.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/detect.py b/detect.py index 01ad797ae6f1..8741e7f7fd62 100644 --- a/detect.py +++ b/detect.py @@ -210,7 +210,7 @@ def run( s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else '' LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}{s}") if update: - strip_optimizer(weights) # update model (to fix SourceChangeWarning) + strip_optimizer(weights[0]) # update model (to fix SourceChangeWarning) def parse_opt(): From e34ae8837b652a35f115d3e780c18abae4bb89ce Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 30 Jul 2022 14:04:03 +0200 Subject: [PATCH 392/661] ci-testing.yml Windows-friendly ENV variables (#8794) Per https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions --- .github/workflows/ci-testing.yml | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index e3359cd3a283..61a527e62ecf 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -85,12 +85,12 @@ jobs: - name: Check environment run: | python -c "import utils; utils.notebook_init()" - echo "RUNNER_OS is $RUNNER_OS" - echo "GITHUB_EVENT_NAME is $GITHUB_EVENT_NAME" - echo "GITHUB_WORKFLOW is $GITHUB_WORKFLOW" - echo "GITHUB_ACTOR is $GITHUB_ACTOR" - echo "GITHUB_REPOSITORY is $GITHUB_REPOSITORY" - echo "GITHUB_REPOSITORY_OWNER is $GITHUB_REPOSITORY_OWNER" + echo "RUNNER_OS is ${{ runner.os }}" + echo "GITHUB_EVENT_NAME is ${{ github.event_name }}" + echo "GITHUB_WORKFLOW is ${{ github.workflow }}" + echo "GITHUB_ACTOR is ${{ github.actor }}" + echo "GITHUB_REPOSITORY is ${{ github.repository }}" + echo "GITHUB_REPOSITORY_OWNER is ${{ github.repository_owner }}" - name: Run tests shell: bash run: | From 9111246208a6f7ada69f2cdc1d5832f22486620a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 30 Jul 2022 21:00:28 +0200 Subject: [PATCH 393/661] Add hubconf.py argparser (#8799) * Add hubconf.py argparser * Add hubconf.py argparser --- .github/workflows/ci-testing.yml | 2 +- hubconf.py | 19 ++++++++++++++----- 2 files changed, 15 insertions(+), 6 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 61a527e62ecf..5b492009d503 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -106,7 +106,7 @@ jobs: # Detect python detect.py --weights $model.pt --device $d python detect.py --weights $best --device $d - python hubconf.py # hub + python hubconf.py --model $model # hub # Export # python models/tf.py --weights $model.pt # build TF model python models/yolo.py --cfg $model.yaml # build PyTorch model diff --git a/hubconf.py b/hubconf.py index 25f9d1b82c14..f579c6471b20 100644 --- a/hubconf.py +++ b/hubconf.py @@ -41,7 +41,6 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo path = name.with_suffix('.pt') if name.suffix == '' and not name.is_dir() else name # checkpoint path try: device = select_device(device) - if pretrained and channels == 3 and classes == 80: model = DetectMultiBackend(path, device=device, fuse=autoshape) # download/load FP32 model # model = models.experimental.attempt_load(path, map_location=device) # download/load FP32 model @@ -123,10 +122,7 @@ def yolov5x6(pretrained=True, channels=3, classes=80, autoshape=True, _verbose=T if __name__ == '__main__': - model = _create(name='yolov5s', pretrained=True, channels=3, classes=80, autoshape=True, verbose=True) - # model = custom(path='path/to/model.pt') # custom - - # Verify inference + import argparse from pathlib import Path import numpy as np @@ -134,6 +130,16 @@ def yolov5x6(pretrained=True, channels=3, classes=80, autoshape=True, _verbose=T from utils.general import cv2 + # Argparser + parser = argparse.ArgumentParser() + parser.add_argument('--model', type=str, default='yolov5s', help='model name') + opt = parser.parse_args() + + # Model + model = _create(name=opt.model, pretrained=True, channels=3, classes=80, autoshape=True, verbose=True) + # model = custom(path='path/to/model.pt') # custom + + # Images imgs = [ 'data/images/zidane.jpg', # filename Path('data/images/zidane.jpg'), # Path @@ -142,6 +148,9 @@ def yolov5x6(pretrained=True, channels=3, classes=80, autoshape=True, _verbose=T Image.open('data/images/bus.jpg'), # PIL np.zeros((320, 640, 3))] # numpy + # Inference results = model(imgs, size=320) # batched inference + + # Results results.print() results.save() From 56f5cb5a28ac8fb5afc49392633763203f37e9bb Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 30 Jul 2022 21:02:26 +0200 Subject: [PATCH 394/661] Print hubconf.py args (#8800) --- hubconf.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/hubconf.py b/hubconf.py index f579c6471b20..08122eaca9dc 100644 --- a/hubconf.py +++ b/hubconf.py @@ -128,12 +128,13 @@ def yolov5x6(pretrained=True, channels=3, classes=80, autoshape=True, _verbose=T import numpy as np from PIL import Image - from utils.general import cv2 + from utils.general import cv2, print_args # Argparser parser = argparse.ArgumentParser() parser.add_argument('--model', type=str, default='yolov5s', help='model name') opt = parser.parse_args() + print_args(vars(opt)) # Model model = _create(name=opt.model, pretrained=True, channels=3, classes=80, autoshape=True, verbose=True) From ec4de43a8aabe497ade56de67bec2b86a22a9c61 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 30 Jul 2022 21:11:19 +0200 Subject: [PATCH 395/661] Update Colab Notebook CI (#8798) * Update Colab Notebook CI * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Created using Colaboratory * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update tutorial.ipynb Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- tutorial.ipynb | 38 ++++++++++++++++++-------------------- 1 file changed, 18 insertions(+), 20 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index bdfba399a883..dcb1162b40af 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -414,7 +414,7 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": 1, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -466,7 +466,7 @@ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", "display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": 2, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -546,7 +546,7 @@ "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip')\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip" ], - "execution_count": 3, + "execution_count": null, "outputs": [ { "output_type": "display_data", @@ -577,7 +577,7 @@ "# Run YOLOv5x on COCO val\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" ], - "execution_count": 4, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -737,7 +737,7 @@ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": 7, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -1032,24 +1032,22 @@ "id": "FGH0ZjkGjejy" }, "source": [ - "# CI Checks\n", + "# YOLOv5 CI\n", "%%shell\n", - "export PYTHONPATH=\"$PWD\" # to run *.py. files in subdirectories\n", "rm -rf runs # remove runs/\n", - "for m in yolov5n; do # models\n", - " python train.py --img 64 --batch 32 --weights $m.pt --epochs 1 --device 0 # train pretrained\n", - " python train.py --img 64 --batch 32 --weights '' --cfg $m.yaml --epochs 1 --device 0 # train scratch\n", - " for d in 0 cpu; do # devices\n", - " python val.py --weights $m.pt --device $d # val official\n", - " python val.py --weights runs/train/exp/weights/best.pt --device $d # val custom\n", - " python detect.py --weights $m.pt --device $d # detect official\n", - " python detect.py --weights runs/train/exp/weights/best.pt --device $d # detect custom\n", + "m=yolov5n # official weights\n", + "b=runs/train/exp/weights/best # best.pt checkpoint\n", + "python train.py --imgsz 64 --batch 32 --weights $m.pt --cfg $m.yaml --epochs 1 --device 0 # train\n", + "for d in 0 cpu; do # devices\n", + " for w in $m $b; do # weights\n", + " python val.py --imgsz 64 --batch 32 --weights $w.pt --device $d # val\n", + " python detect.py --imgsz 64 --weights $w.pt --device $d # detect\n", " done\n", - " python hubconf.py # hub\n", - " python models/yolo.py --cfg $m.yaml # build PyTorch model\n", - " python models/tf.py --weights $m.pt # build TensorFlow model\n", - " python export.py --img 64 --batch 1 --weights $m.pt --include torchscript onnx # export\n", - "done" + "done\n", + "python hubconf.py --model $m # hub\n", + "python models/tf.py --weights $m.pt # build TF model\n", + "python models/yolo.py --cfg $m.yaml # build PyTorch model\n", + "python export.py --weights $m.pt --img 64 --include torchscript # export" ], "execution_count": null, "outputs": [] From 7921351b4e4030a2db9e1488f8ef5a166abff17d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 30 Jul 2022 21:25:16 +0200 Subject: [PATCH 396/661] Deprecate torch 1.6.0 `compat _non_persistent_buffers_set` (#8797) Deprecate torch 1.6.0 compat _non_persistent_buffers_set --- models/experimental.py | 2 -- 1 file changed, 2 deletions(-) diff --git a/models/experimental.py b/models/experimental.py index db8e5b8e1dfd..0317c7526c99 100644 --- a/models/experimental.py +++ b/models/experimental.py @@ -89,8 +89,6 @@ def attempt_load(weights, device=None, inplace=True, fuse=True): if t is Detect and not isinstance(m.anchor_grid, list): delattr(m, 'anchor_grid') setattr(m, 'anchor_grid', [torch.zeros(1)] * m.nl) - elif t is Conv: - m._non_persistent_buffers_set = set() # torch 1.6.0 compatibility elif t is nn.Upsample and not hasattr(m, 'recompute_scale_factor'): m.recompute_scale_factor = None # torch 1.11.0 compatibility From 1e89807d9a208727e3f0e9bf26a1e286d0ce416b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 30 Jul 2022 22:19:40 +0200 Subject: [PATCH 397/661] `Detect.inplace=False` for multithread-safe inference (#8801) Detect.inplace=False for safe multithread inference --- hubconf.py | 1 + models/yolo.py | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/hubconf.py b/hubconf.py index 08122eaca9dc..5bb629005597 100644 --- a/hubconf.py +++ b/hubconf.py @@ -55,6 +55,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo if len(ckpt['model'].names) == classes: model.names = ckpt['model'].names # set class names attribute if autoshape: + model.model.model[-1].inplace = False # Detect.inplace=False for safe multithread inference model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS if not verbose: LOGGER.setLevel(logging.INFO) # reset to default diff --git a/models/yolo.py b/models/yolo.py index 56846815e08a..bc1893ccbc48 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -50,7 +50,7 @@ def __init__(self, nc=80, anchors=(), ch=(), inplace=True): # detection layer self.anchor_grid = [torch.zeros(1)] * self.nl # init anchor grid self.register_buffer('anchors', torch.tensor(anchors).float().view(self.nl, -1, 2)) # shape(nl,na,2) self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv - self.inplace = inplace # use in-place ops (e.g. slice assignment) + self.inplace = inplace # use inplace ops (e.g. slice assignment) def forward(self, x): z = [] # inference output From 59595c136581142766313c25d4fccd09c15a45b2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 31 Jul 2022 04:17:39 +0200 Subject: [PATCH 398/661] Update train.py for `val.run(half=amp)` (#8804) Disable FP16 validation if AMP checks fail or amp=False. --- train.py | 1 + 1 file changed, 1 insertion(+) diff --git a/train.py b/train.py index c298692b7335..dc93c22d621a 100644 --- a/train.py +++ b/train.py @@ -367,6 +367,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio results, maps, _ = val.run(data_dict, batch_size=batch_size // WORLD_SIZE * 2, imgsz=imgsz, + half=amp, model=ema.ema, single_cls=single_cls, dataloader=val_loader, From 34cb277dc5316d8c41cbc7e2020ccf9be5c7dd84 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 31 Jul 2022 14:17:23 +0200 Subject: [PATCH 399/661] Fix val.py 'no labels found bug' (#8806) Resolves https://github.com/ultralytics/yolov5/issues/8791 Bug first introduced in #8782 --- val.py | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/val.py b/val.py index 006ade37d03e..851d679d269b 100644 --- a/val.py +++ b/val.py @@ -182,7 +182,7 @@ def run( seen = 0 confusion_matrix = ConfusionMatrix(nc=nc) - names = {k: v for k, v in enumerate(model.names if hasattr(model, 'names') else model.module.names)} + names = dict(enumerate(model.names if hasattr(model, 'names') else model.module.names)) class_map = coco80_to_coco91_class() if is_coco else list(range(1000)) s = ('%20s' + '%11s' * 6) % ('Class', 'Images', 'Labels', 'P', 'R', 'mAP@.5', 'mAP@.5:.95') dt, p, r, f1, mp, mr, map50, map = [0.0, 0.0, 0.0], 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 @@ -250,7 +250,7 @@ def run( # Save/log if save_txt: - save_one_txt(predn, save_conf, shape, file=save_dir / 'labels' / (path.stem + '.txt')) + save_one_txt(predn, save_conf, shape, file=save_dir / 'labels' / f'{path.stem}.txt') if save_json: save_one_json(predn, jdict, path, class_map) # append to COCO-JSON dictionary callbacks.run('on_val_image_end', pred, predn, path, names, im[si]) @@ -268,9 +268,7 @@ def run( tp, fp, p, r, f1, ap, ap_class = ap_per_class(*stats, plot=plots, save_dir=save_dir, names=names) ap50, ap = ap[:, 0], ap.mean(1) # AP@0.5, AP@0.5:0.95 mp, mr, map50, map = p.mean(), r.mean(), ap50.mean(), ap.mean() - nt = np.bincount(stats[3].astype(int), minlength=nc) # number of targets per class - else: - nt = torch.zeros(1) + nt = np.bincount(stats[3].astype(int), minlength=nc) # number of targets per class # Print results pf = '%20s' + '%11i' * 2 + '%11.3g' * 4 # print format From 9559601b9a24812dc6ae7f3d88a47febef5d0757 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 31 Jul 2022 14:54:55 +0200 Subject: [PATCH 400/661] Update requirements.txt with tf-cpu and tf-aarch64 (#8807) --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 6313cecee578..a7c567a67edf 100644 --- a/requirements.txt +++ b/requirements.txt @@ -29,7 +29,7 @@ seaborn>=0.11.0 # nvidia-pyindex # TensorRT export # nvidia-tensorrt # TensorRT export # scikit-learn==0.19.2 # CoreML quantization -# tensorflow>=2.4.1 # TFLite export +# tensorflow>=2.4.1 # TFLite export (or tensorflow-cpu, tensorflow-aarch64) # tensorflowjs>=3.9.0 # TF.js export # openvino-dev # OpenVINO export From 555976b346b33483984dcd8ff05276bf1107dfc8 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 31 Jul 2022 15:23:57 +0200 Subject: [PATCH 401/661] FROM nvcr.io/nvidia/pytorch:22.07-py3 (#8808) --- utils/docker/Dockerfile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index 312d169d1a76..0e0d82225bc4 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -3,7 +3,7 @@ # Image is CUDA-optimized for YOLOv5 single/multi-GPU training and inference # Start FROM NVIDIA PyTorch image https://ngc.nvidia.com/catalog/containers/nvidia:pytorch -FROM nvcr.io/nvidia/pytorch:22.06-py3 +FROM nvcr.io/nvidia/pytorch:22.07-py3 RUN rm -rf /opt/pytorch # remove 1.2GB dir # Downloads to user config dir From 7b72d9a6071cb39a578362175903f3db00ebcc7a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 31 Jul 2022 16:12:32 +0200 Subject: [PATCH 402/661] Update ci-testing.yml streamlined tests (#8809) * Update ci-testing.yml streamlined tests * Update ci-testing.yml * Update ci-testing.yml --- .github/workflows/ci-testing.yml | 37 ++++++++++++++------------------ 1 file changed, 16 insertions(+), 21 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 5b492009d503..444bab75bbbc 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -95,27 +95,22 @@ jobs: shell: bash run: | # export PYTHONPATH="$PWD" # to run '$ python *.py' files in subdirectories - d=cpu # device - model=${{ matrix.model }} - best=runs/train/exp/weights/best.pt - # Train - python train.py --img 64 --batch 32 --weights $model.pt --cfg $model.yaml --epochs 1 --device $d - # Val - python val.py --img 64 --batch 32 --weights $model.pt --device $d - python val.py --img 64 --batch 32 --weights $best --device $d - # Detect - python detect.py --weights $model.pt --device $d - python detect.py --weights $best --device $d - python hubconf.py --model $model # hub - # Export - # python models/tf.py --weights $model.pt # build TF model - python models/yolo.py --cfg $model.yaml # build PyTorch model - python export.py --weights $model.pt --img 64 --include torchscript # export - # Python + m=${{ matrix.model }} # official weights + b=runs/train/exp/weights/best # best.pt checkpoint + python train.py --imgsz 64 --batch 32 --weights $m.pt --cfg $m.yaml --epochs 1 --device cpu # train + for d in cpu; do # devices + for w in $m $b; do # weights + python val.py --imgsz 64 --batch 32 --weights $w.pt --device $d # val + python detect.py --imgsz 64 --weights $w.pt --device $d # detect + done + done + python hubconf.py --model $m # hub + # python models/tf.py --weights $m.pt # build TF model + python models/yolo.py --cfg $m.yaml # build PyTorch model + python export.py --weights $m.pt --img 64 --include torchscript # export python - < Date: Sun, 31 Jul 2022 20:47:38 +0430 Subject: [PATCH 403/661] Check git status on upstream `ultralytics` or `origin` dynamically (#8694) * Add remote ultralytics and check git status with that * Simplify * Update general.py * Update general.py * s fix Co-authored-by: Glenn Jocher --- utils/general.py | 22 +++++++++++++++------- 1 file changed, 15 insertions(+), 7 deletions(-) diff --git a/utils/general.py b/utils/general.py index 67078338d762..bab0a5d9ab34 100755 --- a/utils/general.py +++ b/utils/general.py @@ -310,20 +310,28 @@ def git_describe(path=ROOT): # path must be a directory @try_except @WorkingDirectory(ROOT) -def check_git_status(): - # Recommend 'git pull' if code is out of date - msg = ', for updates see https://github.com/ultralytics/yolov5' +def check_git_status(repo='ultralytics/yolov5'): + # YOLOv5 status check, recommend 'git pull' if code is out of date + url = f'https://github.com/{repo}' + msg = f', for updates see {url}' s = colorstr('github: ') # string assert Path('.git').exists(), s + 'skipping check (not a git repository)' + msg assert not is_docker(), s + 'skipping check (Docker image)' + msg assert check_online(), s + 'skipping check (offline)' + msg - cmd = 'git fetch && git config --get remote.origin.url' - url = check_output(cmd, shell=True, timeout=5).decode().strip().rstrip('.git') # git fetch + splits = re.split(pattern=r'\s', string=check_output('git remote -v', shell=True).decode()) + matches = [repo in s for s in splits] + if any(matches): + remote = splits[matches.index(True) - 1] + else: + remote = 'ultralytics' + check_output(f'git remote add {remote} {url}', shell=True) + check_output(f'git fetch {remote}', shell=True, timeout=5) # git fetch branch = check_output('git rev-parse --abbrev-ref HEAD', shell=True).decode().strip() # checked out - n = int(check_output(f'git rev-list {branch}..origin/master --count', shell=True)) # commits behind + n = int(check_output(f'git rev-list {branch}..{remote}/master --count', shell=True)) # commits behind if n > 0: - s += f"⚠️ YOLOv5 is out of date by {n} commit{'s' * (n > 1)}. Use `git pull` or `git clone {url}` to update." + pull = 'git pull' if remote == 'origin' else f'git pull {remote} master' + s += f"⚠️ YOLOv5 is out of date by {n} commit{'s' * (n > 1)}. Use `{pull}` or `git clone {url}` to update." else: s += f'up to date with {url} ✅' LOGGER.info(emojis(s)) # emoji-safe From 40c41e42692011f32ce952b44b4bcb4f06e9e0b0 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 31 Jul 2022 19:57:40 +0200 Subject: [PATCH 404/661] Fix Colab-update pre-commit EOF bug (#8810) --- .pre-commit-config.yaml | 1 + 1 file changed, 1 insertion(+) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 9b8f28c77506..97da994e2917 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -16,6 +16,7 @@ repos: rev: v4.3.0 hooks: - id: end-of-file-fixer + stages: [commit] # avoid Colab update EOF issues - id: trailing-whitespace - id: check-case-conflict - id: check-yaml From 685332ede482488cec13a3d6c429d4f1e9b34960 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 31 Jul 2022 20:06:35 +0200 Subject: [PATCH 405/661] Update .pre-commit-config.yaml (#8811) --- .pre-commit-config.yaml | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 97da994e2917..fe26ed5a93a5 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -15,8 +15,9 @@ repos: - repo: https://github.com/pre-commit/pre-commit-hooks rev: v4.3.0 hooks: + - id: no-commit-to-branch + args: ['--branch', 'master'] - id: end-of-file-fixer - stages: [commit] # avoid Colab update EOF issues - id: trailing-whitespace - id: check-case-conflict - id: check-yaml From 0e165c50f79a8ac4286d1920ca7a48220dc5a9db Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 31 Jul 2022 20:34:03 +0200 Subject: [PATCH 406/661] Created using Colaboratory --- tutorial.ipynb | 314 +++++++++++++++++++++++++------------------------ 1 file changed, 160 insertions(+), 154 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index dcb1162b40af..b5cb4964aa6b 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -16,7 +16,7 @@ "accelerator": "GPU", "widgets": { "application/vnd.jupyter.widget-state+json": { - "572de771c7b34c1481def33bd5ed690d": { + "c79427d84662495db06b89a791d61f31": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", @@ -31,14 +31,14 @@ "_view_name": "HBoxView", "box_style": "", "children": [ - "IPY_MODEL_20c89dc0d82a4bdf8756bf5e34152292", - "IPY_MODEL_61026f684725441db2a640e531807675", - "IPY_MODEL_8d2e16d90e13449598d7b3fac75f78a3" + "IPY_MODEL_469c8e5ae4d64adea773341ec22d5851", + "IPY_MODEL_2435573a321341878622d79e1f48f3db", + "IPY_MODEL_a4dcb697b08b4b70ab3ef3ffa54c28e4" ], - "layout": "IPY_MODEL_a09d90f1bd374ece9a29bc6cfe07c072" + "layout": "IPY_MODEL_87495c10d22c4b82bd724a4d7c300df3" } }, - "20c89dc0d82a4bdf8756bf5e34152292": { + "469c8e5ae4d64adea773341ec22d5851": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", @@ -53,13 +53,13 @@ "_view_name": "HTMLView", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_801e720897804703b4d32f99f84cc3b8", + "layout": "IPY_MODEL_098c321358c24cdbb50f6c0e6623bf6c", "placeholder": "​", - "style": "IPY_MODEL_c9fb2e268cc94d508d909b3b72ac9df3", + "style": "IPY_MODEL_20184030ca9d4aef9dac0a149b89e4d3", "value": "100%" } }, - "61026f684725441db2a640e531807675": { + "2435573a321341878622d79e1f48f3db": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", @@ -75,15 +75,15 @@ "bar_style": "success", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_bfbc16e88df24fae93e8c80538e78273", + "layout": "IPY_MODEL_790808c9b4fb448aa136cc1ade0f95b5", "max": 818322941, "min": 0, "orientation": "horizontal", - "style": "IPY_MODEL_d9ffa50bddb7455ca4d67ec220c4a10c", + "style": "IPY_MODEL_99b822fd56b749318b38d8ccbc4ac469", "value": 818322941 } }, - "8d2e16d90e13449598d7b3fac75f78a3": { + "a4dcb697b08b4b70ab3ef3ffa54c28e4": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", @@ -98,13 +98,13 @@ "_view_name": "HTMLView", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_8be83ee30f804775aa55aeb021bf515b", + "layout": "IPY_MODEL_d542739146774953954e92db1666b951", "placeholder": "​", - "style": "IPY_MODEL_78e5b8dba72942bfacfee54ceec53784", - "value": " 780M/780M [01:28<00:00, 9.08MB/s]" + "style": "IPY_MODEL_e11f3a2c51204778832631a5f150b21d", + "value": " 780M/780M [02:31<00:00, 4.89MB/s]" } }, - "a09d90f1bd374ece9a29bc6cfe07c072": { + "87495c10d22c4b82bd724a4d7c300df3": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -156,7 +156,7 @@ "width": null } }, - "801e720897804703b4d32f99f84cc3b8": { + "098c321358c24cdbb50f6c0e6623bf6c": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -208,7 +208,7 @@ "width": null } }, - "c9fb2e268cc94d508d909b3b72ac9df3": { + "20184030ca9d4aef9dac0a149b89e4d3": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", @@ -223,7 +223,7 @@ "description_width": "" } }, - "bfbc16e88df24fae93e8c80538e78273": { + "790808c9b4fb448aa136cc1ade0f95b5": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -275,7 +275,7 @@ "width": null } }, - "d9ffa50bddb7455ca4d67ec220c4a10c": { + "99b822fd56b749318b38d8ccbc4ac469": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", @@ -291,7 +291,7 @@ "description_width": "" } }, - "8be83ee30f804775aa55aeb021bf515b": { + "d542739146774953954e92db1666b951": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -343,7 +343,7 @@ "width": null } }, - "78e5b8dba72942bfacfee54ceec53784": { + "e11f3a2c51204778832631a5f150b21d": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", @@ -403,7 +403,7 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "4bf03330-c2e8-43ec-c5da-b7f5e0b2b123" + "outputId": "7728cbd8-6240-4814-e8fe-a223b9e57ed9" }, "source": [ "!git clone https://github.com/ultralytics/yolov5 # clone\n", @@ -414,20 +414,20 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": null, + "execution_count": 1, "outputs": [ { "output_type": "stream", "name": "stderr", "text": [ - "YOLOv5 🚀 v6.1-257-g669f707 Python-3.7.13 torch-1.11.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" + "YOLOv5 🚀 v6.1-343-g685332e Python-3.7.13 torch-1.12.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" ] }, { "output_type": "stream", "name": "stdout", "text": [ - "Setup complete ✅ (8 CPUs, 51.0 GB RAM, 38.8/166.8 GB disk)\n" + "Setup complete ✅ (8 CPUs, 51.0 GB RAM, 38.6/166.8 GB disk)\n" ] } ] @@ -460,29 +460,29 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "1d1bb361-c8f3-4ddd-8a19-864bb993e7ac" + "outputId": "2d81665e-a0c4-489a-c92e-fe815223adfb" }, "source": [ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", - "display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" + "#display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": null, + "execution_count": 2, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mdetect: \u001b[0mweights=['yolov5s.pt'], source=data/images, data=data/coco128.yaml, imgsz=[640, 640], conf_thres=0.25, iou_thres=0.45, max_det=1000, device=, view_img=False, save_txt=False, save_conf=False, save_crop=False, nosave=False, classes=None, agnostic_nms=False, augment=False, visualize=False, update=False, project=runs/detect, name=exp, exist_ok=False, line_thickness=3, hide_labels=False, hide_conf=False, half=False, dnn=False\n", - "YOLOv5 🚀 v6.1-257-g669f707 Python-3.7.13 torch-1.11.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.1-343-g685332e Python-3.7.13 torch-1.12.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5s.pt to yolov5s.pt...\n", - "100% 14.1M/14.1M [00:00<00:00, 225MB/s]\n", + "100% 14.1M/14.1M [00:02<00:00, 6.87MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5s summary: 213 layers, 7225885 parameters, 0 gradients\n", - "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, Done. (0.013s)\n", - "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, Done. (0.015s)\n", - "Speed: 0.6ms pre-process, 14.1ms inference, 23.9ms NMS per image at shape (1, 3, 640, 640)\n", + "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, Done. (0.014s)\n", + "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, Done. (0.019s)\n", + "Speed: 0.5ms pre-process, 16.3ms inference, 22.1ms NMS per image at shape (1, 3, 640, 640)\n", "Results saved to \u001b[1mruns/detect/exp\u001b[0m\n" ] } @@ -526,27 +526,27 @@ "base_uri": "https://localhost:8080/", "height": 49, "referenced_widgets": [ - "572de771c7b34c1481def33bd5ed690d", - "20c89dc0d82a4bdf8756bf5e34152292", - "61026f684725441db2a640e531807675", - "8d2e16d90e13449598d7b3fac75f78a3", - "a09d90f1bd374ece9a29bc6cfe07c072", - "801e720897804703b4d32f99f84cc3b8", - "c9fb2e268cc94d508d909b3b72ac9df3", - "bfbc16e88df24fae93e8c80538e78273", - "d9ffa50bddb7455ca4d67ec220c4a10c", - "8be83ee30f804775aa55aeb021bf515b", - "78e5b8dba72942bfacfee54ceec53784" + "c79427d84662495db06b89a791d61f31", + "469c8e5ae4d64adea773341ec22d5851", + "2435573a321341878622d79e1f48f3db", + "a4dcb697b08b4b70ab3ef3ffa54c28e4", + "87495c10d22c4b82bd724a4d7c300df3", + "098c321358c24cdbb50f6c0e6623bf6c", + "20184030ca9d4aef9dac0a149b89e4d3", + "790808c9b4fb448aa136cc1ade0f95b5", + "99b822fd56b749318b38d8ccbc4ac469", + "d542739146774953954e92db1666b951", + "e11f3a2c51204778832631a5f150b21d" ] }, - "outputId": "47c358af-138d-42d9-ca89-4364283df9e3" + "outputId": "d880071b-84ce-4567-9e42-a3c3a78bff73" }, "source": [ "# Download COCO val\n", "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip')\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip" ], - "execution_count": null, + "execution_count": 3, "outputs": [ { "output_type": "display_data", @@ -557,7 +557,7 @@ "application/vnd.jupyter.widget-view+json": { "version_major": 2, "version_minor": 0, - "model_id": "572de771c7b34c1481def33bd5ed690d" + "model_id": "c79427d84662495db06b89a791d61f31" } }, "metadata": {} @@ -571,53 +571,53 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "979fe4c2-a058-44de-b401-3cb67878a1b9" + "outputId": "da9456fa-6663-44a8-975b-c99e89d0eb06" }, "source": [ "# Run YOLOv5x on COCO val\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" ], - "execution_count": null, + "execution_count": 4, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mval: \u001b[0mdata=/content/yolov5/data/coco.yaml, weights=['yolov5x.pt'], batch_size=32, imgsz=640, conf_thres=0.001, iou_thres=0.65, task=val, device=, workers=8, single_cls=False, augment=False, verbose=False, save_txt=False, save_hybrid=False, save_conf=False, save_json=True, project=runs/val, name=exp, exist_ok=False, half=True, dnn=False\n", - "YOLOv5 🚀 v6.1-257-g669f707 Python-3.7.13 torch-1.11.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.1-343-g685332e Python-3.7.13 torch-1.12.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5x.pt to yolov5x.pt...\n", - "100% 166M/166M [00:04<00:00, 39.4MB/s]\n", + "100% 166M/166M [00:16<00:00, 10.3MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5x summary: 444 layers, 86705005 parameters, 0 gradients\n", "Downloading https://ultralytics.com/assets/Arial.ttf to /root/.config/Ultralytics/Arial.ttf...\n", - "100% 755k/755k [00:00<00:00, 47.9MB/s]\n", - "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 8742.34it/s]\n", + "100% 755k/755k [00:00<00:00, 14.8MB/s]\n", + "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 11214.34it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mNew cache created: /content/datasets/coco/val2017.cache\n", - " Class Images Labels P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:11<00:00, 2.21it/s]\n", + " Class Images Labels P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:05<00:00, 2.39it/s]\n", " all 5000 36335 0.743 0.625 0.683 0.504\n", - "Speed: 0.1ms pre-process, 4.9ms inference, 1.2ms NMS per image at shape (32, 3, 640, 640)\n", + "Speed: 0.1ms pre-process, 4.7ms inference, 1.1ms NMS per image at shape (32, 3, 640, 640)\n", "\n", "Evaluating pycocotools mAP... saving runs/val/exp/yolov5x_predictions.json...\n", "loading annotations into memory...\n", - "Done (t=0.42s)\n", + "Done (t=0.38s)\n", "creating index...\n", "index created!\n", "Loading and preparing results...\n", - "DONE (t=4.91s)\n", + "DONE (t=5.39s)\n", "creating index...\n", "index created!\n", "Running per image evaluation...\n", "Evaluate annotation type *bbox*\n", - "DONE (t=77.89s).\n", + "DONE (t=71.33s).\n", "Accumulating evaluation results...\n", - "DONE (t=15.36s).\n", + "DONE (t=12.45s).\n", " Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.506\n", " Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.688\n", " Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.549\n", " Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.340\n", - " Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.557\n", + " Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.558\n", " Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.651\n", " Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.382\n", " Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.631\n", @@ -731,26 +731,31 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "be9424b5-34d6-4de0-e951-2c5ae334721e" + "outputId": "9fe5caba-6b0f-4b6e-93a8-4075dae0ee35" }, "source": [ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": null, + "execution_count": 5, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ - "\u001b[34m\u001b[1mtrain: \u001b[0mweights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch-low.yaml, epochs=3, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, noplots=False, evolve=None, bucket=, cache=ram, image_weights=False, device=, multi_scale=False, single_cls=False, optimizer=SGD, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, cos_lr=False, label_smoothing=0.0, patience=100, freeze=[0], save_period=-1, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest\n", - "\u001b[34m\u001b[1mgithub: \u001b[0mup to date with https://github.com/ultralytics/yolov5 ✅\n", - "YOLOv5 🚀 v6.1-257-g669f707 Python-3.7.13 torch-1.11.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mweights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch-low.yaml, epochs=3, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, noplots=False, evolve=None, bucket=, cache=ram, image_weights=False, device=, multi_scale=False, single_cls=False, optimizer=SGD, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, cos_lr=False, label_smoothing=0.0, patience=100, freeze=[0], save_period=-1, seed=0, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest\n", + "\u001b[34m\u001b[1mgithub: \u001b[0mskipping check (Docker image), for updates see https://github.com/ultralytics/yolov5\n", + "YOLOv5 🚀 v6.1-343-g685332e Python-3.7.13 torch-1.12.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "\u001b[34m\u001b[1mhyperparameters: \u001b[0mlr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0\n", "\u001b[34m\u001b[1mWeights & Biases: \u001b[0mrun 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs (RECOMMENDED)\n", "\u001b[34m\u001b[1mTensorBoard: \u001b[0mStart with 'tensorboard --logdir runs/train', view at http://localhost:6006/\n", "\n", + "Dataset not found ⚠️, missing paths ['/content/datasets/coco128/images/train2017']\n", + "Downloading https://ultralytics.com/assets/coco128.zip to coco128.zip...\n", + "100% 6.66M/6.66M [00:00<00:00, 31.8MB/s]\n", + "Dataset download success ✅ (1.5s), saved to \u001b[1m/content/datasets\u001b[0m\n", + "\n", " from n params module arguments \n", " 0 -1 1 3520 models.common.Conv [3, 32, 6, 2, 2] \n", " 1 -1 1 18560 models.common.Conv [32, 64, 3, 2] \n", @@ -777,17 +782,18 @@ " 22 [-1, 10] 1 0 models.common.Concat [1] \n", " 23 -1 1 1182720 models.common.C3 [512, 512, 1, False] \n", " 24 [17, 20, 23] 1 229245 models.yolo.Detect [80, [[10, 13, 16, 30, 33, 23], [30, 61, 62, 45, 59, 119], [116, 90, 156, 198, 373, 326]], [128, 256, 512]]\n", - "Model summary: 270 layers, 7235389 parameters, 7235389 gradients\n", + "Model summary: 270 layers, 7235389 parameters, 7235389 gradients, 16.6 GFLOPs\n", "\n", "Transferred 349/349 items from yolov5s.pt\n", "\u001b[34m\u001b[1mAMP: \u001b[0mchecks passed ✅\n", "Scaled weight_decay = 0.0005\n", "\u001b[34m\u001b[1moptimizer:\u001b[0m SGD with parameter groups 57 weight (no decay), 60 weight, 60 bias\n", "\u001b[34m\u001b[1malbumentations: \u001b[0mversion 1.0.3 required by YOLOv5, but version 0.1.12 is currently installed\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017.cache' images and labels... 128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00 Date: Mon, 1 Aug 2022 02:09:36 +0200 Subject: [PATCH 407/661] Update .pre-commit-config.yaml (#8812) * Update .pre-commit-config.yaml Comment EOF fixer * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- .pre-commit-config.yaml | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index fe26ed5a93a5..76716d160ac1 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -15,9 +15,7 @@ repos: - repo: https://github.com/pre-commit/pre-commit-hooks rev: v4.3.0 hooks: - - id: no-commit-to-branch - args: ['--branch', 'master'] - - id: end-of-file-fixer + # - id: end-of-file-fixer - id: trailing-whitespace - id: check-case-conflict - id: check-yaml From 39ce8ca19a1b97e36c73d86ecc70c2c3e42ac5c0 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 1 Aug 2022 03:01:44 +0200 Subject: [PATCH 408/661] Remove `assert not is_docker()` from GitHub checks (#8813) * Update * Update --- utils/general.py | 1 - 1 file changed, 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index bab0a5d9ab34..22181d3faeb9 100755 --- a/utils/general.py +++ b/utils/general.py @@ -316,7 +316,6 @@ def check_git_status(repo='ultralytics/yolov5'): msg = f', for updates see {url}' s = colorstr('github: ') # string assert Path('.git').exists(), s + 'skipping check (not a git repository)' + msg - assert not is_docker(), s + 'skipping check (Docker image)' + msg assert check_online(), s + 'skipping check (offline)' + msg splits = re.split(pattern=r'\s', string=check_output('git remote -v', shell=True).decode()) From 7b9cc3205ae2cd9fdb0a56ca2818e17c5ae8346e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 1 Aug 2022 03:33:28 +0200 Subject: [PATCH 409/661] Add .git to .dockerignore (#8815) --- .dockerignore | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.dockerignore b/.dockerignore index af51ccc3d8df..3b669254e779 100644 --- a/.dockerignore +++ b/.dockerignore @@ -1,5 +1,5 @@ # Repo-specific DockerIgnore ------------------------------------------------------------------------------------------- -#.git +.git .cache .idea runs From 0669f1b27bbdcbdbb0e2baf4e9f09c6fc8337ec7 Mon Sep 17 00:00:00 2001 From: UnglvKitDe <100289696+UnglvKitDe@users.noreply.github.com> Date: Mon, 1 Aug 2022 12:08:46 +0200 Subject: [PATCH 410/661] Add tensor hooks and 10.0 gradient clipping (#8598) * Add tensor hooks and gradient clipping https://github.com/ultralytics/yolov5/issues/8578 * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Remove retain_grad(), because its not necessary * Update train.py * Simplify * Update train.py * Update train.py * Update train.py * Update train.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- train.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/train.py b/train.py index dc93c22d621a..6ada2a2f121b 100644 --- a/train.py +++ b/train.py @@ -131,6 +131,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio freeze = [f'model.{x}.' for x in (freeze if len(freeze) > 1 else range(freeze[0]))] # layers to freeze for k, v in model.named_parameters(): v.requires_grad = True # train all layers + v.register_hook(lambda x: torch.nan_to_num(x)) # NaN to 0.0 if any(x in k for x in freeze): LOGGER.info(f'freezing {k}') v.requires_grad = False @@ -334,8 +335,10 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # Backward scaler.scale(loss).backward() - # Optimize + # Optimize - https://pytorch.org/docs/master/notes/amp_examples.html if ni - last_opt_step >= accumulate: + scaler.unscale_(optimizer) # unscale gradients + torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm=10.0) # clip gradients scaler.step(optimizer) # optimizer.step scaler.update() optimizer.zero_grad() From 59578f2782cfbf4fe2b270a1c533f45b7cbbd56f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 1 Aug 2022 20:28:24 +0200 Subject: [PATCH 411/661] Update README.md with contributors.png (#8820) * Update README.md with contributors.png Replace dynamic svg from opencollective with static png for improved stability and lighter size (400kB vs 2MB). @AyushExel * Update README.md * Update README.md * Update README_cn.md * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- .github/README_cn.md | 4 +++- README.md | 3 ++- 2 files changed, 5 insertions(+), 2 deletions(-) diff --git a/.github/README_cn.md b/.github/README_cn.md index 7e90336d5157..b653d435cfd1 100644 --- a/.github/README_cn.md +++ b/.github/README_cn.md @@ -249,7 +249,9 @@ We are super excited about our first-ever Ultralytics YOLOv5 🚀 EXPORT Competi ##
贡献
我们重视您的意见! 我们希望给大家提供尽可能的简单和透明的方式对 YOLOv5 做出贡献。开始之前请先点击并查看我们的 [贡献指南](CONTRIBUTING.md)，填写[YOLOv5调查问卷](https://ultralytics.com/survey?utm_source=github&utm_medium=social&utm_campaign=Survey) 来向我们发送您的经验反馈。真诚感谢我们所有的贡献者! - + + + ##
联系
diff --git a/README.md b/README.md index b0ea0a5d814c..b959871211e5 100644 --- a/README.md +++ b/README.md @@ -259,7 +259,8 @@ We are super excited about our first-ever Ultralytics YOLOv5 🚀 EXPORT Competi We love your input! We want to make contributing to YOLOv5 as easy and transparent as possible. Please see our [Contributing Guide](CONTRIBUTING.md) to get started, and fill out the [YOLOv5 Survey](https://ultralytics.com/survey?utm_source=github&utm_medium=social&utm_campaign=Survey) to send us feedback on your experiences. Thank you to all our contributors! - + + ##
Contact
From f3c78a387e9a344b903fbd7bd12bfab2ea292351 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 1 Aug 2022 21:39:04 +0200 Subject: [PATCH 412/661] Remove hook `torch.nan_to_num(x)` (#8826) * Remove hook `torch.nan_to_num(x)` Observed erratic training behavior (green line) with the nan_to_num hook in classifier branch. I'm going to remove it from master. * Update train.py --- train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train.py b/train.py index 6ada2a2f121b..20fef265110c 100644 --- a/train.py +++ b/train.py @@ -131,7 +131,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio freeze = [f'model.{x}.' for x in (freeze if len(freeze) > 1 else range(freeze[0]))] # layers to freeze for k, v in model.named_parameters(): v.requires_grad = True # train all layers - v.register_hook(lambda x: torch.nan_to_num(x)) # NaN to 0.0 + # v.register_hook(lambda x: torch.nan_to_num(x)) # NaN to 0 (commented for erratic training results) if any(x in k for x in freeze): LOGGER.info(f'freezing {k}') v.requires_grad = False From ba140e568555503c54a66c974e15922da9422f1a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 1 Aug 2022 21:45:31 +0200 Subject: [PATCH 413/661] RUN git clone instead of COPY to `/usr/src/app` (#8827) Update --- utils/docker/Dockerfile | 4 ++-- utils/docker/Dockerfile-arm64 | 4 ++-- utils/docker/Dockerfile-cpu | 4 ++-- 3 files changed, 6 insertions(+), 6 deletions(-) diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index 0e0d82225bc4..2280f209e6a1 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -25,8 +25,8 @@ RUN mkdir -p /usr/src/app WORKDIR /usr/src/app # Copy contents -COPY . /usr/src/app -RUN git clone https://github.com/ultralytics/yolov5 /usr/src/yolov5 +# COPY . /usr/src/app (issues as not a .git directory) +RUN git clone https://github.com/ultralytics/yolov5 /usr/src/app # Set environment variables ENV OMP_NUM_THREADS=8 diff --git a/utils/docker/Dockerfile-arm64 b/utils/docker/Dockerfile-arm64 index bca161e67a37..fe92c8d56146 100644 --- a/utils/docker/Dockerfile-arm64 +++ b/utils/docker/Dockerfile-arm64 @@ -29,8 +29,8 @@ RUN mkdir -p /usr/src/app WORKDIR /usr/src/app # Copy contents -COPY . /usr/src/app -RUN git clone https://github.com/ultralytics/yolov5 /usr/src/yolov5 +# COPY . /usr/src/app (issues as not a .git directory) +RUN git clone https://github.com/ultralytics/yolov5 /usr/src/app # Usage Examples ------------------------------------------------------------------------------------------------------- diff --git a/utils/docker/Dockerfile-cpu b/utils/docker/Dockerfile-cpu index f05e920ad53f..d61dfeffe22c 100644 --- a/utils/docker/Dockerfile-cpu +++ b/utils/docker/Dockerfile-cpu @@ -26,8 +26,8 @@ RUN mkdir -p /usr/src/app WORKDIR /usr/src/app # Copy contents -COPY . /usr/src/app -RUN git clone https://github.com/ultralytics/yolov5 /usr/src/yolov5 +# COPY . /usr/src/app (issues as not a .git directory) +RUN git clone https://github.com/ultralytics/yolov5 /usr/src/app # Usage Examples ------------------------------------------------------------------------------------------------------- From b7635efb6ee953615b4ca7d13017d79511ccd3be Mon Sep 17 00:00:00 2001 From: "pre-commit-ci[bot]" <66853113+pre-commit-ci[bot]@users.noreply.github.com> Date: Mon, 1 Aug 2022 21:48:59 +0200 Subject: [PATCH 414/661] [pre-commit.ci] pre-commit suggestions (#8828) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit updates: - [github.com/asottile/pyupgrade: v2.34.0 → v2.37.3](https://github.com/asottile/pyupgrade/compare/v2.34.0...v2.37.3) - [github.com/PyCQA/flake8: 4.0.1 → 5.0.2](https://github.com/PyCQA/flake8/compare/4.0.1...5.0.2) Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- .pre-commit-config.yaml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 76716d160ac1..43aca019feb1 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -24,7 +24,7 @@ repos: - id: check-docstring-first - repo: https://github.com/asottile/pyupgrade - rev: v2.34.0 + rev: v2.37.3 hooks: - id: pyupgrade name: Upgrade code @@ -58,7 +58,7 @@ repos: - id: yesqa - repo: https://github.com/PyCQA/flake8 - rev: 4.0.1 + rev: 5.0.2 hooks: - id: flake8 name: PEP8 From 2e10909905b1e0e7eb7bac086600fe7ee2c0e6a5 Mon Sep 17 00:00:00 2001 From: Jackson Argo Date: Mon, 1 Aug 2022 19:46:08 -0400 Subject: [PATCH 415/661] Fix missing attr model.model when loading custom yolov model (#8830) * Update hubconf.py Loading a custom yolov model causes this line to fail. Adding a test to check if the model actually has a model.model field. With this check, I'm able to load the model no prob. Loading model via ```py model = torch.hub.load( 'ultralytics/yolov5', 'custom', 'models/frozen_backbone_coco_unlabeled_best.onnx', autoshape=True, force_reload=False ) ``` Causes traceback: ``` Traceback (most recent call last): File "/Users/jackson/Documents/GitHub/w210-capstone/.venv/lib/python3.10/site-packages/flask/app.py", line 2077, in wsgi_app response = self.full_dispatch_request() File "/Users/jackson/Documents/GitHub/w210-capstone/.venv/lib/python3.10/site-packages/flask/app.py", line 1525, in full_dispatch_request rv = self.handle_user_exception(e) File "/Users/jackson/Documents/GitHub/w210-capstone/.venv/lib/python3.10/site-packages/flask/app.py", line 1523, in full_dispatch_request rv = self.dispatch_request() File "/Users/jackson/Documents/GitHub/w210-capstone/.venv/lib/python3.10/site-packages/flask/app.py", line 1509, in dispatch_request return self.ensure_sync(self.view_functions[rule.endpoint])(**req.view_args) File "/Users/jackson/Documents/GitHub/w210-capstone/api/endpoints/predictions.py", line 26, in post_predictions yolov_predictions = predict_bounding_boxes_for_collection(collection_id) File "/Users/jackson/Documents/GitHub/w210-capstone/api/predictions/predict_bounding_boxes.py", line 43, in predict_bounding_boxes_for_collection model = torch.hub.load( File "/Users/jackson/Documents/GitHub/w210-capstone/.venv/lib/python3.10/site-packages/torch/hub.py", line 404, in load model = _load_local(repo_or_dir, model, *args, **kwargs) File "/Users/jackson/Documents/GitHub/w210-capstone/.venv/lib/python3.10/site-packages/torch/hub.py", line 433, in _load_local model = entry(*args, **kwargs) File "/Users/jackson/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 72, in custom return _create(path, autoshape=autoshape, verbose=_verbose, device=device) File "/Users/jackson/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 67, in _create raise Exception(s) from e Exception: 'DetectMultiBackend' object has no attribute 'model'. Cache may be out of date, try `force_reload=True` or see https://github.com/ultralytics/yolov5/issues/36 for help. Exception on /api/v1/predictions [POST] Traceback (most recent call last): File "/Users/jackson/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 58, in _create model.model.model[-1].inplace = False # Detect.inplace=False for safe multithread inference File "/Users/jackson/Documents/GitHub/w210-capstone/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1185, in __getattr__ raise AttributeError("'{}' object has no attribute '{}'".format( AttributeError: 'DetectMultiBackend' object has no attribute 'model' ``` * Update hubconf.py * Update common.py Co-authored-by: Glenn Jocher --- hubconf.py | 12 +++++++----- models/common.py | 3 +++ 2 files changed, 10 insertions(+), 5 deletions(-) diff --git a/hubconf.py b/hubconf.py index 5bb629005597..011eaa57ff34 100644 --- a/hubconf.py +++ b/hubconf.py @@ -29,6 +29,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo from pathlib import Path from models.common import AutoShape, DetectMultiBackend + from models.experimental import attempt_load from models.yolo import Model from utils.downloads import attempt_download from utils.general import LOGGER, check_requirements, intersect_dicts, logging @@ -42,8 +43,12 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo try: device = select_device(device) if pretrained and channels == 3 and classes == 80: - model = DetectMultiBackend(path, device=device, fuse=autoshape) # download/load FP32 model - # model = models.experimental.attempt_load(path, map_location=device) # download/load FP32 model + try: + model = DetectMultiBackend(path, device=device, fuse=autoshape) # detection model + if autoshape: + model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS + except Exception: + model = attempt_load(path, device=device, fuse=False) # arbitrary model else: cfg = list((Path(__file__).parent / 'models').rglob(f'{path.stem}.yaml'))[0] # model.yaml path model = Model(cfg, channels, classes) # create model @@ -54,9 +59,6 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo model.load_state_dict(csd, strict=False) # load if len(ckpt['model'].names) == classes: model.names = ckpt['model'].names # set class names attribute - if autoshape: - model.model.model[-1].inplace = False # Detect.inplace=False for safe multithread inference - model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS if not verbose: LOGGER.setLevel(logging.INFO) # reset to default return model.to(device) diff --git a/models/common.py b/models/common.py index 959c965e6002..c898d94a921a 100644 --- a/models/common.py +++ b/models/common.py @@ -562,6 +562,9 @@ def __init__(self, model, verbose=True): self.dmb = isinstance(model, DetectMultiBackend) # DetectMultiBackend() instance self.pt = not self.dmb or model.pt # PyTorch model self.model = model.eval() + if self.pt: + m = self.model.model.model[-1] if self.dmb else self.model.model[-1] # Detect() + m.inplace = False # Detect.inplace=False for safe multithread inference def _apply(self, fn): # Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers From 08c8c3e00a1b0fc7f03a7e76ca3cbf7a0d8542ae Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 2 Aug 2022 15:13:58 +0200 Subject: [PATCH 416/661] New `smart_resume()` (#8838) * New `smart_resume()` * Update torch_utils.py * Update torch_utils.py * Update torch_utils.py * fix --- train.py | 33 ++++++--------------------------- utils/torch_utils.py | 19 +++++++++++++++++++ 2 files changed, 25 insertions(+), 27 deletions(-) diff --git a/train.py b/train.py index 20fef265110c..99a43f8614c4 100644 --- a/train.py +++ b/train.py @@ -54,7 +54,7 @@ from utils.metrics import fitness from utils.plots import plot_evolve, plot_labels from utils.torch_utils import (EarlyStopping, ModelEMA, de_parallel, select_device, smart_DDP, smart_optimizer, - torch_distributed_zero_first) + smart_resume, torch_distributed_zero_first) LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html RANK = int(os.getenv('RANK', -1)) @@ -163,26 +163,9 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio ema = ModelEMA(model) if RANK in {-1, 0} else None # Resume - start_epoch, best_fitness = 0, 0.0 + best_fitness, start_epoch = 0.0, 0 if pretrained: - # Optimizer - if ckpt['optimizer'] is not None: - optimizer.load_state_dict(ckpt['optimizer']) - best_fitness = ckpt['best_fitness'] - - # EMA - if ema and ckpt.get('ema'): - ema.ema.load_state_dict(ckpt['ema'].float().state_dict()) - ema.updates = ckpt['updates'] - - # Epochs - start_epoch = ckpt['epoch'] + 1 - if resume: - assert start_epoch > 0, f'{weights} training to {epochs} epochs is finished, nothing to resume.' - if epochs < start_epoch: - LOGGER.info(f"{weights} has been trained for {ckpt['epoch']} epochs. Fine-tuning for {epochs} more epochs.") - epochs += ckpt['epoch'] # finetune additional epochs - + best_fitness, start_epoch, epochs = smart_resume(ckpt, optimizer, ema, weights, epochs, resume) del ckpt, csd # DP mode @@ -212,8 +195,8 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio quad=opt.quad, prefix=colorstr('train: '), shuffle=True) - mlc = int(np.concatenate(dataset.labels, 0)[:, 0].max()) # max label class - nb = len(train_loader) # number of batches + labels = np.concatenate(dataset.labels, 0) + mlc = int(labels[:, 0].max()) # max label class assert mlc < nc, f'Label class {mlc} exceeds nc={nc} in {data}. Possible class labels are 0-{nc - 1}' # Process 0 @@ -232,10 +215,6 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio prefix=colorstr('val: '))[0] if not resume: - labels = np.concatenate(dataset.labels, 0) - # c = torch.tensor(labels[:, 0]) # classes - # cf = torch.bincount(c.long(), minlength=nc) + 1. # frequency - # model._initialize_biases(cf.to(device)) if plots: plot_labels(labels, names, save_dir) @@ -263,6 +242,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # Start training t0 = time.time() + nb = len(train_loader) # number of batches nw = max(round(hyp['warmup_epochs'] * nb), 100) # number of warmup iterations, max(3 epochs, 100 iterations) # nw = min(nw, (epochs - start_epoch) / 2 * nb) # limit warmup to < 1/2 of training last_opt_step = -1 @@ -510,7 +490,6 @@ def main(opt, callbacks=Callbacks()): with open(Path(ckpt).parent.parent / 'opt.yaml', errors='ignore') as f: opt = argparse.Namespace(**yaml.safe_load(f)) # replace opt.cfg, opt.weights, opt.resume = '', ckpt, True # reinstate - LOGGER.info(f'Resuming training from {ckpt}') else: opt.data, opt.cfg, opt.hyp, opt.weights, opt.project = \ check_file(opt.data), check_yaml(opt.cfg), check_yaml(opt.hyp), str(opt.weights), str(opt.project) # checks diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 5f2a22c36f1a..391ddead2985 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -306,6 +306,25 @@ def smart_optimizer(model, name='Adam', lr=0.001, momentum=0.9, weight_decay=1e- return optimizer +def smart_resume(ckpt, optimizer, ema=None, weights='yolov5s.pt', epochs=300, resume=True): + # Resume training from a partially trained checkpoint + best_fitness = 0.0 + start_epoch = ckpt['epoch'] + 1 + if ckpt['optimizer'] is not None: + optimizer.load_state_dict(ckpt['optimizer']) # optimizer + best_fitness = ckpt['best_fitness'] + if ema and ckpt.get('ema'): + ema.ema.load_state_dict(ckpt['ema'].float().state_dict()) # EMA + ema.updates = ckpt['updates'] + if resume: + assert start_epoch > 0, f'{weights} training to {epochs} epochs is finished, nothing to resume.' + LOGGER.info(f'Resuming training from {weights} for {epochs - start_epoch} more epochs to {epochs} total epochs') + if epochs < start_epoch: + LOGGER.info(f"{weights} has been trained for {ckpt['epoch']} epochs. Fine-tuning for {epochs} more epochs.") + epochs += ckpt['epoch'] # finetune additional epochs + return best_fitness, start_epoch, epochs + + class EarlyStopping: # YOLOv5 simple early stopper def __init__(self, patience=30): From e5991c986725d1229b6d1f5b1533e10f9b41c850 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 3 Aug 2022 00:57:40 +0200 Subject: [PATCH 417/661] Created using Colaboratory --- tutorial.ipynb | 30 +++++++++++++----------------- 1 file changed, 13 insertions(+), 17 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index b5cb4964aa6b..83be1039f22f 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -414,7 +414,7 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": 1, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -466,7 +466,7 @@ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", "#display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": 2, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -546,7 +546,7 @@ "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip')\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip" ], - "execution_count": 3, + "execution_count": null, "outputs": [ { "output_type": "display_data", @@ -577,7 +577,7 @@ "# Run YOLOv5x on COCO val\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" ], - "execution_count": 4, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -737,7 +737,7 @@ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": 5, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -917,13 +917,14 @@ "id": "DLI1JmHU7B0l" }, "source": [ - "## Weights & Biases Logging 🌟 NEW\n", + "## Weights & Biases Logging\n", "\n", - "[Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_notebook) (W&B) is now integrated with YOLOv5 for real-time visualization and cloud logging of training runs. This allows for better run comparison and introspection, as well improved visibility and collaboration for teams. To enable W&B `pip install wandb`, and then train normally (you will be guided through setup on first use). \n", + "[Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_notebook) (W&B) is integrated with YOLOv5 for real-time visualization and cloud logging of training runs. This allows for better run comparison and introspection, as well improved visibility and collaboration for teams. To enable W&B `pip install wandb`, and then train normally (you will be guided through setup on first use). \n", "\n", - "During training you will see live updates at [https://wandb.ai/home](https://wandb.ai/home?utm_campaign=repo_yolo_notebook), and you can create and share detailed [Reports](https://wandb.ai/glenn-jocher/yolov5_tutorial/reports/YOLOv5-COCO128-Tutorial-Results--VmlldzozMDI5OTY) of your results. For more information see the [YOLOv5 Weights & Biases Tutorial](https://github.com/ultralytics/yolov5/issues/1289). \n", + "During training you will see live updates at [https://wandb.ai/home](https://wandb.ai/home?utm_campaign=repo_yolo_notebook), and you can create and share detailed [Reports](https://wandb.ai/glenn-jocher/yolov5_tutorial/reports/YOLOv5-COCO128-Tutorial-Results--VmlldzozMDI5OTY) of your results. For more information see the [YOLOv5 Weights & Biases Tutorial](https://github.com/ultralytics/yolov5/issues/1289). \n", "\n", - "

\"Weights

" + "\n", + "\"Weights" ] }, { @@ -934,16 +935,11 @@ "source": [ "## Local Logging\n", "\n", - "All results are logged by default to `runs/train`, with a new experiment directory created for each new training as `runs/train/exp2`, `runs/train/exp3`, etc. View train and val jpgs to see mosaics, labels, predictions and augmentation effects. Note an Ultralytics **Mosaic Dataloader** is used for training (shown below), which combines 4 images into 1 mosaic during training.\n", - "\n", - "> \n", - "`train_batch0.jpg` shows train batch 0 mosaics and labels\n", + "All results are logged by default to `runs/train`, with a new experiment directory created for each new training as `runs/train/exp2`, `runs/train/exp3`, etc. View train and val statistics, mosaics, labels, predictions and augmentations, as well as metrics and charts including Precision-Recall curves and Confusion Matrices. \n", "\n", - "> \n", - "`test_batch0_labels.jpg` shows val batch 0 labels\n", + "A **Mosaic Dataloader** is used for training (shown in train*.jpg images), which combines 4 images into 1 mosaic during training.\n", "\n", - "> \n", - "`test_batch0_pred.jpg` shows val batch 0 _predictions_\n", + "\"Local\n", "\n", "Training results are automatically logged to [Tensorboard](https://www.tensorflow.org/tensorboard) and [CSV](https://github.com/ultralytics/yolov5/pull/4148) as `results.csv`, which is plotted as `results.png` (below) after training completes. You can also plot any `results.csv` file manually:\n", "\n", From 4d8d84b0ea7147aca64e7c38ce1bdb5fbb9c5a53 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 3 Aug 2022 12:49:02 +0200 Subject: [PATCH 418/661] Created using Colaboratory --- tutorial.ipynb | 17 +++++------------ 1 file changed, 5 insertions(+), 12 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 83be1039f22f..2aaa93b53df6 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -686,6 +686,8 @@ "- **Training Results** are saved to `runs/train/` with incrementing run directories, i.e. `runs/train/exp2`, `runs/train/exp3` etc.\n", "

\n", "\n", + "A **Mosaic Dataloader** is used for training which combines 4 images into 1 mosaic.\n", + "\n", "## Train on Custom Data with Roboflow 🌟 NEW\n", "\n", "[Roboflow](https://roboflow.com/?ref=ultralytics) enables you to easily **organize, label, and prepare** a high quality dataset with your own custom data. Roboflow also makes it easy to establish an active learning pipeline, collaborate with your team on dataset improvement, and integrate directly into your model building workflow with the `roboflow` pip package.\n", @@ -935,20 +937,11 @@ "source": [ "## Local Logging\n", "\n", - "All results are logged by default to `runs/train`, with a new experiment directory created for each new training as `runs/train/exp2`, `runs/train/exp3`, etc. View train and val statistics, mosaics, labels, predictions and augmentations, as well as metrics and charts including Precision-Recall curves and Confusion Matrices. \n", - "\n", - "A **Mosaic Dataloader** is used for training (shown in train*.jpg images), which combines 4 images into 1 mosaic during training.\n", - "\n", - "\"Local\n", - "\n", - "Training results are automatically logged to [Tensorboard](https://www.tensorflow.org/tensorboard) and [CSV](https://github.com/ultralytics/yolov5/pull/4148) as `results.csv`, which is plotted as `results.png` (below) after training completes. You can also plot any `results.csv` file manually:\n", + "Training results are automatically logged with [Tensorboard](https://www.tensorflow.org/tensorboard) and [CSV](https://github.com/ultralytics/yolov5/pull/4148) loggers to `runs/train`, with a new experiment directory created for each new training as `runs/train/exp2`, `runs/train/exp3`, etc.\n", "\n", - "```python\n", - "from utils.plots import plot_results \n", - "plot_results('path/to/results.csv') # plot 'results.csv' as 'results.png'\n", - "```\n", + "This directory contains train and val statistics, mosaics, labels, predictions and augmentated mosaics, as well as metrics and charts including precision-recall (PR) curves and confusion matrices. \n", "\n", - "\"COCO128" + "\"Local\n" ] }, { From a75a1105a1eced888e4b327048775f121436a725 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 3 Aug 2022 21:28:22 +0200 Subject: [PATCH 419/661] Self-contained checkpoint `--resume` (#8839) * Single checkpoint resume * Update train.py * Add hyp * Add hyp * Add hyp * FIX * avoid resume on url data * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * avoid resume on url data * avoid resume on url data * Update Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- train.py | 27 ++++++++++++++++++--------- utils/downloads.py | 10 ++++++---- utils/torch_utils.py | 5 +++-- 3 files changed, 27 insertions(+), 15 deletions(-) diff --git a/train.py b/train.py index 99a43f8614c4..17d16dba1531 100644 --- a/train.py +++ b/train.py @@ -43,7 +43,7 @@ from utils.autobatch import check_train_batch_size from utils.callbacks import Callbacks from utils.dataloaders import create_dataloader -from utils.downloads import attempt_download +from utils.downloads import attempt_download, is_url from utils.general import (LOGGER, check_amp, check_dataset, check_file, check_git_status, check_img_size, check_requirements, check_suffix, check_yaml, colorstr, get_latest_run, increment_path, init_seeds, intersect_dicts, labels_to_class_weights, labels_to_image_weights, methods, @@ -77,6 +77,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio with open(hyp, errors='ignore') as f: hyp = yaml.safe_load(f) # load hyps dict LOGGER.info(colorstr('hyperparameters: ') + ', '.join(f'{k}={v}' for k, v in hyp.items())) + opt.hyp = hyp.copy() # for saving hyps to checkpoints # Save run settings if not evolve: @@ -377,6 +378,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio 'updates': ema.updates, 'optimizer': optimizer.state_dict(), 'wandb_id': loggers.wandb.wandb_run.id if loggers.wandb else None, + 'opt': vars(opt), 'date': datetime.now().isoformat()} # Save last, best and delete @@ -472,8 +474,7 @@ def parse_opt(known=False): parser.add_argument('--bbox_interval', type=int, default=-1, help='W&B: Set bounding-box image logging interval') parser.add_argument('--artifact_alias', type=str, default='latest', help='W&B: Version of dataset artifact to use') - opt = parser.parse_known_args()[0] if known else parser.parse_args() - return opt + return parser.parse_known_args()[0] if known else parser.parse_args() def main(opt, callbacks=Callbacks()): @@ -484,12 +485,20 @@ def main(opt, callbacks=Callbacks()): check_requirements(exclude=['thop']) # Resume - if opt.resume and not check_wandb_resume(opt) and not opt.evolve: # resume an interrupted run - ckpt = opt.resume if isinstance(opt.resume, str) else get_latest_run() # specified or most recent path - assert os.path.isfile(ckpt), 'ERROR: --resume checkpoint does not exist' - with open(Path(ckpt).parent.parent / 'opt.yaml', errors='ignore') as f: - opt = argparse.Namespace(**yaml.safe_load(f)) # replace - opt.cfg, opt.weights, opt.resume = '', ckpt, True # reinstate + if opt.resume and not (check_wandb_resume(opt) or opt.evolve): # resume an interrupted run + last = Path(opt.resume if isinstance(opt.resume, str) else get_latest_run()) # specified or most recent last.pt + assert last.is_file(), f'ERROR: --resume checkpoint {last} does not exist' + opt_yaml = last.parent.parent / 'opt.yaml' # train options yaml + opt_data = opt.data # original dataset + if opt_yaml.is_file(): + with open(opt_yaml, errors='ignore') as f: + d = yaml.safe_load(f) + else: + d = torch.load(last, map_location='cpu')['opt'] + opt = argparse.Namespace(**d) # replace + opt.cfg, opt.weights, opt.resume = '', str(last), True # reinstate + if is_url(opt.data): + opt.data = str(opt_data) # avoid HUB resume auth timeout else: opt.data, opt.cfg, opt.hyp, opt.weights, opt.project = \ check_file(opt.data), check_yaml(opt.cfg), check_yaml(opt.hyp), str(opt.weights), str(opt.project) # checks diff --git a/utils/downloads.py b/utils/downloads.py index ebe5bd36e8ff..9d4780ad28b1 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -16,12 +16,14 @@ import torch -def is_url(url): +def is_url(url, check_online=True): # Check if online file exists try: - r = urllib.request.urlopen(url) # response - return r.getcode() == 200 - except urllib.request.HTTPError: + url = str(url) + result = urllib.parse.urlparse(url) + assert all([result.scheme, result.netloc, result.path]) # check if is url + return (urllib.request.urlopen(url).getcode() == 200) if check_online else True # check if exists online + except (AssertionError, urllib.request.HTTPError): return False diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 391ddead2985..d5615c263e43 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -317,8 +317,9 @@ def smart_resume(ckpt, optimizer, ema=None, weights='yolov5s.pt', epochs=300, re ema.ema.load_state_dict(ckpt['ema'].float().state_dict()) # EMA ema.updates = ckpt['updates'] if resume: - assert start_epoch > 0, f'{weights} training to {epochs} epochs is finished, nothing to resume.' - LOGGER.info(f'Resuming training from {weights} for {epochs - start_epoch} more epochs to {epochs} total epochs') + assert start_epoch > 0, f'{weights} training to {epochs} epochs is finished, nothing to resume.\n' \ + f"Start a new training without --resume, i.e. 'python train.py --weights {weights}'" + LOGGER.info(f'Resuming training from {weights} from epoch {start_epoch} to {epochs} total epochs') if epochs < start_epoch: LOGGER.info(f"{weights} has been trained for {ckpt['epoch']} epochs. Fine-tuning for {epochs} more epochs.") epochs += ckpt['epoch'] # finetune additional epochs From 6884da3a32e97fafcaae5caaddfd13de773cd2dc Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 3 Aug 2022 23:32:31 +0200 Subject: [PATCH 420/661] Add check_file(data) i.e. `--data coco128.yaml` (#8851) * Add check_file(data) i.e. `--data coco128.yaml` * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/export.py b/export.py index 4846624541e4..e3f6af93d1cc 100644 --- a/export.py +++ b/export.py @@ -67,8 +67,8 @@ from models.experimental import attempt_load from models.yolo import Detect from utils.dataloaders import LoadImages -from utils.general import (LOGGER, check_dataset, check_img_size, check_requirements, check_version, colorstr, - file_size, print_args, url2file) +from utils.general import (LOGGER, check_dataset, check_file, check_img_size, check_requirements, check_version, + colorstr, file_size, print_args, url2file) from utils.torch_utils import select_device @@ -371,7 +371,7 @@ def export_tflite(keras_model, im, file, int8, data, nms, agnostic_nms, prefix=c converter.optimizations = [tf.lite.Optimize.DEFAULT] if int8: from models.tf import representative_dataset_gen - dataset = LoadImages(check_dataset(data)['train'], img_size=imgsz, auto=False) # representative data + dataset = LoadImages(check_dataset(check_file(data))['train'], img_size=imgsz, auto=False) converter.representative_dataset = lambda: representative_dataset_gen(dataset, ncalib=100) converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8] converter.target_spec.supported_types = [] From 628c05ca6ff1d7f79d1fc63c298008a1341ba99c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 3 Aug 2022 23:38:36 +0200 Subject: [PATCH 421/661] export.py replace `check_file` -> `check_yaml` (#8852) * export.py replace `check_file` -> `check_yaml` * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/export.py b/export.py index e3f6af93d1cc..546087a4026c 100644 --- a/export.py +++ b/export.py @@ -67,7 +67,7 @@ from models.experimental import attempt_load from models.yolo import Detect from utils.dataloaders import LoadImages -from utils.general import (LOGGER, check_dataset, check_file, check_img_size, check_requirements, check_version, +from utils.general import (LOGGER, check_dataset, check_img_size, check_requirements, check_version, check_yaml, colorstr, file_size, print_args, url2file) from utils.torch_utils import select_device @@ -371,7 +371,7 @@ def export_tflite(keras_model, im, file, int8, data, nms, agnostic_nms, prefix=c converter.optimizations = [tf.lite.Optimize.DEFAULT] if int8: from models.tf import representative_dataset_gen - dataset = LoadImages(check_dataset(check_file(data))['train'], img_size=imgsz, auto=False) + dataset = LoadImages(check_dataset(check_yaml(data))['train'], img_size=imgsz, auto=False) converter.representative_dataset = lambda: representative_dataset_gen(dataset, ncalib=100) converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8] converter.target_spec.supported_types = [] From 84e7748564f83ba04601770f17a38cc55e6be661 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 4 Aug 2022 17:06:08 +0200 Subject: [PATCH 422/661] Update dataloaders.py remove `float64` shapes (#8865) May help https://github.com/ultralytics/yolov5/issues/8862 --- utils/dataloaders.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 9ccfe2545d75..71e7428d4dc1 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -478,7 +478,7 @@ def __init__(self, [cache.pop(k) for k in ('hash', 'version', 'msgs')] # remove items labels, shapes, self.segments = zip(*cache.values()) self.labels = list(labels) - self.shapes = np.array(shapes, dtype=np.float64) + self.shapes = np.array(shapes) self.im_files = list(cache.keys()) # update self.label_files = img2label_paths(cache.keys()) # update n = len(shapes) # number of images From 38a6eb6e99b9e832e7de4a4a57c7b7e4e080fb44 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 4 Aug 2022 23:26:30 +0200 Subject: [PATCH 423/661] Fix TensorRT --dynamic excess outputs bug (#8869) * Fix TensorRT --dynamic excess outputs bug Potential fix for https://github.com/ultralytics/yolov5/issues/8790 * Cleanup * Update common.py * Update common.py * New fix --- models/common.py | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/models/common.py b/models/common.py index c898d94a921a..cfa688ba940b 100644 --- a/models/common.py +++ b/models/common.py @@ -387,13 +387,13 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, context = model.create_execution_context() bindings = OrderedDict() fp16 = False # default updated below - dynamic_input = False + dynamic = False for index in range(model.num_bindings): name = model.get_binding_name(index) dtype = trt.nptype(model.get_binding_dtype(index)) if model.binding_is_input(index): if -1 in tuple(model.get_binding_shape(index)): # dynamic - dynamic_input = True + dynamic = True context.set_binding_shape(index, tuple(model.get_profile_shape(0, index)[2])) if dtype == np.float16: fp16 = True @@ -471,12 +471,14 @@ def forward(self, im, augment=False, visualize=False, val=False): im = im.cpu().numpy() # FP32 y = self.executable_network([im])[self.output_layer] elif self.engine: # TensorRT - if im.shape != self.bindings['images'].shape and self.dynamic_input: - self.context.set_binding_shape(self.model.get_binding_index('images'), im.shape) # reshape if dynamic + if self.dynamic and im.shape != self.bindings['images'].shape: + i_in, i_out = (self.model.get_binding_index(x) for x in ('images', 'output')) + self.context.set_binding_shape(i_in, im.shape) # reshape if dynamic self.bindings['images'] = self.bindings['images']._replace(shape=im.shape) - assert im.shape == self.bindings['images'].shape, ( - f"image shape {im.shape} exceeds model max shape {self.bindings['images'].shape}" if self.dynamic_input - else f"image shape {im.shape} does not match model shape {self.bindings['images'].shape}") + self.bindings['output'].data.resize_(tuple(self.context.get_binding_shape(i_out))) + s = self.bindings['images'].shape + assert im.shape == s, f"image shape {im.shape} " + \ + f"exceeds model max shape {s}" if self.dynamic else f"does not match model shape {s}" self.binding_addrs['images'] = int(im.data_ptr()) self.context.execute_v2(list(self.binding_addrs.values())) y = self.bindings['output'].data From 731a2f8c1ff060bda5e84e34c7cbdd637cfe4d75 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 4 Aug 2022 23:34:15 +0200 Subject: [PATCH 424/661] Single-line TRT dynamic assertion (#8871) --- models/common.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/models/common.py b/models/common.py index cfa688ba940b..a1269c5f3372 100644 --- a/models/common.py +++ b/models/common.py @@ -477,8 +477,7 @@ def forward(self, im, augment=False, visualize=False, val=False): self.bindings['images'] = self.bindings['images']._replace(shape=im.shape) self.bindings['output'].data.resize_(tuple(self.context.get_binding_shape(i_out))) s = self.bindings['images'].shape - assert im.shape == s, f"image shape {im.shape} " + \ - f"exceeds model max shape {s}" if self.dynamic else f"does not match model shape {s}" + assert im.shape == s, f"input size {im.shape} {'>' if self.dynamic else 'not equal to'} max model size {s}" self.binding_addrs['images'] = int(im.data_ptr()) self.context.execute_v2(list(self.binding_addrs.values())) y = self.bindings['output'].data From bc9fcb176734e63d02a1a677c9b2e66f08a2a040 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 5 Aug 2022 14:45:41 +0200 Subject: [PATCH 425/661] HUBDatasetStats() preview images to 50 quality (#8880) @kalenmike should represent a 30% filesize reduction vs 75 quality --- utils/dataloaders.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 71e7428d4dc1..00f6413df7ad 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -1034,7 +1034,7 @@ def _hub_ops(self, f, max_dim=1920): r = max_dim / max(im.height, im.width) # ratio if r < 1.0: # image too large im = im.resize((int(im.width * r), int(im.height * r))) - im.save(f_new, 'JPEG', quality=75, optimize=True) # save + im.save(f_new, 'JPEG', quality=50, optimize=True) # save except Exception as e: # use OpenCV print(f'WARNING: HUB ops PIL failure {f}: {e}') im = cv2.imread(f) From e073658e119dac7bd0bdb209ababc90121c6450d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 5 Aug 2022 16:27:28 +0200 Subject: [PATCH 426/661] `--resume` training from URL weights (#8882) @kalenmike --- train.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/train.py b/train.py index 17d16dba1531..023a39b6c579 100644 --- a/train.py +++ b/train.py @@ -485,9 +485,8 @@ def main(opt, callbacks=Callbacks()): check_requirements(exclude=['thop']) # Resume - if opt.resume and not (check_wandb_resume(opt) or opt.evolve): # resume an interrupted run - last = Path(opt.resume if isinstance(opt.resume, str) else get_latest_run()) # specified or most recent last.pt - assert last.is_file(), f'ERROR: --resume checkpoint {last} does not exist' + if opt.resume and not (check_wandb_resume(opt) or opt.evolve): # resume from specified or most recent last.pt + last = Path(check_file(opt.resume) if isinstance(opt.resume, str) else get_latest_run()) opt_yaml = last.parent.parent / 'opt.yaml' # train options yaml opt_data = opt.data # original dataset if opt_yaml.is_file(): From daed7a844e7f2445b382ca77b0cc5ec84761389b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 5 Aug 2022 16:42:10 +0200 Subject: [PATCH 427/661] `--resume` training from URL weights fix (#8884) --resume training from URL weights fix @kalenmike should fix data error on HUB resume --- train.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/train.py b/train.py index 023a39b6c579..c2f487afe8b0 100644 --- a/train.py +++ b/train.py @@ -496,8 +496,8 @@ def main(opt, callbacks=Callbacks()): d = torch.load(last, map_location='cpu')['opt'] opt = argparse.Namespace(**d) # replace opt.cfg, opt.weights, opt.resume = '', str(last), True # reinstate - if is_url(opt.data): - opt.data = str(opt_data) # avoid HUB resume auth timeout + if is_url(opt_data): + opt.data = check_file(opt_data) # avoid HUB resume auth timeout else: opt.data, opt.cfg, opt.hyp, opt.weights, opt.project = \ check_file(opt.data), check_yaml(opt.cfg), check_yaml(opt.hyp), str(opt.weights), str(opt.project) # checks From 2794483e091d50416289614a1a35f158fd25bee2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 5 Aug 2022 17:10:44 +0200 Subject: [PATCH 428/661] Update CI to default Python 3.10 (#8883) * Update CI to default Python 3.10 * Update ci-testing.yml * Update ci-testing.yml --- .github/workflows/ci-testing.yml | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 444bab75bbbc..0b7fd824d7ea 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -17,7 +17,7 @@ jobs: strategy: matrix: os: [ubuntu-latest] - python-version: [3.9] + python-version: ['3.9'] # requires python<=3.9 model: [yolov5n] steps: - uses: actions/checkout@v3 @@ -48,7 +48,7 @@ jobs: fail-fast: false matrix: os: [ubuntu-latest, macos-latest, windows-latest] - python-version: [3.9] + python-version: ['3.10'] model: [yolov5n] include: - os: ubuntu-latest @@ -58,7 +58,7 @@ jobs: python-version: '3.8' model: yolov5n - os: ubuntu-latest - python-version: '3.10' + python-version: '3.9' model: yolov5n steps: - uses: actions/checkout@v3 From 378bde4bba56b70954d1aa1c75d876164da50d2a Mon Sep 17 00:00:00 2001 From: Victor Sonck Date: Fri, 5 Aug 2022 20:50:49 +0200 Subject: [PATCH 429/661] ClearML experiment tracking integration (#8620) * Add titles to matplotlib plots * Add ClearML Experiment Tracking integration. * Add ClearML Data Version Management automatic download when requested * Add ClearML Hyperparameter Optimization * ClearML save period integration * Fix wandb breaking when used with ClearML dataset * Fix wandb breaking when used with ClearML resume and dataset * Add ClearML documentation * fixed small bug in clearml integration that misreports epoch number * Final ClearMl additions before refactor * Add correct epoch reporting * Add remote execution and autoscaling docs for ClearML integration * Added images to clearml integration docs * fixed logo alignment bug and added hpo screenshot clearml * Fixed small epoch number bug in clearml integration * Remove saved model flush clearml * Cleanup clearml readme section * Cleaned up clearml logger docstring * Remove resume readme section clearml * Clearml integration cleanup * Updated ClearML documentation * Added dark vs light icons ClearML Readme * Clearml Readme styling * Add better gifs * Fixed gif file size * Add better images in tutorial notebook * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Addressed comments in PR #8620 * Fixed circular import * Fixed circular import * Update tutorial.ipynb * Update tutorial.ipynb * Inline comment * Restructured tutorial notebook * Add correct ClearML link to README * Update tutorial.ipynb * Update general.py * Update __init__.py * Update __init__.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update __init__.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update __init__.py * Update README.md * Update __init__.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * spelling * Update tutorial.ipynb * notebook cutt.ly links * Update README.md * Update README.md * cutt.ly links in tutorial * Removed labels as they show up on last subplot only Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- README.md | 21 ++- requirements.txt | 1 + train.py | 2 + tutorial.ipynb | 27 ++- utils/general.py | 4 + utils/loggers/__init__.py | 70 ++++++-- utils/loggers/clearml/README.md | 222 +++++++++++++++++++++++++ utils/loggers/clearml/__init__.py | 0 utils/loggers/clearml/clearml_utils.py | 150 +++++++++++++++++ utils/loggers/clearml/hpo.py | 84 ++++++++++ utils/loggers/wandb/wandb_utils.py | 11 +- utils/metrics.py | 3 + utils/plots.py | 1 + 13 files changed, 575 insertions(+), 21 deletions(-) create mode 100644 utils/loggers/clearml/README.md create mode 100644 utils/loggers/clearml/__init__.py create mode 100644 utils/loggers/clearml/clearml_utils.py create mode 100644 utils/loggers/clearml/hpo.py diff --git a/README.md b/README.md index b959871211e5..5bc3c1c41b93 100644 --- a/README.md +++ b/README.md @@ -151,7 +151,8 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 - [Train Custom Data](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data)  🚀 RECOMMENDED - [Tips for Best Training Results](https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results)  ☘️ RECOMMENDED -- [Weights & Biases Logging](https://github.com/ultralytics/yolov5/issues/1289)  🌟 NEW +- [ClearML Logging](https://github.com/ultralytics/yolov5/tree/master/utils/loggers/clearml) 🌟 NEW +- [Weights & Biases Logging](https://github.com/ultralytics/yolov5/issues/1289) - [Roboflow for Datasets, Labeling, and Active Learning](https://github.com/ultralytics/yolov5/issues/4975)  🌟 NEW - [Multi-GPU Training](https://github.com/ultralytics/yolov5/issues/475) - [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36)  ⭐ NEW @@ -190,17 +191,23 @@ Get started in seconds with our verified environments. Click each icon below for ##
Integrations
-|Weights and Biases|Roboflow ⭐ NEW| -|:-:|:-:| -|Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) | +|ClearML ⭐ NEW|Roboflow|Weights and Biases +|:-:|:-:|:-:| +|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) ##
为什么选择 YOLOv5
@@ -239,6 +236,84 @@ We are super excited about our first-ever Ultralytics YOLOv5 🚀 EXPORT Competi
+ +##
Classification ⭐ NEW
+ +YOLOv5 [release v6.2](https://github.com/ultralytics/yolov5/releases) brings support for classification model training, validation, prediction and export! We've made training classifier models super simple. Click below to get started. + +
+ Classification Checkpoints (click to expand) + +
+ +We trained YOLOv5-cls classification models on ImageNet for 90 epochs using a 4xA100 instance, and we trained ResNet and EfficientNet models alongside with the same default training settings to compare. We exported all models to ONNX FP32 for CPU speed tests and to TensorRT FP16 for GPU speed tests. We ran all speed tests on Google [Colab Pro](https://colab.research.google.com/signup) for easy reproducibility. + +| Model | size
(pixels) | acc
top1 | acc
top5 | Training
90 epochs
4xA100 (hours) | Speed
ONNX CPU
(ms) | Speed
TensorRT V100
(ms) | params
(M) | FLOPs
@224 (B) | +|----------------------------------------------------------------------------------------------------|-----------------------|------------------|------------------|----------------------------------------------|--------------------------------|-------------------------------------|--------------------|------------------------| +| [YOLOv5n-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5n-cls.pt) | 224 | 64.6 | 85.4 | 7:59 | **3.3** | **0.5** | **2.5** | **0.5** | +| [YOLOv5s-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s-cls.pt) | 224 | 71.5 | 90.2 | 8:09 | 6.6 | 0.6 | 5.4 | 1.4 | +| [YOLOv5m-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5m-cls.pt) | 224 | 75.9 | 92.9 | 10:06 | 15.5 | 0.9 | 12.9 | 3.9 | +| [YOLOv5l-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5l-cls.pt) | 224 | 78.0 | 94.0 | 11:56 | 26.9 | 1.4 | 26.5 | 8.5 | +| [YOLOv5x-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5x-cls.pt) | 224 | **79.0** | **94.4** | 15:04 | 54.3 | 1.8 | 48.1 | 15.9 | +| | +| [ResNet18](https://github.com/ultralytics/yolov5/releases/download/v6.2/resnet18.pt) | 224 | 70.3 | 89.5 | **6:47** | 11.2 | 0.5 | 11.7 | 3.7 | +| [ResNet34](https://github.com/ultralytics/yolov5/releases/download/v6.2/resnet34.pt) | 224 | 73.9 | 91.8 | 8:33 | 20.6 | 0.9 | 21.8 | 7.4 | +| [ResNet50](https://github.com/ultralytics/yolov5/releases/download/v6.2/resnet50.pt) | 224 | 76.8 | 93.4 | 11:10 | 23.4 | 1.0 | 25.6 | 8.5 | +| [ResNet101](https://github.com/ultralytics/yolov5/releases/download/v6.2/resnet101.pt) | 224 | 78.5 | 94.3 | 17:10 | 42.1 | 1.9 | 44.5 | 15.9 | +| | +| [EfficientNet_b0](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b0.pt) | 224 | 75.1 | 92.4 | 13:03 | 12.5 | 1.3 | 5.3 | 1.0 | +| [EfficientNet_b1](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b1.pt) | 224 | 76.4 | 93.2 | 17:04 | 14.9 | 1.6 | 7.8 | 1.5 | +| [EfficientNet_b2](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b2.pt) | 224 | 76.6 | 93.4 | 17:10 | 15.9 | 1.6 | 9.1 | 1.7 | +| [EfficientNet_b3](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b3.pt) | 224 | 77.7 | 94.0 | 19:19 | 18.9 | 1.9 | 12.2 | 2.4 | + +
+ Table Notes (click to expand) + +- All checkpoints are trained to 90 epochs with SGD optimizer with lr0=0.001 at image size 224 and all default settings. Runs logged to https://wandb.ai/glenn-jocher/YOLOv5-Classifier-v6-2. +- **Accuracy** values are for single-model single-scale on [ImageNet-1k](https://www.image-net.org/index.php) dataset.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224` +- **Speed** averaged over 100 inference images using a Google [Colab Pro](https://colab.research.google.com/signup) V100 High-RAM instance.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224 --batch 1` +- **Export** to ONNX at FP32 and TensorRT at FP16 done with `export.py`.
Reproduce by `python export.py --weights yolov5s-cls.pt --include engine onnx --imgsz 224` +
+
+ +
+ Classification Usage Examples (click to expand) + +### Train +YOLOv5 classification training supports auto-download of MNIST, Fashion-MNIST, CIFAR10, CIFAR100, Imagenette, Imagewoof, and ImageNet datasets with the `--data` argument. To start training on MNIST for example use `--data mnist`. + +```bash +# Single-GPU +python classify/train.py --model yolov5s-cls.pt --data cifar100 --epochs 5 --img 224 --batch 128 + +# Multi-GPU DDP +python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/train.py --model yolov5s-cls.pt --data imagenet --epochs 5 --img 224 --device 0,1,2,3 +``` + +### Val +Validate accuracy on a pretrained model. To validate YOLOv5s-cls accuracy on ImageNet. +```bash +bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images) +python classify/val.py --weights yolov5s-cls.pt --data ../datasets/imagenet --img 224 +``` + +### Predict +Run a classification prediction on an image. +```bash +python classify/predict.py --weights yolov5s-cls.pt --data data/images/bus.jpg +``` +```python +model = torch.hub.load('ultralytics/yolov5', 'custom', 'yolov5s-cls.pt') # load from PyTorch Hub +``` + +### Export +Export a group of trained YOLOv5-cls, ResNet and EfficientNet models to ONNX and TensorRT. +```bash +python export.py --weights yolov5s-cls.pt resnet50.pt efficientnet_b0.pt --include onnx engine --img 224 +``` +
+ + ##
贡献
我们重视您的意见! 我们希望给大家提供尽可能的简单和透明的方式对 YOLOv5 做出贡献。开始之前请先点击并查看我们的 [贡献指南](CONTRIBUTING.md)，填写[YOLOv5调查问卷](https://ultralytics.com/survey?utm_source=github&utm_medium=social&utm_campaign=Survey) 来向我们发送您的经验反馈。真诚感谢我们所有的贡献者! diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 31d38ead530f..aa797c44d487 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -5,9 +5,9 @@ name: YOLOv5 CI on: push: - branches: [master] + branches: [ master ] pull_request: - branches: [master] + branches: [ master ] schedule: - cron: '0 0 * * *' # runs at 00:00 UTC every day @@ -16,9 +16,9 @@ jobs: runs-on: ${{ matrix.os }} strategy: matrix: - os: [ubuntu-latest] - python-version: ['3.9'] # requires python<=3.9 - model: [yolov5n] + os: [ ubuntu-latest ] + python-version: [ '3.9' ] # requires python<=3.9 + model: [ yolov5n ] steps: - uses: actions/checkout@v3 - uses: actions/setup-python@v4 @@ -47,9 +47,9 @@ jobs: strategy: fail-fast: false matrix: - os: [ubuntu-latest, macos-latest, windows-latest] - python-version: ['3.10'] - model: [yolov5n] + os: [ ubuntu-latest, macos-latest, windows-latest ] + python-version: [ '3.10' ] + model: [ yolov5n ] include: - os: ubuntu-latest python-version: '3.7' # '3.6.8' min @@ -87,7 +87,7 @@ jobs: else pip install -r requirements.txt --extra-index-url https://download.pytorch.org/whl/cpu fi - shell: bash # required for Windows compatibility + shell: bash # for Windows compatibility - name: Check environment run: | python -c "import utils; utils.notebook_init()" @@ -100,8 +100,8 @@ jobs: python --version pip --version pip list - - name: Run tests - shell: bash + - name: Test detection + shell: bash # for Windows compatibility run: | # export PYTHONPATH="$PWD" # to run '$ python *.py' files in subdirectories m=${{ matrix.model }} # official weights @@ -123,3 +123,13 @@ jobs: model = torch.hub.load('.', 'custom', path=path, source='local') print(model('data/images/bus.jpg')) EOF + - name: Test classification + shell: bash # for Windows compatibility + run: | + m=${{ matrix.model }}-cls.pt # official weights + b=runs/train-cls/exp/weights/best.pt # best.pt checkpoint + python classify/train.py --imgsz 32 --model $m --data mnist2560 --epochs 1 # train + python classify/val.py --imgsz 32 --weights $b --data ../datasets/mnist2560 # val + python classify/predict.py --imgsz 32 --weights $b --source ../datasets/mnist2560/test/7/60.png # predict + python classify/predict.py --imgsz 32 --weights $m --source data/images/bus.jpg # predict + python export.py --weights $b --img 64 --imgsz 224 --include torchscript # export diff --git a/README.md b/README.md index 62c7ed4f53e6..b368d1d6e264 100644 --- a/README.md +++ b/README.md @@ -201,14 +201,6 @@ Get started in seconds with our verified environments. Click each icon below for |:-:|:-:|:-:|:-:| |Automatically compile and quantize YOLOv5 for better inference performance in one click at [Deci](https://bit.ly/yolov5-deci-platform)|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) - ##
Why YOLOv5
@@ -254,6 +246,83 @@ We are super excited about our first-ever Ultralytics YOLOv5 🚀 EXPORT Competi +##
Classification ⭐ NEW
+ +YOLOv5 [release v6.2](https://github.com/ultralytics/yolov5/releases) brings support for classification model training, validation, prediction and export! We've made training classifier models super simple. Click below to get started. + +
+ Classification Checkpoints (click to expand) + +
+ +We trained YOLOv5-cls classification models on ImageNet for 90 epochs using a 4xA100 instance, and we trained ResNet and EfficientNet models alongside with the same default training settings to compare. We exported all models to ONNX FP32 for CPU speed tests and to TensorRT FP16 for GPU speed tests. We ran all speed tests on Google [Colab Pro](https://colab.research.google.com/signup) for easy reproducibility. + +| Model | size
(pixels) | acc
top1 | acc
top5 | Training
90 epochs
4xA100 (hours) | Speed
ONNX CPU
(ms) | Speed
TensorRT V100
(ms) | params
(M) | FLOPs
@224 (B) | +|----------------------------------------------------------------------------------------------------|-----------------------|------------------|------------------|----------------------------------------------|--------------------------------|-------------------------------------|--------------------|------------------------| +| [YOLOv5n-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5n-cls.pt) | 224 | 64.6 | 85.4 | 7:59 | **3.3** | **0.5** | **2.5** | **0.5** | +| [YOLOv5s-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s-cls.pt) | 224 | 71.5 | 90.2 | 8:09 | 6.6 | 0.6 | 5.4 | 1.4 | +| [YOLOv5m-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5m-cls.pt) | 224 | 75.9 | 92.9 | 10:06 | 15.5 | 0.9 | 12.9 | 3.9 | +| [YOLOv5l-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5l-cls.pt) | 224 | 78.0 | 94.0 | 11:56 | 26.9 | 1.4 | 26.5 | 8.5 | +| [YOLOv5x-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5x-cls.pt) | 224 | **79.0** | **94.4** | 15:04 | 54.3 | 1.8 | 48.1 | 15.9 | +| | +| [ResNet18](https://github.com/ultralytics/yolov5/releases/download/v6.2/resnet18.pt) | 224 | 70.3 | 89.5 | **6:47** | 11.2 | 0.5 | 11.7 | 3.7 | +| [ResNet34](https://github.com/ultralytics/yolov5/releases/download/v6.2/resnet34.pt) | 224 | 73.9 | 91.8 | 8:33 | 20.6 | 0.9 | 21.8 | 7.4 | +| [ResNet50](https://github.com/ultralytics/yolov5/releases/download/v6.2/resnet50.pt) | 224 | 76.8 | 93.4 | 11:10 | 23.4 | 1.0 | 25.6 | 8.5 | +| [ResNet101](https://github.com/ultralytics/yolov5/releases/download/v6.2/resnet101.pt) | 224 | 78.5 | 94.3 | 17:10 | 42.1 | 1.9 | 44.5 | 15.9 | +| | +| [EfficientNet_b0](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b0.pt) | 224 | 75.1 | 92.4 | 13:03 | 12.5 | 1.3 | 5.3 | 1.0 | +| [EfficientNet_b1](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b1.pt) | 224 | 76.4 | 93.2 | 17:04 | 14.9 | 1.6 | 7.8 | 1.5 | +| [EfficientNet_b2](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b2.pt) | 224 | 76.6 | 93.4 | 17:10 | 15.9 | 1.6 | 9.1 | 1.7 | +| [EfficientNet_b3](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b3.pt) | 224 | 77.7 | 94.0 | 19:19 | 18.9 | 1.9 | 12.2 | 2.4 | + +
+ Table Notes (click to expand) + +- All checkpoints are trained to 90 epochs with SGD optimizer with lr0=0.001 at image size 224 and all default settings. Runs logged to https://wandb.ai/glenn-jocher/YOLOv5-Classifier-v6-2. +- **Accuracy** values are for single-model single-scale on [ImageNet-1k](https://www.image-net.org/index.php) dataset.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224` +- **Speed** averaged over 100 inference images using a Google [Colab Pro](https://colab.research.google.com/signup) V100 High-RAM instance.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224 --batch 1` +- **Export** to ONNX at FP32 and TensorRT at FP16 done with `export.py`.
Reproduce by `python export.py --weights yolov5s-cls.pt --include engine onnx --imgsz 224` +
+
+ +
+ Classification Usage Examples (click to expand) + +### Train +YOLOv5 classification training supports auto-download of MNIST, Fashion-MNIST, CIFAR10, CIFAR100, Imagenette, Imagewoof, and ImageNet datasets with the `--data` argument. To start training on MNIST for example use `--data mnist`. + +```bash +# Single-GPU +python classify/train.py --model yolov5s-cls.pt --data cifar100 --epochs 5 --img 224 --batch 128 + +# Multi-GPU DDP +python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/train.py --model yolov5s-cls.pt --data imagenet --epochs 5 --img 224 --device 0,1,2,3 +``` + +### Val +Validate accuracy on a pretrained model. To validate YOLOv5s-cls accuracy on ImageNet. +```bash +bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images) +python classify/val.py --weights yolov5s-cls.pt --data ../datasets/imagenet --img 224 +``` + +### Predict +Run a classification prediction on an image. +```bash +python classify/predict.py --weights yolov5s-cls.pt --data data/images/bus.jpg +``` +```python +model = torch.hub.load('ultralytics/yolov5', 'custom', 'yolov5s-cls.pt') # load from PyTorch Hub +``` + +### Export +Export a group of trained YOLOv5-cls, ResNet and EfficientNet models to ONNX and TensorRT. +```bash +python export.py --weights yolov5s-cls.pt resnet50.pt efficientnet_b0.pt --include onnx engine --img 224 +``` +
+ + ##
Contribute
We love your input! We want to make contributing to YOLOv5 as easy and transparent as possible. Please see our [Contributing Guide](CONTRIBUTING.md) to get started, and fill out the [YOLOv5 Survey](https://ultralytics.com/survey?utm_source=github&utm_medium=social&utm_campaign=Survey) to send us feedback on your experiences. Thank you to all our contributors! diff --git a/classify/predict.py b/classify/predict.py new file mode 100644 index 000000000000..419830d43952 --- /dev/null +++ b/classify/predict.py @@ -0,0 +1,109 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Run classification inference on images + +Usage: + $ python classify/predict.py --weights yolov5s-cls.pt --source im.jpg +""" + +import argparse +import os +import sys +from pathlib import Path + +import cv2 +import torch.nn.functional as F + +FILE = Path(__file__).resolve() +ROOT = FILE.parents[1] # YOLOv5 root directory +if str(ROOT) not in sys.path: + sys.path.append(str(ROOT)) # add ROOT to PATH +ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative + +from classify.train import imshow_cls +from models.common import DetectMultiBackend +from utils.augmentations import classify_transforms +from utils.general import LOGGER, check_requirements, colorstr, increment_path, print_args +from utils.torch_utils import select_device, smart_inference_mode, time_sync + + +@smart_inference_mode() +def run( + weights=ROOT / 'yolov5s-cls.pt', # model.pt path(s) + source=ROOT / 'data/images/bus.jpg', # file/dir/URL/glob, 0 for webcam + imgsz=224, # inference size + device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu + half=False, # use FP16 half-precision inference + dnn=False, # use OpenCV DNN for ONNX inference + show=True, + project=ROOT / 'runs/predict-cls', # save to project/name + name='exp', # save to project/name + exist_ok=False, # existing project/name ok, do not increment +): + file = str(source) + seen, dt = 1, [0.0, 0.0, 0.0] + device = select_device(device) + + # Directories + save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run + save_dir.mkdir(parents=True, exist_ok=True) # make dir + + # Transforms + transforms = classify_transforms(imgsz) + + # Load model + model = DetectMultiBackend(weights, device=device, dnn=dnn, fp16=half) + model.warmup(imgsz=(1, 3, imgsz, imgsz)) # warmup + + # Image + t1 = time_sync() + im = cv2.cvtColor(cv2.imread(file), cv2.COLOR_BGR2RGB) + im = transforms(im).unsqueeze(0).to(device) + im = im.half() if model.fp16 else im.float() + t2 = time_sync() + dt[0] += t2 - t1 + + # Inference + results = model(im) + t3 = time_sync() + dt[1] += t3 - t2 + + p = F.softmax(results, dim=1) # probabilities + i = p.argsort(1, descending=True)[:, :5].squeeze() # top 5 indices + dt[2] += time_sync() - t3 + LOGGER.info(f"image 1/1 {file}: {imgsz}x{imgsz} {', '.join(f'{model.names[j]} {p[0, j]:.2f}' for j in i)}") + + # Print results + t = tuple(x / seen * 1E3 for x in dt) # speeds per image + shape = (1, 3, imgsz, imgsz) + LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms post-process per image at shape {shape}' % t) + if show: + imshow_cls(im, f=save_dir / Path(file).name, verbose=True) + LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") + return p + + +def parse_opt(): + parser = argparse.ArgumentParser() + parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s-cls.pt', help='model path(s)') + parser.add_argument('--source', type=str, default=ROOT / 'data/images/bus.jpg', help='file') + parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=224, help='train, val image size (pixels)') + parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') + parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') + parser.add_argument('--project', default=ROOT / 'runs/predict-cls', help='save to project/name') + parser.add_argument('--name', default='exp', help='save to project/name') + parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') + opt = parser.parse_args() + print_args(vars(opt)) + return opt + + +def main(opt): + check_requirements(exclude=('tensorboard', 'thop')) + run(**vars(opt)) + + +if __name__ == "__main__": + opt = parse_opt() + main(opt) diff --git a/classify/train.py b/classify/train.py new file mode 100644 index 000000000000..f2b465567446 --- /dev/null +++ b/classify/train.py @@ -0,0 +1,325 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Train a YOLOv5 classifier model on a classification dataset +Datasets: --data mnist, fashion-mnist, cifar10, cifar100, imagenette, imagewoof, imagenet, or 'path/to/custom/dataset' + +Usage: + $ python classify/train.py --model yolov5s-cls.pt --data cifar100 --epochs 5 --img 128 + $ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/train.py --model yolov5s-cls.pt --data imagenet --epochs 5 --img 224 --device 0,1,2,3 +""" + +import argparse +import os +import subprocess +import sys +import time +from copy import deepcopy +from datetime import datetime +from pathlib import Path + +import torch +import torch.distributed as dist +import torch.hub as hub +import torch.optim.lr_scheduler as lr_scheduler +import torchvision +from torch.cuda import amp +from tqdm import tqdm + +FILE = Path(__file__).resolve() +ROOT = FILE.parents[1] # YOLOv5 root directory +if str(ROOT) not in sys.path: + sys.path.append(str(ROOT)) # add ROOT to PATH +ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative + +from classify import val as validate +from models.experimental import attempt_load +from models.yolo import ClassificationModel, DetectionModel +from utils.dataloaders import create_classification_dataloader +from utils.general import (DATASETS_DIR, LOGGER, WorkingDirectory, check_git_status, check_requirements, colorstr, + download, increment_path, init_seeds, print_args, yaml_save) +from utils.loggers import GenericLogger +from utils.plots import imshow_cls +from utils.torch_utils import (ModelEMA, model_info, reshape_classifier_output, select_device, smart_DDP, + smart_optimizer, smartCrossEntropyLoss, torch_distributed_zero_first) + +LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html +RANK = int(os.getenv('RANK', -1)) +WORLD_SIZE = int(os.getenv('WORLD_SIZE', 1)) + + +def train(opt, device): + init_seeds(opt.seed + 1 + RANK, deterministic=True) + save_dir, data, bs, epochs, nw, imgsz, pretrained = \ + opt.save_dir, Path(opt.data), opt.batch_size, opt.epochs, min(os.cpu_count() - 1, opt.workers), \ + opt.imgsz, str(opt.pretrained).lower() == 'true' + cuda = device.type != 'cpu' + + # Directories + wdir = save_dir / 'weights' + wdir.mkdir(parents=True, exist_ok=True) # make dir + last, best = wdir / 'last.pt', wdir / 'best.pt' + + # Save run settings + yaml_save(save_dir / 'opt.yaml', vars(opt)) + + # Logger + logger = GenericLogger(opt=opt, console_logger=LOGGER) if RANK in {-1, 0} else None + + # Download Dataset + with torch_distributed_zero_first(LOCAL_RANK), WorkingDirectory(ROOT): + data_dir = data if data.is_dir() else (DATASETS_DIR / data) + if not data_dir.is_dir(): + LOGGER.info(f'\nDataset not found ⚠️, missing path {data_dir}, attempting download...') + t = time.time() + if str(data) == 'imagenet': + subprocess.run(f"bash {ROOT / 'data/scripts/get_imagenet.sh'}", shell=True, check=True) + else: + url = f'https://github.com/ultralytics/yolov5/releases/download/v1.0/{data}.zip' + download(url, dir=data_dir.parent) + s = f"Dataset download success ✅ ({time.time() - t:.1f}s), saved to {colorstr('bold', data_dir)}\n" + LOGGER.info(s) + + # Dataloaders + nc = len([x for x in (data_dir / 'train').glob('*') if x.is_dir()]) # number of classes + trainloader = create_classification_dataloader(path=data_dir / 'train', + imgsz=imgsz, + batch_size=bs // WORLD_SIZE, + augment=True, + cache=opt.cache, + rank=LOCAL_RANK, + workers=nw) + + test_dir = data_dir / 'test' if (data_dir / 'test').exists() else data_dir / 'val' # data/test or data/val + if RANK in {-1, 0}: + testloader = create_classification_dataloader(path=test_dir, + imgsz=imgsz, + batch_size=bs // WORLD_SIZE * 2, + augment=False, + cache=opt.cache, + rank=-1, + workers=nw) + + # Model + with torch_distributed_zero_first(LOCAL_RANK), WorkingDirectory(ROOT): + if Path(opt.model).is_file() or opt.model.endswith('.pt'): + model = attempt_load(opt.model, device='cpu', fuse=False) + elif opt.model in torchvision.models.__dict__: # TorchVision models i.e. resnet50, efficientnet_b0 + model = torchvision.models.__dict__[opt.model](weights='IMAGENET1K_V1' if pretrained else None) + else: + m = hub.list('ultralytics/yolov5') # + hub.list('pytorch/vision') # models + raise ModuleNotFoundError(f'--model {opt.model} not found. Available models are: \n' + '\n'.join(m)) + if isinstance(model, DetectionModel): + LOGGER.warning("WARNING: pass YOLOv5 classifier model with '-cls' suffix, i.e. '--model yolov5s-cls.pt'") + model = ClassificationModel(model=model, nc=nc, cutoff=opt.cutoff or 10) # convert to classification model + reshape_classifier_output(model, nc) # update class count + for p in model.parameters(): + p.requires_grad = True # for training + for m in model.modules(): + if not pretrained and hasattr(m, 'reset_parameters'): + m.reset_parameters() + if isinstance(m, torch.nn.Dropout) and opt.dropout is not None: + m.p = opt.dropout # set dropout + model = model.to(device) + names = trainloader.dataset.classes # class names + model.names = names # attach class names + + # Info + if RANK in {-1, 0}: + model_info(model) + if opt.verbose: + LOGGER.info(model) + images, labels = next(iter(trainloader)) + file = imshow_cls(images[:25], labels[:25], names=names, f=save_dir / 'train_images.jpg') + logger.log_images(file, name='Train Examples') + logger.log_graph(model, imgsz) # log model + + # Optimizer + optimizer = smart_optimizer(model, opt.optimizer, opt.lr0, momentum=0.9, decay=5e-5) + + # Scheduler + lrf = 0.01 # final lr (fraction of lr0) + # lf = lambda x: ((1 + math.cos(x * math.pi / epochs)) / 2) * (1 - lrf) + lrf # cosine + lf = lambda x: (1 - x / epochs) * (1 - lrf) + lrf # linear + scheduler = lr_scheduler.LambdaLR(optimizer, lr_lambda=lf) + # scheduler = lr_scheduler.OneCycleLR(optimizer, max_lr=lr0, total_steps=epochs, pct_start=0.1, + # final_div_factor=1 / 25 / lrf) + + # EMA + ema = ModelEMA(model) if RANK in {-1, 0} else None + + # DDP mode + if cuda and RANK != -1: + model = smart_DDP(model) + + # Train + t0 = time.time() + criterion = smartCrossEntropyLoss(label_smoothing=opt.label_smoothing) # loss function + best_fitness = 0.0 + scaler = amp.GradScaler(enabled=cuda) + val = test_dir.stem # 'val' or 'test' + LOGGER.info(f'Image sizes {imgsz} train, {imgsz} test\n' + f'Using {nw * WORLD_SIZE} dataloader workers\n' + f"Logging results to {colorstr('bold', save_dir)}\n" + f'Starting {opt.model} training on {data} dataset with {nc} classes for {epochs} epochs...\n\n' + f"{'Epoch':>10}{'GPU_mem':>10}{'train_loss':>12}{f'{val}_loss':>12}{'top1_acc':>12}{'top5_acc':>12}") + for epoch in range(epochs): # loop over the dataset multiple times + tloss, vloss, fitness = 0.0, 0.0, 0.0 # train loss, val loss, fitness + model.train() + if RANK != -1: + trainloader.sampler.set_epoch(epoch) + pbar = enumerate(trainloader) + if RANK in {-1, 0}: + pbar = tqdm(enumerate(trainloader), total=len(trainloader), bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') + for i, (images, labels) in pbar: # progress bar + images, labels = images.to(device, non_blocking=True), labels.to(device) + + # Forward + with amp.autocast(enabled=cuda): # stability issues when enabled + loss = criterion(model(images), labels) + + # Backward + scaler.scale(loss).backward() + + # Optimize + scaler.unscale_(optimizer) # unscale gradients + torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm=10.0) # clip gradients + scaler.step(optimizer) + scaler.update() + optimizer.zero_grad() + if ema: + ema.update(model) + + if RANK in {-1, 0}: + # Print + tloss = (tloss * i + loss.item()) / (i + 1) # update mean losses + mem = '%.3gG' % (torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0) # (GB) + pbar.desc = f"{f'{epoch + 1}/{epochs}':>10}{mem:>10}{tloss:>12.3g}" + ' ' * 36 + + # Test + if i == len(pbar) - 1: # last batch + top1, top5, vloss = validate.run(model=ema.ema, + dataloader=testloader, + criterion=criterion, + pbar=pbar) # test accuracy, loss + fitness = top1 # define fitness as top1 accuracy + + # Scheduler + scheduler.step() + + # Log metrics + if RANK in {-1, 0}: + # Best fitness + if fitness > best_fitness: + best_fitness = fitness + + # Log + metrics = { + "train/loss": tloss, + f"{val}/loss": vloss, + "metrics/accuracy_top1": top1, + "metrics/accuracy_top5": top5, + "lr/0": optimizer.param_groups[0]['lr']} # learning rate + logger.log_metrics(metrics, epoch) + + # Save model + final_epoch = epoch + 1 == epochs + if (not opt.nosave) or final_epoch: + ckpt = { + 'epoch': epoch, + 'best_fitness': best_fitness, + 'model': deepcopy(ema.ema).half(), # deepcopy(de_parallel(model)).half(), + 'ema': None, # deepcopy(ema.ema).half(), + 'updates': ema.updates, + 'optimizer': None, # optimizer.state_dict(), + 'opt': vars(opt), + 'date': datetime.now().isoformat()} + + # Save last, best and delete + torch.save(ckpt, last) + if best_fitness == fitness: + torch.save(ckpt, best) + del ckpt + + # Train complete + if RANK in {-1, 0} and final_epoch: + LOGGER.info(f'\nTraining complete ({(time.time() - t0) / 3600:.3f} hours)' + f"\nResults saved to {colorstr('bold', save_dir)}" + f"\nPredict: python classify/predict.py --weights {best} --source im.jpg" + f"\nValidate: python classify/val.py --weights {best} --data {data_dir}" + f"\nExport: python export.py --weights {best} --include onnx" + f"\nPyTorch Hub: model = torch.hub.load('ultralytics/yolov5', 'custom', '{best}')" + f"\nVisualize: https://netron.app\n") + + # Plot examples + images, labels = (x[:25] for x in next(iter(testloader))) # first 25 images and labels + pred = torch.max(ema.ema((images.half() if cuda else images.float()).to(device)), 1)[1] + file = imshow_cls(images, labels, pred, names, verbose=False, f=save_dir / 'test_images.jpg') + + # Log results + meta = {"epochs": epochs, "top1_acc": best_fitness, "date": datetime.now().isoformat()} + logger.log_images(file, name='Test Examples (true-predicted)', epoch=epoch) + logger.log_model(best, epochs, metadata=meta) + + +def parse_opt(known=False): + parser = argparse.ArgumentParser() + parser.add_argument('--model', type=str, default='yolov5s-cls.pt', help='initial weights path') + parser.add_argument('--data', type=str, default='mnist', help='cifar10, cifar100, mnist, imagenet, etc.') + parser.add_argument('--epochs', type=int, default=10) + parser.add_argument('--batch-size', type=int, default=64, help='total batch size for all GPUs') + parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=128, help='train, val image size (pixels)') + parser.add_argument('--nosave', action='store_true', help='only save final checkpoint') + parser.add_argument('--cache', type=str, nargs='?', const='ram', help='--cache images in "ram" (default) or "disk"') + parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--workers', type=int, default=8, help='max dataloader workers (per RANK in DDP mode)') + parser.add_argument('--project', default=ROOT / 'runs/train-cls', help='save to project/name') + parser.add_argument('--name', default='exp', help='save to project/name') + parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') + parser.add_argument('--pretrained', nargs='?', const=True, default=True, help='start from i.e. --pretrained False') + parser.add_argument('--optimizer', choices=['SGD', 'Adam', 'AdamW', 'RMSProp'], default='Adam', help='optimizer') + parser.add_argument('--lr0', type=float, default=0.001, help='initial learning rate') + parser.add_argument('--label-smoothing', type=float, default=0.1, help='Label smoothing epsilon') + parser.add_argument('--cutoff', type=int, default=None, help='Model layer cutoff index for Classify() head') + parser.add_argument('--dropout', type=float, default=None, help='Dropout (fraction)') + parser.add_argument('--verbose', action='store_true', help='Verbose mode') + parser.add_argument('--seed', type=int, default=0, help='Global training seed') + parser.add_argument('--local_rank', type=int, default=-1, help='Automatic DDP Multi-GPU argument, do not modify') + return parser.parse_known_args()[0] if known else parser.parse_args() + + +def main(opt): + # Checks + if RANK in {-1, 0}: + print_args(vars(opt)) + check_git_status() + check_requirements() + + # DDP mode + device = select_device(opt.device, batch_size=opt.batch_size) + if LOCAL_RANK != -1: + assert opt.batch_size != -1, 'AutoBatch is coming soon for classification, please pass a valid --batch-size' + assert opt.batch_size % WORLD_SIZE == 0, f'--batch-size {opt.batch_size} must be multiple of WORLD_SIZE' + assert torch.cuda.device_count() > LOCAL_RANK, 'insufficient CUDA devices for DDP command' + torch.cuda.set_device(LOCAL_RANK) + device = torch.device('cuda', LOCAL_RANK) + dist.init_process_group(backend="nccl" if dist.is_nccl_available() else "gloo") + + # Parameters + opt.save_dir = increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok) # increment run + + # Train + train(opt, device) + + +def run(**kwargs): + # Usage: from yolov5 import classify; classify.train.run(data=mnist, imgsz=320, model='yolov5m') + opt = parse_opt(True) + for k, v in kwargs.items(): + setattr(opt, k, v) + main(opt) + return opt + + +if __name__ == "__main__": + opt = parse_opt() + main(opt) diff --git a/classify/val.py b/classify/val.py new file mode 100644 index 000000000000..0930ba8c9c51 --- /dev/null +++ b/classify/val.py @@ -0,0 +1,158 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Validate a classification model on a dataset + +Usage: + $ python classify/val.py --weights yolov5s-cls.pt --data ../datasets/imagenet +""" + +import argparse +import os +import sys +from pathlib import Path + +import torch +from tqdm import tqdm + +FILE = Path(__file__).resolve() +ROOT = FILE.parents[1] # YOLOv5 root directory +if str(ROOT) not in sys.path: + sys.path.append(str(ROOT)) # add ROOT to PATH +ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative + +from models.common import DetectMultiBackend +from utils.dataloaders import create_classification_dataloader +from utils.general import LOGGER, check_img_size, check_requirements, colorstr, increment_path, print_args +from utils.torch_utils import select_device, smart_inference_mode, time_sync + + +@smart_inference_mode() +def run( + data=ROOT / '../datasets/mnist', # dataset dir + weights=ROOT / 'yolov5s-cls.pt', # model.pt path(s) + batch_size=128, # batch size + imgsz=224, # inference size (pixels) + device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu + workers=8, # max dataloader workers (per RANK in DDP mode) + verbose=False, # verbose output + project=ROOT / 'runs/val-cls', # save to project/name + name='exp', # save to project/name + exist_ok=False, # existing project/name ok, do not increment + half=True, # use FP16 half-precision inference + dnn=False, # use OpenCV DNN for ONNX inference + model=None, + dataloader=None, + criterion=None, + pbar=None, +): + # Initialize/load model and set device + training = model is not None + if training: # called by train.py + device, pt, jit, engine = next(model.parameters()).device, True, False, False # get model device, PyTorch model + half &= device.type != 'cpu' # half precision only supported on CUDA + model.half() if half else model.float() + else: # called directly + device = select_device(device, batch_size=batch_size) + + # Directories + save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run + save_dir.mkdir(parents=True, exist_ok=True) # make dir + + # Load model + model = DetectMultiBackend(weights, device=device, dnn=dnn, fp16=half) + stride, pt, jit, engine = model.stride, model.pt, model.jit, model.engine + imgsz = check_img_size(imgsz, s=stride) # check image size + half = model.fp16 # FP16 supported on limited backends with CUDA + if engine: + batch_size = model.batch_size + else: + device = model.device + if not (pt or jit): + batch_size = 1 # export.py models default to batch-size 1 + LOGGER.info(f'Forcing --batch-size 1 square inference (1,3,{imgsz},{imgsz}) for non-PyTorch models') + + # Dataloader + data = Path(data) + test_dir = data / 'test' if (data / 'test').exists() else data / 'val' # data/test or data/val + dataloader = create_classification_dataloader(path=test_dir, + imgsz=imgsz, + batch_size=batch_size, + augment=False, + rank=-1, + workers=workers) + + model.eval() + pred, targets, loss, dt = [], [], 0, [0.0, 0.0, 0.0] + n = len(dataloader) # number of batches + action = 'validating' if dataloader.dataset.root.stem == 'val' else 'testing' + desc = f"{pbar.desc[:-36]}{action:>36}" if pbar else f"{action}" + bar = tqdm(dataloader, desc, n, not training, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}', position=0) + with torch.cuda.amp.autocast(enabled=device.type != 'cpu'): + for images, labels in bar: + t1 = time_sync() + images, labels = images.to(device, non_blocking=True), labels.to(device) + t2 = time_sync() + dt[0] += t2 - t1 + + y = model(images) + t3 = time_sync() + dt[1] += t3 - t2 + + pred.append(y.argsort(1, descending=True)[:, :5]) + targets.append(labels) + if criterion: + loss += criterion(y, labels) + dt[2] += time_sync() - t3 + + loss /= n + pred, targets = torch.cat(pred), torch.cat(targets) + correct = (targets[:, None] == pred).float() + acc = torch.stack((correct[:, 0], correct.max(1).values), dim=1) # (top1, top5) accuracy + top1, top5 = acc.mean(0).tolist() + + if pbar: + pbar.desc = f"{pbar.desc[:-36]}{loss:>12.3g}{top1:>12.3g}{top5:>12.3g}" + if verbose: # all classes + LOGGER.info(f"{'Class':>24}{'Images':>12}{'top1_acc':>12}{'top5_acc':>12}") + LOGGER.info(f"{'all':>24}{targets.shape[0]:>12}{top1:>12.3g}{top5:>12.3g}") + for i, c in enumerate(model.names): + aci = acc[targets == i] + top1i, top5i = aci.mean(0).tolist() + LOGGER.info(f"{c:>24}{aci.shape[0]:>12}{top1i:>12.3g}{top5i:>12.3g}") + + # Print results + t = tuple(x / len(dataloader.dataset.samples) * 1E3 for x in dt) # speeds per image + shape = (1, 3, imgsz, imgsz) + LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms post-process per image at shape {shape}' % t) + LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") + + return top1, top5, loss + + +def parse_opt(): + parser = argparse.ArgumentParser() + parser.add_argument('--data', type=str, default=ROOT / '../datasets/mnist', help='dataset path') + parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s-cls.pt', help='model.pt path(s)') + parser.add_argument('--batch-size', type=int, default=128, help='batch size') + parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=224, help='inference size (pixels)') + parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--workers', type=int, default=8, help='max dataloader workers (per RANK in DDP mode)') + parser.add_argument('--verbose', nargs='?', const=True, default=True, help='verbose output') + parser.add_argument('--project', default=ROOT / 'runs/val-cls', help='save to project/name') + parser.add_argument('--name', default='exp', help='save to project/name') + parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') + parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') + parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') + opt = parser.parse_args() + print_args(vars(opt)) + return opt + + +def main(opt): + check_requirements(exclude=('tensorboard', 'thop')) + run(**vars(opt)) + + +if __name__ == "__main__": + opt = parse_opt() + main(opt) diff --git a/data/ImageNet.yaml b/data/ImageNet.yaml new file mode 100644 index 000000000000..9f89b4268aff --- /dev/null +++ b/data/ImageNet.yaml @@ -0,0 +1,156 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +# ImageNet-1k dataset https://www.image-net.org/index.php by Stanford University +# Simplified class names from https://github.com/anishathalye/imagenet-simple-labels +# Example usage: python classify/train.py --data imagenet +# parent +# ├── yolov5 +# └── datasets +# └── imagenet ← downloads here (144 GB) + + +# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] +path: ../datasets/imagenet # dataset root dir +train: train # train images (relative to 'path') 1281167 images +val: val # val images (relative to 'path') 50000 images +test: # test images (optional) + +# Classes +nc: 1000 # number of classes +names: ['tench', 'goldfish', 'great white shark', 'tiger shark', 'hammerhead shark', 'electric ray', 'stingray', 'cock', + 'hen', 'ostrich', 'brambling', 'goldfinch', 'house finch', 'junco', 'indigo bunting', 'American robin', + 'bulbul', 'jay', 'magpie', 'chickadee', 'American dipper', 'kite', 'bald eagle', 'vulture', 'great grey owl', + 'fire salamander', 'smooth newt', 'newt', 'spotted salamander', 'axolotl', 'American bullfrog', 'tree frog', + 'tailed frog', 'loggerhead sea turtle', 'leatherback sea turtle', 'mud turtle', 'terrapin', 'box turtle', + 'banded gecko', 'green iguana', 'Carolina anole', 'desert grassland whiptail lizard', 'agama', + 'frilled-necked lizard', 'alligator lizard', 'Gila monster', 'European green lizard', 'chameleon', + 'Komodo dragon', 'Nile crocodile', 'American alligator', 'triceratops', 'worm snake', 'ring-necked snake', + 'eastern hog-nosed snake', 'smooth green snake', 'kingsnake', 'garter snake', 'water snake', 'vine snake', + 'night snake', 'boa constrictor', 'African rock python', 'Indian cobra', 'green mamba', 'sea snake', + 'Saharan horned viper', 'eastern diamondback rattlesnake', 'sidewinder', 'trilobite', 'harvestman', 'scorpion', + 'yellow garden spider', 'barn spider', 'European garden spider', 'southern black widow', 'tarantula', + 'wolf spider', 'tick', 'centipede', 'black grouse', 'ptarmigan', 'ruffed grouse', 'prairie grouse', 'peacock', + 'quail', 'partridge', 'grey parrot', 'macaw', 'sulphur-crested cockatoo', 'lorikeet', 'coucal', 'bee eater', + 'hornbill', 'hummingbird', 'jacamar', 'toucan', 'duck', 'red-breasted merganser', 'goose', 'black swan', + 'tusker', 'echidna', 'platypus', 'wallaby', 'koala', 'wombat', 'jellyfish', 'sea anemone', 'brain coral', + 'flatworm', 'nematode', 'conch', 'snail', 'slug', 'sea slug', 'chiton', 'chambered nautilus', 'Dungeness crab', + 'rock crab', 'fiddler crab', 'red king crab', 'American lobster', 'spiny lobster', 'crayfish', 'hermit crab', + 'isopod', 'white stork', 'black stork', 'spoonbill', 'flamingo', 'little blue heron', 'great egret', 'bittern', + 'crane (bird)', 'limpkin', 'common gallinule', 'American coot', 'bustard', 'ruddy turnstone', 'dunlin', + 'common redshank', 'dowitcher', 'oystercatcher', 'pelican', 'king penguin', 'albatross', 'grey whale', + 'killer whale', 'dugong', 'sea lion', 'Chihuahua', 'Japanese Chin', 'Maltese', 'Pekingese', 'Shih Tzu', + 'King Charles Spaniel', 'Papillon', 'toy terrier', 'Rhodesian Ridgeback', 'Afghan Hound', 'Basset Hound', + 'Beagle', 'Bloodhound', 'Bluetick Coonhound', 'Black and Tan Coonhound', 'Treeing Walker Coonhound', + 'English foxhound', 'Redbone Coonhound', 'borzoi', 'Irish Wolfhound', 'Italian Greyhound', 'Whippet', + 'Ibizan Hound', 'Norwegian Elkhound', 'Otterhound', 'Saluki', 'Scottish Deerhound', 'Weimaraner', + 'Staffordshire Bull Terrier', 'American Staffordshire Terrier', 'Bedlington Terrier', 'Border Terrier', + 'Kerry Blue Terrier', 'Irish Terrier', 'Norfolk Terrier', 'Norwich Terrier', 'Yorkshire Terrier', + 'Wire Fox Terrier', 'Lakeland Terrier', 'Sealyham Terrier', 'Airedale Terrier', 'Cairn Terrier', + 'Australian Terrier', 'Dandie Dinmont Terrier', 'Boston Terrier', 'Miniature Schnauzer', 'Giant Schnauzer', + 'Standard Schnauzer', 'Scottish Terrier', 'Tibetan Terrier', 'Australian Silky Terrier', + 'Soft-coated Wheaten Terrier', 'West Highland White Terrier', 'Lhasa Apso', 'Flat-Coated Retriever', + 'Curly-coated Retriever', 'Golden Retriever', 'Labrador Retriever', 'Chesapeake Bay Retriever', + 'German Shorthaired Pointer', 'Vizsla', 'English Setter', 'Irish Setter', 'Gordon Setter', 'Brittany', + 'Clumber Spaniel', 'English Springer Spaniel', 'Welsh Springer Spaniel', 'Cocker Spaniels', 'Sussex Spaniel', + 'Irish Water Spaniel', 'Kuvasz', 'Schipperke', 'Groenendael', 'Malinois', 'Briard', 'Australian Kelpie', + 'Komondor', 'Old English Sheepdog', 'Shetland Sheepdog', 'collie', 'Border Collie', 'Bouvier des Flandres', + 'Rottweiler', 'German Shepherd Dog', 'Dobermann', 'Miniature Pinscher', 'Greater Swiss Mountain Dog', + 'Bernese Mountain Dog', 'Appenzeller Sennenhund', 'Entlebucher Sennenhund', 'Boxer', 'Bullmastiff', + 'Tibetan Mastiff', 'French Bulldog', 'Great Dane', 'St. Bernard', 'husky', 'Alaskan Malamute', 'Siberian Husky', + 'Dalmatian', 'Affenpinscher', 'Basenji', 'pug', 'Leonberger', 'Newfoundland', 'Pyrenean Mountain Dog', + 'Samoyed', 'Pomeranian', 'Chow Chow', 'Keeshond', 'Griffon Bruxellois', 'Pembroke Welsh Corgi', + 'Cardigan Welsh Corgi', 'Toy Poodle', 'Miniature Poodle', 'Standard Poodle', 'Mexican hairless dog', + 'grey wolf', 'Alaskan tundra wolf', 'red wolf', 'coyote', 'dingo', 'dhole', 'African wild dog', 'hyena', + 'red fox', 'kit fox', 'Arctic fox', 'grey fox', 'tabby cat', 'tiger cat', 'Persian cat', 'Siamese cat', + 'Egyptian Mau', 'cougar', 'lynx', 'leopard', 'snow leopard', 'jaguar', 'lion', 'tiger', 'cheetah', 'brown bear', + 'American black bear', 'polar bear', 'sloth bear', 'mongoose', 'meerkat', 'tiger beetle', 'ladybug', + 'ground beetle', 'longhorn beetle', 'leaf beetle', 'dung beetle', 'rhinoceros beetle', 'weevil', 'fly', 'bee', + 'ant', 'grasshopper', 'cricket', 'stick insect', 'cockroach', 'mantis', 'cicada', 'leafhopper', 'lacewing', + 'dragonfly', 'damselfly', 'red admiral', 'ringlet', 'monarch butterfly', 'small white', 'sulphur butterfly', + 'gossamer-winged butterfly', 'starfish', 'sea urchin', 'sea cucumber', 'cottontail rabbit', 'hare', + 'Angora rabbit', 'hamster', 'porcupine', 'fox squirrel', 'marmot', 'beaver', 'guinea pig', 'common sorrel', + 'zebra', 'pig', 'wild boar', 'warthog', 'hippopotamus', 'ox', 'water buffalo', 'bison', 'ram', 'bighorn sheep', + 'Alpine ibex', 'hartebeest', 'impala', 'gazelle', 'dromedary', 'llama', 'weasel', 'mink', 'European polecat', + 'black-footed ferret', 'otter', 'skunk', 'badger', 'armadillo', 'three-toed sloth', 'orangutan', 'gorilla', + 'chimpanzee', 'gibbon', 'siamang', 'guenon', 'patas monkey', 'baboon', 'macaque', 'langur', + 'black-and-white colobus', 'proboscis monkey', 'marmoset', 'white-headed capuchin', 'howler monkey', 'titi', + "Geoffroy's spider monkey", 'common squirrel monkey', 'ring-tailed lemur', 'indri', 'Asian elephant', + 'African bush elephant', 'red panda', 'giant panda', 'snoek', 'eel', 'coho salmon', 'rock beauty', 'clownfish', + 'sturgeon', 'garfish', 'lionfish', 'pufferfish', 'abacus', 'abaya', 'academic gown', 'accordion', + 'acoustic guitar', 'aircraft carrier', 'airliner', 'airship', 'altar', 'ambulance', 'amphibious vehicle', + 'analog clock', 'apiary', 'apron', 'waste container', 'assault rifle', 'backpack', 'bakery', 'balance beam', + 'balloon', 'ballpoint pen', 'Band-Aid', 'banjo', 'baluster', 'barbell', 'barber chair', 'barbershop', 'barn', + 'barometer', 'barrel', 'wheelbarrow', 'baseball', 'basketball', 'bassinet', 'bassoon', 'swimming cap', + 'bath towel', 'bathtub', 'station wagon', 'lighthouse', 'beaker', 'military cap', 'beer bottle', 'beer glass', + 'bell-cot', 'bib', 'tandem bicycle', 'bikini', 'ring binder', 'binoculars', 'birdhouse', 'boathouse', + 'bobsleigh', 'bolo tie', 'poke bonnet', 'bookcase', 'bookstore', 'bottle cap', 'bow', 'bow tie', 'brass', 'bra', + 'breakwater', 'breastplate', 'broom', 'bucket', 'buckle', 'bulletproof vest', 'high-speed train', + 'butcher shop', 'taxicab', 'cauldron', 'candle', 'cannon', 'canoe', 'can opener', 'cardigan', 'car mirror', + 'carousel', 'tool kit', 'carton', 'car wheel', 'automated teller machine', 'cassette', 'cassette player', + 'castle', 'catamaran', 'CD player', 'cello', 'mobile phone', 'chain', 'chain-link fence', 'chain mail', + 'chainsaw', 'chest', 'chiffonier', 'chime', 'china cabinet', 'Christmas stocking', 'church', 'movie theater', + 'cleaver', 'cliff dwelling', 'cloak', 'clogs', 'cocktail shaker', 'coffee mug', 'coffeemaker', 'coil', + 'combination lock', 'computer keyboard', 'confectionery store', 'container ship', 'convertible', 'corkscrew', + 'cornet', 'cowboy boot', 'cowboy hat', 'cradle', 'crane (machine)', 'crash helmet', 'crate', 'infant bed', + 'Crock Pot', 'croquet ball', 'crutch', 'cuirass', 'dam', 'desk', 'desktop computer', 'rotary dial telephone', + 'diaper', 'digital clock', 'digital watch', 'dining table', 'dishcloth', 'dishwasher', 'disc brake', 'dock', + 'dog sled', 'dome', 'doormat', 'drilling rig', 'drum', 'drumstick', 'dumbbell', 'Dutch oven', 'electric fan', + 'electric guitar', 'electric locomotive', 'entertainment center', 'envelope', 'espresso machine', 'face powder', + 'feather boa', 'filing cabinet', 'fireboat', 'fire engine', 'fire screen sheet', 'flagpole', 'flute', + 'folding chair', 'football helmet', 'forklift', 'fountain', 'fountain pen', 'four-poster bed', 'freight car', + 'French horn', 'frying pan', 'fur coat', 'garbage truck', 'gas mask', 'gas pump', 'goblet', 'go-kart', + 'golf ball', 'golf cart', 'gondola', 'gong', 'gown', 'grand piano', 'greenhouse', 'grille', 'grocery store', + 'guillotine', 'barrette', 'hair spray', 'half-track', 'hammer', 'hamper', 'hair dryer', 'hand-held computer', + 'handkerchief', 'hard disk drive', 'harmonica', 'harp', 'harvester', 'hatchet', 'holster', 'home theater', + 'honeycomb', 'hook', 'hoop skirt', 'horizontal bar', 'horse-drawn vehicle', 'hourglass', 'iPod', 'clothes iron', + "jack-o'-lantern", 'jeans', 'jeep', 'T-shirt', 'jigsaw puzzle', 'pulled rickshaw', 'joystick', 'kimono', + 'knee pad', 'knot', 'lab coat', 'ladle', 'lampshade', 'laptop computer', 'lawn mower', 'lens cap', + 'paper knife', 'library', 'lifeboat', 'lighter', 'limousine', 'ocean liner', 'lipstick', 'slip-on shoe', + 'lotion', 'speaker', 'loupe', 'sawmill', 'magnetic compass', 'mail bag', 'mailbox', 'tights', 'tank suit', + 'manhole cover', 'maraca', 'marimba', 'mask', 'match', 'maypole', 'maze', 'measuring cup', 'medicine chest', + 'megalith', 'microphone', 'microwave oven', 'military uniform', 'milk can', 'minibus', 'miniskirt', 'minivan', + 'missile', 'mitten', 'mixing bowl', 'mobile home', 'Model T', 'modem', 'monastery', 'monitor', 'moped', + 'mortar', 'square academic cap', 'mosque', 'mosquito net', 'scooter', 'mountain bike', 'tent', 'computer mouse', + 'mousetrap', 'moving van', 'muzzle', 'nail', 'neck brace', 'necklace', 'nipple', 'notebook computer', 'obelisk', + 'oboe', 'ocarina', 'odometer', 'oil filter', 'organ', 'oscilloscope', 'overskirt', 'bullock cart', + 'oxygen mask', 'packet', 'paddle', 'paddle wheel', 'padlock', 'paintbrush', 'pajamas', 'palace', 'pan flute', + 'paper towel', 'parachute', 'parallel bars', 'park bench', 'parking meter', 'passenger car', 'patio', + 'payphone', 'pedestal', 'pencil case', 'pencil sharpener', 'perfume', 'Petri dish', 'photocopier', 'plectrum', + 'Pickelhaube', 'picket fence', 'pickup truck', 'pier', 'piggy bank', 'pill bottle', 'pillow', 'ping-pong ball', + 'pinwheel', 'pirate ship', 'pitcher', 'hand plane', 'planetarium', 'plastic bag', 'plate rack', 'plow', + 'plunger', 'Polaroid camera', 'pole', 'police van', 'poncho', 'billiard table', 'soda bottle', 'pot', + "potter's wheel", 'power drill', 'prayer rug', 'printer', 'prison', 'projectile', 'projector', 'hockey puck', + 'punching bag', 'purse', 'quill', 'quilt', 'race car', 'racket', 'radiator', 'radio', 'radio telescope', + 'rain barrel', 'recreational vehicle', 'reel', 'reflex camera', 'refrigerator', 'remote control', 'restaurant', + 'revolver', 'rifle', 'rocking chair', 'rotisserie', 'eraser', 'rugby ball', 'ruler', 'running shoe', 'safe', + 'safety pin', 'salt shaker', 'sandal', 'sarong', 'saxophone', 'scabbard', 'weighing scale', 'school bus', + 'schooner', 'scoreboard', 'CRT screen', 'screw', 'screwdriver', 'seat belt', 'sewing machine', 'shield', + 'shoe store', 'shoji', 'shopping basket', 'shopping cart', 'shovel', 'shower cap', 'shower curtain', 'ski', + 'ski mask', 'sleeping bag', 'slide rule', 'sliding door', 'slot machine', 'snorkel', 'snowmobile', 'snowplow', + 'soap dispenser', 'soccer ball', 'sock', 'solar thermal collector', 'sombrero', 'soup bowl', 'space bar', + 'space heater', 'space shuttle', 'spatula', 'motorboat', 'spider web', 'spindle', 'sports car', 'spotlight', + 'stage', 'steam locomotive', 'through arch bridge', 'steel drum', 'stethoscope', 'scarf', 'stone wall', + 'stopwatch', 'stove', 'strainer', 'tram', 'stretcher', 'couch', 'stupa', 'submarine', 'suit', 'sundial', + 'sunglass', 'sunglasses', 'sunscreen', 'suspension bridge', 'mop', 'sweatshirt', 'swimsuit', 'swing', 'switch', + 'syringe', 'table lamp', 'tank', 'tape player', 'teapot', 'teddy bear', 'television', 'tennis ball', + 'thatched roof', 'front curtain', 'thimble', 'threshing machine', 'throne', 'tile roof', 'toaster', + 'tobacco shop', 'toilet seat', 'torch', 'totem pole', 'tow truck', 'toy store', 'tractor', 'semi-trailer truck', + 'tray', 'trench coat', 'tricycle', 'trimaran', 'tripod', 'triumphal arch', 'trolleybus', 'trombone', 'tub', + 'turnstile', 'typewriter keyboard', 'umbrella', 'unicycle', 'upright piano', 'vacuum cleaner', 'vase', 'vault', + 'velvet', 'vending machine', 'vestment', 'viaduct', 'violin', 'volleyball', 'waffle iron', 'wall clock', + 'wallet', 'wardrobe', 'military aircraft', 'sink', 'washing machine', 'water bottle', 'water jug', + 'water tower', 'whiskey jug', 'whistle', 'wig', 'window screen', 'window shade', 'Windsor tie', 'wine bottle', + 'wing', 'wok', 'wooden spoon', 'wool', 'split-rail fence', 'shipwreck', 'yawl', 'yurt', 'website', 'comic book', + 'crossword', 'traffic sign', 'traffic light', 'dust jacket', 'menu', 'plate', 'guacamole', 'consomme', + 'hot pot', 'trifle', 'ice cream', 'ice pop', 'baguette', 'bagel', 'pretzel', 'cheeseburger', 'hot dog', + 'mashed potato', 'cabbage', 'broccoli', 'cauliflower', 'zucchini', 'spaghetti squash', 'acorn squash', + 'butternut squash', 'cucumber', 'artichoke', 'bell pepper', 'cardoon', 'mushroom', 'Granny Smith', 'strawberry', + 'orange', 'lemon', 'fig', 'pineapple', 'banana', 'jackfruit', 'custard apple', 'pomegranate', 'hay', + 'carbonara', 'chocolate syrup', 'dough', 'meatloaf', 'pizza', 'pot pie', 'burrito', 'red wine', 'espresso', + 'cup', 'eggnog', 'alp', 'bubble', 'cliff', 'coral reef', 'geyser', 'lakeshore', 'promontory', 'shoal', + 'seashore', 'valley', 'volcano', 'baseball player', 'bridegroom', 'scuba diver', 'rapeseed', 'daisy', + "yellow lady's slipper", 'corn', 'acorn', 'rose hip', 'horse chestnut seed', 'coral fungus', 'agaric', + 'gyromitra', 'stinkhorn mushroom', 'earth star', 'hen-of-the-woods', 'bolete', 'ear', + 'toilet paper'] # class names + +# Download script/URL (optional) +download: data/scripts/get_imagenet.sh diff --git a/data/scripts/download_weights.sh b/data/scripts/download_weights.sh index e9fa65394178..a4f3becfdbeb 100755 --- a/data/scripts/download_weights.sh +++ b/data/scripts/download_weights.sh @@ -1,7 +1,7 @@ #!/bin/bash # YOLOv5 🚀 by Ultralytics, GPL-3.0 license # Download latest models from https://github.com/ultralytics/yolov5/releases -# Example usage: bash path/to/download_weights.sh +# Example usage: bash data/scripts/download_weights.sh # parent # └── yolov5 # ├── yolov5s.pt ← downloads here @@ -11,10 +11,11 @@ python - <=7.0.0 + if device.type == 'cpu': + device = torch.device('cuda:0') Binding = namedtuple('Binding', ('name', 'dtype', 'shape', 'data', 'ptr')) logger = trt.Logger(trt.Logger.INFO) with open(w, 'rb') as f, trt.Runtime(logger) as runtime: @@ -398,8 +396,8 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, if dtype == np.float16: fp16 = True shape = tuple(context.get_binding_shape(index)) - data = torch.from_numpy(np.empty(shape, dtype=np.dtype(dtype))).to(device) - bindings[name] = Binding(name, dtype, shape, data, int(data.data_ptr())) + im = torch.from_numpy(np.empty(shape, dtype=dtype)).to(device) + bindings[name] = Binding(name, dtype, shape, im, int(im.data_ptr())) binding_addrs = OrderedDict((n, d.ptr) for n, d in bindings.items()) batch_size = bindings['images'].shape[0] # if dynamic, this is instead max batch size elif coreml: # CoreML @@ -445,9 +443,16 @@ def wrap_frozen_graph(gd, inputs, outputs): input_details = interpreter.get_input_details() # inputs output_details = interpreter.get_output_details() # outputs elif tfjs: - raise Exception('ERROR: YOLOv5 TF.js inference is not supported') + raise NotImplementedError('ERROR: YOLOv5 TF.js inference is not supported') else: - raise Exception(f'ERROR: {w} is not a supported format') + raise NotImplementedError(f'ERROR: {w} is not a supported format') + + # class names + if 'names' not in locals(): + names = yaml_load(data)['names'] if data else [f'class{i}' for i in range(999)] + if names[0] == 'n01440764' and len(names) == 1000: # ImageNet + names = yaml_load(ROOT / 'data/ImageNet.yaml')['names'] # human-readable names + self.__dict__.update(locals()) # assign all variables to self def forward(self, im, augment=False, visualize=False, val=False): @@ -457,7 +462,9 @@ def forward(self, im, augment=False, visualize=False, val=False): im = im.half() # to FP16 if self.pt: # PyTorch - y = self.model(im, augment=augment, visualize=visualize)[0] + y = self.model(im, augment=augment, visualize=visualize) if augment or visualize else self.model(im) + if isinstance(y, tuple): + y = y[0] elif self.jit: # TorchScript y = self.model(im)[0] elif self.dnn: # ONNX OpenCV DNN @@ -526,7 +533,7 @@ def warmup(self, imgsz=(1, 3, 640, 640)): self.forward(im) # warmup @staticmethod - def model_type(p='path/to/model.pt'): + def _model_type(p='path/to/model.pt'): # Return model type from model path, i.e. path='path/to/model.onnx' -> type=onnx from export import export_formats suffixes = list(export_formats().Suffix) + ['.xml'] # export suffixes @@ -540,8 +547,7 @@ def model_type(p='path/to/model.pt'): @staticmethod def _load_metadata(f='path/to/meta.yaml'): # Load metadata from meta.yaml if it exists - with open(f, errors='ignore') as f: - d = yaml.safe_load(f) + d = yaml_load(f) return d['stride'], d['names'] # assign stride, names @@ -753,10 +759,13 @@ class Classify(nn.Module): # Classification head, i.e. x(b,c1,20,20) to x(b,c2) def __init__(self, c1, c2, k=1, s=1, p=None, g=1): # ch_in, ch_out, kernel, stride, padding, groups super().__init__() - self.aap = nn.AdaptiveAvgPool2d(1) # to x(b,c1,1,1) - self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g) # to x(b,c2,1,1) - self.flat = nn.Flatten() + c_ = 1280 # efficientnet_b0 size + self.conv = Conv(c1, c_, k, s, autopad(k, p), g) + self.pool = nn.AdaptiveAvgPool2d(1) # to x(b,c_,1,1) + self.drop = nn.Dropout(p=0.0, inplace=True) + self.linear = nn.Linear(c_, c2) # to x(b,c2) def forward(self, x): - z = torch.cat([self.aap(y) for y in (x if isinstance(x, list) else [x])], 1) # cat if list - return self.flat(self.conv(z)) # flatten to x(b,c2) + if isinstance(x, list): + x = torch.cat(x, 1) + return self.linear(self.drop(self.pool(self.conv(x)).flatten(1))) diff --git a/models/experimental.py b/models/experimental.py index 0317c7526c99..cb32d01ba46a 100644 --- a/models/experimental.py +++ b/models/experimental.py @@ -79,7 +79,9 @@ def attempt_load(weights, device=None, inplace=True, fuse=True): for w in weights if isinstance(weights, list) else [weights]: ckpt = torch.load(attempt_download(w), map_location='cpu') # load ckpt = (ckpt.get('ema') or ckpt['model']).to(device).float() # FP32 model - model.append(ckpt.fuse().eval() if fuse else ckpt.eval()) # fused or un-fused model in eval mode + if not hasattr(ckpt, 'stride'): + ckpt.stride = torch.tensor([32.]) # compatibility update for ResNet etc. + model.append(ckpt.fuse().eval() if fuse and hasattr(ckpt, 'fuse') else ckpt.eval()) # model in eval mode # Compatibility updates for m in model.modules(): @@ -92,11 +94,14 @@ def attempt_load(weights, device=None, inplace=True, fuse=True): elif t is nn.Upsample and not hasattr(m, 'recompute_scale_factor'): m.recompute_scale_factor = None # torch 1.11.0 compatibility + # Return model if len(model) == 1: - return model[-1] # return model + return model[-1] + + # Return detection ensemble print(f'Ensemble created with {weights}\n') for k in 'names', 'nc', 'yaml': setattr(model, k, getattr(model[0], k)) model.stride = model[torch.argmax(torch.tensor([m.stride.max() for m in model])).int()].stride # max stride assert all(model[0].nc == m.nc for m in model), f'Models have different class counts: {[m.nc for m in model]}' - return model # return ensemble + return model diff --git a/models/yolo.py b/models/yolo.py index 307b74844ca0..df4209726e0d 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -90,8 +90,64 @@ def _make_grid(self, nx=20, ny=20, i=0, torch_1_10=check_version(torch.__version return grid, anchor_grid -class Model(nn.Module): - # YOLOv5 model +class BaseModel(nn.Module): + # YOLOv5 base model + def forward(self, x, profile=False, visualize=False): + return self._forward_once(x, profile, visualize) # single-scale inference, train + + def _forward_once(self, x, profile=False, visualize=False): + y, dt = [], [] # outputs + for m in self.model: + if m.f != -1: # if not from previous layer + x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers + if profile: + self._profile_one_layer(m, x, dt) + x = m(x) # run + y.append(x if m.i in self.save else None) # save output + if visualize: + feature_visualization(x, m.type, m.i, save_dir=visualize) + return x + + def _profile_one_layer(self, m, x, dt): + c = m == self.model[-1] # is final layer, copy input as inplace fix + o = thop.profile(m, inputs=(x.copy() if c else x,), verbose=False)[0] / 1E9 * 2 if thop else 0 # FLOPs + t = time_sync() + for _ in range(10): + m(x.copy() if c else x) + dt.append((time_sync() - t) * 100) + if m == self.model[0]: + LOGGER.info(f"{'time (ms)':>10s} {'GFLOPs':>10s} {'params':>10s} module") + LOGGER.info(f'{dt[-1]:10.2f} {o:10.2f} {m.np:10.0f} {m.type}') + if c: + LOGGER.info(f"{sum(dt):10.2f} {'-':>10s} {'-':>10s} Total") + + def fuse(self): # fuse model Conv2d() + BatchNorm2d() layers + LOGGER.info('Fusing layers... ') + for m in self.model.modules(): + if isinstance(m, (Conv, DWConv)) and hasattr(m, 'bn'): + m.conv = fuse_conv_and_bn(m.conv, m.bn) # update conv + delattr(m, 'bn') # remove batchnorm + m.forward = m.forward_fuse # update forward + self.info() + return self + + def info(self, verbose=False, img_size=640): # print model information + model_info(self, verbose, img_size) + + def _apply(self, fn): + # Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers + self = super()._apply(fn) + m = self.model[-1] # Detect() + if isinstance(m, Detect): + m.stride = fn(m.stride) + m.grid = list(map(fn, m.grid)) + if isinstance(m.anchor_grid, list): + m.anchor_grid = list(map(fn, m.anchor_grid)) + return self + + +class DetectionModel(BaseModel): + # YOLOv5 detection model def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, anchors=None): # model, input channels, number of classes super().__init__() if isinstance(cfg, dict): @@ -149,19 +205,6 @@ def _forward_augment(self, x): y = self._clip_augmented(y) # clip augmented tails return torch.cat(y, 1), None # augmented inference, train - def _forward_once(self, x, profile=False, visualize=False): - y, dt = [], [] # outputs - for m in self.model: - if m.f != -1: # if not from previous layer - x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers - if profile: - self._profile_one_layer(m, x, dt) - x = m(x) # run - y.append(x if m.i in self.save else None) # save output - if visualize: - feature_visualization(x, m.type, m.i, save_dir=visualize) - return x - def _descale_pred(self, p, flips, scale, img_size): # de-scale predictions following augmented inference (inverse operation) if self.inplace: @@ -190,19 +233,6 @@ def _clip_augmented(self, y): y[-1] = y[-1][:, i:] # small return y - def _profile_one_layer(self, m, x, dt): - c = isinstance(m, Detect) # is final layer, copy input as inplace fix - o = thop.profile(m, inputs=(x.copy() if c else x,), verbose=False)[0] / 1E9 * 2 if thop else 0 # FLOPs - t = time_sync() - for _ in range(10): - m(x.copy() if c else x) - dt.append((time_sync() - t) * 100) - if m == self.model[0]: - LOGGER.info(f"{'time (ms)':>10s} {'GFLOPs':>10s} {'params':>10s} module") - LOGGER.info(f'{dt[-1]:10.2f} {o:10.2f} {m.np:10.0f} {m.type}') - if c: - LOGGER.info(f"{sum(dt):10.2f} {'-':>10s} {'-':>10s} Total") - def _initialize_biases(self, cf=None): # initialize biases into Detect(), cf is class frequency # https://arxiv.org/abs/1708.02002 section 3.3 # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1. @@ -213,41 +243,34 @@ def _initialize_biases(self, cf=None): # initialize biases into Detect(), cf is b[:, 5:] += math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum()) # cls mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True) - def _print_biases(self): - m = self.model[-1] # Detect() module - for mi in m.m: # from - b = mi.bias.detach().view(m.na, -1).T # conv.bias(255) to (3,85) - LOGGER.info( - ('%6g Conv2d.bias:' + '%10.3g' * 6) % (mi.weight.shape[1], *b[:5].mean(1).tolist(), b[5:].mean())) - # def _print_weights(self): - # for m in self.model.modules(): - # if type(m) is Bottleneck: - # LOGGER.info('%10.3g' % (m.w.detach().sigmoid() * 2)) # shortcut weights +Model = DetectionModel # retain YOLOv5 'Model' class for backwards compatibility - def fuse(self): # fuse model Conv2d() + BatchNorm2d() layers - LOGGER.info('Fusing layers... ') - for m in self.model.modules(): - if isinstance(m, (Conv, DWConv)) and hasattr(m, 'bn'): - m.conv = fuse_conv_and_bn(m.conv, m.bn) # update conv - delattr(m, 'bn') # remove batchnorm - m.forward = m.forward_fuse # update forward - self.info() - return self - def info(self, verbose=False, img_size=640): # print model information - model_info(self, verbose, img_size) - - def _apply(self, fn): - # Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers - self = super()._apply(fn) - m = self.model[-1] # Detect() - if isinstance(m, Detect): - m.stride = fn(m.stride) - m.grid = list(map(fn, m.grid)) - if isinstance(m.anchor_grid, list): - m.anchor_grid = list(map(fn, m.anchor_grid)) - return self +class ClassificationModel(BaseModel): + # YOLOv5 classification model + def __init__(self, cfg=None, model=None, nc=1000, cutoff=10): # yaml, model, number of classes, cutoff index + super().__init__() + self._from_detection_model(model, nc, cutoff) if model is not None else self._from_yaml(cfg) + + def _from_detection_model(self, model, nc=1000, cutoff=10): + # Create a YOLOv5 classification model from a YOLOv5 detection model + if isinstance(model, DetectMultiBackend): + model = model.model # unwrap DetectMultiBackend + model.model = model.model[:cutoff] # backbone + m = model.model[-1] # last layer + ch = m.conv.in_channels if hasattr(m, 'conv') else m.cv1.conv.in_channels # ch into module + c = Classify(ch, nc) # Classify() + c.i, c.f, c.type = m.i, m.f, 'models.common.Classify' # index, from, type + model.model[-1] = c # replace + self.model = model.model + self.stride = model.stride + self.save = [] + self.nc = nc + + def _from_yaml(self, cfg): + # Create a YOLOv5 classification model from a *.yaml file + self.model = None def parse_model(d, ch): # model_dict, input_channels(3) @@ -321,7 +344,7 @@ def parse_model(d, ch): # model_dict, input_channels(3) # Options if opt.line_profile: # profile layer by layer - _ = model(im, profile=True) + model(im, profile=True) elif opt.profile: # profile forward-backward results = profile(input=im, ops=[model], n=3) diff --git a/train.py b/train.py index d24ac57df23d..bbb26cdeafeb 100644 --- a/train.py +++ b/train.py @@ -47,7 +47,7 @@ from utils.general import (LOGGER, check_amp, check_dataset, check_file, check_git_status, check_img_size, check_requirements, check_suffix, check_yaml, colorstr, get_latest_run, increment_path, init_seeds, intersect_dicts, labels_to_class_weights, labels_to_image_weights, methods, - one_cycle, print_args, print_mutation, strip_optimizer) + one_cycle, print_args, print_mutation, strip_optimizer, yaml_save) from utils.loggers import Loggers from utils.loggers.wandb.wandb_utils import check_wandb_resume from utils.loss import ComputeLoss @@ -81,10 +81,8 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # Save run settings if not evolve: - with open(save_dir / 'hyp.yaml', 'w') as f: - yaml.safe_dump(hyp, f, sort_keys=False) - with open(save_dir / 'opt.yaml', 'w') as f: - yaml.safe_dump(vars(opt), f, sort_keys=False) + yaml_save(save_dir / 'hyp.yaml', hyp) + yaml_save(save_dir / 'opt.yaml', vars(opt)) # Loggers data_dict = None @@ -484,7 +482,7 @@ def main(opt, callbacks=Callbacks()): if RANK in {-1, 0}: print_args(vars(opt)) check_git_status() - check_requirements(exclude=['thop']) + check_requirements() # Resume if opt.resume and not (check_wandb_resume(opt) or opt.evolve): # resume from specified or most recent last.pt diff --git a/utils/augmentations.py b/utils/augmentations.py index 3f764c06ae3b..a55fefa68a76 100644 --- a/utils/augmentations.py +++ b/utils/augmentations.py @@ -8,15 +8,21 @@ import cv2 import numpy as np +import torchvision.transforms as T +import torchvision.transforms.functional as TF from utils.general import LOGGER, check_version, colorstr, resample_segments, segment2box from utils.metrics import bbox_ioa +IMAGENET_MEAN = 0.485, 0.456, 0.406 # RGB mean +IMAGENET_STD = 0.229, 0.224, 0.225 # RGB standard deviation + class Albumentations: # YOLOv5 Albumentations class (optional, only used if package is installed) def __init__(self): self.transform = None + prefix = colorstr('albumentations: ') try: import albumentations as A check_version(A.__version__, '1.0.3', hard=True) # version requirement @@ -31,11 +37,11 @@ def __init__(self): A.ImageCompression(quality_lower=75, p=0.0)] # transforms self.transform = A.Compose(T, bbox_params=A.BboxParams(format='yolo', label_fields=['class_labels'])) - LOGGER.info(colorstr('albumentations: ') + ', '.join(f'{x}' for x in self.transform.transforms if x.p)) + LOGGER.info(prefix + ', '.join(f'{x}'.replace('always_apply=False, ', '') for x in T if x.p)) except ImportError: # package not installed, skip pass except Exception as e: - LOGGER.info(colorstr('albumentations: ') + f'{e}') + LOGGER.info(f'{prefix}{e}') def __call__(self, im, labels, p=1.0): if self.transform and random.random() < p: @@ -44,6 +50,18 @@ def __call__(self, im, labels, p=1.0): return im, labels +def normalize(x, mean=IMAGENET_MEAN, std=IMAGENET_STD, inplace=False): + # Denormalize RGB images x per ImageNet stats in BCHW format, i.e. = (x - mean) / std + return TF.normalize(x, mean, std, inplace=inplace) + + +def denormalize(x, mean=IMAGENET_MEAN, std=IMAGENET_STD): + # Denormalize RGB images x per ImageNet stats in BCHW format, i.e. = x * std + mean + for i in range(3): + x[:, i] = x[:, i] * std[i] + mean[i] + return x + + def augment_hsv(im, hgain=0.5, sgain=0.5, vgain=0.5): # HSV color-space augmentation if hgain or sgain or vgain: @@ -282,3 +300,48 @@ def box_candidates(box1, box2, wh_thr=2, ar_thr=100, area_thr=0.1, eps=1e-16): w2, h2 = box2[2] - box2[0], box2[3] - box2[1] ar = np.maximum(w2 / (h2 + eps), h2 / (w2 + eps)) # aspect ratio return (w2 > wh_thr) & (h2 > wh_thr) & (w2 * h2 / (w1 * h1 + eps) > area_thr) & (ar < ar_thr) # candidates + + +def classify_albumentations(augment=True, + size=224, + scale=(0.08, 1.0), + hflip=0.5, + vflip=0.0, + jitter=0.4, + mean=IMAGENET_MEAN, + std=IMAGENET_STD, + auto_aug=False): + # YOLOv5 classification Albumentations (optional, only used if package is installed) + prefix = colorstr('albumentations: ') + try: + import albumentations as A + from albumentations.pytorch import ToTensorV2 + check_version(A.__version__, '1.0.3', hard=True) # version requirement + if augment: # Resize and crop + T = [A.RandomResizedCrop(height=size, width=size, scale=scale)] + if auto_aug: + # TODO: implement AugMix, AutoAug & RandAug in albumentation + LOGGER.info(f'{prefix}auto augmentations are currently not supported') + else: + if hflip > 0: + T += [A.HorizontalFlip(p=hflip)] + if vflip > 0: + T += [A.VerticalFlip(p=vflip)] + if jitter > 0: + color_jitter = (float(jitter),) * 3 # repeat value for brightness, contrast, satuaration, 0 hue + T += [A.ColorJitter(*color_jitter, 0)] + else: # Use fixed crop for eval set (reproducibility) + T = [A.SmallestMaxSize(max_size=size), A.CenterCrop(height=size, width=size)] + T += [A.Normalize(mean=mean, std=std), ToTensorV2()] # Normalize and convert to Tensor + LOGGER.info(prefix + ', '.join(f'{x}'.replace('always_apply=False, ', '') for x in T if x.p)) + return A.Compose(T) + + except ImportError: # package not installed, skip + pass + except Exception as e: + LOGGER.info(f'{prefix}{e}') + + +def classify_transforms(size=224): + # Transforms to apply if albumentations not installed + return T.Compose([T.ToTensor(), T.Resize(size), T.CenterCrop(size), T.Normalize(IMAGENET_MEAN, IMAGENET_STD)]) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 00f6413df7ad..2c04040bf25d 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -22,12 +22,14 @@ import numpy as np import torch import torch.nn.functional as F +import torchvision import yaml from PIL import ExifTags, Image, ImageOps from torch.utils.data import DataLoader, Dataset, dataloader, distributed from tqdm import tqdm -from utils.augmentations import Albumentations, augment_hsv, copy_paste, letterbox, mixup, random_perspective +from utils.augmentations import (Albumentations, augment_hsv, classify_albumentations, classify_transforms, copy_paste, + letterbox, mixup, random_perspective) from utils.general import (DATASETS_DIR, LOGGER, NUM_THREADS, check_dataset, check_requirements, check_yaml, clean_str, cv2, is_colab, is_kaggle, segments2boxes, xyn2xy, xywh2xyxy, xywhn2xyxy, xyxy2xywhn) from utils.torch_utils import torch_distributed_zero_first @@ -870,7 +872,7 @@ def flatten_recursive(path=DATASETS_DIR / 'coco128'): def extract_boxes(path=DATASETS_DIR / 'coco128'): # from utils.dataloaders import *; extract_boxes() # Convert detection dataset into classification dataset, with one directory per class path = Path(path) # images dir - shutil.rmtree(path / 'classifier') if (path / 'classifier').is_dir() else None # remove existing + shutil.rmtree(path / 'classification') if (path / 'classification').is_dir() else None # remove existing files = list(path.rglob('*.*')) n = len(files) # number of files for im_file in tqdm(files, total=n): @@ -1090,3 +1092,65 @@ def process_images(self): pass print(f'Done. All images saved to {self.im_dir}') return self.im_dir + + +# Classification dataloaders ------------------------------------------------------------------------------------------- +class ClassificationDataset(torchvision.datasets.ImageFolder): + """ + YOLOv5 Classification Dataset. + Arguments + root: Dataset path + transform: torchvision transforms, used by default + album_transform: Albumentations transforms, used if installed + """ + + def __init__(self, root, augment, imgsz, cache=False): + super().__init__(root=root) + self.torch_transforms = classify_transforms(imgsz) + self.album_transforms = classify_albumentations(augment, imgsz) if augment else None + self.cache_ram = cache is True or cache == 'ram' + self.cache_disk = cache == 'disk' + self.samples = [list(x) + [Path(x[0]).with_suffix('.npy'), None] for x in self.samples] # file, index, npy, im + + def __getitem__(self, i): + f, j, fn, im = self.samples[i] # filename, index, filename.with_suffix('.npy'), image + if self.album_transforms: + if self.cache_ram and im is None: + im = self.samples[i][3] = cv2.imread(f) + elif self.cache_disk: + if not fn.exists(): # load npy + np.save(fn.as_posix(), cv2.imread(f)) + im = np.load(fn) + else: # read image + im = cv2.imread(f) # BGR + sample = self.album_transforms(image=cv2.cvtColor(im, cv2.COLOR_BGR2RGB))["image"] + else: + sample = self.torch_transforms(self.loader(f)) + return sample, j + + +def create_classification_dataloader(path, + imgsz=224, + batch_size=16, + augment=True, + cache=False, + rank=-1, + workers=8, + shuffle=True): + # Returns Dataloader object to be used with YOLOv5 Classifier + with torch_distributed_zero_first(rank): # init dataset *.cache only once if DDP + dataset = ClassificationDataset(root=path, imgsz=imgsz, augment=augment, cache=cache) + batch_size = min(batch_size, len(dataset)) + nd = torch.cuda.device_count() + nw = min([os.cpu_count() // max(nd, 1), batch_size if batch_size > 1 else 0, workers]) + sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle) + generator = torch.Generator() + generator.manual_seed(0) + return InfiniteDataLoader(dataset, + batch_size=batch_size, + shuffle=shuffle and sampler is None, + num_workers=nw, + sampler=sampler, + pin_memory=True, + worker_init_fn=seed_worker, + generator=generator) # or DataLoader(persistent_workers=True) diff --git a/utils/general.py b/utils/general.py index 2a3ce37cd853..1c525c45f649 100755 --- a/utils/general.py +++ b/utils/general.py @@ -217,7 +217,11 @@ def print_args(args: Optional[dict] = None, show_file=True, show_fcn=False): if args is None: # get args automatically args, _, _, frm = inspect.getargvalues(x) args = {k: v for k, v in frm.items() if k in args} - s = (f'{Path(file).stem}: ' if show_file else '') + (f'{fcn}: ' if show_fcn else '') + try: + file = Path(file).resolve().relative_to(ROOT).with_suffix('') + except ValueError: + file = Path(file).stem + s = (f'{file}: ' if show_file else '') + (f'{fcn}: ' if show_fcn else '') LOGGER.info(colorstr(s) + ', '.join(f'{k}={v}' for k, v in args.items())) @@ -345,7 +349,7 @@ def check_version(current='0.0.0', minimum='0.0.0', name='version ', pinned=Fals @try_except def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), install=True, cmds=()): - # Check installed dependencies meet requirements (pass *.txt file or list of packages) + # Check installed dependencies meet YOLOv5 requirements (pass *.txt file or list of packages) prefix = colorstr('red', 'bold', 'requirements:') check_python() # check python version if isinstance(requirements, (str, Path)): # requirements.txt file @@ -549,6 +553,18 @@ def amp_allclose(model, im): return False +def yaml_load(file='data.yaml'): + # Single-line safe yaml loading + with open(file, errors='ignore') as f: + return yaml.safe_load(f) + + +def yaml_save(file='data.yaml', data={}): + # Single-line safe yaml saving + with open(file, 'w') as f: + yaml.safe_dump({k: str(v) if isinstance(v, Path) else v for k, v in data.items()}, f, sort_keys=False) + + def url2file(url): # Convert URL to filename, i.e. https://url.com/file.txt?auth -> file.txt url = str(Path(url)).replace(':/', '://') # Pathlib turns :// -> :/ diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 0f3eceafd0db..8ec846f8cfac 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -5,6 +5,7 @@ import os import warnings +from pathlib import Path import pkg_resources as pkg import torch @@ -76,7 +77,7 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, self.logger.info(s) if not clearml: prefix = colorstr('ClearML: ') - s = f"{prefix}run 'pip install clearml' to automatically track, visualize and remotely train YOLOv5 🚀 runs in ClearML" + s = f"{prefix}run 'pip install clearml' to automatically track, visualize and remotely train YOLOv5 🚀 in ClearML" self.logger.info(s) # TensorBoard @@ -121,11 +122,8 @@ def on_train_batch_end(self, ni, model, imgs, targets, paths, plots): # Callback runs on train batch end # ni: number integrated batches (since train start) if plots: - if ni == 0: - if self.tb and not self.opt.sync_bn: # --sync known issue https://github.com/ultralytics/yolov5/issues/3754 - with warnings.catch_warnings(): - warnings.simplefilter('ignore') # suppress jit trace warning - self.tb.add_graph(torch.jit.trace(de_parallel(model), imgs[0:1], strict=False), []) + if ni == 0 and not self.opt.sync_bn and self.tb: + log_tensorboard_graph(self.tb, model, imgsz=list(imgs.shape[2:4])) if ni < 3: f = self.save_dir / f'train_batch{ni}.jpg' # filename plot_images(imgs, targets, paths, f) @@ -233,3 +231,78 @@ def on_params_update(self, params): # params: A dict containing {param: value} pairs if self.wandb: self.wandb.wandb_run.config.update(params, allow_val_change=True) + + +class GenericLogger: + """ + YOLOv5 General purpose logger for non-task specific logging + Usage: from utils.loggers import GenericLogger; logger = GenericLogger(...) + Arguments + opt: Run arguments + console_logger: Console logger + include: loggers to include + """ + + def __init__(self, opt, console_logger, include=('tb', 'wandb')): + # init default loggers + self.save_dir = opt.save_dir + self.include = include + self.console_logger = console_logger + if 'tb' in self.include: + prefix = colorstr('TensorBoard: ') + self.console_logger.info( + f"{prefix}Start with 'tensorboard --logdir {self.save_dir.parent}', view at http://localhost:6006/") + self.tb = SummaryWriter(str(self.save_dir)) + + if wandb and 'wandb' in self.include: + self.wandb = wandb.init(project="YOLOv5-Classifier" if opt.project == "runs/train" else opt.project, + name=None if opt.name == "exp" else opt.name, + config=opt) + else: + self.wandb = None + + def log_metrics(self, metrics_dict, epoch): + # Log metrics dictionary to all loggers + if self.tb: + for k, v in metrics_dict.items(): + self.tb.add_scalar(k, v, epoch) + + if self.wandb: + self.wandb.log(metrics_dict, step=epoch) + + def log_images(self, files, name='Images', epoch=0): + # Log images to all loggers + files = [Path(f) for f in (files if isinstance(files, (tuple, list)) else [files])] # to Path + files = [f for f in files if f.exists()] # filter by exists + + if self.tb: + for f in files: + self.tb.add_image(f.stem, cv2.imread(str(f))[..., ::-1], epoch, dataformats='HWC') + + if self.wandb: + self.wandb.log({name: [wandb.Image(str(f), caption=f.name) for f in files]}, step=epoch) + + def log_graph(self, model, imgsz=(640, 640)): + # Log model graph to all loggers + if self.tb: + log_tensorboard_graph(self.tb, model, imgsz) + + def log_model(self, model_path, epoch=0, metadata={}): + # Log model to all loggers + if self.wandb: + art = wandb.Artifact(name=f"run_{wandb.run.id}_model", type="model", metadata=metadata) + art.add_file(str(model_path)) + wandb.log_artifact(art) + + +def log_tensorboard_graph(tb, model, imgsz=(640, 640)): + # Log model graph to TensorBoard + try: + p = next(model.parameters()) # for device, type + imgsz = (imgsz, imgsz) if isinstance(imgsz, int) else imgsz # expand + im = torch.zeros((1, 3, *imgsz)).to(p.device).type_as(p) # input image + with warnings.catch_warnings(): + warnings.simplefilter('ignore') # suppress jit trace warning + tb.add_graph(torch.jit.trace(de_parallel(model), im, strict=False), []) + except Exception: + print('WARNING: TensorBoard graph visualization failure') diff --git a/utils/plots.py b/utils/plots.py index d050f5d36aba..7417308c4d82 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -388,6 +388,35 @@ def plot_labels(labels, names=(), save_dir=Path('')): plt.close() +def imshow_cls(im, labels=None, pred=None, names=None, nmax=25, verbose=False, f=Path('images.jpg')): + # Show classification image grid with labels (optional) and predictions (optional) + from utils.augmentations import denormalize + + names = names or [f'class{i}' for i in range(1000)] + blocks = torch.chunk(denormalize(im.clone()).cpu().float(), len(im), + dim=0) # select batch index 0, block by channels + n = min(len(blocks), nmax) # number of plots + m = min(8, round(n ** 0.5)) # 8 x 8 default + fig, ax = plt.subplots(math.ceil(n / m), m) # 8 rows x n/8 cols + ax = ax.ravel() if m > 1 else [ax] + # plt.subplots_adjust(wspace=0.05, hspace=0.05) + for i in range(n): + ax[i].imshow(blocks[i].squeeze().permute((1, 2, 0)).numpy().clip(0.0, 1.0)) + ax[i].axis('off') + if labels is not None: + s = names[labels[i]] + (f'—{names[pred[i]]}' if pred is not None else '') + ax[i].set_title(s, fontsize=8, verticalalignment='top') + plt.savefig(f, dpi=300, bbox_inches='tight') + plt.close() + if verbose: + LOGGER.info(f"Saving {f}") + if labels is not None: + LOGGER.info('True: ' + ' '.join(f'{names[i]:3s}' for i in labels[:nmax])) + if pred is not None: + LOGGER.info('Predicted:' + ' '.join(f'{names[i]:3s}' for i in pred[:nmax])) + return f + + def plot_evolve(evolve_csv='path/to/evolve.csv'): # from utils.plots import *; plot_evolve() # Plot evolve.csv hyp evolution results evolve_csv = Path(evolve_csv) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 1ceb0aa346e9..1cdbe20f8670 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -42,6 +42,16 @@ def decorate(fn): return decorate +def smartCrossEntropyLoss(label_smoothing=0.0): + # Returns nn.CrossEntropyLoss with label smoothing enabled for torch>=1.10.0 + if check_version(torch.__version__, '1.10.0'): + return nn.CrossEntropyLoss(label_smoothing=label_smoothing) # loss function + else: + if label_smoothing > 0: + LOGGER.warning(f'WARNING: label smoothing {label_smoothing} requires torch>=1.10.0') + return nn.CrossEntropyLoss() # loss function + + def smart_DDP(model): # Model DDP creation with checks assert not check_version(torch.__version__, '1.12.0', pinned=True), \ @@ -53,6 +63,28 @@ def smart_DDP(model): return DDP(model, device_ids=[LOCAL_RANK], output_device=LOCAL_RANK) +def reshape_classifier_output(model, n=1000): + # Update a TorchVision classification model to class count 'n' if required + from models.common import Classify + name, m = list((model.model if hasattr(model, 'model') else model).named_children())[-1] # last module + if isinstance(m, Classify): # YOLOv5 Classify() head + if m.linear.out_features != n: + m.linear = nn.Linear(m.linear.in_features, n) + elif isinstance(m, nn.Linear): # ResNet, EfficientNet + if m.out_features != n: + setattr(model, name, nn.Linear(m.in_features, n)) + elif isinstance(m, nn.Sequential): + types = [type(x) for x in m] + if nn.Linear in types: + i = types.index(nn.Linear) # nn.Linear index + if m[i].out_features != n: + m[i] = nn.Linear(m[i].in_features, n) + elif nn.Conv2d in types: + i = types.index(nn.Conv2d) # nn.Conv2d index + if m[i].out_channels != n: + m[i] = nn.Conv2d(m[i].in_channels, n, m[i].kernel_size, m[i].stride, bias=m[i].bias) + + @contextmanager def torch_distributed_zero_first(local_rank: int): # Decorator to make all processes in distributed training wait for each local_master to do something @@ -117,14 +149,13 @@ def time_sync(): def profile(input, ops, n=10, device=None): - # YOLOv5 speed/memory/FLOPs profiler - # - # Usage: - # input = torch.randn(16, 3, 640, 640) - # m1 = lambda x: x * torch.sigmoid(x) - # m2 = nn.SiLU() - # profile(input, [m1, m2], n=100) # profile over 100 iterations - + """ YOLOv5 speed/memory/FLOPs profiler + Usage: + input = torch.randn(16, 3, 640, 640) + m1 = lambda x: x * torch.sigmoid(x) + m2 = nn.SiLU() + profile(input, [m1, m2], n=100) # profile over 100 iterations + """ results = [] if not isinstance(device, torch.device): device = select_device(device) @@ -313,6 +344,18 @@ def smart_optimizer(model, name='Adam', lr=0.001, momentum=0.9, decay=1e-5): return optimizer +def smart_hub_load(repo='ultralytics/yolov5', model='yolov5s', **kwargs): + # YOLOv5 torch.hub.load() wrapper with smart error/issue handling + if check_version(torch.__version__, '1.9.1'): + kwargs['skip_validation'] = True # validation causes GitHub API rate limit errors + if check_version(torch.__version__, '1.12.0'): + kwargs['trust_repo'] = True # argument required starting in torch 0.12 + try: + return torch.hub.load(repo, model, **kwargs) + except Exception: + return torch.hub.load(repo, model, force_reload=True, **kwargs) + + def smart_resume(ckpt, optimizer, ema=None, weights='yolov5s.pt', epochs=300, resume=True): # Resume training from a partially trained checkpoint best_fitness = 0.0 From e61756910758f59406255269921e55992ca0b64b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 17 Aug 2022 15:33:37 +0200 Subject: [PATCH 465/661] Improve classification comments (#8997) --- .github/README_cn.md | 10 +++++----- README.md | 10 +++++----- classify/predict.py | 2 +- classify/train.py | 4 +++- classify/val.py | 3 ++- 5 files changed, 16 insertions(+), 13 deletions(-) diff --git a/.github/README_cn.md b/.github/README_cn.md index 86b502df61f7..816adf6b0449 100644 --- a/.github/README_cn.md +++ b/.github/README_cn.md @@ -269,7 +269,7 @@ We trained YOLOv5-cls classification models on ImageNet for 90 epochs using a 4x
Table Notes (click to expand) -- All checkpoints are trained to 90 epochs with SGD optimizer with lr0=0.001 at image size 224 and all default settings. Runs logged to https://wandb.ai/glenn-jocher/YOLOv5-Classifier-v6-2. +- All checkpoints are trained to 90 epochs with SGD optimizer with `lr0=0.001` and `weight_decay=5e-5` at image size 224 and all default settings.
Runs logged to https://wandb.ai/glenn-jocher/YOLOv5-Classifier-v6-2 - **Accuracy** values are for single-model single-scale on [ImageNet-1k](https://www.image-net.org/index.php) dataset.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224` - **Speed** averaged over 100 inference images using a Google [Colab Pro](https://colab.research.google.com/signup) V100 High-RAM instance.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224 --batch 1` - **Export** to ONNX at FP32 and TensorRT at FP16 done with `export.py`.
Reproduce by `python export.py --weights yolov5s-cls.pt --include engine onnx --imgsz 224` @@ -291,14 +291,14 @@ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/trai ``` ### Val -Validate accuracy on a pretrained model. To validate YOLOv5s-cls accuracy on ImageNet. +Validate YOLOv5m-cls accuracy on ImageNet-1k dataset: ```bash bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images) -python classify/val.py --weights yolov5s-cls.pt --data ../datasets/imagenet --img 224 +python classify/val.py --weights yolov5m-cls.pt --data ../datasets/imagenet --img 224 # validate ``` ### Predict -Run a classification prediction on an image. +Use pretrained YOLOv5s-cls.pt to predict bus.jpg: ```bash python classify/predict.py --weights yolov5s-cls.pt --data data/images/bus.jpg ``` @@ -307,7 +307,7 @@ model = torch.hub.load('ultralytics/yolov5', 'custom', 'yolov5s-cls.pt') # load ``` ### Export -Export a group of trained YOLOv5-cls, ResNet and EfficientNet models to ONNX and TensorRT. +Export a group of trained YOLOv5s-cls, ResNet and EfficientNet models to ONNX and TensorRT: ```bash python export.py --weights yolov5s-cls.pt resnet50.pt efficientnet_b0.pt --include onnx engine --img 224 ``` diff --git a/README.md b/README.md index b368d1d6e264..7335394402da 100644 --- a/README.md +++ b/README.md @@ -278,7 +278,7 @@ We trained YOLOv5-cls classification models on ImageNet for 90 epochs using a 4x
Table Notes (click to expand) -- All checkpoints are trained to 90 epochs with SGD optimizer with lr0=0.001 at image size 224 and all default settings. Runs logged to https://wandb.ai/glenn-jocher/YOLOv5-Classifier-v6-2. +- All checkpoints are trained to 90 epochs with SGD optimizer with `lr0=0.001` and `weight_decay=5e-5` at image size 224 and all default settings.
Runs logged to https://wandb.ai/glenn-jocher/YOLOv5-Classifier-v6-2 - **Accuracy** values are for single-model single-scale on [ImageNet-1k](https://www.image-net.org/index.php) dataset.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224` - **Speed** averaged over 100 inference images using a Google [Colab Pro](https://colab.research.google.com/signup) V100 High-RAM instance.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224 --batch 1` - **Export** to ONNX at FP32 and TensorRT at FP16 done with `export.py`.
Reproduce by `python export.py --weights yolov5s-cls.pt --include engine onnx --imgsz 224` @@ -300,14 +300,14 @@ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/trai ``` ### Val -Validate accuracy on a pretrained model. To validate YOLOv5s-cls accuracy on ImageNet. +Validate YOLOv5m-cls accuracy on ImageNet-1k dataset: ```bash bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images) -python classify/val.py --weights yolov5s-cls.pt --data ../datasets/imagenet --img 224 +python classify/val.py --weights yolov5m-cls.pt --data ../datasets/imagenet --img 224 # validate ``` ### Predict -Run a classification prediction on an image. +Use pretrained YOLOv5s-cls.pt to predict bus.jpg: ```bash python classify/predict.py --weights yolov5s-cls.pt --data data/images/bus.jpg ``` @@ -316,7 +316,7 @@ model = torch.hub.load('ultralytics/yolov5', 'custom', 'yolov5s-cls.pt') # load ``` ### Export -Export a group of trained YOLOv5-cls, ResNet and EfficientNet models to ONNX and TensorRT. +Export a group of trained YOLOv5s-cls, ResNet and EfficientNet models to ONNX and TensorRT: ```bash python export.py --weights yolov5s-cls.pt resnet50.pt efficientnet_b0.pt --include onnx engine --img 224 ``` diff --git a/classify/predict.py b/classify/predict.py index 419830d43952..4247e3c8e7fa 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -3,7 +3,7 @@ Run classification inference on images Usage: - $ python classify/predict.py --weights yolov5s-cls.pt --source im.jpg + $ python classify/predict.py --weights yolov5s-cls.pt --source data/images/bus.jpg """ import argparse diff --git a/classify/train.py b/classify/train.py index f2b465567446..b85f14236039 100644 --- a/classify/train.py +++ b/classify/train.py @@ -2,8 +2,10 @@ """ Train a YOLOv5 classifier model on a classification dataset Datasets: --data mnist, fashion-mnist, cifar10, cifar100, imagenette, imagewoof, imagenet, or 'path/to/custom/dataset' +YOLOv5-cls models: --model yolov5n-cls.pt, yolov5s-cls.pt, yolov5m-cls.pt, yolov5l-cls.pt, yolov5x-cls.pt +Torchvision models: --model resnet50, efficientnet_b0, etc. See https://pytorch.org/vision/stable/models.html -Usage: +Usage - Single-GPU and Multi-GPU DDP $ python classify/train.py --model yolov5s-cls.pt --data cifar100 --epochs 5 --img 128 $ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/train.py --model yolov5s-cls.pt --data imagenet --epochs 5 --img 224 --device 0,1,2,3 """ diff --git a/classify/val.py b/classify/val.py index 0930ba8c9c51..9d965d9f1fdc 100644 --- a/classify/val.py +++ b/classify/val.py @@ -3,7 +3,8 @@ Validate a classification model on a dataset Usage: - $ python classify/val.py --weights yolov5s-cls.pt --data ../datasets/imagenet + $ bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images) + $ python classify/val.py --weights yolov5m-cls.pt --data ../datasets/imagenet --img 224 # validate """ import argparse From 7c9486e16f6a2c35bf5cfca892898a11a81009fc Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 17 Aug 2022 15:48:17 +0200 Subject: [PATCH 466/661] Update `attempt_download(release='v6.2')` (#8998) * Update attempt_download(release='v6.2') Signed-off-by: Glenn Jocher * Update README.md Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- README.md | 20 ++++++++++---------- utils/downloads.py | 8 ++++---- 2 files changed, 14 insertions(+), 14 deletions(-) diff --git a/README.md b/README.md index 7335394402da..89e4f1199cde 100644 --- a/README.md +++ b/README.md @@ -224,17 +224,17 @@ Get started in seconds with our verified environments. Click each icon below for | Model | size
(pixels) | mAPval
0.5:0.95 | mAPval
0.5 | Speed
CPU b1
(ms) | Speed
V100 b1
(ms) | Speed
V100 b32
(ms) | params
(M) | FLOPs
@640 (B) | |------------------------------------------------------------------------------------------------------|-----------------------|-------------------------|--------------------|------------------------------|-------------------------------|--------------------------------|--------------------|------------------------| -| [YOLOv5n](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5n.pt) | 640 | 28.0 | 45.7 | **45** | **6.3** | **0.6** | **1.9** | **4.5** | -| [YOLOv5s](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5s.pt) | 640 | 37.4 | 56.8 | 98 | 6.4 | 0.9 | 7.2 | 16.5 | -| [YOLOv5m](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5m.pt) | 640 | 45.4 | 64.1 | 224 | 8.2 | 1.7 | 21.2 | 49.0 | -| [YOLOv5l](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5l.pt) | 640 | 49.0 | 67.3 | 430 | 10.1 | 2.7 | 46.5 | 109.1 | -| [YOLOv5x](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5x.pt) | 640 | 50.7 | 68.9 | 766 | 12.1 | 4.8 | 86.7 | 205.7 | +| [YOLOv5n](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5n.pt) | 640 | 28.0 | 45.7 | **45** | **6.3** | **0.6** | **1.9** | **4.5** | +| [YOLOv5s](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s.pt) | 640 | 37.4 | 56.8 | 98 | 6.4 | 0.9 | 7.2 | 16.5 | +| [YOLOv5m](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5m.pt) | 640 | 45.4 | 64.1 | 224 | 8.2 | 1.7 | 21.2 | 49.0 | +| [YOLOv5l](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5l.pt) | 640 | 49.0 | 67.3 | 430 | 10.1 | 2.7 | 46.5 | 109.1 | +| [YOLOv5x](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5x.pt) | 640 | 50.7 | 68.9 | 766 | 12.1 | 4.8 | 86.7 | 205.7 | | | | | | | | | | | -| [YOLOv5n6](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5n6.pt) | 1280 | 36.0 | 54.4 | 153 | 8.1 | 2.1 | 3.2 | 4.6 | -| [YOLOv5s6](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5s6.pt) | 1280 | 44.8 | 63.7 | 385 | 8.2 | 3.6 | 12.6 | 16.8 | -| [YOLOv5m6](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5m6.pt) | 1280 | 51.3 | 69.3 | 887 | 11.1 | 6.8 | 35.7 | 50.0 | -| [YOLOv5l6](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5l6.pt) | 1280 | 53.7 | 71.3 | 1784 | 15.8 | 10.5 | 76.8 | 111.4 | -| [YOLOv5x6](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5x6.pt)
+ [TTA][TTA] | 1280
1536 | 55.0
**55.8** | 72.7
**72.7** | 3136
- | 26.2
- | 19.4
- | 140.7
- | 209.8
- | +| [YOLOv5n6](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5n6.pt) | 1280 | 36.0 | 54.4 | 153 | 8.1 | 2.1 | 3.2 | 4.6 | +| [YOLOv5s6](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s6.pt) | 1280 | 44.8 | 63.7 | 385 | 8.2 | 3.6 | 12.6 | 16.8 | +| [YOLOv5m6](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5m6.pt) | 1280 | 51.3 | 69.3 | 887 | 11.1 | 6.8 | 35.7 | 50.0 | +| [YOLOv5l6](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5l6.pt) | 1280 | 53.7 | 71.3 | 1784 | 15.8 | 10.5 | 76.8 | 111.4 | +| [YOLOv5x6](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5x6.pt)
+ [TTA][TTA] | 1280
1536 | 55.0
**55.8** | 72.7
**72.7** | 3136
- | 26.2
- | 19.4
- | 140.7
- | 209.8
- |
Table Notes (click to expand) diff --git a/utils/downloads.py b/utils/downloads.py index 9d4780ad28b1..c4d4a85c38ae 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -54,14 +54,14 @@ def safe_download(file, url, url2=None, min_bytes=1E0, error_msg=''): LOGGER.info('') -def attempt_download(file, repo='ultralytics/yolov5', release='v6.1'): - # Attempt file download from GitHub release assets if not found locally. release = 'latest', 'v6.1', etc. +def attempt_download(file, repo='ultralytics/yolov5', release='v6.2'): + # Attempt file download from GitHub release assets if not found locally. release = 'latest', 'v6.2', etc. from utils.general import LOGGER def github_assets(repository, version='latest'): - # Return GitHub repo tag (i.e. 'v6.1') and assets (i.e. ['yolov5s.pt', 'yolov5m.pt', ...]) + # Return GitHub repo tag (i.e. 'v6.2') and assets (i.e. ['yolov5s.pt', 'yolov5m.pt', ...]) if version != 'latest': - version = f'tags/{version}' # i.e. tags/v6.1 + version = f'tags/{version}' # i.e. tags/v6.2 response = requests.get(f'https://api.github.com/repos/{repository}/releases/{version}').json() # github api return response['tag_name'], [x['name'] for x in response['assets']] # tag, assets From fe809b8dad5236d86d5acbe047b5e0e6895b2b8a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 17 Aug 2022 16:18:06 +0200 Subject: [PATCH 467/661] Created using Colaboratory --- tutorial.ipynb | 304 ++++++++++++++++++++++++------------------------- 1 file changed, 152 insertions(+), 152 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 61641bab1833..1438924e4112 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -17,7 +17,7 @@ "accelerator": "GPU", "widgets": { "application/vnd.jupyter.widget-state+json": { - "c31d2039ccf74c22b67841f4877d1186": { + "57c562894aed45cd9a107d0455e3e3f4": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", @@ -32,14 +32,14 @@ "_view_name": "HBoxView", "box_style": "", "children": [ - "IPY_MODEL_d4bba1727c714d94ad58a72bffa07c4c", - "IPY_MODEL_9aeff9f1780b45f892422fdc96e56913", - "IPY_MODEL_bf55a7c71d074d3fa88b10b997820825" + "IPY_MODEL_040d53c6cc924350bcb656cd21a7c713", + "IPY_MODEL_e029890942a74c098408ce5a9a566d51", + "IPY_MODEL_8fb991c03e434566a4297b6ab9446f89" ], - "layout": "IPY_MODEL_d8b66044e2fb4f5b916696834d880c81" + "layout": "IPY_MODEL_a9a376923a7742d89fb335db709c7a7e" } }, - "d4bba1727c714d94ad58a72bffa07c4c": { + "040d53c6cc924350bcb656cd21a7c713": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", @@ -54,13 +54,13 @@ "_view_name": "HTMLView", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_102e1deda239436fa72751c58202fa0f", + "layout": "IPY_MODEL_8b4276ac834c4735bf60ee9b761b9962", "placeholder": "​", - "style": "IPY_MODEL_4fd4431ced6c42368e18424912b877e4", + "style": "IPY_MODEL_52cc8da75b724198856617247541cb1e", "value": "100%" } }, - "9aeff9f1780b45f892422fdc96e56913": { + "e029890942a74c098408ce5a9a566d51": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", @@ -76,15 +76,15 @@ "bar_style": "success", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_cdd709c4f40941bea1b2053523c9fac8", + "layout": "IPY_MODEL_b6652f46480243c4adf60e6440043d6f", "max": 818322941, "min": 0, "orientation": "horizontal", - "style": "IPY_MODEL_a1ef2d8de2b741c78ca5d938e2ddbcdf", + "style": "IPY_MODEL_e502754177ff4ea8abf82a6e9ac77a4a", "value": 818322941 } }, - "bf55a7c71d074d3fa88b10b997820825": { + "8fb991c03e434566a4297b6ab9446f89": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", @@ -99,13 +99,13 @@ "_view_name": "HTMLView", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_0dbce99bb6184238842cbec0587d564a", + "layout": "IPY_MODEL_447398becdb04836b5ffb5915318db07", "placeholder": "​", - "style": "IPY_MODEL_91ff5f93f2a24c5790ab29e347965946", - "value": " 780M/780M [01:10<00:00, 10.5MB/s]" + "style": "IPY_MODEL_2fddcb27ad4a4caa81ff51111f8d0ed6", + "value": " 780M/780M [01:17<00:00, 12.3MB/s]" } }, - "d8b66044e2fb4f5b916696834d880c81": { + "a9a376923a7742d89fb335db709c7a7e": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -157,7 +157,7 @@ "width": null } }, - "102e1deda239436fa72751c58202fa0f": { + "8b4276ac834c4735bf60ee9b761b9962": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -209,7 +209,7 @@ "width": null } }, - "4fd4431ced6c42368e18424912b877e4": { + "52cc8da75b724198856617247541cb1e": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", @@ -224,7 +224,7 @@ "description_width": "" } }, - "cdd709c4f40941bea1b2053523c9fac8": { + "b6652f46480243c4adf60e6440043d6f": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -276,7 +276,7 @@ "width": null } }, - "a1ef2d8de2b741c78ca5d938e2ddbcdf": { + "e502754177ff4ea8abf82a6e9ac77a4a": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", @@ -292,7 +292,7 @@ "description_width": "" } }, - "0dbce99bb6184238842cbec0587d564a": { + "447398becdb04836b5ffb5915318db07": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -344,7 +344,7 @@ "width": null } }, - "91ff5f93f2a24c5790ab29e347965946": { + "2fddcb27ad4a4caa81ff51111f8d0ed6": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", @@ -404,7 +404,7 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "185d0979-edcd-4860-e6fb-b8a27dbf5096" + "outputId": "e0f693e4-413b-4cc8-ae7e-91537da370b0" }, "source": [ "!git clone https://github.com/ultralytics/yolov5 # clone\n", @@ -415,13 +415,13 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": null, + "execution_count": 1, "outputs": [ { "output_type": "stream", "name": "stderr", "text": [ - "YOLOv5 🚀 v6.1-370-g20f1b7e Python-3.7.13 torch-1.12.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" + "YOLOv5 🚀 v6.2-2-g7c9486e Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" ] }, { @@ -461,29 +461,29 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "4b13989f-32a4-4ef0-b403-06ff3aac255c" + "outputId": "941d625b-01a1-4f1b-dfd2-d9ef1c945715" }, "source": [ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", - "#display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" + "# display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": null, + "execution_count": 2, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mdetect: \u001b[0mweights=['yolov5s.pt'], source=data/images, data=data/coco128.yaml, imgsz=[640, 640], conf_thres=0.25, iou_thres=0.45, max_det=1000, device=, view_img=False, save_txt=False, save_conf=False, save_crop=False, nosave=False, classes=None, agnostic_nms=False, augment=False, visualize=False, update=False, project=runs/detect, name=exp, exist_ok=False, line_thickness=3, hide_labels=False, hide_conf=False, half=False, dnn=False\n", - "YOLOv5 🚀 v6.1-370-g20f1b7e Python-3.7.13 torch-1.12.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-2-g7c9486e Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", - "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5s.pt to yolov5s.pt...\n", - "100% 14.1M/14.1M [00:00<00:00, 53.9MB/s]\n", + "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s.pt to yolov5s.pt...\n", + "100% 14.1M/14.1M [00:00<00:00, 50.5MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5s summary: 213 layers, 7225885 parameters, 0 gradients\n", - "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, Done. (0.016s)\n", - "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, Done. (0.021s)\n", - "Speed: 0.6ms pre-process, 18.6ms inference, 25.0ms NMS per image at shape (1, 3, 640, 640)\n", + "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, Done. (0.014s)\n", + "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, Done. (0.020s)\n", + "Speed: 0.6ms pre-process, 17.0ms inference, 20.2ms NMS per image at shape (1, 3, 640, 640)\n", "Results saved to \u001b[1mruns/detect/exp\u001b[0m\n" ] } @@ -527,27 +527,27 @@ "base_uri": "https://localhost:8080/", "height": 49, "referenced_widgets": [ - "c31d2039ccf74c22b67841f4877d1186", - "d4bba1727c714d94ad58a72bffa07c4c", - "9aeff9f1780b45f892422fdc96e56913", - "bf55a7c71d074d3fa88b10b997820825", - "d8b66044e2fb4f5b916696834d880c81", - "102e1deda239436fa72751c58202fa0f", - "4fd4431ced6c42368e18424912b877e4", - "cdd709c4f40941bea1b2053523c9fac8", - "a1ef2d8de2b741c78ca5d938e2ddbcdf", - "0dbce99bb6184238842cbec0587d564a", - "91ff5f93f2a24c5790ab29e347965946" + "57c562894aed45cd9a107d0455e3e3f4", + "040d53c6cc924350bcb656cd21a7c713", + "e029890942a74c098408ce5a9a566d51", + "8fb991c03e434566a4297b6ab9446f89", + "a9a376923a7742d89fb335db709c7a7e", + "8b4276ac834c4735bf60ee9b761b9962", + "52cc8da75b724198856617247541cb1e", + "b6652f46480243c4adf60e6440043d6f", + "e502754177ff4ea8abf82a6e9ac77a4a", + "447398becdb04836b5ffb5915318db07", + "2fddcb27ad4a4caa81ff51111f8d0ed6" ] }, - "outputId": "a9004b06-37a6-41ed-a1f2-ac956f3963b3" + "outputId": "d593b41a-55e7-48a5-e285-5df449edc8c0" }, "source": [ "# Download COCO val\n", "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip')\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip" ], - "execution_count": null, + "execution_count": 3, "outputs": [ { "output_type": "display_data", @@ -558,7 +558,7 @@ "application/vnd.jupyter.widget-view+json": { "version_major": 2, "version_minor": 0, - "model_id": "c31d2039ccf74c22b67841f4877d1186" + "model_id": "57c562894aed45cd9a107d0455e3e3f4" } }, "metadata": {} @@ -572,48 +572,48 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "c0f29758-4ec8-4def-893d-0efd6ed5b7f4" + "outputId": "701132a6-9ca8-4e1f-c89f-5d38893a6fc4" }, "source": [ "# Run YOLOv5x on COCO val\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" ], - "execution_count": null, + "execution_count": 4, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mval: \u001b[0mdata=/content/yolov5/data/coco.yaml, weights=['yolov5x.pt'], batch_size=32, imgsz=640, conf_thres=0.001, iou_thres=0.65, task=val, device=, workers=8, single_cls=False, augment=False, verbose=False, save_txt=False, save_hybrid=False, save_conf=False, save_json=True, project=runs/val, name=exp, exist_ok=False, half=True, dnn=False\n", - "YOLOv5 🚀 v6.1-370-g20f1b7e Python-3.7.13 torch-1.12.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-2-g7c9486e Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", - "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5x.pt to yolov5x.pt...\n", - "100% 166M/166M [00:35<00:00, 4.97MB/s]\n", + "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5x.pt to yolov5x.pt...\n", + "100% 166M/166M [00:11<00:00, 15.1MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5x summary: 444 layers, 86705005 parameters, 0 gradients\n", "Downloading https://ultralytics.com/assets/Arial.ttf to /root/.config/Ultralytics/Arial.ttf...\n", - "100% 755k/755k [00:00<00:00, 49.4MB/s]\n", - "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 10716.86it/s]\n", + "100% 755k/755k [00:00<00:00, 48.6MB/s]\n", + "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 10889.87it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mNew cache created: /content/datasets/coco/val2017.cache\n", - " Class Images Labels P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:08<00:00, 2.28it/s]\n", + " Class Images Labels P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:05<00:00, 2.38it/s]\n", " all 5000 36335 0.743 0.625 0.683 0.504\n", - "Speed: 0.1ms pre-process, 4.6ms inference, 1.2ms NMS per image at shape (32, 3, 640, 640)\n", + "Speed: 0.1ms pre-process, 4.7ms inference, 1.0ms NMS per image at shape (32, 3, 640, 640)\n", "\n", "Evaluating pycocotools mAP... saving runs/val/exp/yolov5x_predictions.json...\n", "loading annotations into memory...\n", - "Done (t=0.41s)\n", + "Done (t=0.39s)\n", "creating index...\n", "index created!\n", "Loading and preparing results...\n", - "DONE (t=5.64s)\n", + "DONE (t=5.53s)\n", "creating index...\n", "index created!\n", "Running per image evaluation...\n", "Evaluate annotation type *bbox*\n", - "DONE (t=72.86s).\n", + "DONE (t=73.01s).\n", "Accumulating evaluation results...\n", - "DONE (t=14.20s).\n", + "DONE (t=15.27s).\n", " Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.506\n", " Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.688\n", " Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.549\n", @@ -745,13 +745,13 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "bce1b4bd-1a14-4c07-aebd-6c11e91ad24b" + "outputId": "50a9318f-d438-41d5-db95-928f1842c057" }, "source": [ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": null, + "execution_count": 5, "outputs": [ { "output_type": "stream", @@ -759,17 +759,17 @@ "text": [ "\u001b[34m\u001b[1mtrain: \u001b[0mweights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch-low.yaml, epochs=3, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, noplots=False, evolve=None, bucket=, cache=ram, image_weights=False, device=, multi_scale=False, single_cls=False, optimizer=SGD, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, cos_lr=False, label_smoothing=0.0, patience=100, freeze=[0], save_period=-1, seed=0, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest\n", "\u001b[34m\u001b[1mgithub: \u001b[0mup to date with https://github.com/ultralytics/yolov5 ✅\n", - "YOLOv5 🚀 v6.1-370-g20f1b7e Python-3.7.13 torch-1.12.0+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-2-g7c9486e Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "\u001b[34m\u001b[1mhyperparameters: \u001b[0mlr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0\n", "\u001b[34m\u001b[1mWeights & Biases: \u001b[0mrun 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs in Weights & Biases\n", - "\u001b[34m\u001b[1mClearML: \u001b[0mrun 'pip install clearml' to automatically track, visualize and remotely train YOLOv5 🚀 runs in ClearML\n", + "\u001b[34m\u001b[1mClearML: \u001b[0mrun 'pip install clearml' to automatically track, visualize and remotely train YOLOv5 🚀 in ClearML\n", "\u001b[34m\u001b[1mTensorBoard: \u001b[0mStart with 'tensorboard --logdir runs/train', view at http://localhost:6006/\n", "\n", "Dataset not found ⚠️, missing paths ['/content/datasets/coco128/images/train2017']\n", "Downloading https://ultralytics.com/assets/coco128.zip to coco128.zip...\n", - "100% 6.66M/6.66M [00:00<00:00, 75.2MB/s]\n", - "Dataset download success ✅ (0.7s), saved to \u001b[1m/content/datasets\u001b[0m\n", + "100% 6.66M/6.66M [00:00<00:00, 12.4MB/s]\n", + "Dataset download success ✅ (1.3s), saved to \u001b[1m/content/datasets\u001b[0m\n", "\n", " from n params module arguments \n", " 0 -1 1 3520 models.common.Conv [3, 32, 6, 2, 2] \n", @@ -802,12 +802,12 @@ "Transferred 349/349 items from yolov5s.pt\n", "\u001b[34m\u001b[1mAMP: \u001b[0mchecks passed ✅\n", "\u001b[34m\u001b[1moptimizer:\u001b[0m SGD(lr=0.01) with parameter groups 57 weight(decay=0.0), 60 weight(decay=0.0005), 60 bias\n", - "\u001b[34m\u001b[1malbumentations: \u001b[0mBlur(always_apply=False, p=0.01, blur_limit=(3, 7)), MedianBlur(always_apply=False, p=0.01, blur_limit=(3, 7)), ToGray(always_apply=False, p=0.01), CLAHE(always_apply=False, p=0.01, clip_limit=(1, 4.0), tile_grid_size=(8, 8))\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017' images and labels...128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00<00:00, 7926.40it/s]\n", + "\u001b[34m\u001b[1malbumentations: \u001b[0mBlur(p=0.01, blur_limit=(3, 7)), MedianBlur(p=0.01, blur_limit=(3, 7)), ToGray(p=0.01), CLAHE(p=0.01, clip_limit=(1, 4.0), tile_grid_size=(8, 8))\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017' images and labels...128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00<00:00, 8516.89it/s]\n", "\u001b[34m\u001b[1mtrain: \u001b[0mNew cache created: /content/datasets/coco128/labels/train2017.cache\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mCaching images (0.1GB ram): 100% 128/128 [00:00<00:00, 975.81it/s]\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mCaching images (0.1GB ram): 100% 128/128 [00:00<00:00, 1043.44it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco128/labels/train2017.cache' images and labels... 128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00 Date: Wed, 17 Aug 2022 17:50:32 +0200 Subject: [PATCH 468/661] Update README_cn.md (#9001) Includes v6.2 updates Signed-off-by: KieraMengru0907 <108015280+KieraMengru0907@users.noreply.github.com> Signed-off-by: KieraMengru0907 <108015280+KieraMengru0907@users.noreply.github.com> --- .github/README_cn.md | 65 +++++++++++++++++++++++--------------------- 1 file changed, 34 insertions(+), 31 deletions(-) diff --git a/.github/README_cn.md b/.github/README_cn.md index 816adf6b0449..46aafd86ec9b 100644 --- a/.github/README_cn.md +++ b/.github/README_cn.md @@ -130,19 +130,22 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12
教程 -- [训练自定义数据](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data) 🚀 推荐 -- [获得最佳训练效果的技巧](https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results) ☘️ 推荐 -- [使用 Weights & Biases 记录实验](https://github.com/ultralytics/yolov5/issues/1289) 🌟 新 -- [Roboflow：数据集、标签和主动学习](https://github.com/ultralytics/yolov5/issues/4975) 🌟 新 +- [训练自定义数据集](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data) 🚀 推荐 +- [获得最佳训练效果的技巧](https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results) ☘️ + 推荐 - [多GPU训练](https://github.com/ultralytics/yolov5/issues/475) -- [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36) ⭐ 新 -- [TFLite, ONNX, CoreML, TensorRT 导出](https://github.com/ultralytics/yolov5/issues/251) 🚀 +- [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36) 🌟 新 +- [TFLite, ONNX, CoreML, TensorRT 输出](https://github.com/ultralytics/yolov5/issues/251) 🚀 - [测试时数据增强 (TTA)](https://github.com/ultralytics/yolov5/issues/303) - [模型集成](https://github.com/ultralytics/yolov5/issues/318) - [模型剪枝/稀疏性](https://github.com/ultralytics/yolov5/issues/304) - [超参数进化](https://github.com/ultralytics/yolov5/issues/607) -- [带有冻结层的迁移学习](https://github.com/ultralytics/yolov5/issues/1314) ⭐ 新 -- [架构概要](https://github.com/ultralytics/yolov5/issues/6998) ⭐ 新 +- [带有冻结层的迁移学习](https://github.com/ultralytics/yolov5/issues/1314) +- [架构概要](https://github.com/ultralytics/yolov5/issues/6998) 🌟 新 +- [使用Weights & Biases 记录实验](https://github.com/ultralytics/yolov5/issues/1289) +- [Roboflow：数据集，标签和主动学习](https://github.com/ultralytics/yolov5/issues/4975) 🌟 新 +- [使用ClearML 记录实验](https://github.com/ultralytics/yolov5/tree/master/utils/loggers/clearml) 🌟 新 +- [Deci 平台](https://github.com/ultralytics/yolov5/wiki/Deci-Platform) 🌟 新
@@ -186,7 +189,7 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 |Deci ⭐ NEW|ClearML ⭐ NEW|Roboflow|Weights & Biases |:-:|:-:|:-:|:-:| -|Automatically compile and quantize YOLOv5 for better inference performance in one click at [Deci](https://bit.ly/yolov5-deci-platform)|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) +|在[Deci](https://bit.ly/yolov5-deci-platform)一键自动编译和量化YOLOv5以提高推理性能|使用[ClearML](https://cutt.ly/yolov5-readme-clearml) (开源!)自动追踪，可视化，以及远程训练YOLOv5|标记并将您的自定义数据直接导出到YOLOv5后，用[Roboflow](https://roboflow.com/?ref=ultralytics)进行训练 |通过[Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme)自动跟踪以及可视化你在云端所有的YOLOv5训练运行情况 ##
为什么选择 YOLOv5
@@ -209,7 +212,7 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 ### 预训练检查点 -| Model | size
(pixels) | mAPval
0.5:0.95 | mAPval
0.5 | Speed
CPU b1
(ms) | Speed
V100 b1
(ms) | Speed
V100 b32
(ms) | params
(M) | FLOPs
@640 (B) | +| 模型 | 规模
(像素) | mAP验证
0.5:0.95 | mAP验证
0.5 | 速度
CPU b1
(ms) | 速度
V100 b1
(ms) | 速度
V100 b32
(ms) | 参数
(M) | 浮点运算
@640 (B) | |------------------------------------------------------------------------------------------------------|-----------------------|-------------------------|--------------------|------------------------------|-------------------------------|--------------------------------|--------------------|------------------------| | [YOLOv5n](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5n.pt) | 640 | 28.0 | 45.7 | **45** | **6.3** | **0.6** | **1.9** | **4.5** | | [YOLOv5s](https://github.com/ultralytics/yolov5/releases/download/v6.1/yolov5s.pt) | 640 | 37.4 | 56.8 | 98 | 6.4 | 0.9 | 7.2 | 16.5 | @@ -237,18 +240,18 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12
-##
Classification ⭐ NEW
+##
分类 ⭐ 新
-YOLOv5 [release v6.2](https://github.com/ultralytics/yolov5/releases) brings support for classification model training, validation, prediction and export! We've made training classifier models super simple. Click below to get started. +YOLOv5发布的[v6.2版本](https://github.com/ultralytics/yolov5/releases) 支持训练，验证，预测和输出分类模型！这使得训练分类器模型非常简单。点击下面开始尝试！
- Classification Checkpoints (click to expand) + 分类检查点 (点击展开)
-We trained YOLOv5-cls classification models on ImageNet for 90 epochs using a 4xA100 instance, and we trained ResNet and EfficientNet models alongside with the same default training settings to compare. We exported all models to ONNX FP32 for CPU speed tests and to TensorRT FP16 for GPU speed tests. We ran all speed tests on Google [Colab Pro](https://colab.research.google.com/signup) for easy reproducibility. +我们在ImageNet上使用了4xA100的实例训练YOLOv5-cls分类模型90个epochs，并以相同的默认设置同时训练了ResNet和EfficientNet模型来进行比较。我们将所有的模型导出到ONNX FP32进行CPU速度测试，又导出到TensorRT FP16进行GPU速度测试。最后，为了方便重现，我们在[Google Colab Pro](https://colab.research.google.com/signup)上进行了所有的速度测试。 -| Model | size
(pixels) | acc
top1 | acc
top5 | Training
90 epochs
4xA100 (hours) | Speed
ONNX CPU
(ms) | Speed
TensorRT V100
(ms) | params
(M) | FLOPs
@224 (B) | +| 模型 | 规模
(像素) | 准确度
第一 | 准确度
前五 | 训练
90 epochs
4xA100 (小时) | 速度
ONNX CPU
(ms) | 速度
TensorRT V100
(ms) | 参数
(M) | 浮点运算
@224 (B) | |----------------------------------------------------------------------------------------------------|-----------------------|------------------|------------------|----------------------------------------------|--------------------------------|-------------------------------------|--------------------|------------------------| | [YOLOv5n-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5n-cls.pt) | 224 | 64.6 | 85.4 | 7:59 | **3.3** | **0.5** | **2.5** | **0.5** | | [YOLOv5s-cls](https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s-cls.pt) | 224 | 71.5 | 90.2 | 8:09 | 6.6 | 0.6 | 5.4 | 1.4 | @@ -267,38 +270,38 @@ We trained YOLOv5-cls classification models on ImageNet for 90 epochs using a 4x | [EfficientNet_b3](https://github.com/ultralytics/yolov5/releases/download/v6.2/efficientnet_b3.pt) | 224 | 77.7 | 94.0 | 19:19 | 18.9 | 1.9 | 12.2 | 2.4 |
- Table Notes (click to expand) + 表格注释 (点击扩展) -- All checkpoints are trained to 90 epochs with SGD optimizer with `lr0=0.001` and `weight_decay=5e-5` at image size 224 and all default settings.
Runs logged to https://wandb.ai/glenn-jocher/YOLOv5-Classifier-v6-2 -- **Accuracy** values are for single-model single-scale on [ImageNet-1k](https://www.image-net.org/index.php) dataset.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224` -- **Speed** averaged over 100 inference images using a Google [Colab Pro](https://colab.research.google.com/signup) V100 High-RAM instance.
Reproduce by `python classify/val.py --data ../datasets/imagenet --img 224 --batch 1` -- **Export** to ONNX at FP32 and TensorRT at FP16 done with `export.py`.
Reproduce by `python export.py --weights yolov5s-cls.pt --include engine onnx --imgsz 224` +- 所有检查点都被SGD优化器训练到90 epochs, `lr0=0.001` 和 `weight_decay=5e-5`， 图像大小为224，全为默认设置。
运行数据记录于 https://wandb.ai/glenn-jocher/YOLOv5-Classifier-v6-2。 +- **准确度** 值为[ImageNet-1k](https://www.image-net.org/index.php)数据集上的单模型单尺度。
通过`python classify/val.py --data ../datasets/imagenet --img 224`进行复制。 +- 使用Google [Colab Pro](https://colab.research.google.com/signup) V100 High-RAM实例得出的100张推理图像的平均**速度**。
通过 `python classify/val.py --data ../datasets/imagenet --img 224 --batch 1`进行复制。 +- 用`export.py`**导出**到FP32的ONNX和FP16的TensorRT。
通过 `python export.py --weights yolov5s-cls.pt --include engine onnx --imgsz 224`进行复制。
- Classification Usage Examples (click to expand) + 分类使用实例 (点击展开) -### Train -YOLOv5 classification training supports auto-download of MNIST, Fashion-MNIST, CIFAR10, CIFAR100, Imagenette, Imagewoof, and ImageNet datasets with the `--data` argument. To start training on MNIST for example use `--data mnist`. +### 训练 +YOLOv5分类训练支持自动下载MNIST, Fashion-MNIST, CIFAR10, CIFAR100, Imagenette, Imagewoof和ImageNet数据集，并使用`--data` 参数. 打个比方，在MNIST上使用`--data mnist`开始训练。 ```bash -# Single-GPU +# 单GPU python classify/train.py --model yolov5s-cls.pt --data cifar100 --epochs 5 --img 224 --batch 128 -# Multi-GPU DDP +# 多-GPU DDP python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/train.py --model yolov5s-cls.pt --data imagenet --epochs 5 --img 224 --device 0,1,2,3 ``` -### Val -Validate YOLOv5m-cls accuracy on ImageNet-1k dataset: +### 验证 +在ImageNet-1k数据集上验证YOLOv5m-cl的准确性: ```bash bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images) python classify/val.py --weights yolov5m-cls.pt --data ../datasets/imagenet --img 224 # validate ``` -### Predict -Use pretrained YOLOv5s-cls.pt to predict bus.jpg: +### 预测 +用提前训练好的YOLOv5s-cls.pt去预测bus.jpg: ```bash python classify/predict.py --weights yolov5s-cls.pt --data data/images/bus.jpg ``` @@ -306,8 +309,8 @@ python classify/predict.py --weights yolov5s-cls.pt --data data/images/bus.jpg model = torch.hub.load('ultralytics/yolov5', 'custom', 'yolov5s-cls.pt') # load from PyTorch Hub ``` -### Export -Export a group of trained YOLOv5s-cls, ResNet and EfficientNet models to ONNX and TensorRT: +### 导出 +导出一组训练好的YOLOv5s-cls, ResNet和EfficientNet模型到ONNX和TensorRT: ```bash python export.py --weights yolov5s-cls.pt resnet50.pt efficientnet_b0.pt --include onnx engine --img 224 ``` From e83b422a69bbd69628687b2dc50102c08877505c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 17 Aug 2022 17:52:53 +0200 Subject: [PATCH 469/661] Update dataset `names` from array to dictionary (#9000) * Migrate dataset names to dictionary * fix check * backwards compat * predict fix * val fix * Keep dataset stats behavior identical Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- classify/predict.py | 2 +- data/Argoverse.yaml | 11 +- data/GlobalWheat2020.yaml | 4 +- data/ImageNet.yaml | 1138 ++++++++++++++++++++++++++++++++----- data/Objects365.yaml | 408 +++++++++++-- data/SKU-110K.yaml | 4 +- data/VOC.yaml | 24 +- data/VisDrone.yaml | 13 +- data/coco.yaml | 91 ++- data/coco128.yaml | 91 ++- data/xView.yaml | 71 ++- models/common.py | 2 +- utils/dataloaders.py | 2 +- utils/general.py | 8 +- val.py | 4 +- 15 files changed, 1646 insertions(+), 227 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 4247e3c8e7fa..87379e42159b 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -71,7 +71,7 @@ def run( p = F.softmax(results, dim=1) # probabilities i = p.argsort(1, descending=True)[:, :5].squeeze() # top 5 indices dt[2] += time_sync() - t3 - LOGGER.info(f"image 1/1 {file}: {imgsz}x{imgsz} {', '.join(f'{model.names[j]} {p[0, j]:.2f}' for j in i)}") + LOGGER.info(f"image 1/1 {file}: {imgsz}x{imgsz} {', '.join(f'{model.names[j]} {p[0, j]:.2f}' for j in i.tolist())}") # Print results t = tuple(x / seen * 1E3 for x in dt) # speeds per image diff --git a/data/Argoverse.yaml b/data/Argoverse.yaml index 9d21296e3291..e3e9ba161ed0 100644 --- a/data/Argoverse.yaml +++ b/data/Argoverse.yaml @@ -14,8 +14,15 @@ val: Argoverse-1.1/images/val/ # val images (relative to 'path') 15062 images test: Argoverse-1.1/images/test/ # test images (optional) https://eval.ai/web/challenges/challenge-page/800/overview # Classes -nc: 8 # number of classes -names: ['person', 'bicycle', 'car', 'motorcycle', 'bus', 'truck', 'traffic_light', 'stop_sign'] # class names +names: + 0: person + 1: bicycle + 2: car + 3: motorcycle + 4: bus + 5: truck + 6: traffic_light + 7: stop_sign # Download script/URL (optional) --------------------------------------------------------------------------------------- diff --git a/data/GlobalWheat2020.yaml b/data/GlobalWheat2020.yaml index 4c43693f1d82..01812d031bc5 100644 --- a/data/GlobalWheat2020.yaml +++ b/data/GlobalWheat2020.yaml @@ -26,8 +26,8 @@ test: # test images (optional) 1276 images - images/uq_1 # Classes -nc: 1 # number of classes -names: ['wheat_head'] # class names +names: + 0: wheat_head # Download script/URL (optional) --------------------------------------------------------------------------------------- diff --git a/data/ImageNet.yaml b/data/ImageNet.yaml index 9f89b4268aff..14f12950605f 100644 --- a/data/ImageNet.yaml +++ b/data/ImageNet.yaml @@ -15,142 +15,1008 @@ val: val # val images (relative to 'path') 50000 images test: # test images (optional) # Classes -nc: 1000 # number of classes -names: ['tench', 'goldfish', 'great white shark', 'tiger shark', 'hammerhead shark', 'electric ray', 'stingray', 'cock', - 'hen', 'ostrich', 'brambling', 'goldfinch', 'house finch', 'junco', 'indigo bunting', 'American robin', - 'bulbul', 'jay', 'magpie', 'chickadee', 'American dipper', 'kite', 'bald eagle', 'vulture', 'great grey owl', - 'fire salamander', 'smooth newt', 'newt', 'spotted salamander', 'axolotl', 'American bullfrog', 'tree frog', - 'tailed frog', 'loggerhead sea turtle', 'leatherback sea turtle', 'mud turtle', 'terrapin', 'box turtle', - 'banded gecko', 'green iguana', 'Carolina anole', 'desert grassland whiptail lizard', 'agama', - 'frilled-necked lizard', 'alligator lizard', 'Gila monster', 'European green lizard', 'chameleon', - 'Komodo dragon', 'Nile crocodile', 'American alligator', 'triceratops', 'worm snake', 'ring-necked snake', - 'eastern hog-nosed snake', 'smooth green snake', 'kingsnake', 'garter snake', 'water snake', 'vine snake', - 'night snake', 'boa constrictor', 'African rock python', 'Indian cobra', 'green mamba', 'sea snake', - 'Saharan horned viper', 'eastern diamondback rattlesnake', 'sidewinder', 'trilobite', 'harvestman', 'scorpion', - 'yellow garden spider', 'barn spider', 'European garden spider', 'southern black widow', 'tarantula', - 'wolf spider', 'tick', 'centipede', 'black grouse', 'ptarmigan', 'ruffed grouse', 'prairie grouse', 'peacock', - 'quail', 'partridge', 'grey parrot', 'macaw', 'sulphur-crested cockatoo', 'lorikeet', 'coucal', 'bee eater', - 'hornbill', 'hummingbird', 'jacamar', 'toucan', 'duck', 'red-breasted merganser', 'goose', 'black swan', - 'tusker', 'echidna', 'platypus', 'wallaby', 'koala', 'wombat', 'jellyfish', 'sea anemone', 'brain coral', - 'flatworm', 'nematode', 'conch', 'snail', 'slug', 'sea slug', 'chiton', 'chambered nautilus', 'Dungeness crab', - 'rock crab', 'fiddler crab', 'red king crab', 'American lobster', 'spiny lobster', 'crayfish', 'hermit crab', - 'isopod', 'white stork', 'black stork', 'spoonbill', 'flamingo', 'little blue heron', 'great egret', 'bittern', - 'crane (bird)', 'limpkin', 'common gallinule', 'American coot', 'bustard', 'ruddy turnstone', 'dunlin', - 'common redshank', 'dowitcher', 'oystercatcher', 'pelican', 'king penguin', 'albatross', 'grey whale', - 'killer whale', 'dugong', 'sea lion', 'Chihuahua', 'Japanese Chin', 'Maltese', 'Pekingese', 'Shih Tzu', - 'King Charles Spaniel', 'Papillon', 'toy terrier', 'Rhodesian Ridgeback', 'Afghan Hound', 'Basset Hound', - 'Beagle', 'Bloodhound', 'Bluetick Coonhound', 'Black and Tan Coonhound', 'Treeing Walker Coonhound', - 'English foxhound', 'Redbone Coonhound', 'borzoi', 'Irish Wolfhound', 'Italian Greyhound', 'Whippet', - 'Ibizan Hound', 'Norwegian Elkhound', 'Otterhound', 'Saluki', 'Scottish Deerhound', 'Weimaraner', - 'Staffordshire Bull Terrier', 'American Staffordshire Terrier', 'Bedlington Terrier', 'Border Terrier', - 'Kerry Blue Terrier', 'Irish Terrier', 'Norfolk Terrier', 'Norwich Terrier', 'Yorkshire Terrier', - 'Wire Fox Terrier', 'Lakeland Terrier', 'Sealyham Terrier', 'Airedale Terrier', 'Cairn Terrier', - 'Australian Terrier', 'Dandie Dinmont Terrier', 'Boston Terrier', 'Miniature Schnauzer', 'Giant Schnauzer', - 'Standard Schnauzer', 'Scottish Terrier', 'Tibetan Terrier', 'Australian Silky Terrier', - 'Soft-coated Wheaten Terrier', 'West Highland White Terrier', 'Lhasa Apso', 'Flat-Coated Retriever', - 'Curly-coated Retriever', 'Golden Retriever', 'Labrador Retriever', 'Chesapeake Bay Retriever', - 'German Shorthaired Pointer', 'Vizsla', 'English Setter', 'Irish Setter', 'Gordon Setter', 'Brittany', - 'Clumber Spaniel', 'English Springer Spaniel', 'Welsh Springer Spaniel', 'Cocker Spaniels', 'Sussex Spaniel', - 'Irish Water Spaniel', 'Kuvasz', 'Schipperke', 'Groenendael', 'Malinois', 'Briard', 'Australian Kelpie', - 'Komondor', 'Old English Sheepdog', 'Shetland Sheepdog', 'collie', 'Border Collie', 'Bouvier des Flandres', - 'Rottweiler', 'German Shepherd Dog', 'Dobermann', 'Miniature Pinscher', 'Greater Swiss Mountain Dog', - 'Bernese Mountain Dog', 'Appenzeller Sennenhund', 'Entlebucher Sennenhund', 'Boxer', 'Bullmastiff', - 'Tibetan Mastiff', 'French Bulldog', 'Great Dane', 'St. Bernard', 'husky', 'Alaskan Malamute', 'Siberian Husky', - 'Dalmatian', 'Affenpinscher', 'Basenji', 'pug', 'Leonberger', 'Newfoundland', 'Pyrenean Mountain Dog', - 'Samoyed', 'Pomeranian', 'Chow Chow', 'Keeshond', 'Griffon Bruxellois', 'Pembroke Welsh Corgi', - 'Cardigan Welsh Corgi', 'Toy Poodle', 'Miniature Poodle', 'Standard Poodle', 'Mexican hairless dog', - 'grey wolf', 'Alaskan tundra wolf', 'red wolf', 'coyote', 'dingo', 'dhole', 'African wild dog', 'hyena', - 'red fox', 'kit fox', 'Arctic fox', 'grey fox', 'tabby cat', 'tiger cat', 'Persian cat', 'Siamese cat', - 'Egyptian Mau', 'cougar', 'lynx', 'leopard', 'snow leopard', 'jaguar', 'lion', 'tiger', 'cheetah', 'brown bear', - 'American black bear', 'polar bear', 'sloth bear', 'mongoose', 'meerkat', 'tiger beetle', 'ladybug', - 'ground beetle', 'longhorn beetle', 'leaf beetle', 'dung beetle', 'rhinoceros beetle', 'weevil', 'fly', 'bee', - 'ant', 'grasshopper', 'cricket', 'stick insect', 'cockroach', 'mantis', 'cicada', 'leafhopper', 'lacewing', - 'dragonfly', 'damselfly', 'red admiral', 'ringlet', 'monarch butterfly', 'small white', 'sulphur butterfly', - 'gossamer-winged butterfly', 'starfish', 'sea urchin', 'sea cucumber', 'cottontail rabbit', 'hare', - 'Angora rabbit', 'hamster', 'porcupine', 'fox squirrel', 'marmot', 'beaver', 'guinea pig', 'common sorrel', - 'zebra', 'pig', 'wild boar', 'warthog', 'hippopotamus', 'ox', 'water buffalo', 'bison', 'ram', 'bighorn sheep', - 'Alpine ibex', 'hartebeest', 'impala', 'gazelle', 'dromedary', 'llama', 'weasel', 'mink', 'European polecat', - 'black-footed ferret', 'otter', 'skunk', 'badger', 'armadillo', 'three-toed sloth', 'orangutan', 'gorilla', - 'chimpanzee', 'gibbon', 'siamang', 'guenon', 'patas monkey', 'baboon', 'macaque', 'langur', - 'black-and-white colobus', 'proboscis monkey', 'marmoset', 'white-headed capuchin', 'howler monkey', 'titi', - "Geoffroy's spider monkey", 'common squirrel monkey', 'ring-tailed lemur', 'indri', 'Asian elephant', - 'African bush elephant', 'red panda', 'giant panda', 'snoek', 'eel', 'coho salmon', 'rock beauty', 'clownfish', - 'sturgeon', 'garfish', 'lionfish', 'pufferfish', 'abacus', 'abaya', 'academic gown', 'accordion', - 'acoustic guitar', 'aircraft carrier', 'airliner', 'airship', 'altar', 'ambulance', 'amphibious vehicle', - 'analog clock', 'apiary', 'apron', 'waste container', 'assault rifle', 'backpack', 'bakery', 'balance beam', - 'balloon', 'ballpoint pen', 'Band-Aid', 'banjo', 'baluster', 'barbell', 'barber chair', 'barbershop', 'barn', - 'barometer', 'barrel', 'wheelbarrow', 'baseball', 'basketball', 'bassinet', 'bassoon', 'swimming cap', - 'bath towel', 'bathtub', 'station wagon', 'lighthouse', 'beaker', 'military cap', 'beer bottle', 'beer glass', - 'bell-cot', 'bib', 'tandem bicycle', 'bikini', 'ring binder', 'binoculars', 'birdhouse', 'boathouse', - 'bobsleigh', 'bolo tie', 'poke bonnet', 'bookcase', 'bookstore', 'bottle cap', 'bow', 'bow tie', 'brass', 'bra', - 'breakwater', 'breastplate', 'broom', 'bucket', 'buckle', 'bulletproof vest', 'high-speed train', - 'butcher shop', 'taxicab', 'cauldron', 'candle', 'cannon', 'canoe', 'can opener', 'cardigan', 'car mirror', - 'carousel', 'tool kit', 'carton', 'car wheel', 'automated teller machine', 'cassette', 'cassette player', - 'castle', 'catamaran', 'CD player', 'cello', 'mobile phone', 'chain', 'chain-link fence', 'chain mail', - 'chainsaw', 'chest', 'chiffonier', 'chime', 'china cabinet', 'Christmas stocking', 'church', 'movie theater', - 'cleaver', 'cliff dwelling', 'cloak', 'clogs', 'cocktail shaker', 'coffee mug', 'coffeemaker', 'coil', - 'combination lock', 'computer keyboard', 'confectionery store', 'container ship', 'convertible', 'corkscrew', - 'cornet', 'cowboy boot', 'cowboy hat', 'cradle', 'crane (machine)', 'crash helmet', 'crate', 'infant bed', - 'Crock Pot', 'croquet ball', 'crutch', 'cuirass', 'dam', 'desk', 'desktop computer', 'rotary dial telephone', - 'diaper', 'digital clock', 'digital watch', 'dining table', 'dishcloth', 'dishwasher', 'disc brake', 'dock', - 'dog sled', 'dome', 'doormat', 'drilling rig', 'drum', 'drumstick', 'dumbbell', 'Dutch oven', 'electric fan', - 'electric guitar', 'electric locomotive', 'entertainment center', 'envelope', 'espresso machine', 'face powder', - 'feather boa', 'filing cabinet', 'fireboat', 'fire engine', 'fire screen sheet', 'flagpole', 'flute', - 'folding chair', 'football helmet', 'forklift', 'fountain', 'fountain pen', 'four-poster bed', 'freight car', - 'French horn', 'frying pan', 'fur coat', 'garbage truck', 'gas mask', 'gas pump', 'goblet', 'go-kart', - 'golf ball', 'golf cart', 'gondola', 'gong', 'gown', 'grand piano', 'greenhouse', 'grille', 'grocery store', - 'guillotine', 'barrette', 'hair spray', 'half-track', 'hammer', 'hamper', 'hair dryer', 'hand-held computer', - 'handkerchief', 'hard disk drive', 'harmonica', 'harp', 'harvester', 'hatchet', 'holster', 'home theater', - 'honeycomb', 'hook', 'hoop skirt', 'horizontal bar', 'horse-drawn vehicle', 'hourglass', 'iPod', 'clothes iron', - "jack-o'-lantern", 'jeans', 'jeep', 'T-shirt', 'jigsaw puzzle', 'pulled rickshaw', 'joystick', 'kimono', - 'knee pad', 'knot', 'lab coat', 'ladle', 'lampshade', 'laptop computer', 'lawn mower', 'lens cap', - 'paper knife', 'library', 'lifeboat', 'lighter', 'limousine', 'ocean liner', 'lipstick', 'slip-on shoe', - 'lotion', 'speaker', 'loupe', 'sawmill', 'magnetic compass', 'mail bag', 'mailbox', 'tights', 'tank suit', - 'manhole cover', 'maraca', 'marimba', 'mask', 'match', 'maypole', 'maze', 'measuring cup', 'medicine chest', - 'megalith', 'microphone', 'microwave oven', 'military uniform', 'milk can', 'minibus', 'miniskirt', 'minivan', - 'missile', 'mitten', 'mixing bowl', 'mobile home', 'Model T', 'modem', 'monastery', 'monitor', 'moped', - 'mortar', 'square academic cap', 'mosque', 'mosquito net', 'scooter', 'mountain bike', 'tent', 'computer mouse', - 'mousetrap', 'moving van', 'muzzle', 'nail', 'neck brace', 'necklace', 'nipple', 'notebook computer', 'obelisk', - 'oboe', 'ocarina', 'odometer', 'oil filter', 'organ', 'oscilloscope', 'overskirt', 'bullock cart', - 'oxygen mask', 'packet', 'paddle', 'paddle wheel', 'padlock', 'paintbrush', 'pajamas', 'palace', 'pan flute', - 'paper towel', 'parachute', 'parallel bars', 'park bench', 'parking meter', 'passenger car', 'patio', - 'payphone', 'pedestal', 'pencil case', 'pencil sharpener', 'perfume', 'Petri dish', 'photocopier', 'plectrum', - 'Pickelhaube', 'picket fence', 'pickup truck', 'pier', 'piggy bank', 'pill bottle', 'pillow', 'ping-pong ball', - 'pinwheel', 'pirate ship', 'pitcher', 'hand plane', 'planetarium', 'plastic bag', 'plate rack', 'plow', - 'plunger', 'Polaroid camera', 'pole', 'police van', 'poncho', 'billiard table', 'soda bottle', 'pot', - "potter's wheel", 'power drill', 'prayer rug', 'printer', 'prison', 'projectile', 'projector', 'hockey puck', - 'punching bag', 'purse', 'quill', 'quilt', 'race car', 'racket', 'radiator', 'radio', 'radio telescope', - 'rain barrel', 'recreational vehicle', 'reel', 'reflex camera', 'refrigerator', 'remote control', 'restaurant', - 'revolver', 'rifle', 'rocking chair', 'rotisserie', 'eraser', 'rugby ball', 'ruler', 'running shoe', 'safe', - 'safety pin', 'salt shaker', 'sandal', 'sarong', 'saxophone', 'scabbard', 'weighing scale', 'school bus', - 'schooner', 'scoreboard', 'CRT screen', 'screw', 'screwdriver', 'seat belt', 'sewing machine', 'shield', - 'shoe store', 'shoji', 'shopping basket', 'shopping cart', 'shovel', 'shower cap', 'shower curtain', 'ski', - 'ski mask', 'sleeping bag', 'slide rule', 'sliding door', 'slot machine', 'snorkel', 'snowmobile', 'snowplow', - 'soap dispenser', 'soccer ball', 'sock', 'solar thermal collector', 'sombrero', 'soup bowl', 'space bar', - 'space heater', 'space shuttle', 'spatula', 'motorboat', 'spider web', 'spindle', 'sports car', 'spotlight', - 'stage', 'steam locomotive', 'through arch bridge', 'steel drum', 'stethoscope', 'scarf', 'stone wall', - 'stopwatch', 'stove', 'strainer', 'tram', 'stretcher', 'couch', 'stupa', 'submarine', 'suit', 'sundial', - 'sunglass', 'sunglasses', 'sunscreen', 'suspension bridge', 'mop', 'sweatshirt', 'swimsuit', 'swing', 'switch', - 'syringe', 'table lamp', 'tank', 'tape player', 'teapot', 'teddy bear', 'television', 'tennis ball', - 'thatched roof', 'front curtain', 'thimble', 'threshing machine', 'throne', 'tile roof', 'toaster', - 'tobacco shop', 'toilet seat', 'torch', 'totem pole', 'tow truck', 'toy store', 'tractor', 'semi-trailer truck', - 'tray', 'trench coat', 'tricycle', 'trimaran', 'tripod', 'triumphal arch', 'trolleybus', 'trombone', 'tub', - 'turnstile', 'typewriter keyboard', 'umbrella', 'unicycle', 'upright piano', 'vacuum cleaner', 'vase', 'vault', - 'velvet', 'vending machine', 'vestment', 'viaduct', 'violin', 'volleyball', 'waffle iron', 'wall clock', - 'wallet', 'wardrobe', 'military aircraft', 'sink', 'washing machine', 'water bottle', 'water jug', - 'water tower', 'whiskey jug', 'whistle', 'wig', 'window screen', 'window shade', 'Windsor tie', 'wine bottle', - 'wing', 'wok', 'wooden spoon', 'wool', 'split-rail fence', 'shipwreck', 'yawl', 'yurt', 'website', 'comic book', - 'crossword', 'traffic sign', 'traffic light', 'dust jacket', 'menu', 'plate', 'guacamole', 'consomme', - 'hot pot', 'trifle', 'ice cream', 'ice pop', 'baguette', 'bagel', 'pretzel', 'cheeseburger', 'hot dog', - 'mashed potato', 'cabbage', 'broccoli', 'cauliflower', 'zucchini', 'spaghetti squash', 'acorn squash', - 'butternut squash', 'cucumber', 'artichoke', 'bell pepper', 'cardoon', 'mushroom', 'Granny Smith', 'strawberry', - 'orange', 'lemon', 'fig', 'pineapple', 'banana', 'jackfruit', 'custard apple', 'pomegranate', 'hay', - 'carbonara', 'chocolate syrup', 'dough', 'meatloaf', 'pizza', 'pot pie', 'burrito', 'red wine', 'espresso', - 'cup', 'eggnog', 'alp', 'bubble', 'cliff', 'coral reef', 'geyser', 'lakeshore', 'promontory', 'shoal', - 'seashore', 'valley', 'volcano', 'baseball player', 'bridegroom', 'scuba diver', 'rapeseed', 'daisy', - "yellow lady's slipper", 'corn', 'acorn', 'rose hip', 'horse chestnut seed', 'coral fungus', 'agaric', - 'gyromitra', 'stinkhorn mushroom', 'earth star', 'hen-of-the-woods', 'bolete', 'ear', - 'toilet paper'] # class names +names: + 0: tench + 1: goldfish + 2: great white shark + 3: tiger shark + 4: hammerhead shark + 5: electric ray + 6: stingray + 7: cock + 8: hen + 9: ostrich + 10: brambling + 11: goldfinch + 12: house finch + 13: junco + 14: indigo bunting + 15: American robin + 16: bulbul + 17: jay + 18: magpie + 19: chickadee + 20: American dipper + 21: kite + 22: bald eagle + 23: vulture + 24: great grey owl + 25: fire salamander + 26: smooth newt + 27: newt + 28: spotted salamander + 29: axolotl + 30: American bullfrog + 31: tree frog + 32: tailed frog + 33: loggerhead sea turtle + 34: leatherback sea turtle + 35: mud turtle + 36: terrapin + 37: box turtle + 38: banded gecko + 39: green iguana + 40: Carolina anole + 41: desert grassland whiptail lizard + 42: agama + 43: frilled-necked lizard + 44: alligator lizard + 45: Gila monster + 46: European green lizard + 47: chameleon + 48: Komodo dragon + 49: Nile crocodile + 50: American alligator + 51: triceratops + 52: worm snake + 53: ring-necked snake + 54: eastern hog-nosed snake + 55: smooth green snake + 56: kingsnake + 57: garter snake + 58: water snake + 59: vine snake + 60: night snake + 61: boa constrictor + 62: African rock python + 63: Indian cobra + 64: green mamba + 65: sea snake + 66: Saharan horned viper + 67: eastern diamondback rattlesnake + 68: sidewinder + 69: trilobite + 70: harvestman + 71: scorpion + 72: yellow garden spider + 73: barn spider + 74: European garden spider + 75: southern black widow + 76: tarantula + 77: wolf spider + 78: tick + 79: centipede + 80: black grouse + 81: ptarmigan + 82: ruffed grouse + 83: prairie grouse + 84: peacock + 85: quail + 86: partridge + 87: grey parrot + 88: macaw + 89: sulphur-crested cockatoo + 90: lorikeet + 91: coucal + 92: bee eater + 93: hornbill + 94: hummingbird + 95: jacamar + 96: toucan + 97: duck + 98: red-breasted merganser + 99: goose + 100: black swan + 101: tusker + 102: echidna + 103: platypus + 104: wallaby + 105: koala + 106: wombat + 107: jellyfish + 108: sea anemone + 109: brain coral + 110: flatworm + 111: nematode + 112: conch + 113: snail + 114: slug + 115: sea slug + 116: chiton + 117: chambered nautilus + 118: Dungeness crab + 119: rock crab + 120: fiddler crab + 121: red king crab + 122: American lobster + 123: spiny lobster + 124: crayfish + 125: hermit crab + 126: isopod + 127: white stork + 128: black stork + 129: spoonbill + 130: flamingo + 131: little blue heron + 132: great egret + 133: bittern + 134: crane (bird) + 135: limpkin + 136: common gallinule + 137: American coot + 138: bustard + 139: ruddy turnstone + 140: dunlin + 141: common redshank + 142: dowitcher + 143: oystercatcher + 144: pelican + 145: king penguin + 146: albatross + 147: grey whale + 148: killer whale + 149: dugong + 150: sea lion + 151: Chihuahua + 152: Japanese Chin + 153: Maltese + 154: Pekingese + 155: Shih Tzu + 156: King Charles Spaniel + 157: Papillon + 158: toy terrier + 159: Rhodesian Ridgeback + 160: Afghan Hound + 161: Basset Hound + 162: Beagle + 163: Bloodhound + 164: Bluetick Coonhound + 165: Black and Tan Coonhound + 166: Treeing Walker Coonhound + 167: English foxhound + 168: Redbone Coonhound + 169: borzoi + 170: Irish Wolfhound + 171: Italian Greyhound + 172: Whippet + 173: Ibizan Hound + 174: Norwegian Elkhound + 175: Otterhound + 176: Saluki + 177: Scottish Deerhound + 178: Weimaraner + 179: Staffordshire Bull Terrier + 180: American Staffordshire Terrier + 181: Bedlington Terrier + 182: Border Terrier + 183: Kerry Blue Terrier + 184: Irish Terrier + 185: Norfolk Terrier + 186: Norwich Terrier + 187: Yorkshire Terrier + 188: Wire Fox Terrier + 189: Lakeland Terrier + 190: Sealyham Terrier + 191: Airedale Terrier + 192: Cairn Terrier + 193: Australian Terrier + 194: Dandie Dinmont Terrier + 195: Boston Terrier + 196: Miniature Schnauzer + 197: Giant Schnauzer + 198: Standard Schnauzer + 199: Scottish Terrier + 200: Tibetan Terrier + 201: Australian Silky Terrier + 202: Soft-coated Wheaten Terrier + 203: West Highland White Terrier + 204: Lhasa Apso + 205: Flat-Coated Retriever + 206: Curly-coated Retriever + 207: Golden Retriever + 208: Labrador Retriever + 209: Chesapeake Bay Retriever + 210: German Shorthaired Pointer + 211: Vizsla + 212: English Setter + 213: Irish Setter + 214: Gordon Setter + 215: Brittany + 216: Clumber Spaniel + 217: English Springer Spaniel + 218: Welsh Springer Spaniel + 219: Cocker Spaniels + 220: Sussex Spaniel + 221: Irish Water Spaniel + 222: Kuvasz + 223: Schipperke + 224: Groenendael + 225: Malinois + 226: Briard + 227: Australian Kelpie + 228: Komondor + 229: Old English Sheepdog + 230: Shetland Sheepdog + 231: collie + 232: Border Collie + 233: Bouvier des Flandres + 234: Rottweiler + 235: German Shepherd Dog + 236: Dobermann + 237: Miniature Pinscher + 238: Greater Swiss Mountain Dog + 239: Bernese Mountain Dog + 240: Appenzeller Sennenhund + 241: Entlebucher Sennenhund + 242: Boxer + 243: Bullmastiff + 244: Tibetan Mastiff + 245: French Bulldog + 246: Great Dane + 247: St. Bernard + 248: husky + 249: Alaskan Malamute + 250: Siberian Husky + 251: Dalmatian + 252: Affenpinscher + 253: Basenji + 254: pug + 255: Leonberger + 256: Newfoundland + 257: Pyrenean Mountain Dog + 258: Samoyed + 259: Pomeranian + 260: Chow Chow + 261: Keeshond + 262: Griffon Bruxellois + 263: Pembroke Welsh Corgi + 264: Cardigan Welsh Corgi + 265: Toy Poodle + 266: Miniature Poodle + 267: Standard Poodle + 268: Mexican hairless dog + 269: grey wolf + 270: Alaskan tundra wolf + 271: red wolf + 272: coyote + 273: dingo + 274: dhole + 275: African wild dog + 276: hyena + 277: red fox + 278: kit fox + 279: Arctic fox + 280: grey fox + 281: tabby cat + 282: tiger cat + 283: Persian cat + 284: Siamese cat + 285: Egyptian Mau + 286: cougar + 287: lynx + 288: leopard + 289: snow leopard + 290: jaguar + 291: lion + 292: tiger + 293: cheetah + 294: brown bear + 295: American black bear + 296: polar bear + 297: sloth bear + 298: mongoose + 299: meerkat + 300: tiger beetle + 301: ladybug + 302: ground beetle + 303: longhorn beetle + 304: leaf beetle + 305: dung beetle + 306: rhinoceros beetle + 307: weevil + 308: fly + 309: bee + 310: ant + 311: grasshopper + 312: cricket + 313: stick insect + 314: cockroach + 315: mantis + 316: cicada + 317: leafhopper + 318: lacewing + 319: dragonfly + 320: damselfly + 321: red admiral + 322: ringlet + 323: monarch butterfly + 324: small white + 325: sulphur butterfly + 326: gossamer-winged butterfly + 327: starfish + 328: sea urchin + 329: sea cucumber + 330: cottontail rabbit + 331: hare + 332: Angora rabbit + 333: hamster + 334: porcupine + 335: fox squirrel + 336: marmot + 337: beaver + 338: guinea pig + 339: common sorrel + 340: zebra + 341: pig + 342: wild boar + 343: warthog + 344: hippopotamus + 345: ox + 346: water buffalo + 347: bison + 348: ram + 349: bighorn sheep + 350: Alpine ibex + 351: hartebeest + 352: impala + 353: gazelle + 354: dromedary + 355: llama + 356: weasel + 357: mink + 358: European polecat + 359: black-footed ferret + 360: otter + 361: skunk + 362: badger + 363: armadillo + 364: three-toed sloth + 365: orangutan + 366: gorilla + 367: chimpanzee + 368: gibbon + 369: siamang + 370: guenon + 371: patas monkey + 372: baboon + 373: macaque + 374: langur + 375: black-and-white colobus + 376: proboscis monkey + 377: marmoset + 378: white-headed capuchin + 379: howler monkey + 380: titi + 381: Geoffroy's spider monkey + 382: common squirrel monkey + 383: ring-tailed lemur + 384: indri + 385: Asian elephant + 386: African bush elephant + 387: red panda + 388: giant panda + 389: snoek + 390: eel + 391: coho salmon + 392: rock beauty + 393: clownfish + 394: sturgeon + 395: garfish + 396: lionfish + 397: pufferfish + 398: abacus + 399: abaya + 400: academic gown + 401: accordion + 402: acoustic guitar + 403: aircraft carrier + 404: airliner + 405: airship + 406: altar + 407: ambulance + 408: amphibious vehicle + 409: analog clock + 410: apiary + 411: apron + 412: waste container + 413: assault rifle + 414: backpack + 415: bakery + 416: balance beam + 417: balloon + 418: ballpoint pen + 419: Band-Aid + 420: banjo + 421: baluster + 422: barbell + 423: barber chair + 424: barbershop + 425: barn + 426: barometer + 427: barrel + 428: wheelbarrow + 429: baseball + 430: basketball + 431: bassinet + 432: bassoon + 433: swimming cap + 434: bath towel + 435: bathtub + 436: station wagon + 437: lighthouse + 438: beaker + 439: military cap + 440: beer bottle + 441: beer glass + 442: bell-cot + 443: bib + 444: tandem bicycle + 445: bikini + 446: ring binder + 447: binoculars + 448: birdhouse + 449: boathouse + 450: bobsleigh + 451: bolo tie + 452: poke bonnet + 453: bookcase + 454: bookstore + 455: bottle cap + 456: bow + 457: bow tie + 458: brass + 459: bra + 460: breakwater + 461: breastplate + 462: broom + 463: bucket + 464: buckle + 465: bulletproof vest + 466: high-speed train + 467: butcher shop + 468: taxicab + 469: cauldron + 470: candle + 471: cannon + 472: canoe + 473: can opener + 474: cardigan + 475: car mirror + 476: carousel + 477: tool kit + 478: carton + 479: car wheel + 480: automated teller machine + 481: cassette + 482: cassette player + 483: castle + 484: catamaran + 485: CD player + 486: cello + 487: mobile phone + 488: chain + 489: chain-link fence + 490: chain mail + 491: chainsaw + 492: chest + 493: chiffonier + 494: chime + 495: china cabinet + 496: Christmas stocking + 497: church + 498: movie theater + 499: cleaver + 500: cliff dwelling + 501: cloak + 502: clogs + 503: cocktail shaker + 504: coffee mug + 505: coffeemaker + 506: coil + 507: combination lock + 508: computer keyboard + 509: confectionery store + 510: container ship + 511: convertible + 512: corkscrew + 513: cornet + 514: cowboy boot + 515: cowboy hat + 516: cradle + 517: crane (machine) + 518: crash helmet + 519: crate + 520: infant bed + 521: Crock Pot + 522: croquet ball + 523: crutch + 524: cuirass + 525: dam + 526: desk + 527: desktop computer + 528: rotary dial telephone + 529: diaper + 530: digital clock + 531: digital watch + 532: dining table + 533: dishcloth + 534: dishwasher + 535: disc brake + 536: dock + 537: dog sled + 538: dome + 539: doormat + 540: drilling rig + 541: drum + 542: drumstick + 543: dumbbell + 544: Dutch oven + 545: electric fan + 546: electric guitar + 547: electric locomotive + 548: entertainment center + 549: envelope + 550: espresso machine + 551: face powder + 552: feather boa + 553: filing cabinet + 554: fireboat + 555: fire engine + 556: fire screen sheet + 557: flagpole + 558: flute + 559: folding chair + 560: football helmet + 561: forklift + 562: fountain + 563: fountain pen + 564: four-poster bed + 565: freight car + 566: French horn + 567: frying pan + 568: fur coat + 569: garbage truck + 570: gas mask + 571: gas pump + 572: goblet + 573: go-kart + 574: golf ball + 575: golf cart + 576: gondola + 577: gong + 578: gown + 579: grand piano + 580: greenhouse + 581: grille + 582: grocery store + 583: guillotine + 584: barrette + 585: hair spray + 586: half-track + 587: hammer + 588: hamper + 589: hair dryer + 590: hand-held computer + 591: handkerchief + 592: hard disk drive + 593: harmonica + 594: harp + 595: harvester + 596: hatchet + 597: holster + 598: home theater + 599: honeycomb + 600: hook + 601: hoop skirt + 602: horizontal bar + 603: horse-drawn vehicle + 604: hourglass + 605: iPod + 606: clothes iron + 607: jack-o'-lantern + 608: jeans + 609: jeep + 610: T-shirt + 611: jigsaw puzzle + 612: pulled rickshaw + 613: joystick + 614: kimono + 615: knee pad + 616: knot + 617: lab coat + 618: ladle + 619: lampshade + 620: laptop computer + 621: lawn mower + 622: lens cap + 623: paper knife + 624: library + 625: lifeboat + 626: lighter + 627: limousine + 628: ocean liner + 629: lipstick + 630: slip-on shoe + 631: lotion + 632: speaker + 633: loupe + 634: sawmill + 635: magnetic compass + 636: mail bag + 637: mailbox + 638: tights + 639: tank suit + 640: manhole cover + 641: maraca + 642: marimba + 643: mask + 644: match + 645: maypole + 646: maze + 647: measuring cup + 648: medicine chest + 649: megalith + 650: microphone + 651: microwave oven + 652: military uniform + 653: milk can + 654: minibus + 655: miniskirt + 656: minivan + 657: missile + 658: mitten + 659: mixing bowl + 660: mobile home + 661: Model T + 662: modem + 663: monastery + 664: monitor + 665: moped + 666: mortar + 667: square academic cap + 668: mosque + 669: mosquito net + 670: scooter + 671: mountain bike + 672: tent + 673: computer mouse + 674: mousetrap + 675: moving van + 676: muzzle + 677: nail + 678: neck brace + 679: necklace + 680: nipple + 681: notebook computer + 682: obelisk + 683: oboe + 684: ocarina + 685: odometer + 686: oil filter + 687: organ + 688: oscilloscope + 689: overskirt + 690: bullock cart + 691: oxygen mask + 692: packet + 693: paddle + 694: paddle wheel + 695: padlock + 696: paintbrush + 697: pajamas + 698: palace + 699: pan flute + 700: paper towel + 701: parachute + 702: parallel bars + 703: park bench + 704: parking meter + 705: passenger car + 706: patio + 707: payphone + 708: pedestal + 709: pencil case + 710: pencil sharpener + 711: perfume + 712: Petri dish + 713: photocopier + 714: plectrum + 715: Pickelhaube + 716: picket fence + 717: pickup truck + 718: pier + 719: piggy bank + 720: pill bottle + 721: pillow + 722: ping-pong ball + 723: pinwheel + 724: pirate ship + 725: pitcher + 726: hand plane + 727: planetarium + 728: plastic bag + 729: plate rack + 730: plow + 731: plunger + 732: Polaroid camera + 733: pole + 734: police van + 735: poncho + 736: billiard table + 737: soda bottle + 738: pot + 739: potter's wheel + 740: power drill + 741: prayer rug + 742: printer + 743: prison + 744: projectile + 745: projector + 746: hockey puck + 747: punching bag + 748: purse + 749: quill + 750: quilt + 751: race car + 752: racket + 753: radiator + 754: radio + 755: radio telescope + 756: rain barrel + 757: recreational vehicle + 758: reel + 759: reflex camera + 760: refrigerator + 761: remote control + 762: restaurant + 763: revolver + 764: rifle + 765: rocking chair + 766: rotisserie + 767: eraser + 768: rugby ball + 769: ruler + 770: running shoe + 771: safe + 772: safety pin + 773: salt shaker + 774: sandal + 775: sarong + 776: saxophone + 777: scabbard + 778: weighing scale + 779: school bus + 780: schooner + 781: scoreboard + 782: CRT screen + 783: screw + 784: screwdriver + 785: seat belt + 786: sewing machine + 787: shield + 788: shoe store + 789: shoji + 790: shopping basket + 791: shopping cart + 792: shovel + 793: shower cap + 794: shower curtain + 795: ski + 796: ski mask + 797: sleeping bag + 798: slide rule + 799: sliding door + 800: slot machine + 801: snorkel + 802: snowmobile + 803: snowplow + 804: soap dispenser + 805: soccer ball + 806: sock + 807: solar thermal collector + 808: sombrero + 809: soup bowl + 810: space bar + 811: space heater + 812: space shuttle + 813: spatula + 814: motorboat + 815: spider web + 816: spindle + 817: sports car + 818: spotlight + 819: stage + 820: steam locomotive + 821: through arch bridge + 822: steel drum + 823: stethoscope + 824: scarf + 825: stone wall + 826: stopwatch + 827: stove + 828: strainer + 829: tram + 830: stretcher + 831: couch + 832: stupa + 833: submarine + 834: suit + 835: sundial + 836: sunglass + 837: sunglasses + 838: sunscreen + 839: suspension bridge + 840: mop + 841: sweatshirt + 842: swimsuit + 843: swing + 844: switch + 845: syringe + 846: table lamp + 847: tank + 848: tape player + 849: teapot + 850: teddy bear + 851: television + 852: tennis ball + 853: thatched roof + 854: front curtain + 855: thimble + 856: threshing machine + 857: throne + 858: tile roof + 859: toaster + 860: tobacco shop + 861: toilet seat + 862: torch + 863: totem pole + 864: tow truck + 865: toy store + 866: tractor + 867: semi-trailer truck + 868: tray + 869: trench coat + 870: tricycle + 871: trimaran + 872: tripod + 873: triumphal arch + 874: trolleybus + 875: trombone + 876: tub + 877: turnstile + 878: typewriter keyboard + 879: umbrella + 880: unicycle + 881: upright piano + 882: vacuum cleaner + 883: vase + 884: vault + 885: velvet + 886: vending machine + 887: vestment + 888: viaduct + 889: violin + 890: volleyball + 891: waffle iron + 892: wall clock + 893: wallet + 894: wardrobe + 895: military aircraft + 896: sink + 897: washing machine + 898: water bottle + 899: water jug + 900: water tower + 901: whiskey jug + 902: whistle + 903: wig + 904: window screen + 905: window shade + 906: Windsor tie + 907: wine bottle + 908: wing + 909: wok + 910: wooden spoon + 911: wool + 912: split-rail fence + 913: shipwreck + 914: yawl + 915: yurt + 916: website + 917: comic book + 918: crossword + 919: traffic sign + 920: traffic light + 921: dust jacket + 922: menu + 923: plate + 924: guacamole + 925: consomme + 926: hot pot + 927: trifle + 928: ice cream + 929: ice pop + 930: baguette + 931: bagel + 932: pretzel + 933: cheeseburger + 934: hot dog + 935: mashed potato + 936: cabbage + 937: broccoli + 938: cauliflower + 939: zucchini + 940: spaghetti squash + 941: acorn squash + 942: butternut squash + 943: cucumber + 944: artichoke + 945: bell pepper + 946: cardoon + 947: mushroom + 948: Granny Smith + 949: strawberry + 950: orange + 951: lemon + 952: fig + 953: pineapple + 954: banana + 955: jackfruit + 956: custard apple + 957: pomegranate + 958: hay + 959: carbonara + 960: chocolate syrup + 961: dough + 962: meatloaf + 963: pizza + 964: pot pie + 965: burrito + 966: red wine + 967: espresso + 968: cup + 969: eggnog + 970: alp + 971: bubble + 972: cliff + 973: coral reef + 974: geyser + 975: lakeshore + 976: promontory + 977: shoal + 978: seashore + 979: valley + 980: volcano + 981: baseball player + 982: bridegroom + 983: scuba diver + 984: rapeseed + 985: daisy + 986: yellow lady's slipper + 987: corn + 988: acorn + 989: rose hip + 990: horse chestnut seed + 991: coral fungus + 992: agaric + 993: gyromitra + 994: stinkhorn mushroom + 995: earth star + 996: hen-of-the-woods + 997: bolete + 998: ear + 999: toilet paper + # Download script/URL (optional) download: data/scripts/get_imagenet.sh diff --git a/data/Objects365.yaml b/data/Objects365.yaml index 4cc94753f530..05b26a1f4796 100644 --- a/data/Objects365.yaml +++ b/data/Objects365.yaml @@ -14,48 +14,372 @@ val: images/val # val images (relative to 'path') 80000 images test: # test images (optional) # Classes -nc: 365 # number of classes -names: ['Person', 'Sneakers', 'Chair', 'Other Shoes', 'Hat', 'Car', 'Lamp', 'Glasses', 'Bottle', 'Desk', 'Cup', - 'Street Lights', 'Cabinet/shelf', 'Handbag/Satchel', 'Bracelet', 'Plate', 'Picture/Frame', 'Helmet', 'Book', - 'Gloves', 'Storage box', 'Boat', 'Leather Shoes', 'Flower', 'Bench', 'Potted Plant', 'Bowl/Basin', 'Flag', - 'Pillow', 'Boots', 'Vase', 'Microphone', 'Necklace', 'Ring', 'SUV', 'Wine Glass', 'Belt', 'Monitor/TV', - 'Backpack', 'Umbrella', 'Traffic Light', 'Speaker', 'Watch', 'Tie', 'Trash bin Can', 'Slippers', 'Bicycle', - 'Stool', 'Barrel/bucket', 'Van', 'Couch', 'Sandals', 'Basket', 'Drum', 'Pen/Pencil', 'Bus', 'Wild Bird', - 'High Heels', 'Motorcycle', 'Guitar', 'Carpet', 'Cell Phone', 'Bread', 'Camera', 'Canned', 'Truck', - 'Traffic cone', 'Cymbal', 'Lifesaver', 'Towel', 'Stuffed Toy', 'Candle', 'Sailboat', 'Laptop', 'Awning', - 'Bed', 'Faucet', 'Tent', 'Horse', 'Mirror', 'Power outlet', 'Sink', 'Apple', 'Air Conditioner', 'Knife', - 'Hockey Stick', 'Paddle', 'Pickup Truck', 'Fork', 'Traffic Sign', 'Balloon', 'Tripod', 'Dog', 'Spoon', 'Clock', - 'Pot', 'Cow', 'Cake', 'Dinning Table', 'Sheep', 'Hanger', 'Blackboard/Whiteboard', 'Napkin', 'Other Fish', - 'Orange/Tangerine', 'Toiletry', 'Keyboard', 'Tomato', 'Lantern', 'Machinery Vehicle', 'Fan', - 'Green Vegetables', 'Banana', 'Baseball Glove', 'Airplane', 'Mouse', 'Train', 'Pumpkin', 'Soccer', 'Skiboard', - 'Luggage', 'Nightstand', 'Tea pot', 'Telephone', 'Trolley', 'Head Phone', 'Sports Car', 'Stop Sign', - 'Dessert', 'Scooter', 'Stroller', 'Crane', 'Remote', 'Refrigerator', 'Oven', 'Lemon', 'Duck', 'Baseball Bat', - 'Surveillance Camera', 'Cat', 'Jug', 'Broccoli', 'Piano', 'Pizza', 'Elephant', 'Skateboard', 'Surfboard', - 'Gun', 'Skating and Skiing shoes', 'Gas stove', 'Donut', 'Bow Tie', 'Carrot', 'Toilet', 'Kite', 'Strawberry', - 'Other Balls', 'Shovel', 'Pepper', 'Computer Box', 'Toilet Paper', 'Cleaning Products', 'Chopsticks', - 'Microwave', 'Pigeon', 'Baseball', 'Cutting/chopping Board', 'Coffee Table', 'Side Table', 'Scissors', - 'Marker', 'Pie', 'Ladder', 'Snowboard', 'Cookies', 'Radiator', 'Fire Hydrant', 'Basketball', 'Zebra', 'Grape', - 'Giraffe', 'Potato', 'Sausage', 'Tricycle', 'Violin', 'Egg', 'Fire Extinguisher', 'Candy', 'Fire Truck', - 'Billiards', 'Converter', 'Bathtub', 'Wheelchair', 'Golf Club', 'Briefcase', 'Cucumber', 'Cigar/Cigarette', - 'Paint Brush', 'Pear', 'Heavy Truck', 'Hamburger', 'Extractor', 'Extension Cord', 'Tong', 'Tennis Racket', - 'Folder', 'American Football', 'earphone', 'Mask', 'Kettle', 'Tennis', 'Ship', 'Swing', 'Coffee Machine', - 'Slide', 'Carriage', 'Onion', 'Green beans', 'Projector', 'Frisbee', 'Washing Machine/Drying Machine', - 'Chicken', 'Printer', 'Watermelon', 'Saxophone', 'Tissue', 'Toothbrush', 'Ice cream', 'Hot-air balloon', - 'Cello', 'French Fries', 'Scale', 'Trophy', 'Cabbage', 'Hot dog', 'Blender', 'Peach', 'Rice', 'Wallet/Purse', - 'Volleyball', 'Deer', 'Goose', 'Tape', 'Tablet', 'Cosmetics', 'Trumpet', 'Pineapple', 'Golf Ball', - 'Ambulance', 'Parking meter', 'Mango', 'Key', 'Hurdle', 'Fishing Rod', 'Medal', 'Flute', 'Brush', 'Penguin', - 'Megaphone', 'Corn', 'Lettuce', 'Garlic', 'Swan', 'Helicopter', 'Green Onion', 'Sandwich', 'Nuts', - 'Speed Limit Sign', 'Induction Cooker', 'Broom', 'Trombone', 'Plum', 'Rickshaw', 'Goldfish', 'Kiwi fruit', - 'Router/modem', 'Poker Card', 'Toaster', 'Shrimp', 'Sushi', 'Cheese', 'Notepaper', 'Cherry', 'Pliers', 'CD', - 'Pasta', 'Hammer', 'Cue', 'Avocado', 'Hamimelon', 'Flask', 'Mushroom', 'Screwdriver', 'Soap', 'Recorder', - 'Bear', 'Eggplant', 'Board Eraser', 'Coconut', 'Tape Measure/Ruler', 'Pig', 'Showerhead', 'Globe', 'Chips', - 'Steak', 'Crosswalk Sign', 'Stapler', 'Camel', 'Formula 1', 'Pomegranate', 'Dishwasher', 'Crab', - 'Hoverboard', 'Meat ball', 'Rice Cooker', 'Tuba', 'Calculator', 'Papaya', 'Antelope', 'Parrot', 'Seal', - 'Butterfly', 'Dumbbell', 'Donkey', 'Lion', 'Urinal', 'Dolphin', 'Electric Drill', 'Hair Dryer', 'Egg tart', - 'Jellyfish', 'Treadmill', 'Lighter', 'Grapefruit', 'Game board', 'Mop', 'Radish', 'Baozi', 'Target', 'French', - 'Spring Rolls', 'Monkey', 'Rabbit', 'Pencil Case', 'Yak', 'Red Cabbage', 'Binoculars', 'Asparagus', 'Barbell', - 'Scallop', 'Noddles', 'Comb', 'Dumpling', 'Oyster', 'Table Tennis paddle', 'Cosmetics Brush/Eyeliner Pencil', - 'Chainsaw', 'Eraser', 'Lobster', 'Durian', 'Okra', 'Lipstick', 'Cosmetics Mirror', 'Curling', 'Table Tennis'] +names: + 0: Person + 1: Sneakers + 2: Chair + 3: Other Shoes + 4: Hat + 5: Car + 6: Lamp + 7: Glasses + 8: Bottle + 9: Desk + 10: Cup + 11: Street Lights + 12: Cabinet/shelf + 13: Handbag/Satchel + 14: Bracelet + 15: Plate + 16: Picture/Frame + 17: Helmet + 18: Book + 19: Gloves + 20: Storage box + 21: Boat + 22: Leather Shoes + 23: Flower + 24: Bench + 25: Potted Plant + 26: Bowl/Basin + 27: Flag + 28: Pillow + 29: Boots + 30: Vase + 31: Microphone + 32: Necklace + 33: Ring + 34: SUV + 35: Wine Glass + 36: Belt + 37: Monitor/TV + 38: Backpack + 39: Umbrella + 40: Traffic Light + 41: Speaker + 42: Watch + 43: Tie + 44: Trash bin Can + 45: Slippers + 46: Bicycle + 47: Stool + 48: Barrel/bucket + 49: Van + 50: Couch + 51: Sandals + 52: Basket + 53: Drum + 54: Pen/Pencil + 55: Bus + 56: Wild Bird + 57: High Heels + 58: Motorcycle + 59: Guitar + 60: Carpet + 61: Cell Phone + 62: Bread + 63: Camera + 64: Canned + 65: Truck + 66: Traffic cone + 67: Cymbal + 68: Lifesaver + 69: Towel + 70: Stuffed Toy + 71: Candle + 72: Sailboat + 73: Laptop + 74: Awning + 75: Bed + 76: Faucet + 77: Tent + 78: Horse + 79: Mirror + 80: Power outlet + 81: Sink + 82: Apple + 83: Air Conditioner + 84: Knife + 85: Hockey Stick + 86: Paddle + 87: Pickup Truck + 88: Fork + 89: Traffic Sign + 90: Balloon + 91: Tripod + 92: Dog + 93: Spoon + 94: Clock + 95: Pot + 96: Cow + 97: Cake + 98: Dinning Table + 99: Sheep + 100: Hanger + 101: Blackboard/Whiteboard + 102: Napkin + 103: Other Fish + 104: Orange/Tangerine + 105: Toiletry + 106: Keyboard + 107: Tomato + 108: Lantern + 109: Machinery Vehicle + 110: Fan + 111: Green Vegetables + 112: Banana + 113: Baseball Glove + 114: Airplane + 115: Mouse + 116: Train + 117: Pumpkin + 118: Soccer + 119: Skiboard + 120: Luggage + 121: Nightstand + 122: Tea pot + 123: Telephone + 124: Trolley + 125: Head Phone + 126: Sports Car + 127: Stop Sign + 128: Dessert + 129: Scooter + 130: Stroller + 131: Crane + 132: Remote + 133: Refrigerator + 134: Oven + 135: Lemon + 136: Duck + 137: Baseball Bat + 138: Surveillance Camera + 139: Cat + 140: Jug + 141: Broccoli + 142: Piano + 143: Pizza + 144: Elephant + 145: Skateboard + 146: Surfboard + 147: Gun + 148: Skating and Skiing shoes + 149: Gas stove + 150: Donut + 151: Bow Tie + 152: Carrot + 153: Toilet + 154: Kite + 155: Strawberry + 156: Other Balls + 157: Shovel + 158: Pepper + 159: Computer Box + 160: Toilet Paper + 161: Cleaning Products + 162: Chopsticks + 163: Microwave + 164: Pigeon + 165: Baseball + 166: Cutting/chopping Board + 167: Coffee Table + 168: Side Table + 169: Scissors + 170: Marker + 171: Pie + 172: Ladder + 173: Snowboard + 174: Cookies + 175: Radiator + 176: Fire Hydrant + 177: Basketball + 178: Zebra + 179: Grape + 180: Giraffe + 181: Potato + 182: Sausage + 183: Tricycle + 184: Violin + 185: Egg + 186: Fire Extinguisher + 187: Candy + 188: Fire Truck + 189: Billiards + 190: Converter + 191: Bathtub + 192: Wheelchair + 193: Golf Club + 194: Briefcase + 195: Cucumber + 196: Cigar/Cigarette + 197: Paint Brush + 198: Pear + 199: Heavy Truck + 200: Hamburger + 201: Extractor + 202: Extension Cord + 203: Tong + 204: Tennis Racket + 205: Folder + 206: American Football + 207: earphone + 208: Mask + 209: Kettle + 210: Tennis + 211: Ship + 212: Swing + 213: Coffee Machine + 214: Slide + 215: Carriage + 216: Onion + 217: Green beans + 218: Projector + 219: Frisbee + 220: Washing Machine/Drying Machine + 221: Chicken + 222: Printer + 223: Watermelon + 224: Saxophone + 225: Tissue + 226: Toothbrush + 227: Ice cream + 228: Hot-air balloon + 229: Cello + 230: French Fries + 231: Scale + 232: Trophy + 233: Cabbage + 234: Hot dog + 235: Blender + 236: Peach + 237: Rice + 238: Wallet/Purse + 239: Volleyball + 240: Deer + 241: Goose + 242: Tape + 243: Tablet + 244: Cosmetics + 245: Trumpet + 246: Pineapple + 247: Golf Ball + 248: Ambulance + 249: Parking meter + 250: Mango + 251: Key + 252: Hurdle + 253: Fishing Rod + 254: Medal + 255: Flute + 256: Brush + 257: Penguin + 258: Megaphone + 259: Corn + 260: Lettuce + 261: Garlic + 262: Swan + 263: Helicopter + 264: Green Onion + 265: Sandwich + 266: Nuts + 267: Speed Limit Sign + 268: Induction Cooker + 269: Broom + 270: Trombone + 271: Plum + 272: Rickshaw + 273: Goldfish + 274: Kiwi fruit + 275: Router/modem + 276: Poker Card + 277: Toaster + 278: Shrimp + 279: Sushi + 280: Cheese + 281: Notepaper + 282: Cherry + 283: Pliers + 284: CD + 285: Pasta + 286: Hammer + 287: Cue + 288: Avocado + 289: Hamimelon + 290: Flask + 291: Mushroom + 292: Screwdriver + 293: Soap + 294: Recorder + 295: Bear + 296: Eggplant + 297: Board Eraser + 298: Coconut + 299: Tape Measure/Ruler + 300: Pig + 301: Showerhead + 302: Globe + 303: Chips + 304: Steak + 305: Crosswalk Sign + 306: Stapler + 307: Camel + 308: Formula 1 + 309: Pomegranate + 310: Dishwasher + 311: Crab + 312: Hoverboard + 313: Meat ball + 314: Rice Cooker + 315: Tuba + 316: Calculator + 317: Papaya + 318: Antelope + 319: Parrot + 320: Seal + 321: Butterfly + 322: Dumbbell + 323: Donkey + 324: Lion + 325: Urinal + 326: Dolphin + 327: Electric Drill + 328: Hair Dryer + 329: Egg tart + 330: Jellyfish + 331: Treadmill + 332: Lighter + 333: Grapefruit + 334: Game board + 335: Mop + 336: Radish + 337: Baozi + 338: Target + 339: French + 340: Spring Rolls + 341: Monkey + 342: Rabbit + 343: Pencil Case + 344: Yak + 345: Red Cabbage + 346: Binoculars + 347: Asparagus + 348: Barbell + 349: Scallop + 350: Noddles + 351: Comb + 352: Dumpling + 353: Oyster + 354: Table Tennis paddle + 355: Cosmetics Brush/Eyeliner Pencil + 356: Chainsaw + 357: Eraser + 358: Lobster + 359: Durian + 360: Okra + 361: Lipstick + 362: Cosmetics Mirror + 363: Curling + 364: Table Tennis # Download script/URL (optional) --------------------------------------------------------------------------------------- diff --git a/data/SKU-110K.yaml b/data/SKU-110K.yaml index 2acf34d155bd..edae7171c660 100644 --- a/data/SKU-110K.yaml +++ b/data/SKU-110K.yaml @@ -14,8 +14,8 @@ val: val.txt # val images (relative to 'path') 588 images test: test.txt # test images (optional) 2936 images # Classes -nc: 1 # number of classes -names: ['object'] # class names +names: + 0: object # Download script/URL (optional) --------------------------------------------------------------------------------------- diff --git a/data/VOC.yaml b/data/VOC.yaml index 636ddc42d46c..bbe5cf90a838 100644 --- a/data/VOC.yaml +++ b/data/VOC.yaml @@ -20,9 +20,27 @@ test: # test images (optional) - images/test2007 # Classes -nc: 20 # number of classes -names: ['aeroplane', 'bicycle', 'bird', 'boat', 'bottle', 'bus', 'car', 'cat', 'chair', 'cow', 'diningtable', 'dog', - 'horse', 'motorbike', 'person', 'pottedplant', 'sheep', 'sofa', 'train', 'tvmonitor'] # class names +names: + 0: aeroplane + 1: bicycle + 2: bird + 3: boat + 4: bottle + 5: bus + 6: car + 7: cat + 8: chair + 9: cow + 10: diningtable + 11: dog + 12: horse + 13: motorbike + 14: person + 15: pottedplant + 16: sheep + 17: sofa + 18: train + 19: tvmonitor # Download script/URL (optional) --------------------------------------------------------------------------------------- diff --git a/data/VisDrone.yaml b/data/VisDrone.yaml index 10337b46f104..a8bcf8e628ec 100644 --- a/data/VisDrone.yaml +++ b/data/VisDrone.yaml @@ -14,8 +14,17 @@ val: VisDrone2019-DET-val/images # val images (relative to 'path') 548 images test: VisDrone2019-DET-test-dev/images # test images (optional) 1610 images # Classes -nc: 10 # number of classes -names: ['pedestrian', 'people', 'bicycle', 'car', 'van', 'truck', 'tricycle', 'awning-tricycle', 'bus', 'motor'] +names: + 0: pedestrian + 1: people + 2: bicycle + 3: car + 4: van + 5: truck + 6: tricycle + 7: awning-tricycle + 8: bus + 9: motor # Download script/URL (optional) --------------------------------------------------------------------------------------- diff --git a/data/coco.yaml b/data/coco.yaml index 0c0c4adab05d..d64dfc7fed76 100644 --- a/data/coco.yaml +++ b/data/coco.yaml @@ -14,16 +14,87 @@ val: val2017.txt # val images (relative to 'path') 5000 images test: test-dev2017.txt # 20288 of 40670 images, submit to https://competitions.codalab.org/competitions/20794 # Classes -nc: 80 # number of classes -names: ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light', - 'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow', - 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee', - 'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard', - 'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple', - 'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch', - 'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', - 'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear', - 'hair drier', 'toothbrush'] # class names +names: + 0: person + 1: bicycle + 2: car + 3: motorcycle + 4: airplane + 5: bus + 6: train + 7: truck + 8: boat + 9: traffic light + 10: fire hydrant + 11: stop sign + 12: parking meter + 13: bench + 14: bird + 15: cat + 16: dog + 17: horse + 18: sheep + 19: cow + 20: elephant + 21: bear + 22: zebra + 23: giraffe + 24: backpack + 25: umbrella + 26: handbag + 27: tie + 28: suitcase + 29: frisbee + 30: skis + 31: snowboard + 32: sports ball + 33: kite + 34: baseball bat + 35: baseball glove + 36: skateboard + 37: surfboard + 38: tennis racket + 39: bottle + 40: wine glass + 41: cup + 42: fork + 43: knife + 44: spoon + 45: bowl + 46: banana + 47: apple + 48: sandwich + 49: orange + 50: broccoli + 51: carrot + 52: hot dog + 53: pizza + 54: donut + 55: cake + 56: chair + 57: couch + 58: potted plant + 59: bed + 60: dining table + 61: toilet + 62: tv + 63: laptop + 64: mouse + 65: remote + 66: keyboard + 67: cell phone + 68: microwave + 69: oven + 70: toaster + 71: sink + 72: refrigerator + 73: book + 74: clock + 75: vase + 76: scissors + 77: teddy bear + 78: hair drier + 79: toothbrush # Download script/URL (optional) diff --git a/data/coco128.yaml b/data/coco128.yaml index 2517d2079257..12556736a571 100644 --- a/data/coco128.yaml +++ b/data/coco128.yaml @@ -14,16 +14,87 @@ val: images/train2017 # val images (relative to 'path') 128 images test: # test images (optional) # Classes -nc: 80 # number of classes -names: ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light', - 'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow', - 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee', - 'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard', - 'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple', - 'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch', - 'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', - 'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear', - 'hair drier', 'toothbrush'] # class names +names: + 0: person + 1: bicycle + 2: car + 3: motorcycle + 4: airplane + 5: bus + 6: train + 7: truck + 8: boat + 9: traffic light + 10: fire hydrant + 11: stop sign + 12: parking meter + 13: bench + 14: bird + 15: cat + 16: dog + 17: horse + 18: sheep + 19: cow + 20: elephant + 21: bear + 22: zebra + 23: giraffe + 24: backpack + 25: umbrella + 26: handbag + 27: tie + 28: suitcase + 29: frisbee + 30: skis + 31: snowboard + 32: sports ball + 33: kite + 34: baseball bat + 35: baseball glove + 36: skateboard + 37: surfboard + 38: tennis racket + 39: bottle + 40: wine glass + 41: cup + 42: fork + 43: knife + 44: spoon + 45: bowl + 46: banana + 47: apple + 48: sandwich + 49: orange + 50: broccoli + 51: carrot + 52: hot dog + 53: pizza + 54: donut + 55: cake + 56: chair + 57: couch + 58: potted plant + 59: bed + 60: dining table + 61: toilet + 62: tv + 63: laptop + 64: mouse + 65: remote + 66: keyboard + 67: cell phone + 68: microwave + 69: oven + 70: toaster + 71: sink + 72: refrigerator + 73: book + 74: clock + 75: vase + 76: scissors + 77: teddy bear + 78: hair drier + 79: toothbrush # Download script/URL (optional) diff --git a/data/xView.yaml b/data/xView.yaml index 3b38f1ff4439..b134ceac8164 100644 --- a/data/xView.yaml +++ b/data/xView.yaml @@ -14,16 +14,67 @@ train: images/autosplit_train.txt # train images (relative to 'path') 90% of 84 val: images/autosplit_val.txt # train images (relative to 'path') 10% of 847 train images # Classes -nc: 60 # number of classes -names: ['Fixed-wing Aircraft', 'Small Aircraft', 'Cargo Plane', 'Helicopter', 'Passenger Vehicle', 'Small Car', 'Bus', - 'Pickup Truck', 'Utility Truck', 'Truck', 'Cargo Truck', 'Truck w/Box', 'Truck Tractor', 'Trailer', - 'Truck w/Flatbed', 'Truck w/Liquid', 'Crane Truck', 'Railway Vehicle', 'Passenger Car', 'Cargo Car', - 'Flat Car', 'Tank car', 'Locomotive', 'Maritime Vessel', 'Motorboat', 'Sailboat', 'Tugboat', 'Barge', - 'Fishing Vessel', 'Ferry', 'Yacht', 'Container Ship', 'Oil Tanker', 'Engineering Vehicle', 'Tower crane', - 'Container Crane', 'Reach Stacker', 'Straddle Carrier', 'Mobile Crane', 'Dump Truck', 'Haul Truck', - 'Scraper/Tractor', 'Front loader/Bulldozer', 'Excavator', 'Cement Mixer', 'Ground Grader', 'Hut/Tent', 'Shed', - 'Building', 'Aircraft Hangar', 'Damaged Building', 'Facility', 'Construction Site', 'Vehicle Lot', 'Helipad', - 'Storage Tank', 'Shipping container lot', 'Shipping Container', 'Pylon', 'Tower'] # class names +names: + 0: Fixed-wing Aircraft + 1: Small Aircraft + 2: Cargo Plane + 3: Helicopter + 4: Passenger Vehicle + 5: Small Car + 6: Bus + 7: Pickup Truck + 8: Utility Truck + 9: Truck + 10: Cargo Truck + 11: Truck w/Box + 12: Truck Tractor + 13: Trailer + 14: Truck w/Flatbed + 15: Truck w/Liquid + 16: Crane Truck + 17: Railway Vehicle + 18: Passenger Car + 19: Cargo Car + 20: Flat Car + 21: Tank car + 22: Locomotive + 23: Maritime Vessel + 24: Motorboat + 25: Sailboat + 26: Tugboat + 27: Barge + 28: Fishing Vessel + 29: Ferry + 30: Yacht + 31: Container Ship + 32: Oil Tanker + 33: Engineering Vehicle + 34: Tower crane + 35: Container Crane + 36: Reach Stacker + 37: Straddle Carrier + 38: Mobile Crane + 39: Dump Truck + 40: Haul Truck + 41: Scraper/Tractor + 42: Front loader/Bulldozer + 43: Excavator + 44: Cement Mixer + 45: Ground Grader + 46: Hut/Tent + 47: Shed + 48: Building + 49: Aircraft Hangar + 50: Damaged Building + 51: Facility + 52: Construction Site + 53: Vehicle Lot + 54: Helipad + 55: Storage Tank + 56: Shipping container lot + 57: Shipping Container + 58: Pylon + 59: Tower # Download script/URL (optional) --------------------------------------------------------------------------------------- diff --git a/models/common.py b/models/common.py index 17e40e60d7d7..30202ca1abd7 100644 --- a/models/common.py +++ b/models/common.py @@ -449,7 +449,7 @@ def wrap_frozen_graph(gd, inputs, outputs): # class names if 'names' not in locals(): - names = yaml_load(data)['names'] if data else [f'class{i}' for i in range(999)] + names = yaml_load(data)['names'] if data else {i: f'class{i}' for i in range(999)} if names[0] == 'n01440764' and len(names) == 1000: # ImageNet names = yaml_load(ROOT / 'data/ImageNet.yaml')['names'] # human-readable names diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 2c04040bf25d..33e84ce4056e 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -1004,7 +1004,7 @@ def __init__(self, path='coco128.yaml', autodownload=False): self.hub_dir = Path(data['path'] + '-hub') self.im_dir = self.hub_dir / 'images' self.im_dir.mkdir(parents=True, exist_ok=True) # makes /images - self.stats = {'nc': data['nc'], 'names': data['names']} # statistics dictionary + self.stats = {'nc': data['nc'], 'names': list(data['names'].values())} # statistics dictionary self.data = data @staticmethod diff --git a/utils/general.py b/utils/general.py index 1c525c45f649..76872b696d79 100755 --- a/utils/general.py +++ b/utils/general.py @@ -481,11 +481,11 @@ def check_dataset(data, autodownload=True): data = yaml.safe_load(f) # dictionary # Checks - for k in 'train', 'val', 'nc': + for k in 'train', 'val', 'names': assert k in data, f"data.yaml '{k}:' field missing ❌" - if 'names' not in data: - LOGGER.warning("data.yaml 'names:' field missing ⚠️, assigning default names 'class0', 'class1', etc.") - data['names'] = [f'class{i}' for i in range(data['nc'])] # default names + if isinstance(data['names'], (list, tuple)): # old array format + data['names'] = dict(enumerate(data['names'])) # convert to dict + data['nc'] = len(data['names']) # Resolve paths path = Path(extract_dir or data.get('path') or '') # optional 'path' default to '.' diff --git a/val.py b/val.py index 130496233467..ce743b506aff 100644 --- a/val.py +++ b/val.py @@ -182,7 +182,9 @@ def run( seen = 0 confusion_matrix = ConfusionMatrix(nc=nc) - names = dict(enumerate(model.names if hasattr(model, 'names') else model.module.names)) + names = model.names if hasattr(model, 'names') else model.module.names # get class names + if isinstance(names, (list, tuple)): # old format + names = dict(enumerate(names)) class_map = coco80_to_coco91_class() if is_coco else list(range(1000)) s = ('%20s' + '%11s' * 6) % ('Class', 'Images', 'Labels', 'P', 'R', 'mAP@.5', 'mAP@.5:.95') dt, p, r, f1, mp, mr, map50, map = [0.0, 0.0, 0.0], 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 From 64e0757edffc6b2e927e16c8e2aa26439aceb4ce Mon Sep 17 00:00:00 2001 From: Ayush Chaurasia Date: Thu, 18 Aug 2022 02:11:43 +0530 Subject: [PATCH 470/661] [Classify]: Allow inference on dirs and videos (#9003) * allow image dirs * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update predict.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update dataloaders.py * Update predict.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update predict.py * Update predict.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- classify/predict.py | 64 +++++++++++++++++++++++--------------------- utils/dataloaders.py | 25 ++++++++--------- 2 files changed, 46 insertions(+), 43 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 87379e42159b..7af5f60a2b9d 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -1,6 +1,6 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ -Run classification inference on images +Run classification inference on file/dir/URL/glob Usage: $ python classify/predict.py --weights yolov5s-cls.pt --source data/images/bus.jpg @@ -11,7 +11,6 @@ import sys from pathlib import Path -import cv2 import torch.nn.functional as F FILE = Path(__file__).resolve() @@ -20,27 +19,31 @@ sys.path.append(str(ROOT)) # add ROOT to PATH ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative -from classify.train import imshow_cls from models.common import DetectMultiBackend from utils.augmentations import classify_transforms -from utils.general import LOGGER, check_requirements, colorstr, increment_path, print_args +from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages +from utils.general import LOGGER, check_file, check_requirements, colorstr, increment_path, print_args from utils.torch_utils import select_device, smart_inference_mode, time_sync @smart_inference_mode() def run( weights=ROOT / 'yolov5s-cls.pt', # model.pt path(s) - source=ROOT / 'data/images/bus.jpg', # file/dir/URL/glob, 0 for webcam + source=ROOT / 'data/images', # file/dir/URL/glob imgsz=224, # inference size device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu half=False, # use FP16 half-precision inference dnn=False, # use OpenCV DNN for ONNX inference - show=True, project=ROOT / 'runs/predict-cls', # save to project/name name='exp', # save to project/name exist_ok=False, # existing project/name ok, do not increment ): - file = str(source) + source = str(source) + is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS) + is_url = source.lower().startswith(('rtsp://', 'rtmp://', 'http://', 'https://')) + if is_url and is_file: + source = check_file(source) # download + seen, dt = 1, [0.0, 0.0, 0.0] device = select_device(device) @@ -48,37 +51,36 @@ def run( save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run save_dir.mkdir(parents=True, exist_ok=True) # make dir - # Transforms - transforms = classify_transforms(imgsz) - # Load model model = DetectMultiBackend(weights, device=device, dnn=dnn, fp16=half) model.warmup(imgsz=(1, 3, imgsz, imgsz)) # warmup - - # Image - t1 = time_sync() - im = cv2.cvtColor(cv2.imread(file), cv2.COLOR_BGR2RGB) - im = transforms(im).unsqueeze(0).to(device) - im = im.half() if model.fp16 else im.float() - t2 = time_sync() - dt[0] += t2 - t1 - - # Inference - results = model(im) - t3 = time_sync() - dt[1] += t3 - t2 - - p = F.softmax(results, dim=1) # probabilities - i = p.argsort(1, descending=True)[:, :5].squeeze() # top 5 indices - dt[2] += time_sync() - t3 - LOGGER.info(f"image 1/1 {file}: {imgsz}x{imgsz} {', '.join(f'{model.names[j]} {p[0, j]:.2f}' for j in i.tolist())}") + dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz)) + for path, im, im0s, vid_cap, s in dataset: + # Image + t1 = time_sync() + im = im.unsqueeze(0).to(device) + im = im.half() if model.fp16 else im.float() + t2 = time_sync() + dt[0] += t2 - t1 + + # Inference + results = model(im) + t3 = time_sync() + dt[1] += t3 - t2 + + # Post-process + p = F.softmax(results, dim=1) # probabilities + i = p.argsort(1, descending=True)[:, :5].squeeze().tolist() # top 5 indices + dt[2] += time_sync() - t3 + # if save: + # imshow_cls(im, f=save_dir / Path(path).name, verbose=True) + seen += 1 + LOGGER.info(f"{s}{imgsz}x{imgsz} {', '.join(f'{model.names[j]} {p[0, j]:.2f}' for j in i)}") # Print results t = tuple(x / seen * 1E3 for x in dt) # speeds per image shape = (1, 3, imgsz, imgsz) LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms post-process per image at shape {shape}' % t) - if show: - imshow_cls(im, f=save_dir / Path(file).name, verbose=True) LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") return p @@ -86,7 +88,7 @@ def run( def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s-cls.pt', help='model path(s)') - parser.add_argument('--source', type=str, default=ROOT / 'data/images/bus.jpg', help='file') + parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob') parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=224, help='train, val image size (pixels)') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 33e84ce4056e..3f26be2cd32d 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -186,7 +186,7 @@ def __iter__(self): class LoadImages: # YOLOv5 image/video dataloader, i.e. `python detect.py --source image.jpg/vid.mp4` - def __init__(self, path, img_size=640, stride=32, auto=True): + def __init__(self, path, img_size=640, stride=32, auto=True, transforms=None): files = [] for p in sorted(path) if isinstance(path, (list, tuple)) else [path]: p = str(Path(p).resolve()) @@ -210,6 +210,7 @@ def __init__(self, path, img_size=640, stride=32, auto=True): self.video_flag = [False] * ni + [True] * nv self.mode = 'image' self.auto = auto + self.transforms = transforms # optional if any(videos): self.new_video(videos[0]) # new video else: @@ -229,7 +230,7 @@ def __next__(self): if self.video_flag[self.count]: # Read video self.mode = 'video' - ret_val, img0 = self.cap.read() + ret_val, im0 = self.cap.read() while not ret_val: self.count += 1 self.cap.release() @@ -237,7 +238,7 @@ def __next__(self): raise StopIteration path = self.files[self.count] self.new_video(path) - ret_val, img0 = self.cap.read() + ret_val, im0 = self.cap.read() self.frame += 1 s = f'video {self.count + 1}/{self.nf} ({self.frame}/{self.frames}) {path}: ' @@ -245,18 +246,18 @@ def __next__(self): else: # Read image self.count += 1 - img0 = cv2.imread(path) # BGR - assert img0 is not None, f'Image Not Found {path}' + im0 = cv2.imread(path) # BGR + assert im0 is not None, f'Image Not Found {path}' s = f'image {self.count}/{self.nf} {path}: ' - # Padded resize - img = letterbox(img0, self.img_size, stride=self.stride, auto=self.auto)[0] - - # Convert - img = img.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB - img = np.ascontiguousarray(img) + if self.transforms: + im = self.transforms(cv2.cvtColor(im0, cv2.COLOR_BGR2RGB)) # classify transforms + else: + im = letterbox(im0, self.img_size, stride=self.stride, auto=self.auto)[0] # padded resize + im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB + im = np.ascontiguousarray(im) # contiguous - return path, img, img0, self.cap, s + return path, im, im0, self.cap, s def new_video(self, path): self.frame = 0 From 0922bc2082d8c754bbd733d90bd1ccd2aea79ee9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 17 Aug 2022 22:50:08 +0200 Subject: [PATCH 471/661] DockerHub tag update Usage example (#9005) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/docker/Dockerfile | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index 2280f209e6a1..cf2c1c5cb3cb 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -49,11 +49,8 @@ ENV OMP_NUM_THREADS=8 # Kill all image-based # sudo docker kill $(sudo docker ps -qa --filter ancestor=ultralytics/yolov5:latest) -# Bash into running container -# sudo docker exec -it 5a9b5863d93d bash - -# Bash into stopped container -# id=$(sudo docker ps -qa) && sudo docker start $id && sudo docker exec -it $id bash +# DockerHub tag update +# t=ultralytics/yolov5:latest tnew=ultralytics/yolov5:v6.2 && sudo docker pull $t && sudo docker tag $t $tnew && sudo docker push $tnew # Clean up # docker system prune -a --volumes From 6728dad76df8d62ed3c08e39c224a773d20582a0 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 17 Aug 2022 22:57:55 +0200 Subject: [PATCH 472/661] Add weight `decay` to argparser (#9006) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- classify/train.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/classify/train.py b/classify/train.py index b85f14236039..d55dc066d7a3 100644 --- a/classify/train.py +++ b/classify/train.py @@ -136,7 +136,7 @@ def train(opt, device): logger.log_graph(model, imgsz) # log model # Optimizer - optimizer = smart_optimizer(model, opt.optimizer, opt.lr0, momentum=0.9, decay=5e-5) + optimizer = smart_optimizer(model, opt.optimizer, opt.lr0, momentum=0.9, decay=opt.decay) # Scheduler lrf = 0.01 # final lr (fraction of lr0) @@ -280,6 +280,7 @@ def parse_opt(known=False): parser.add_argument('--pretrained', nargs='?', const=True, default=True, help='start from i.e. --pretrained False') parser.add_argument('--optimizer', choices=['SGD', 'Adam', 'AdamW', 'RMSProp'], default='Adam', help='optimizer') parser.add_argument('--lr0', type=float, default=0.001, help='initial learning rate') + parser.add_argument('--decay', type=float, default=5e-5, help='weight decay') parser.add_argument('--label-smoothing', type=float, default=0.1, help='Label smoothing epsilon') parser.add_argument('--cutoff', type=int, default=None, help='Model layer cutoff index for Classify() head') parser.add_argument('--dropout', type=float, default=None, help='Dropout (fraction)') From e08d568d39a8b1c24ec7eb54da80cf3b22f64f07 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 18 Aug 2022 01:08:52 +0200 Subject: [PATCH 473/661] Add glob quotes to detect.py usage example (#9007) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- detect.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/detect.py b/detect.py index c699a749a09f..dd60b87ca33a 100644 --- a/detect.py +++ b/detect.py @@ -7,7 +7,7 @@ img.jpg # image vid.mp4 # video path/ # directory - path/*.jpg # glob + 'path/*.jpg' # glob 'https://youtu.be/Zgi9g1ksQHc' # YouTube 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP stream From 5c854fab5e43df82ebfd51197c2dc58e5212c5a6 Mon Sep 17 00:00:00 2001 From: glennjocher Date: Thu, 18 Aug 2022 02:44:50 +0200 Subject: [PATCH 474/661] requires grad after reset params --- classify/train.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/classify/train.py b/classify/train.py index d55dc066d7a3..9fb7c52b545a 100644 --- a/classify/train.py +++ b/classify/train.py @@ -114,13 +114,13 @@ def train(opt, device): LOGGER.warning("WARNING: pass YOLOv5 classifier model with '-cls' suffix, i.e. '--model yolov5s-cls.pt'") model = ClassificationModel(model=model, nc=nc, cutoff=opt.cutoff or 10) # convert to classification model reshape_classifier_output(model, nc) # update class count - for p in model.parameters(): - p.requires_grad = True # for training for m in model.modules(): if not pretrained and hasattr(m, 'reset_parameters'): m.reset_parameters() if isinstance(m, torch.nn.Dropout) and opt.dropout is not None: m.p = opt.dropout # set dropout + for p in model.parameters(): + p.requires_grad = True # for training model = model.to(device) names = trainloader.dataset.classes # class names model.names = names # attach class names From 529aafd737053264cf8676b29c37f5d5300460eb Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 18 Aug 2022 11:50:24 +0200 Subject: [PATCH 475/661] Fix TorchScript JSON string key bug (#9015) * Fix TorchScript JSON string key bug Resolves https://github.com/ultralytics/yolov5/issues/9011 Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- models/common.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/models/common.py b/models/common.py index 30202ca1abd7..4f93887c55e0 100644 --- a/models/common.py +++ b/models/common.py @@ -337,8 +337,10 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, extra_files = {'config.txt': ''} # model metadata model = torch.jit.load(w, _extra_files=extra_files) model.half() if fp16 else model.float() - if extra_files['config.txt']: - d = json.loads(extra_files['config.txt']) # extra_files dict + if extra_files['config.txt']: # load metadata dict + d = json.loads(extra_files['config.txt'], + object_hook=lambda d: {int(k) if k.isdigit() else k: v + for k, v in d.items()}) stride, names = int(d['stride']), d['names'] elif dnn: # ONNX OpenCV DNN LOGGER.info(f'Loading {w} for ONNX OpenCV DNN inference...') From 20049be2e7dc6f330e3620dd82761bc3f4d02e36 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 18 Aug 2022 14:06:15 +0200 Subject: [PATCH 476/661] EMA FP32 assert classification bug fix (#9016) * Return EMA float on classification val * verbose val fix * EMA check --- classify/val.py | 3 ++- export.py | 2 +- models/experimental.py | 10 +++++++--- train.py | 3 +-- utils/torch_utils.py | 7 +++---- 5 files changed, 14 insertions(+), 11 deletions(-) diff --git a/classify/val.py b/classify/val.py index 9d965d9f1fdc..b76fb5147ecd 100644 --- a/classify/val.py +++ b/classify/val.py @@ -116,7 +116,7 @@ def run( if verbose: # all classes LOGGER.info(f"{'Class':>24}{'Images':>12}{'top1_acc':>12}{'top5_acc':>12}") LOGGER.info(f"{'all':>24}{targets.shape[0]:>12}{top1:>12.3g}{top5:>12.3g}") - for i, c in enumerate(model.names): + for i, c in model.names.items(): aci = acc[targets == i] top1i, top5i = aci.mean(0).tolist() LOGGER.info(f"{c:>24}{aci.shape[0]:>12}{top1i:>12.3g}{top5i:>12.3g}") @@ -127,6 +127,7 @@ def run( LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms post-process per image at shape {shape}' % t) LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") + model.float() # for training return top1, top5, loss diff --git a/export.py b/export.py index 595039b24bce..7b398fdc4d93 100644 --- a/export.py +++ b/export.py @@ -599,7 +599,7 @@ def parse_opt(): parser.add_argument('--conf-thres', type=float, default=0.25, help='TF.js NMS: confidence threshold') parser.add_argument('--include', nargs='+', - default=['torchscript', 'onnx'], + default=['torchscript'], help='torchscript, onnx, openvino, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs') opt = parser.parse_args() print_args(vars(opt)) diff --git a/models/experimental.py b/models/experimental.py index cb32d01ba46a..02d35b9ebd11 100644 --- a/models/experimental.py +++ b/models/experimental.py @@ -8,7 +8,6 @@ import torch import torch.nn as nn -from models.common import Conv from utils.downloads import attempt_download @@ -79,11 +78,16 @@ def attempt_load(weights, device=None, inplace=True, fuse=True): for w in weights if isinstance(weights, list) else [weights]: ckpt = torch.load(attempt_download(w), map_location='cpu') # load ckpt = (ckpt.get('ema') or ckpt['model']).to(device).float() # FP32 model + + # Model compatibility updates if not hasattr(ckpt, 'stride'): - ckpt.stride = torch.tensor([32.]) # compatibility update for ResNet etc. + ckpt.stride = torch.tensor([32.]) + if hasattr(ckpt, 'names') and isinstance(ckpt.names, (list, tuple)): + ckpt.names = dict(enumerate(ckpt.names)) # convert to dict + model.append(ckpt.fuse().eval() if fuse and hasattr(ckpt, 'fuse') else ckpt.eval()) # model in eval mode - # Compatibility updates + # Module compatibility updates for m in model.modules(): t = type(m) if t in (nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6, nn.SiLU, Detect, Model): diff --git a/train.py b/train.py index bbb26cdeafeb..10a3bdb56002 100644 --- a/train.py +++ b/train.py @@ -107,8 +107,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio data_dict = data_dict or check_dataset(data) # check if None train_path, val_path = data_dict['train'], data_dict['val'] nc = 1 if single_cls else int(data_dict['nc']) # number of classes - names = ['item'] if single_cls and len(data_dict['names']) != 1 else data_dict['names'] # class names - assert len(names) == nc, f'{len(names)} names found for nc={nc} dataset in {data}' # check + names = {0: 'item'} if single_cls and len(data_dict['names']) != 1 else data_dict['names'] # class names is_coco = isinstance(val_path, str) and val_path.endswith('coco/val2017.txt') # COCO dataset # Model diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 1cdbe20f8670..ed56064ce02e 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -408,8 +408,6 @@ class ModelEMA: def __init__(self, model, decay=0.9999, tau=2000, updates=0): # Create EMA self.ema = deepcopy(de_parallel(model)).eval() # FP32 EMA - # if next(model.parameters()).device.type != 'cpu': - # self.ema.half() # FP16 EMA self.updates = updates # number of EMA updates self.decay = lambda x: decay * (1 - math.exp(-x / tau)) # decay exponential ramp (to help early epochs) for p in self.ema.parameters(): @@ -423,9 +421,10 @@ def update(self, model): msd = de_parallel(model).state_dict() # model state_dict for k, v in self.ema.state_dict().items(): - if v.dtype.is_floating_point: + if v.dtype.is_floating_point: # true for FP16 and FP32 v *= d - v += (1 - d) * msd[k].detach() + v += (1 - d) * msd[k] + assert v.dtype == msd[k].dtype == torch.float32, f'EMA {v.dtype} and model {msd[k]} must be updated in FP32' def update_attr(self, model, include=(), exclude=('process_group', 'reducer')): # Update EMA attributes From c0e7a776cd55e8c01b63714f6f7fea3d53f6bf5b Mon Sep 17 00:00:00 2001 From: cher-liang <88578531+cher-liang@users.noreply.github.com> Date: Thu, 18 Aug 2022 20:18:02 +0800 Subject: [PATCH 477/661] Faster pre-processing for gray image input (#9009) * faster 1 channel to 3 channels image conversion * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- models/common.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/common.py b/models/common.py index 4f93887c55e0..f914c9d60fdb 100644 --- a/models/common.py +++ b/models/common.py @@ -617,7 +617,7 @@ def forward(self, imgs, size=640, augment=False, profile=False): files.append(Path(f).with_suffix('.jpg').name) if im.shape[0] < 5: # image in CHW im = im.transpose((1, 2, 0)) # reverse dataloader .transpose(2, 0, 1) - im = im[..., :3] if im.ndim == 3 else np.tile(im[..., None], 3) # enforce 3ch input + im = im[..., :3] if im.ndim == 3 else cv2.cvtColor(im, cv2.COLOR_GRAY2BGR) # enforce 3ch input s = im.shape[:2] # HWC shape0.append(s) # image shape g = (size / max(s)) # gain From d40cd0d454dcc34312cb5c40f45f64b76665c40c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 18 Aug 2022 19:55:38 +0200 Subject: [PATCH 478/661] Improved `Profile()` inference timing (#9024) * Improved `Profile()` class * Update predict.py * Update val.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update val.py * Update AutoShape Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- classify/predict.py | 37 +++++++------- classify/val.py | 29 +++++------ detect.py | 35 ++++++------- models/common.py | 117 ++++++++++++++++++++++---------------------- utils/general.py | 18 +++++-- val.py | 31 ++++++------ 6 files changed, 133 insertions(+), 134 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 7af5f60a2b9d..0bf99140b8e3 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -22,8 +22,8 @@ from models.common import DetectMultiBackend from utils.augmentations import classify_transforms from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages -from utils.general import LOGGER, check_file, check_requirements, colorstr, increment_path, print_args -from utils.torch_utils import select_device, smart_inference_mode, time_sync +from utils.general import LOGGER, Profile, check_file, check_requirements, colorstr, increment_path, print_args +from utils.torch_utils import select_device, smart_inference_mode @smart_inference_mode() @@ -44,7 +44,7 @@ def run( if is_url and is_file: source = check_file(source) # download - seen, dt = 1, [0.0, 0.0, 0.0] + dt = Profile(), Profile(), Profile() device = select_device(device) # Directories @@ -55,30 +55,27 @@ def run( model = DetectMultiBackend(weights, device=device, dnn=dnn, fp16=half) model.warmup(imgsz=(1, 3, imgsz, imgsz)) # warmup dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz)) - for path, im, im0s, vid_cap, s in dataset: + for seen, (path, im, im0s, vid_cap, s) in enumerate(dataset): # Image - t1 = time_sync() - im = im.unsqueeze(0).to(device) - im = im.half() if model.fp16 else im.float() - t2 = time_sync() - dt[0] += t2 - t1 + with dt[0]: + im = im.unsqueeze(0).to(device) + im = im.half() if model.fp16 else im.float() # Inference - results = model(im) - t3 = time_sync() - dt[1] += t3 - t2 + with dt[1]: + results = model(im) # Post-process - p = F.softmax(results, dim=1) # probabilities - i = p.argsort(1, descending=True)[:, :5].squeeze().tolist() # top 5 indices - dt[2] += time_sync() - t3 - # if save: - # imshow_cls(im, f=save_dir / Path(path).name, verbose=True) - seen += 1 - LOGGER.info(f"{s}{imgsz}x{imgsz} {', '.join(f'{model.names[j]} {p[0, j]:.2f}' for j in i)}") + with dt[2]: + p = F.softmax(results, dim=1) # probabilities + i = p.argsort(1, descending=True)[:, :5].squeeze().tolist() # top 5 indices + # if save: + # imshow_cls(im, f=save_dir / Path(path).name, verbose=True) + LOGGER.info( + f"{s}{imgsz}x{imgsz} {', '.join(f'{model.names[j]} {p[0, j]:.2f}' for j in i)}, {dt[1].dt * 1E3:.1f}ms") # Print results - t = tuple(x / seen * 1E3 for x in dt) # speeds per image + t = tuple(x.t / (seen + 1) * 1E3 for x in dt) # speeds per image shape = (1, 3, imgsz, imgsz) LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms post-process per image at shape {shape}' % t) LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") diff --git a/classify/val.py b/classify/val.py index b76fb5147ecd..c91e2cf82c81 100644 --- a/classify/val.py +++ b/classify/val.py @@ -23,8 +23,8 @@ from models.common import DetectMultiBackend from utils.dataloaders import create_classification_dataloader -from utils.general import LOGGER, check_img_size, check_requirements, colorstr, increment_path, print_args -from utils.torch_utils import select_device, smart_inference_mode, time_sync +from utils.general import LOGGER, Profile, check_img_size, check_requirements, colorstr, increment_path, print_args +from utils.torch_utils import select_device, smart_inference_mode @smart_inference_mode() @@ -83,27 +83,24 @@ def run( workers=workers) model.eval() - pred, targets, loss, dt = [], [], 0, [0.0, 0.0, 0.0] + pred, targets, loss, dt = [], [], 0, (Profile(), Profile(), Profile()) n = len(dataloader) # number of batches action = 'validating' if dataloader.dataset.root.stem == 'val' else 'testing' desc = f"{pbar.desc[:-36]}{action:>36}" if pbar else f"{action}" bar = tqdm(dataloader, desc, n, not training, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}', position=0) with torch.cuda.amp.autocast(enabled=device.type != 'cpu'): for images, labels in bar: - t1 = time_sync() - images, labels = images.to(device, non_blocking=True), labels.to(device) - t2 = time_sync() - dt[0] += t2 - t1 + with dt[0]: + images, labels = images.to(device, non_blocking=True), labels.to(device) - y = model(images) - t3 = time_sync() - dt[1] += t3 - t2 + with dt[1]: + y = model(images) - pred.append(y.argsort(1, descending=True)[:, :5]) - targets.append(labels) - if criterion: - loss += criterion(y, labels) - dt[2] += time_sync() - t3 + with dt[2]: + pred.append(y.argsort(1, descending=True)[:, :5]) + targets.append(labels) + if criterion: + loss += criterion(y, labels) loss /= n pred, targets = torch.cat(pred), torch.cat(targets) @@ -122,7 +119,7 @@ def run( LOGGER.info(f"{c:>24}{aci.shape[0]:>12}{top1i:>12.3g}{top5i:>12.3g}") # Print results - t = tuple(x / len(dataloader.dataset.samples) * 1E3 for x in dt) # speeds per image + t = tuple(x.t / len(dataloader.dataset.samples) * 1E3 for x in dt) # speeds per image shape = (1, 3, imgsz, imgsz) LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms post-process per image at shape {shape}' % t) LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") diff --git a/detect.py b/detect.py index dd60b87ca33a..93ae0baccd13 100644 --- a/detect.py +++ b/detect.py @@ -41,10 +41,10 @@ from models.common import DetectMultiBackend from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadStreams -from utils.general import (LOGGER, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, +from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, increment_path, non_max_suppression, print_args, scale_coords, strip_optimizer, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box -from utils.torch_utils import select_device, smart_inference_mode, time_sync +from utils.torch_utils import select_device, smart_inference_mode @smart_inference_mode() @@ -107,26 +107,23 @@ def run( # Run inference model.warmup(imgsz=(1 if pt else bs, 3, *imgsz)) # warmup - seen, windows, dt = 0, [], [0.0, 0.0, 0.0] + seen, windows, dt = 0, [], (Profile(), Profile(), Profile()) for path, im, im0s, vid_cap, s in dataset: - t1 = time_sync() - im = torch.from_numpy(im).to(device) - im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 - im /= 255 # 0 - 255 to 0.0 - 1.0 - if len(im.shape) == 3: - im = im[None] # expand for batch dim - t2 = time_sync() - dt[0] += t2 - t1 + with dt[0]: + im = torch.from_numpy(im).to(device) + im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 + im /= 255 # 0 - 255 to 0.0 - 1.0 + if len(im.shape) == 3: + im = im[None] # expand for batch dim # Inference - visualize = increment_path(save_dir / Path(path).stem, mkdir=True) if visualize else False - pred = model(im, augment=augment, visualize=visualize) - t3 = time_sync() - dt[1] += t3 - t2 + with dt[1]: + visualize = increment_path(save_dir / Path(path).stem, mkdir=True) if visualize else False + pred = model(im, augment=augment, visualize=visualize) # NMS - pred = non_max_suppression(pred, conf_thres, iou_thres, classes, agnostic_nms, max_det=max_det) - dt[2] += time_sync() - t3 + with dt[2]: + pred = non_max_suppression(pred, conf_thres, iou_thres, classes, agnostic_nms, max_det=max_det) # Second-stage classifier (optional) # pred = utils.general.apply_classifier(pred, classifier_model, im, im0s) @@ -201,10 +198,10 @@ def run( vid_writer[i].write(im0) # Print time (inference-only) - LOGGER.info(f'{s}Done. ({t3 - t2:.3f}s)') + LOGGER.info(f"{s}{'' if len(det) else '(no detections), '}{dt[1].dt * 1E3:.1f}ms") # Print results - t = tuple(x / seen * 1E3 for x in dt) # speeds per image + t = tuple(x.t / seen * 1E3 for x in dt) # speeds per image LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {(1, 3, *imgsz)}' % t) if save_txt or save_img: s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else '' diff --git a/models/common.py b/models/common.py index f914c9d60fdb..33aa2ac12465 100644 --- a/models/common.py +++ b/models/common.py @@ -21,10 +21,11 @@ from torch.cuda import amp from utils.dataloaders import exif_transpose, letterbox -from utils.general import (LOGGER, ROOT, check_requirements, check_suffix, check_version, colorstr, increment_path, - make_divisible, non_max_suppression, scale_coords, xywh2xyxy, xyxy2xywh, yaml_load) +from utils.general import (LOGGER, ROOT, Profile, check_requirements, check_suffix, check_version, colorstr, + increment_path, make_divisible, non_max_suppression, scale_coords, xywh2xyxy, xyxy2xywh, + yaml_load) from utils.plots import Annotator, colors, save_one_box -from utils.torch_utils import copy_attr, smart_inference_mode, time_sync +from utils.torch_utils import copy_attr, smart_inference_mode def autopad(k, p=None): # kernel, padding @@ -587,9 +588,9 @@ def _apply(self, fn): return self @smart_inference_mode() - def forward(self, imgs, size=640, augment=False, profile=False): + def forward(self, ims, size=640, augment=False, profile=False): # Inference from various sources. For height=640, width=1280, RGB images example inputs are: - # file: imgs = 'data/images/zidane.jpg' # str or PosixPath + # file: ims = 'data/images/zidane.jpg' # str or PosixPath # URI: = 'https://ultralytics.com/images/zidane.jpg' # OpenCV: = cv2.imread('image.jpg')[:,:,::-1] # HWC BGR to RGB x(640,1280,3) # PIL: = Image.open('image.jpg') or ImageGrab.grab() # HWC x(640,1280,3) @@ -597,65 +598,65 @@ def forward(self, imgs, size=640, augment=False, profile=False): # torch: = torch.zeros(16,3,320,640) # BCHW (scaled to size=640, 0-1 values) # multiple: = [Image.open('image1.jpg'), Image.open('image2.jpg'), ...] # list of images - t = [time_sync()] - p = next(self.model.parameters()) if self.pt else torch.zeros(1, device=self.model.device) # for device, type - autocast = self.amp and (p.device.type != 'cpu') # Automatic Mixed Precision (AMP) inference - if isinstance(imgs, torch.Tensor): # torch - with amp.autocast(autocast): - return self.model(imgs.to(p.device).type_as(p), augment, profile) # inference - - # Pre-process - n, imgs = (len(imgs), list(imgs)) if isinstance(imgs, (list, tuple)) else (1, [imgs]) # number, list of images - shape0, shape1, files = [], [], [] # image and inference shapes, filenames - for i, im in enumerate(imgs): - f = f'image{i}' # filename - if isinstance(im, (str, Path)): # filename or uri - im, f = Image.open(requests.get(im, stream=True).raw if str(im).startswith('http') else im), im - im = np.asarray(exif_transpose(im)) - elif isinstance(im, Image.Image): # PIL Image - im, f = np.asarray(exif_transpose(im)), getattr(im, 'filename', f) or f - files.append(Path(f).with_suffix('.jpg').name) - if im.shape[0] < 5: # image in CHW - im = im.transpose((1, 2, 0)) # reverse dataloader .transpose(2, 0, 1) - im = im[..., :3] if im.ndim == 3 else cv2.cvtColor(im, cv2.COLOR_GRAY2BGR) # enforce 3ch input - s = im.shape[:2] # HWC - shape0.append(s) # image shape - g = (size / max(s)) # gain - shape1.append([y * g for y in s]) - imgs[i] = im if im.data.contiguous else np.ascontiguousarray(im) # update - shape1 = [make_divisible(x, self.stride) if self.pt else size for x in np.array(shape1).max(0)] # inf shape - x = [letterbox(im, shape1, auto=False)[0] for im in imgs] # pad - x = np.ascontiguousarray(np.array(x).transpose((0, 3, 1, 2))) # stack and BHWC to BCHW - x = torch.from_numpy(x).to(p.device).type_as(p) / 255 # uint8 to fp16/32 - t.append(time_sync()) + dt = (Profile(), Profile(), Profile()) + with dt[0]: + p = next(self.model.parameters()) if self.pt else torch.zeros(1, device=self.model.device) # param + autocast = self.amp and (p.device.type != 'cpu') # Automatic Mixed Precision (AMP) inference + if isinstance(ims, torch.Tensor): # torch + with amp.autocast(autocast): + return self.model(ims.to(p.device).type_as(p), augment, profile) # inference + + # Pre-process + n, ims = (len(ims), list(ims)) if isinstance(ims, (list, tuple)) else (1, [ims]) # number, list of images + shape0, shape1, files = [], [], [] # image and inference shapes, filenames + for i, im in enumerate(ims): + f = f'image{i}' # filename + if isinstance(im, (str, Path)): # filename or uri + im, f = Image.open(requests.get(im, stream=True).raw if str(im).startswith('http') else im), im + im = np.asarray(exif_transpose(im)) + elif isinstance(im, Image.Image): # PIL Image + im, f = np.asarray(exif_transpose(im)), getattr(im, 'filename', f) or f + files.append(Path(f).with_suffix('.jpg').name) + if im.shape[0] < 5: # image in CHW + im = im.transpose((1, 2, 0)) # reverse dataloader .transpose(2, 0, 1) + im = im[..., :3] if im.ndim == 3 else cv2.cvtColor(im, cv2.COLOR_GRAY2BGR) # enforce 3ch input + s = im.shape[:2] # HWC + shape0.append(s) # image shape + g = (size / max(s)) # gain + shape1.append([y * g for y in s]) + ims[i] = im if im.data.contiguous else np.ascontiguousarray(im) # update + shape1 = [make_divisible(x, self.stride) if self.pt else size for x in np.array(shape1).max(0)] # inf shape + x = [letterbox(im, shape1, auto=False)[0] for im in ims] # pad + x = np.ascontiguousarray(np.array(x).transpose((0, 3, 1, 2))) # stack and BHWC to BCHW + x = torch.from_numpy(x).to(p.device).type_as(p) / 255 # uint8 to fp16/32 with amp.autocast(autocast): # Inference - y = self.model(x, augment, profile) # forward - t.append(time_sync()) + with dt[1]: + y = self.model(x, augment, profile) # forward # Post-process - y = non_max_suppression(y if self.dmb else y[0], - self.conf, - self.iou, - self.classes, - self.agnostic, - self.multi_label, - max_det=self.max_det) # NMS - for i in range(n): - scale_coords(shape1, y[i][:, :4], shape0[i]) + with dt[2]: + y = non_max_suppression(y if self.dmb else y[0], + self.conf, + self.iou, + self.classes, + self.agnostic, + self.multi_label, + max_det=self.max_det) # NMS + for i in range(n): + scale_coords(shape1, y[i][:, :4], shape0[i]) - t.append(time_sync()) - return Detections(imgs, y, files, t, self.names, x.shape) + return Detections(ims, y, files, dt, self.names, x.shape) class Detections: # YOLOv5 detections class for inference results - def __init__(self, imgs, pred, files, times=(0, 0, 0, 0), names=None, shape=None): + def __init__(self, ims, pred, files, times=(0, 0, 0), names=None, shape=None): super().__init__() d = pred[0].device # device - gn = [torch.tensor([*(im.shape[i] for i in [1, 0, 1, 0]), 1, 1], device=d) for im in imgs] # normalizations - self.imgs = imgs # list of images as numpy arrays + gn = [torch.tensor([*(im.shape[i] for i in [1, 0, 1, 0]), 1, 1], device=d) for im in ims] # normalizations + self.ims = ims # list of images as numpy arrays self.pred = pred # list of tensors pred[0] = (xyxy, conf, cls) self.names = names # class names self.files = files # image filenames @@ -665,12 +666,12 @@ def __init__(self, imgs, pred, files, times=(0, 0, 0, 0), names=None, shape=None self.xyxyn = [x / g for x, g in zip(self.xyxy, gn)] # xyxy normalized self.xywhn = [x / g for x, g in zip(self.xywh, gn)] # xywh normalized self.n = len(self.pred) # number of images (batch size) - self.t = tuple((times[i + 1] - times[i]) * 1000 / self.n for i in range(3)) # timestamps (ms) + self.t = tuple(x.t / self.n * 1E3 for x in times) # timestamps (ms) self.s = shape # inference BCHW shape def display(self, pprint=False, show=False, save=False, crop=False, render=False, labels=True, save_dir=Path('')): crops = [] - for i, (im, pred) in enumerate(zip(self.imgs, self.pred)): + for i, (im, pred) in enumerate(zip(self.ims, self.pred)): s = f'image {i + 1}/{len(self.pred)}: {im.shape[0]}x{im.shape[1]} ' # string if pred.shape[0]: for c in pred[:, -1].unique(): @@ -705,7 +706,7 @@ def display(self, pprint=False, show=False, save=False, crop=False, render=False if i == self.n - 1: LOGGER.info(f"Saved {self.n} image{'s' * (self.n > 1)} to {colorstr('bold', save_dir)}") if render: - self.imgs[i] = np.asarray(im) + self.ims[i] = np.asarray(im) if crop: if save: LOGGER.info(f'Saved results to {save_dir}\n') @@ -728,7 +729,7 @@ def crop(self, save=True, save_dir='runs/detect/exp'): def render(self, labels=True): self.display(render=True, labels=labels) # render results - return self.imgs + return self.ims def pandas(self): # return detections as pandas DataFrames, i.e. print(results.pandas().xyxy[0]) @@ -743,9 +744,9 @@ def pandas(self): def tolist(self): # return a list of Detections objects, i.e. 'for result in results.tolist():' r = range(self.n) # iterable - x = [Detections([self.imgs[i]], [self.pred[i]], [self.files[i]], self.times, self.names, self.s) for i in r] + x = [Detections([self.ims[i]], [self.pred[i]], [self.files[i]], self.times, self.names, self.s) for i in r] # for d in x: - # for k in ['imgs', 'pred', 'xyxy', 'xyxyn', 'xywh', 'xywhn']: + # for k in ['ims', 'pred', 'xyxy', 'xyxyn', 'xywh', 'xywhn']: # setattr(d, k, getattr(d, k)[0]) # pop out of list return x diff --git a/utils/general.py b/utils/general.py index 76872b696d79..42d000918c13 100755 --- a/utils/general.py +++ b/utils/general.py @@ -141,16 +141,26 @@ def user_config_dir(dir='Ultralytics', env_var='YOLOV5_CONFIG_DIR'): class Profile(contextlib.ContextDecorator): - # Usage: @Profile() decorator or 'with Profile():' context manager + # YOLOv5 Profile class. Usage: @Profile() decorator or 'with Profile():' context manager + def __init__(self, t=0.0): + self.t = t + self.cuda = torch.cuda.is_available() + def __enter__(self): - self.start = time.time() + self.start = self.time() def __exit__(self, type, value, traceback): - print(f'Profile results: {time.time() - self.start:.5f}s') + self.dt = self.time() - self.start # delta-time + self.t += self.dt # accumulate dt + + def time(self): + if self.cuda: + torch.cuda.synchronize() + return time.time() class Timeout(contextlib.ContextDecorator): - # Usage: @Timeout(seconds) decorator or 'with Timeout(seconds):' context manager + # YOLOv5 Timeout class. Usage: @Timeout(seconds) decorator or 'with Timeout(seconds):' context manager def __init__(self, seconds, *, timeout_msg='', suppress_timeout_errors=True): self.seconds = int(seconds) self.timeout_message = timeout_msg diff --git a/val.py b/val.py index ce743b506aff..876fc5bf50bb 100644 --- a/val.py +++ b/val.py @@ -37,7 +37,7 @@ from models.common import DetectMultiBackend from utils.callbacks import Callbacks from utils.dataloaders import create_dataloader -from utils.general import (LOGGER, check_dataset, check_img_size, check_requirements, check_yaml, +from utils.general import (LOGGER, Profile, check_dataset, check_img_size, check_requirements, check_yaml, coco80_to_coco91_class, colorstr, increment_path, non_max_suppression, print_args, scale_coords, xywh2xyxy, xyxy2xywh) from utils.metrics import ConfusionMatrix, ap_per_class, box_iou @@ -187,26 +187,24 @@ def run( names = dict(enumerate(names)) class_map = coco80_to_coco91_class() if is_coco else list(range(1000)) s = ('%20s' + '%11s' * 6) % ('Class', 'Images', 'Labels', 'P', 'R', 'mAP@.5', 'mAP@.5:.95') - dt, p, r, f1, mp, mr, map50, map = [0.0, 0.0, 0.0], 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 + dt, p, r, f1, mp, mr, map50, map = (Profile(), Profile(), Profile()), 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 loss = torch.zeros(3, device=device) jdict, stats, ap, ap_class = [], [], [], [] callbacks.run('on_val_start') pbar = tqdm(dataloader, desc=s, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar for batch_i, (im, targets, paths, shapes) in enumerate(pbar): callbacks.run('on_val_batch_start') - t1 = time_sync() - if cuda: - im = im.to(device, non_blocking=True) - targets = targets.to(device) - im = im.half() if half else im.float() # uint8 to fp16/32 - im /= 255 # 0 - 255 to 0.0 - 1.0 - nb, _, height, width = im.shape # batch size, channels, height, width - t2 = time_sync() - dt[0] += t2 - t1 + with dt[0]: + if cuda: + im = im.to(device, non_blocking=True) + targets = targets.to(device) + im = im.half() if half else im.float() # uint8 to fp16/32 + im /= 255 # 0 - 255 to 0.0 - 1.0 + nb, _, height, width = im.shape # batch size, channels, height, width # Inference - out, train_out = model(im) if training else model(im, augment=augment, val=True) # inference, loss outputs - dt[1] += time_sync() - t2 + with dt[1]: + out, train_out = model(im) if training else model(im, augment=augment, val=True) # inference, loss outputs # Loss if compute_loss: @@ -215,9 +213,8 @@ def run( # NMS targets[:, 2:] *= torch.tensor((width, height, width, height), device=device) # to pixels lb = [targets[targets[:, 0] == i, 1:] for i in range(nb)] if save_hybrid else [] # for autolabelling - t3 = time_sync() - out = non_max_suppression(out, conf_thres, iou_thres, labels=lb, multi_label=True, agnostic=single_cls) - dt[2] += time_sync() - t3 + with dt[2]: + out = non_max_suppression(out, conf_thres, iou_thres, labels=lb, multi_label=True, agnostic=single_cls) # Metrics for si, pred in enumerate(out): @@ -284,7 +281,7 @@ def run( LOGGER.info(pf % (names[c], seen, nt[c], p[i], r[i], ap50[i], ap[i])) # Print speeds - t = tuple(x / seen * 1E3 for x in dt) # speeds per image + t = tuple(x.t / seen * 1E3 for x in dt) # speeds per image if not training: shape = (batch_size, 3, imgsz, imgsz) LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {shape}' % t) From 61adf017f231f470afca2636f1f13e4cce13914b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 18 Aug 2022 20:12:33 +0200 Subject: [PATCH 479/661] `torch.empty()` for speed improvements (#9025) `torch.empty()` for speed improvement Signed-off-by: Glenn Jocher --- models/common.py | 4 ++-- models/yolo.py | 6 +++--- utils/autobatch.py | 2 +- utils/loggers/__init__.py | 2 +- utils/torch_utils.py | 2 +- 5 files changed, 8 insertions(+), 8 deletions(-) diff --git a/models/common.py b/models/common.py index 33aa2ac12465..44192e622bb5 100644 --- a/models/common.py +++ b/models/common.py @@ -531,7 +531,7 @@ def warmup(self, imgsz=(1, 3, 640, 640)): # Warmup model by running inference once warmup_types = self.pt, self.jit, self.onnx, self.engine, self.saved_model, self.pb if any(warmup_types) and self.device.type != 'cpu': - im = torch.zeros(*imgsz, dtype=torch.half if self.fp16 else torch.float, device=self.device) # input + im = torch.empty(*imgsz, dtype=torch.half if self.fp16 else torch.float, device=self.device) # input for _ in range(2 if self.jit else 1): # self.forward(im) # warmup @@ -600,7 +600,7 @@ def forward(self, ims, size=640, augment=False, profile=False): dt = (Profile(), Profile(), Profile()) with dt[0]: - p = next(self.model.parameters()) if self.pt else torch.zeros(1, device=self.model.device) # param + p = next(self.model.parameters()) if self.pt else torch.empty(1, device=self.model.device) # param autocast = self.amp and (p.device.type != 'cpu') # Automatic Mixed Precision (AMP) inference if isinstance(ims, torch.Tensor): # torch with amp.autocast(autocast): diff --git a/models/yolo.py b/models/yolo.py index df4209726e0d..32a47e9591da 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -46,8 +46,8 @@ def __init__(self, nc=80, anchors=(), ch=(), inplace=True): # detection layer self.no = nc + 5 # number of outputs per anchor self.nl = len(anchors) # number of detection layers self.na = len(anchors[0]) // 2 # number of anchors - self.grid = [torch.zeros(1)] * self.nl # init grid - self.anchor_grid = [torch.zeros(1)] * self.nl # init anchor grid + self.grid = [torch.empty(1)] * self.nl # init grid + self.anchor_grid = [torch.empty(1)] * self.nl # init anchor grid self.register_buffer('anchors', torch.tensor(anchors).float().view(self.nl, -1, 2)) # shape(nl,na,2) self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv self.inplace = inplace # use inplace ops (e.g. slice assignment) @@ -175,7 +175,7 @@ def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, anchors=None): # model, i if isinstance(m, Detect): s = 256 # 2x min stride m.inplace = self.inplace - m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))]) # forward + m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.empty(1, ch, s, s))]) # forward check_anchor_order(m) # must be in pixel-space (not grid-space) m.anchors /= m.stride.view(-1, 1, 1) self.stride = m.stride diff --git a/utils/autobatch.py b/utils/autobatch.py index c231d24c0706..07cddc99f400 100644 --- a/utils/autobatch.py +++ b/utils/autobatch.py @@ -47,7 +47,7 @@ def autobatch(model, imgsz=640, fraction=0.9, batch_size=16): # Profile batch sizes batch_sizes = [1, 2, 4, 8, 16] try: - img = [torch.zeros(b, 3, imgsz, imgsz) for b in batch_sizes] + img = [torch.empty(b, 3, imgsz, imgsz) for b in batch_sizes] results = profile(img, model, n=3, device=device) except Exception as e: LOGGER.warning(f'{prefix}{e}') diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 8ec846f8cfac..34704b625294 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -300,7 +300,7 @@ def log_tensorboard_graph(tb, model, imgsz=(640, 640)): try: p = next(model.parameters()) # for device, type imgsz = (imgsz, imgsz) if isinstance(imgsz, int) else imgsz # expand - im = torch.zeros((1, 3, *imgsz)).to(p.device).type_as(p) # input image + im = torch.empty((1, 3, *imgsz)).to(p.device).type_as(p) # input image with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress jit trace warning tb.add_graph(torch.jit.trace(de_parallel(model), im, strict=False), []) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index ed56064ce02e..4de2520b26a2 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -282,7 +282,7 @@ def model_info(model, verbose=False, imgsz=640): try: # FLOPs p = next(model.parameters()) stride = max(int(model.stride.max()), 32) if hasattr(model, 'stride') else 32 # max stride - im = torch.zeros((1, p.shape[1], stride, stride), device=p.device) # input image in BCHW format + im = torch.empty((1, p.shape[1], stride, stride), device=p.device) # input image in BCHW format flops = thop.profile(deepcopy(model), inputs=(im,), verbose=False)[0] / 1E9 * 2 # stride GFLOPs imgsz = imgsz if isinstance(imgsz, list) else [imgsz, imgsz] # expand if int/float fs = f', {flops * imgsz[0] / stride * imgsz[1] / stride:.1f} GFLOPs' # 640x640 GFLOPs From de6e6c0110adbb41f829c1288d5cdab7105892ae Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 18 Aug 2022 20:23:14 +0200 Subject: [PATCH 480/661] Created using Colaboratory --- tutorial.ipynb | 136 ++++++++++++++++++++++++------------------------- 1 file changed, 68 insertions(+), 68 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 1438924e4112..97e572798427 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -17,7 +17,7 @@ "accelerator": "GPU", "widgets": { "application/vnd.jupyter.widget-state+json": { - "57c562894aed45cd9a107d0455e3e3f4": { + "6d6b90ead2db49b3bdf624b6ba9b44e9": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", @@ -32,14 +32,14 @@ "_view_name": "HBoxView", "box_style": "", "children": [ - "IPY_MODEL_040d53c6cc924350bcb656cd21a7c713", - "IPY_MODEL_e029890942a74c098408ce5a9a566d51", - "IPY_MODEL_8fb991c03e434566a4297b6ab9446f89" + "IPY_MODEL_cb77443edb9e43328a56aaa4413a0df3", + "IPY_MODEL_954c8b8699e143bf92be6bfc02fc52f6", + "IPY_MODEL_a64775946e13477f83d8bba6086385b9" ], - "layout": "IPY_MODEL_a9a376923a7742d89fb335db709c7a7e" + "layout": "IPY_MODEL_1413611b7f4f4ef99e4f541f5ca35ed6" } }, - "040d53c6cc924350bcb656cd21a7c713": { + "cb77443edb9e43328a56aaa4413a0df3": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", @@ -54,13 +54,13 @@ "_view_name": "HTMLView", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_8b4276ac834c4735bf60ee9b761b9962", + "layout": "IPY_MODEL_00737f5558eb4fbd968172acb978e54a", "placeholder": "​", - "style": "IPY_MODEL_52cc8da75b724198856617247541cb1e", + "style": "IPY_MODEL_f03e5ddfd1c04bedaf68ab02c3f6f0ea", "value": "100%" } }, - "e029890942a74c098408ce5a9a566d51": { + "954c8b8699e143bf92be6bfc02fc52f6": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", @@ -76,15 +76,15 @@ "bar_style": "success", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_b6652f46480243c4adf60e6440043d6f", + "layout": "IPY_MODEL_6926db7e0035455f99e1dd4508c4b19c", "max": 818322941, "min": 0, "orientation": "horizontal", - "style": "IPY_MODEL_e502754177ff4ea8abf82a6e9ac77a4a", + "style": "IPY_MODEL_a6a52c9f828b458e97ddf7a11ae9275f", "value": 818322941 } }, - "8fb991c03e434566a4297b6ab9446f89": { + "a64775946e13477f83d8bba6086385b9": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", @@ -99,13 +99,13 @@ "_view_name": "HTMLView", "description": "", "description_tooltip": null, - "layout": "IPY_MODEL_447398becdb04836b5ffb5915318db07", + "layout": "IPY_MODEL_c4c7dc45a1c24dc4b2c709e21271a37e", "placeholder": "​", - "style": "IPY_MODEL_2fddcb27ad4a4caa81ff51111f8d0ed6", - "value": " 780M/780M [01:17<00:00, 12.3MB/s]" + "style": "IPY_MODEL_09c43ffe2c7e4bdc9489e83f9d82ab73", + "value": " 780M/780M [01:12<00:00, 23.8MB/s]" } }, - "a9a376923a7742d89fb335db709c7a7e": { + "1413611b7f4f4ef99e4f541f5ca35ed6": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -157,7 +157,7 @@ "width": null } }, - "8b4276ac834c4735bf60ee9b761b9962": { + "00737f5558eb4fbd968172acb978e54a": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -209,7 +209,7 @@ "width": null } }, - "52cc8da75b724198856617247541cb1e": { + "f03e5ddfd1c04bedaf68ab02c3f6f0ea": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", @@ -224,7 +224,7 @@ "description_width": "" } }, - "b6652f46480243c4adf60e6440043d6f": { + "6926db7e0035455f99e1dd4508c4b19c": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -276,7 +276,7 @@ "width": null } }, - "e502754177ff4ea8abf82a6e9ac77a4a": { + "a6a52c9f828b458e97ddf7a11ae9275f": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", @@ -292,7 +292,7 @@ "description_width": "" } }, - "447398becdb04836b5ffb5915318db07": { + "c4c7dc45a1c24dc4b2c709e21271a37e": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", @@ -344,7 +344,7 @@ "width": null } }, - "2fddcb27ad4a4caa81ff51111f8d0ed6": { + "09c43ffe2c7e4bdc9489e83f9d82ab73": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", @@ -404,7 +404,7 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "e0f693e4-413b-4cc8-ae7e-91537da370b0" + "outputId": "508de90c-846e-495d-c7d6-50681af62a98" }, "source": [ "!git clone https://github.com/ultralytics/yolov5 # clone\n", @@ -421,7 +421,7 @@ "output_type": "stream", "name": "stderr", "text": [ - "YOLOv5 🚀 v6.2-2-g7c9486e Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" + "YOLOv5 🚀 v6.2-15-g61adf01 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" ] }, { @@ -461,7 +461,7 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "941d625b-01a1-4f1b-dfd2-d9ef1c945715" + "outputId": "93881540-331e-4890-cd38-4c2776933238" }, "source": [ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", @@ -474,16 +474,16 @@ "name": "stdout", "text": [ "\u001b[34m\u001b[1mdetect: \u001b[0mweights=['yolov5s.pt'], source=data/images, data=data/coco128.yaml, imgsz=[640, 640], conf_thres=0.25, iou_thres=0.45, max_det=1000, device=, view_img=False, save_txt=False, save_conf=False, save_crop=False, nosave=False, classes=None, agnostic_nms=False, augment=False, visualize=False, update=False, project=runs/detect, name=exp, exist_ok=False, line_thickness=3, hide_labels=False, hide_conf=False, half=False, dnn=False\n", - "YOLOv5 🚀 v6.2-2-g7c9486e Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-15-g61adf01 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s.pt to yolov5s.pt...\n", - "100% 14.1M/14.1M [00:00<00:00, 50.5MB/s]\n", + "100% 14.1M/14.1M [00:00<00:00, 39.3MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5s summary: 213 layers, 7225885 parameters, 0 gradients\n", - "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, Done. (0.014s)\n", - "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, Done. (0.020s)\n", - "Speed: 0.6ms pre-process, 17.0ms inference, 20.2ms NMS per image at shape (1, 3, 640, 640)\n", + "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, 14.9ms\n", + "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, 22.0ms\n", + "Speed: 0.6ms pre-process, 18.4ms inference, 24.1ms NMS per image at shape (1, 3, 640, 640)\n", "Results saved to \u001b[1mruns/detect/exp\u001b[0m\n" ] } @@ -527,20 +527,20 @@ "base_uri": "https://localhost:8080/", "height": 49, "referenced_widgets": [ - "57c562894aed45cd9a107d0455e3e3f4", - "040d53c6cc924350bcb656cd21a7c713", - "e029890942a74c098408ce5a9a566d51", - "8fb991c03e434566a4297b6ab9446f89", - "a9a376923a7742d89fb335db709c7a7e", - "8b4276ac834c4735bf60ee9b761b9962", - "52cc8da75b724198856617247541cb1e", - "b6652f46480243c4adf60e6440043d6f", - "e502754177ff4ea8abf82a6e9ac77a4a", - "447398becdb04836b5ffb5915318db07", - "2fddcb27ad4a4caa81ff51111f8d0ed6" + "6d6b90ead2db49b3bdf624b6ba9b44e9", + "cb77443edb9e43328a56aaa4413a0df3", + "954c8b8699e143bf92be6bfc02fc52f6", + "a64775946e13477f83d8bba6086385b9", + "1413611b7f4f4ef99e4f541f5ca35ed6", + "00737f5558eb4fbd968172acb978e54a", + "f03e5ddfd1c04bedaf68ab02c3f6f0ea", + "6926db7e0035455f99e1dd4508c4b19c", + "a6a52c9f828b458e97ddf7a11ae9275f", + "c4c7dc45a1c24dc4b2c709e21271a37e", + "09c43ffe2c7e4bdc9489e83f9d82ab73" ] }, - "outputId": "d593b41a-55e7-48a5-e285-5df449edc8c0" + "outputId": "ed2ca46e-a1a9-4a16-c449-859278d8aa18" }, "source": [ "# Download COCO val\n", @@ -558,7 +558,7 @@ "application/vnd.jupyter.widget-view+json": { "version_major": 2, "version_minor": 0, - "model_id": "57c562894aed45cd9a107d0455e3e3f4" + "model_id": "6d6b90ead2db49b3bdf624b6ba9b44e9" } }, "metadata": {} @@ -572,7 +572,7 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "701132a6-9ca8-4e1f-c89f-5d38893a6fc4" + "outputId": "19a590ef-363e-424c-d9ce-78bbe0593cd5" }, "source": [ "# Run YOLOv5x on COCO val\n", @@ -585,35 +585,35 @@ "name": "stdout", "text": [ "\u001b[34m\u001b[1mval: \u001b[0mdata=/content/yolov5/data/coco.yaml, weights=['yolov5x.pt'], batch_size=32, imgsz=640, conf_thres=0.001, iou_thres=0.65, task=val, device=, workers=8, single_cls=False, augment=False, verbose=False, save_txt=False, save_hybrid=False, save_conf=False, save_json=True, project=runs/val, name=exp, exist_ok=False, half=True, dnn=False\n", - "YOLOv5 🚀 v6.2-2-g7c9486e Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-15-g61adf01 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5x.pt to yolov5x.pt...\n", - "100% 166M/166M [00:11<00:00, 15.1MB/s]\n", + "100% 166M/166M [00:06<00:00, 28.1MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5x summary: 444 layers, 86705005 parameters, 0 gradients\n", "Downloading https://ultralytics.com/assets/Arial.ttf to /root/.config/Ultralytics/Arial.ttf...\n", - "100% 755k/755k [00:00<00:00, 48.6MB/s]\n", - "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 10889.87it/s]\n", + "100% 755k/755k [00:00<00:00, 47.3MB/s]\n", + "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 10756.32it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mNew cache created: /content/datasets/coco/val2017.cache\n", - " Class Images Labels P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:05<00:00, 2.38it/s]\n", + " Class Images Labels P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:07<00:00, 2.33it/s]\n", " all 5000 36335 0.743 0.625 0.683 0.504\n", - "Speed: 0.1ms pre-process, 4.7ms inference, 1.0ms NMS per image at shape (32, 3, 640, 640)\n", + "Speed: 0.1ms pre-process, 4.6ms inference, 1.2ms NMS per image at shape (32, 3, 640, 640)\n", "\n", "Evaluating pycocotools mAP... saving runs/val/exp/yolov5x_predictions.json...\n", "loading annotations into memory...\n", - "Done (t=0.39s)\n", + "Done (t=0.41s)\n", "creating index...\n", "index created!\n", "Loading and preparing results...\n", - "DONE (t=5.53s)\n", + "DONE (t=5.64s)\n", "creating index...\n", "index created!\n", "Running per image evaluation...\n", "Evaluate annotation type *bbox*\n", - "DONE (t=73.01s).\n", + "DONE (t=76.80s).\n", "Accumulating evaluation results...\n", - "DONE (t=15.27s).\n", + "DONE (t=14.61s).\n", " Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.506\n", " Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.688\n", " Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.549\n", @@ -745,7 +745,7 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "50a9318f-d438-41d5-db95-928f1842c057" + "outputId": "47759d5e-34f0-4a6a-c714-ff533391cfff" }, "source": [ "# Train YOLOv5s on COCO128 for 3 epochs\n", @@ -759,7 +759,7 @@ "text": [ "\u001b[34m\u001b[1mtrain: \u001b[0mweights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch-low.yaml, epochs=3, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, noplots=False, evolve=None, bucket=, cache=ram, image_weights=False, device=, multi_scale=False, single_cls=False, optimizer=SGD, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, cos_lr=False, label_smoothing=0.0, patience=100, freeze=[0], save_period=-1, seed=0, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest\n", "\u001b[34m\u001b[1mgithub: \u001b[0mup to date with https://github.com/ultralytics/yolov5 ✅\n", - "YOLOv5 🚀 v6.2-2-g7c9486e Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-15-g61adf01 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "\u001b[34m\u001b[1mhyperparameters: \u001b[0mlr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0\n", "\u001b[34m\u001b[1mWeights & Biases: \u001b[0mrun 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs in Weights & Biases\n", @@ -768,8 +768,8 @@ "\n", "Dataset not found ⚠️, missing paths ['/content/datasets/coco128/images/train2017']\n", "Downloading https://ultralytics.com/assets/coco128.zip to coco128.zip...\n", - "100% 6.66M/6.66M [00:00<00:00, 12.4MB/s]\n", - "Dataset download success ✅ (1.3s), saved to \u001b[1m/content/datasets\u001b[0m\n", + "100% 6.66M/6.66M [00:00<00:00, 75.3MB/s]\n", + "Dataset download success ✅ (0.7s), saved to \u001b[1m/content/datasets\u001b[0m\n", "\n", " from n params module arguments \n", " 0 -1 1 3520 models.common.Conv [3, 32, 6, 2, 2] \n", @@ -803,11 +803,11 @@ "\u001b[34m\u001b[1mAMP: \u001b[0mchecks passed ✅\n", "\u001b[34m\u001b[1moptimizer:\u001b[0m SGD(lr=0.01) with parameter groups 57 weight(decay=0.0), 60 weight(decay=0.0005), 60 bias\n", "\u001b[34m\u001b[1malbumentations: \u001b[0mBlur(p=0.01, blur_limit=(3, 7)), MedianBlur(p=0.01, blur_limit=(3, 7)), ToGray(p=0.01), CLAHE(p=0.01, clip_limit=(1, 4.0), tile_grid_size=(8, 8))\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017' images and labels...128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00<00:00, 8516.89it/s]\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017' images and labels...128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00<00:00, 7246.20it/s]\n", "\u001b[34m\u001b[1mtrain: \u001b[0mNew cache created: /content/datasets/coco128/labels/train2017.cache\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mCaching images (0.1GB ram): 100% 128/128 [00:00<00:00, 1043.44it/s]\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mCaching images (0.1GB ram): 100% 128/128 [00:00<00:00, 986.21it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco128/labels/train2017.cache' images and labels... 128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00 Date: Thu, 18 Aug 2022 20:26:18 +0200 Subject: [PATCH 481/661] Remove unused `time_sync` import (#9026) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- val.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/val.py b/val.py index 876fc5bf50bb..7b4fab4c63be 100644 --- a/val.py +++ b/val.py @@ -42,7 +42,7 @@ scale_coords, xywh2xyxy, xyxy2xywh) from utils.metrics import ConfusionMatrix, ap_per_class, box_iou from utils.plots import output_to_target, plot_images, plot_val_study -from utils.torch_utils import select_device, smart_inference_mode, time_sync +from utils.torch_utils import select_device, smart_inference_mode def save_one_txt(predn, save_conf, shape, file): From eb359c3a226f55c9b51efcfeae2e31c820e6e08a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 18 Aug 2022 21:45:11 +0200 Subject: [PATCH 482/661] Add PyTorch Hub classification CI checks (#9027) * Add PyTorch Hub classification CI checks Add PyTorch Hub loading of official and custom trained classification models to CI checks. May help resolve https://github.com/ultralytics/yolov5/issues/8790#issuecomment-1219840718 Signed-off-by: Glenn Jocher * Update hubconf.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/workflows/ci-testing.yml | 5 +++++ hubconf.py | 6 +++--- 2 files changed, 8 insertions(+), 3 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index aa797c44d487..fde6fffe92f4 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -133,3 +133,8 @@ jobs: python classify/predict.py --imgsz 32 --weights $b --source ../datasets/mnist2560/test/7/60.png # predict python classify/predict.py --imgsz 32 --weights $m --source data/images/bus.jpg # predict python export.py --weights $b --img 64 --imgsz 224 --include torchscript # export + python - < Date: Fri, 19 Aug 2022 01:30:14 +0200 Subject: [PATCH 483/661] Created using Colaboratory --- tutorial.ipynb | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 97e572798427..7a1edf7ef86a 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -415,7 +415,7 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": 1, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -467,7 +467,7 @@ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", "# display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": 2, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -547,7 +547,7 @@ "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip')\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip" ], - "execution_count": 3, + "execution_count": null, "outputs": [ { "output_type": "display_data", @@ -578,7 +578,7 @@ "# Run YOLOv5x on COCO val\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" ], - "execution_count": 4, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -751,7 +751,7 @@ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": 5, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -1113,7 +1113,7 @@ "cell_type": "code", "source": [ "# Classification\n", - "for m in [*(f'yolov5{x}.pt' for x in 'nsmlx'), 'resnet50.pt', 'efficientnet_b0.pt']:\n", + "for m in [*(f'yolov5{x}-cls.pt' for x in 'nsmlx'), 'resnet50.pt', 'efficientnet_b0.pt']:\n", " for d in 'mnist', 'fashion-mnist', 'cifar10', 'cifar100', 'imagenette160', 'imagenette320', 'imagenette', 'imagewoof160', 'imagewoof320', 'imagewoof':\n", " !python classify/train.py --model {m} --data {d} --epochs 10 --project YOLOv5-cls --name {m}-{d}" ], From 840b7232dbaff6296e6e2519895c3065e937fdcf Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 19 Aug 2022 01:59:51 +0200 Subject: [PATCH 484/661] Attach transforms to model (#9028) * Attach transforms to model Signed-off-by: Glenn Jocher * Update val.py Signed-off-by: Glenn Jocher * Update train.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- classify/train.py | 10 +++++----- classify/val.py | 3 +-- 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/classify/train.py b/classify/train.py index 9fb7c52b545a..5881e16e47db 100644 --- a/classify/train.py +++ b/classify/train.py @@ -122,16 +122,16 @@ def train(opt, device): for p in model.parameters(): p.requires_grad = True # for training model = model.to(device) - names = trainloader.dataset.classes # class names - model.names = names # attach class names # Info if RANK in {-1, 0}: + model.names = trainloader.dataset.classes # attach class names + model.transforms = testloader.dataset.torch_transforms # attach inference transforms model_info(model) if opt.verbose: LOGGER.info(model) images, labels = next(iter(trainloader)) - file = imshow_cls(images[:25], labels[:25], names=names, f=save_dir / 'train_images.jpg') + file = imshow_cls(images[:25], labels[:25], names=model.names, f=save_dir / 'train_images.jpg') logger.log_images(file, name='Train Examples') logger.log_graph(model, imgsz) # log model @@ -254,8 +254,8 @@ def train(opt, device): # Plot examples images, labels = (x[:25] for x in next(iter(testloader))) # first 25 images and labels - pred = torch.max(ema.ema((images.half() if cuda else images.float()).to(device)), 1)[1] - file = imshow_cls(images, labels, pred, names, verbose=False, f=save_dir / 'test_images.jpg') + pred = torch.max(ema.ema(images.to(device)), 1)[1] + file = imshow_cls(images, labels, pred, model.names, verbose=False, f=save_dir / 'test_images.jpg') # Log results meta = {"epochs": epochs, "top1_acc": best_fitness, "date": datetime.now().isoformat()} diff --git a/classify/val.py b/classify/val.py index c91e2cf82c81..2353737957d3 100644 --- a/classify/val.py +++ b/classify/val.py @@ -39,7 +39,7 @@ def run( project=ROOT / 'runs/val-cls', # save to project/name name='exp', # save to project/name exist_ok=False, # existing project/name ok, do not increment - half=True, # use FP16 half-precision inference + half=False, # use FP16 half-precision inference dnn=False, # use OpenCV DNN for ONNX inference model=None, dataloader=None, @@ -124,7 +124,6 @@ def run( LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms post-process per image at shape {shape}' % t) LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") - model.float() # for training return top1, top5, loss From 1cd3e752def0ecbcb39a95d75e3c93fad3114ab9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 19 Aug 2022 02:01:40 +0200 Subject: [PATCH 485/661] Created using Colaboratory --- tutorial.ipynb | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 7a1edf7ef86a..a70887e97360 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -1112,8 +1112,8 @@ { "cell_type": "code", "source": [ - "# Classification\n", - "for m in [*(f'yolov5{x}-cls.pt' for x in 'nsmlx'), 'resnet50.pt', 'efficientnet_b0.pt']:\n", + "# Classification train\n", + "for m in [*(f'yolov5{x}-cls.pt' for x in 'nsmlx'), 'resnet50.pt', 'resnet101.pt', 'efficientnet_b0.pt', 'efficientnet_b1.pt']:\n", " for d in 'mnist', 'fashion-mnist', 'cifar10', 'cifar100', 'imagenette160', 'imagenette320', 'imagenette', 'imagewoof160', 'imagewoof320', 'imagewoof':\n", " !python classify/train.py --model {m} --data {d} --epochs 10 --project YOLOv5-cls --name {m}-{d}" ], @@ -1123,6 +1123,19 @@ "execution_count": null, "outputs": [] }, + { + "cell_type": "code", + "source": [ + "# Classification val\n", + "!bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images)\n", + "!python classify/val.py --weights yolov5m-cls.pt --data ../datasets/imagenet --img 224 # validate" + ], + "metadata": { + "id": "yYgOiFNHZx-1" + }, + "execution_count": null, + "outputs": [] + }, { "cell_type": "code", "metadata": { From 781401ec70bc481b789b214003b722174e4b99e0 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 19 Aug 2022 15:06:37 +0200 Subject: [PATCH 486/661] Default --data `imagenette160` training (fastest) (#9033) * Default --data `imagenette160` training (fastest) Signed-off-by: Glenn Jocher * Update train.py Signed-off-by: Glenn Jocher * Update train.py Signed-off-by: Glenn Jocher * Update train.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- classify/train.py | 6 +++--- train.py | 2 +- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/classify/train.py b/classify/train.py index 5881e16e47db..8fe90c1b19eb 100644 --- a/classify/train.py +++ b/classify/train.py @@ -6,7 +6,7 @@ Torchvision models: --model resnet50, efficientnet_b0, etc. See https://pytorch.org/vision/stable/models.html Usage - Single-GPU and Multi-GPU DDP - $ python classify/train.py --model yolov5s-cls.pt --data cifar100 --epochs 5 --img 128 + $ python classify/train.py --model yolov5s-cls.pt --data imagenette160 --epochs 5 --img 128 $ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/train.py --model yolov5s-cls.pt --data imagenet --epochs 5 --img 224 --device 0,1,2,3 """ @@ -266,8 +266,8 @@ def train(opt, device): def parse_opt(known=False): parser = argparse.ArgumentParser() parser.add_argument('--model', type=str, default='yolov5s-cls.pt', help='initial weights path') - parser.add_argument('--data', type=str, default='mnist', help='cifar10, cifar100, mnist, imagenet, etc.') - parser.add_argument('--epochs', type=int, default=10) + parser.add_argument('--data', type=str, default='imagenette160', help='cifar10, cifar100, mnist, imagenet, ...') + parser.add_argument('--epochs', type=int, default=10, help='total training epochs') parser.add_argument('--batch-size', type=int, default=64, help='total batch size for all GPUs') parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=128, help='train, val image size (pixels)') parser.add_argument('--nosave', action='store_true', help='only save final checkpoint') diff --git a/train.py b/train.py index 10a3bdb56002..279d52de6d74 100644 --- a/train.py +++ b/train.py @@ -436,7 +436,7 @@ def parse_opt(known=False): parser.add_argument('--cfg', type=str, default='', help='model.yaml path') parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path') parser.add_argument('--hyp', type=str, default=ROOT / 'data/hyps/hyp.scratch-low.yaml', help='hyperparameters path') - parser.add_argument('--epochs', type=int, default=300) + parser.add_argument('--epochs', type=int, default=300, help='total training epochs') parser.add_argument('--batch-size', type=int, default=16, help='total batch size for all GPUs, -1 for autobatch') parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='train, val image size (pixels)') parser.add_argument('--rect', action='store_true', help='rectangular training') From 4a8ab3bc42d32f3e2e9c026b87dc29fba6143064 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 19 Aug 2022 15:07:04 +0200 Subject: [PATCH 487/661] VOC `names` dictionary fix (#9034) * VOC names dictionary fix Signed-off-by: Glenn Jocher * Update dataloaders.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- data/VOC.yaml | 5 +++-- utils/dataloaders.py | 10 ++++++---- 2 files changed, 9 insertions(+), 6 deletions(-) diff --git a/data/VOC.yaml b/data/VOC.yaml index bbe5cf90a838..27d38109c53a 100644 --- a/data/VOC.yaml +++ b/data/VOC.yaml @@ -65,12 +65,13 @@ download: | w = int(size.find('width').text) h = int(size.find('height').text) + names = list(yaml['names'].values()) # names list for obj in root.iter('object'): cls = obj.find('name').text - if cls in yaml['names'] and not int(obj.find('difficult').text) == 1: + if cls in names and int(obj.find('difficult').text) != 1: xmlbox = obj.find('bndbox') bb = convert_box((w, h), [float(xmlbox.find(x).text) for x in ('xmin', 'xmax', 'ymin', 'ymax')]) - cls_id = yaml['names'].index(cls) # class id + cls_id = names.index(cls) # class id out_file.write(" ".join([str(a) for a in (cls_id, *bb)]) + '\n') diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 3f26be2cd32d..e73b20a58915 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -35,7 +35,7 @@ from utils.torch_utils import torch_distributed_zero_first # Parameters -HELP_URL = 'https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' +HELP_URL = 'See https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' IMG_FORMATS = 'bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp' # include image suffixes VID_FORMATS = 'asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'ts', 'wmv' # include video suffixes BAR_FORMAT = '{l_bar}{bar:10}{r_bar}{bar:-10b}' # tqdm bar format @@ -456,7 +456,7 @@ def __init__(self, # self.img_files = sorted([x for x in f if x.suffix[1:].lower() in IMG_FORMATS]) # pathlib assert self.im_files, f'{prefix}No images found' except Exception as e: - raise Exception(f'{prefix}Error loading data from {path}: {e}\nSee {HELP_URL}') + raise Exception(f'{prefix}Error loading data from {path}: {e}\n{HELP_URL}') # Check cache self.label_files = img2label_paths(self.im_files) # labels @@ -475,11 +475,13 @@ def __init__(self, tqdm(None, desc=prefix + d, total=n, initial=n, bar_format=BAR_FORMAT) # display cache results if cache['msgs']: LOGGER.info('\n'.join(cache['msgs'])) # display warnings - assert nf > 0 or not augment, f'{prefix}No labels in {cache_path}. Can not train without labels. See {HELP_URL}' + assert nf > 0 or not augment, f'{prefix}No labels found in {cache_path}, can not start training. {HELP_URL}' # Read cache [cache.pop(k) for k in ('hash', 'version', 'msgs')] # remove items labels, shapes, self.segments = zip(*cache.values()) + nl = len(np.concatenate(labels, 0)) # number of labels + assert nl > 0 or not augment, f'{prefix}All labels empty in {cache_path}, can not start training. {HELP_URL}' self.labels = list(labels) self.shapes = np.array(shapes) self.im_files = list(cache.keys()) # update @@ -572,7 +574,7 @@ def cache_labels(self, path=Path('./labels.cache'), prefix=''): if msgs: LOGGER.info('\n'.join(msgs)) if nf == 0: - LOGGER.warning(f'{prefix}WARNING: No labels found in {path}. See {HELP_URL}') + LOGGER.warning(f'{prefix}WARNING: No labels found in {path}. {HELP_URL}') x['hash'] = get_hash(self.label_files + self.im_files) x['results'] = nf, nm, ne, nc, len(self.im_files) x['msgs'] = msgs # warnings From fdcb92a938ef27d1b277a156af7f7922400279e3 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 19 Aug 2022 16:54:58 +0200 Subject: [PATCH 488/661] Update train.py `import val as validate` (#9037) * Update train.py `import val as validate` Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- train.py | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/train.py b/train.py index 279d52de6d74..665b4f5b609e 100644 --- a/train.py +++ b/train.py @@ -36,7 +36,7 @@ sys.path.append(str(ROOT)) # add ROOT to PATH ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative -import val # for end-of-epoch mAP +import val as validate # for end-of-epoch mAP from models.experimental import attempt_load from models.yolo import Model from utils.autoanchor import check_anchors @@ -347,17 +347,17 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio ema.update_attr(model, include=['yaml', 'nc', 'hyp', 'names', 'stride', 'class_weights']) final_epoch = (epoch + 1 == epochs) or stopper.possible_stop if not noval or final_epoch: # Calculate mAP - results, maps, _ = val.run(data_dict, - batch_size=batch_size // WORLD_SIZE * 2, - imgsz=imgsz, - half=amp, - model=ema.ema, - single_cls=single_cls, - dataloader=val_loader, - save_dir=save_dir, - plots=False, - callbacks=callbacks, - compute_loss=compute_loss) + results, maps, _ = validate.run(data_dict, + batch_size=batch_size // WORLD_SIZE * 2, + imgsz=imgsz, + half=amp, + model=ema.ema, + single_cls=single_cls, + dataloader=val_loader, + save_dir=save_dir, + plots=False, + callbacks=callbacks, + compute_loss=compute_loss) # Update best mAP fi = fitness(np.array(results).reshape(1, -1)) # weighted combination of [P, R, mAP@.5, mAP@.5-.95] @@ -407,12 +407,12 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio strip_optimizer(f) # strip optimizers if f is best: LOGGER.info(f'\nValidating {f}...') - results, _, _ = val.run( + results, _, _ = validate.run( data_dict, batch_size=batch_size // WORLD_SIZE * 2, imgsz=imgsz, model=attempt_load(f, device).half(), - iou_thres=0.65 if is_coco else 0.60, # best pycocotools results at 0.65 + iou_thres=0.65 if is_coco else 0.60, # best pycocotools at iou 0.65 single_cls=single_cls, dataloader=val_loader, save_dir=save_dir, From aed88848a25fe0f4d98e70e79f0ee876265b48fd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 19 Aug 2022 18:00:38 +0200 Subject: [PATCH 489/661] Simplified notebook --- tutorial.ipynb | 67 ++++++++++++++------------------------------------ 1 file changed, 18 insertions(+), 49 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index a70887e97360..1c5d77813f15 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -439,7 +439,7 @@ "id": "4JnkELT0cIJg" }, "source": [ - "# 1. Inference\n", + "# 1. Detect\n", "\n", "`detect.py` runs YOLOv5 inference on a variety of sources, downloading models automatically from the [latest YOLOv5 release](https://github.com/ultralytics/yolov5/releases), and saving results to `runs/detect`. Example inference sources are:\n", "\n", @@ -506,17 +506,7 @@ }, "source": [ "# 2. Validate\n", - "Validate a model's accuracy on [COCO](https://cocodataset.org/#home) val or test-dev datasets. Models are downloaded automatically from the [latest YOLOv5 release](https://github.com/ultralytics/yolov5/releases). To show results by class use the `--verbose` flag. Note that `pycocotools` metrics may be ~1% better than the equivalent repo metrics, as is visible below, due to slight differences in mAP computation." - ] - }, - { - "cell_type": "markdown", - "metadata": { - "id": "eyTZYGgRjnMc" - }, - "source": [ - "## COCO val\n", - "Download [COCO val 2017](https://github.com/ultralytics/yolov5/blob/74b34872fdf41941cddcf243951cdb090fbac17b/data/coco.yaml#L14) dataset (1GB - 5000 images), and test model accuracy." + "Validate a model's accuracy on the [COCO](https://cocodataset.org/#home) dataset's `val` or `test` splits. Models are downloaded automatically from the [latest YOLOv5 release](https://github.com/ultralytics/yolov5/releases). To show results by class use the `--verbose` flag." ] }, { @@ -544,8 +534,8 @@ }, "source": [ "# Download COCO val\n", - "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip')\n", - "!unzip -q tmp.zip -d ../datasets && rm tmp.zip" + "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip') # download COCO val (1GB - 5000 images)\n", + "!unzip -q tmp.zip -d ../datasets && rm tmp.zip # unzip" ], "execution_count": null, "outputs": [ @@ -575,7 +565,7 @@ "outputId": "19a590ef-363e-424c-d9ce-78bbe0593cd5" }, "source": [ - "# Run YOLOv5x on COCO val\n", + "# Validate YOLOv5x on COCO val\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" ], "execution_count": null, @@ -631,40 +621,6 @@ } ] }, - { - "cell_type": "markdown", - "metadata": { - "id": "rc_KbFk0juX2" - }, - "source": [ - "## COCO test\n", - "Download [COCO test2017](https://github.com/ultralytics/yolov5/blob/74b34872fdf41941cddcf243951cdb090fbac17b/data/coco.yaml#L15) dataset (7GB - 40,000 images), to test model accuracy on test-dev set (**20,000 images, no labels**). Results are saved to a `*.json` file which should be **zipped** and submitted to the evaluation server at https://competitions.codalab.org/competitions/20794." - ] - }, - { - "cell_type": "code", - "metadata": { - "id": "V0AJnSeCIHyJ" - }, - "source": [ - "# Download COCO test-dev2017\n", - "!bash data/scripts/get_coco.sh --test" - ], - "execution_count": null, - "outputs": [] - }, - { - "cell_type": "code", - "metadata": { - "id": "29GJXAP_lPrt" - }, - "source": [ - "# Run YOLOv5x on COCO test\n", - "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half --task test" - ], - "execution_count": null, - "outputs": [] - }, { "cell_type": "markdown", "metadata": { @@ -1136,6 +1092,19 @@ "execution_count": null, "outputs": [] }, + { + "cell_type": "code", + "source": [ + "# Validate on COCO test. Zip results.json and submit to eval server at https://competitions.codalab.org/competitions/20794\n", + "!bash data/scripts/get_coco.sh --test # download COCO test-dev2017 (7GB - 40,000 images, test 20,000)\n", + "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half --task test" + ], + "metadata": { + "id": "aq4DPWGu0Bl1" + }, + "execution_count": null, + "outputs": [] + }, { "cell_type": "code", "metadata": { From ba1c6773c2691943a355ad956105a4cb3aeedbca Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 19 Aug 2022 18:41:30 +0200 Subject: [PATCH 490/661] Created using Colaboratory --- tutorial.ipynb | 45 ++++++++++++++------------------------------- 1 file changed, 14 insertions(+), 31 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 1c5d77813f15..91e2d7e75eab 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -654,46 +654,29 @@ "

Label images lightning fast (including with model-assisted labeling)" ] }, - { - "cell_type": "code", - "metadata": { - "id": "bOy5KI2ncnWd" - }, - "source": [ - "# Tensorboard (optional)\n", - "%load_ext tensorboard\n", - "%tensorboard --logdir runs/train" - ], - "execution_count": null, - "outputs": [] - }, { "cell_type": "code", "source": [ - "# ClearML (optional)\n", - "%pip install -q clearml\n", - "!clearml-init" + "#@title Select YOLOv5 🚀 logger\n", + "logger = 'TensorBoard' #@param ['TensorBoard', 'ClearML', 'W&B']\n", + "\n", + "if logger == 'Tensorboard':\n", + " %load_ext tensorboard\n", + " %tensorboard --logdir runs/train\n", + "elif logger == 'ClearML':\n", + " %pip install -q clearml\n", + " !clearml-init\n", + "elif logger == 'W&B':\n", + " %pip install -q wandb\n", + " import wandb\n", + " wandb.login()" ], "metadata": { - "id": "DQhI6vvaRWjR" + "id": "i3oKtE4g-aNn" }, "execution_count": null, "outputs": [] }, - { - "cell_type": "code", - "metadata": { - "id": "2fLAV42oNb7M" - }, - "source": [ - "# Weights & Biases (optional)\n", - "%pip install -q wandb\n", - "import wandb\n", - "wandb.login()" - ], - "execution_count": null, - "outputs": [] - }, { "cell_type": "code", "metadata": { From a409ec7953e1c5dd572fc73f633de38efe0c101a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 20 Aug 2022 16:29:08 +0200 Subject: [PATCH 491/661] AutoBatch protect from negative batch sizes (#9048) * AutoBatch protect from negative batch sizes Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * cleanup * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/autobatch.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/utils/autobatch.py b/utils/autobatch.py index 07cddc99f400..8d12e46f0f09 100644 --- a/utils/autobatch.py +++ b/utils/autobatch.py @@ -60,6 +60,9 @@ def autobatch(model, imgsz=640, fraction=0.9, batch_size=16): i = results.index(None) # first fail index if b >= batch_sizes[i]: # y intercept above failure point b = batch_sizes[max(i - 1, 0)] # select prior safe point + if b < 1: # zero or negative batch size + b = 16 + LOGGER.warning(f'{prefix}WARNING: ⚠️ CUDA anomaly detected, recommend restart environment and retry command.') fraction = np.polyval(p, b) / t # actual fraction predicted LOGGER.info(f'{prefix}Using batch-size {b} for {d} {t * fraction:.2f}G/{t:.2f}G ({fraction * 100:.0f}%) ✅') From fc8758a49bd30526fb21d0683359e86be3a292a8 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 20 Aug 2022 16:45:11 +0200 Subject: [PATCH 492/661] Temporarily remove `macos-latest` from CI (#9049) * Temporarily remove macos-latest from CI macos-latest causing many failed CI runs that resolve after manually re-running 2 or 3 times. I don't know what the cause is. Will replace at a later date. Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/workflows/ci-testing.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index fde6fffe92f4..4ef930c61233 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -47,7 +47,7 @@ jobs: strategy: fail-fast: false matrix: - os: [ ubuntu-latest, macos-latest, windows-latest ] + os: [ ubuntu-latest, windows-latest ] # macos-latest bug https://github.com/ultralytics/yolov5/pull/9049 python-version: [ '3.10' ] model: [ yolov5n ] include: From f258cf8b37aeb3062230d43e1e9a4bf3b9874588 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 20 Aug 2022 17:17:35 +0200 Subject: [PATCH 493/661] Add `--save-hybrid` mAP warning (#9050) * Add `--save-hybrid` mAP warning Signed-off-by: Glenn Jocher * Update val.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- val.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/val.py b/val.py index 7b4fab4c63be..fcaca889d7e2 100644 --- a/val.py +++ b/val.py @@ -365,6 +365,8 @@ def main(opt): if opt.task in ('train', 'val', 'test'): # run normally if opt.conf_thres > 0.001: # https://github.com/ultralytics/yolov5/issues/1466 LOGGER.info(f'WARNING: confidence threshold {opt.conf_thres} > 0.001 produces invalid results ⚠️') + if opt.save_hybrid: + LOGGER.info('WARNING: --save-hybrid will return high mAP from hybrid labels, not from predictions alone ⚠️') run(**vars(opt)) else: From c725511bfc14eb86daf6edefa0d257084aa24c85 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 01:34:03 +0200 Subject: [PATCH 494/661] Refactor for simplification (#9054) * Refactor for simplification * cleanup * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/downloads.py | 2 +- utils/general.py | 5 +++-- utils/metrics.py | 2 +- utils/plots.py | 8 +++----- utils/torch_utils.py | 11 +++++------ 5 files changed, 13 insertions(+), 15 deletions(-) diff --git a/utils/downloads.py b/utils/downloads.py index c4d4a85c38ae..69887a579966 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -46,7 +46,7 @@ def safe_download(file, url, url2=None, min_bytes=1E0, error_msg=''): except Exception as e: # url2 file.unlink(missing_ok=True) # remove partial downloads LOGGER.info(f'ERROR: {e}\nRe-attempting {url2 or url} to {file}...') - os.system(f"curl -L '{url2 or url}' -o '{file}' --retry 3 -C -") # curl download, retry and resume on fail + os.system(f"curl -# -L '{url2 or url}' -o '{file}' --retry 3 -C -") # curl download, retry and resume on fail finally: if not file.exists() or file.stat().st_size < min_bytes: # check file.unlink(missing_ok=True) # remove partial downloads diff --git a/utils/general.py b/utils/general.py index 42d000918c13..d9f436a36359 100755 --- a/utils/general.py +++ b/utils/general.py @@ -582,7 +582,7 @@ def url2file(url): def download(url, dir='.', unzip=True, delete=True, curl=False, threads=1, retry=3): - # Multi-threaded file download and unzip function, used in data.yaml for autodownload + # Multithreaded file download and unzip function, used in data.yaml for autodownload def download_one(url, dir): # Download 1 file success = True @@ -594,7 +594,8 @@ def download_one(url, dir): for i in range(retry + 1): if curl: s = 'sS' if threads > 1 else '' # silent - r = os.system(f'curl -{s}L "{url}" -o "{f}" --retry 9 -C -') # curl download with retry, continue + r = os.system( + f'curl -# -{s}L "{url}" -o "{f}" --retry 9 -C -') # curl download with retry, continue success = r == 0 else: torch.hub.download_url_to_file(url, f, progress=threads == 1) # torch download diff --git a/utils/metrics.py b/utils/metrics.py index 08880cd3f212..8fa3c7e217c7 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -141,7 +141,7 @@ def process_batch(self, detections, labels): """ if detections is None: gt_classes = labels.int() - for i, gc in enumerate(gt_classes): + for gc in gt_classes: self.matrix[self.nc, gc] += 1 # background FN return diff --git a/utils/plots.py b/utils/plots.py index 7417308c4d82..2c7a80b4c872 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -3,6 +3,7 @@ Plotting utils """ +import contextlib import math import os from copy import copy @@ -180,8 +181,7 @@ def output_to_target(output): # Convert model output to target format [batch_id, class_id, x, y, w, h, conf] targets = [] for i, o in enumerate(output): - for *box, conf, cls in o.cpu().numpy(): - targets.append([i, cls, *list(*xyxy2xywh(np.array(box)[None])), conf]) + targets.extend([i, cls, *list(*xyxy2xywh(np.array(box)[None])), conf] for *box, conf, cls in o.cpu().numpy()) return np.array(targets) @@ -357,10 +357,8 @@ def plot_labels(labels, names=(), save_dir=Path('')): matplotlib.use('svg') # faster ax = plt.subplots(2, 2, figsize=(8, 8), tight_layout=True)[1].ravel() y = ax[0].hist(c, bins=np.linspace(0, nc, nc + 1) - 0.5, rwidth=0.8) - try: # color histogram bars by class + with contextlib.suppress(Exception): # color histogram bars by class [y[2].patches[i].set_color([x / 255 for x in colors(i)]) for i in range(nc)] # known issue #3195 - except Exception: - pass ax[0].set_ylabel('instances') if 0 < len(names) < 30: ax[0].set_xticks(range(len(names))) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 4de2520b26a2..88108906bfd3 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -45,11 +45,10 @@ def decorate(fn): def smartCrossEntropyLoss(label_smoothing=0.0): # Returns nn.CrossEntropyLoss with label smoothing enabled for torch>=1.10.0 if check_version(torch.__version__, '1.10.0'): - return nn.CrossEntropyLoss(label_smoothing=label_smoothing) # loss function - else: - if label_smoothing > 0: - LOGGER.warning(f'WARNING: label smoothing {label_smoothing} requires torch>=1.10.0') - return nn.CrossEntropyLoss() # loss function + return nn.CrossEntropyLoss(label_smoothing=label_smoothing) + if label_smoothing > 0: + LOGGER.warning(f'WARNING: label smoothing {label_smoothing} requires torch>=1.10.0') + return nn.CrossEntropyLoss() def smart_DDP(model): @@ -118,7 +117,7 @@ def select_device(device='', batch_size=0, newline=True): assert torch.cuda.is_available() and torch.cuda.device_count() >= len(device.replace(',', '')), \ f"Invalid CUDA '--device {device}' requested, use '--device cpu' or pass valid CUDA device(s)" - if not (cpu or mps) and torch.cuda.is_available(): # prefer GPU if available + if not cpu and not mps and torch.cuda.is_available(): # prefer GPU if available devices = device.split(',') if device else '0' # range(torch.cuda.device_count()) # i.e. 0,1,6,7 n = len(devices) # device count if n > 1 and batch_size > 0: # check batch_size is divisible by device_count From 93f63ee33f2dd2fe9e61268464c9a79f30aa7549 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 02:00:35 +0200 Subject: [PATCH 495/661] Refactor for simplification 2 (#9055) * Refactor for simplification 2 * Update __init__.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- export.py | 3 +-- utils/loggers/__init__.py | 20 +++++++------------- 2 files changed, 8 insertions(+), 15 deletions(-) diff --git a/export.py b/export.py index 7b398fdc4d93..166b5f406a20 100644 --- a/export.py +++ b/export.py @@ -436,8 +436,7 @@ def export_tfjs(file, prefix=colorstr('TensorFlow.js:')): f'--output_node_names=Identity,Identity_1,Identity_2,Identity_3 {f_pb} {f}' subprocess.run(cmd.split()) - with open(f_json) as j: - json = j.read() + json = Path(f_json).read_text() with open(f_json, 'w') as j: # sort JSON Identity_* in ascending order subst = re.sub( r'{"outputs": {"Identity.?.?": {"name": "Identity.?.?"}, ' diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 34704b625294..b95a463717f8 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -187,18 +187,16 @@ def on_fit_epoch_end(self, vals, epoch, best_fitness, fi): def on_model_save(self, last, epoch, final_epoch, best_fitness, fi): # Callback runs on model save event - if self.wandb: - if ((epoch + 1) % self.opt.save_period == 0 and not final_epoch) and self.opt.save_period != -1: + if (epoch + 1) % self.opt.save_period == 0 and not final_epoch and self.opt.save_period != -1: + if self.wandb: self.wandb.log_model(last.parent, self.opt, epoch, fi, best_model=best_fitness == fi) - - if self.clearml: - if ((epoch + 1) % self.opt.save_period == 0 and not final_epoch) and self.opt.save_period != -1: + if self.clearml: self.clearml.task.update_output_model(model_path=str(last), model_name='Latest Model', auto_delete_file=False) def on_train_end(self, last, best, plots, epoch, results): - # Callback runs on training end + # Callback runs on training end, i.e. saving best model if plots: plot_results(file=self.save_dir / 'results.csv') # save results.png files = ['results.png', 'confusion_matrix.png', *(f'{x}_curve.png' for x in ('F1', 'PR', 'P', 'R'))] @@ -220,15 +218,11 @@ def on_train_end(self, last, best, plots, epoch, results): aliases=['latest', 'best', 'stripped']) self.wandb.finish_run() - if self.clearml: - # Save the best model here - if not self.opt.evolve: - self.clearml.task.update_output_model(model_path=str(best if best.exists() else last), - name='Best Model') + if self.clearml and not self.opt.evolve: + self.clearml.task.update_output_model(model_path=str(best if best.exists() else last), name='Best Model') - def on_params_update(self, params): + def on_params_update(self, params: dict): # Update hyperparams or configs of the experiment - # params: A dict containing {param: value} pairs if self.wandb: self.wandb.wandb_run.config.update(params, allow_val_change=True) From 841f312f9384d3ab8f2ff2ae287441ecfba03740 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 02:23:39 +0200 Subject: [PATCH 496/661] zero-mAP fix return `.detach()` to EMA (#9056) Resolves https://github.com/ultralytics/hub/issues/82 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/torch_utils.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 88108906bfd3..b934248dee43 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -422,7 +422,7 @@ def update(self, model): for k, v in self.ema.state_dict().items(): if v.dtype.is_floating_point: # true for FP16 and FP32 v *= d - v += (1 - d) * msd[k] + v += (1 - d) * msd[k].detach() assert v.dtype == msd[k].dtype == torch.float32, f'EMA {v.dtype} and model {msd[k]} must be updated in FP32' def update_attr(self, model, include=(), exclude=('process_group', 'reducer')): From 27fb6fd8fc21c20290041f38046d7a60ae8c6e3a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 03:22:24 +0200 Subject: [PATCH 497/661] zero-mAP fix 3 (#9058) * zero-mAP fix 3 Signed-off-by: Glenn Jocher * Update torch_utils.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update torch_utils.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/torch_utils.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index b934248dee43..5fbe8bbf10f6 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -412,7 +412,6 @@ def __init__(self, model, decay=0.9999, tau=2000, updates=0): for p in self.ema.parameters(): p.requires_grad_(False) - @smart_inference_mode() def update(self, model): # Update EMA parameters self.updates += 1 @@ -423,7 +422,7 @@ def update(self, model): if v.dtype.is_floating_point: # true for FP16 and FP32 v *= d v += (1 - d) * msd[k].detach() - assert v.dtype == msd[k].dtype == torch.float32, f'EMA {v.dtype} and model {msd[k]} must be updated in FP32' + assert v.dtype == msd[k].detach().dtype == torch.float32, f'EMA {v.dtype} and model {msd[k]} must both be FP32' def update_attr(self, model, include=(), exclude=('process_group', 'reducer')): # Update EMA attributes From e0700cce776c557e7cee51103c53032b766f224a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 03:47:37 +0200 Subject: [PATCH 498/661] Daemon `plot_labels()` for faster start (#9057) * Daemon `plot_labels()` for faster start * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update train.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- train.py | 10 +++------- utils/callbacks.py | 13 +++++++++---- utils/general.py | 2 +- utils/loggers/__init__.py | 12 +++++++----- utils/plots.py | 1 - 5 files changed, 20 insertions(+), 18 deletions(-) diff --git a/train.py b/train.py index 665b4f5b609e..0bfcaffc16db 100644 --- a/train.py +++ b/train.py @@ -52,7 +52,7 @@ from utils.loggers.wandb.wandb_utils import check_wandb_resume from utils.loss import ComputeLoss from utils.metrics import fitness -from utils.plots import plot_evolve, plot_labels +from utils.plots import plot_evolve from utils.torch_utils import (EarlyStopping, ModelEMA, de_parallel, select_device, smart_DDP, smart_optimizer, smart_resume, torch_distributed_zero_first) @@ -215,15 +215,11 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio prefix=colorstr('val: '))[0] if not resume: - if plots: - plot_labels(labels, names, save_dir) - - # Anchors if not opt.noautoanchor: - check_anchors(dataset, model=model, thr=hyp['anchor_t'], imgsz=imgsz) + check_anchors(dataset, model=model, thr=hyp['anchor_t'], imgsz=imgsz) # run AutoAnchor model.half().float() # pre-reduce anchor precision - callbacks.run('on_pretrain_routine_end') + callbacks.run('on_pretrain_routine_end', labels, names, plots) # DDP mode if cuda and RANK != -1: diff --git a/utils/callbacks.py b/utils/callbacks.py index 2b32df0bf1c1..166d8938322d 100644 --- a/utils/callbacks.py +++ b/utils/callbacks.py @@ -3,6 +3,8 @@ Callback utils """ +import threading + class Callbacks: """" @@ -55,17 +57,20 @@ def get_registered_actions(self, hook=None): """ return self._callbacks[hook] if hook else self._callbacks - def run(self, hook, *args, **kwargs): + def run(self, hook, *args, thread=False, **kwargs): """ - Loop through the registered actions and fire all callbacks + Loop through the registered actions and fire all callbacks on main thread Args: hook: The name of the hook to check, defaults to all args: Arguments to receive from YOLOv5 + thread: (boolean) Run callbacks in daemon thread kwargs: Keyword Arguments to receive from YOLOv5 """ assert hook in self._callbacks, f"hook '{hook}' not found in callbacks {self._callbacks}" - for logger in self._callbacks[hook]: - logger['callback'](*args, **kwargs) + if thread: + threading.Thread(target=logger['callback'], args=args, kwargs=kwargs, daemon=True).start() + else: + logger['callback'](*args, **kwargs) diff --git a/utils/general.py b/utils/general.py index d9f436a36359..3bc6fbc22d57 100755 --- a/utils/general.py +++ b/utils/general.py @@ -622,7 +622,7 @@ def download_one(url, dir): dir.mkdir(parents=True, exist_ok=True) # make directory if threads > 1: pool = ThreadPool(threads) - pool.imap(lambda x: download_one(*x), zip(url, repeat(dir))) # multi-threaded + pool.imap(lambda x: download_one(*x), zip(url, repeat(dir))) # multithreaded pool.close() pool.join() else: diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index b95a463717f8..c5cdd92772f2 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -11,10 +11,10 @@ import torch from torch.utils.tensorboard import SummaryWriter -from utils.general import colorstr, cv2 +from utils.general import colorstr, cv2, threaded from utils.loggers.clearml.clearml_utils import ClearmlLogger from utils.loggers.wandb.wandb_utils import WandbLogger -from utils.plots import plot_images, plot_results +from utils.plots import plot_images, plot_labels, plot_results from utils.torch_utils import de_parallel LOGGERS = ('csv', 'tb', 'wandb', 'clearml') # *.csv, TensorBoard, Weights & Biases, ClearML @@ -110,13 +110,15 @@ def on_train_start(self): # Callback runs on train start pass - def on_pretrain_routine_end(self): + def on_pretrain_routine_end(self, labels, names, plots): # Callback runs on pre-train routine end + if plots: + plot_labels(labels, names, self.save_dir) paths = self.save_dir.glob('*labels*.jpg') # training labels if self.wandb: self.wandb.log({"Labels": [wandb.Image(str(x), caption=x.name) for x in paths]}) - if self.clearml: - pass # ClearML saves these images automatically using hooks + # if self.clearml: + # pass # ClearML saves these images automatically using hooks def on_train_batch_end(self, ni, model, imgs, targets, paths, plots): # Callback runs on train batch end diff --git a/utils/plots.py b/utils/plots.py index 2c7a80b4c872..7e1de43aba1b 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -340,7 +340,6 @@ def plot_val_study(file='', dir='', x=None): # from utils.plots import *; plot_ @try_except # known issue https://github.com/ultralytics/yolov5/issues/5395 -@Timeout(30) # known issue https://github.com/ultralytics/yolov5/issues/5611 def plot_labels(labels, names=(), save_dir=Path('')): # plot dataset labels LOGGER.info(f"Plotting labels to {save_dir / 'labels.jpg'}... ") From 6077bf032aa67b8b849b755aa29c66b2eaaee59e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 13:10:02 +0200 Subject: [PATCH 499/661] TensorBoard fix in tutorial.ipynb (#9064) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- tutorial.ipynb | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 91e2d7e75eab..55e423d72833 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -660,7 +660,7 @@ "#@title Select YOLOv5 🚀 logger\n", "logger = 'TensorBoard' #@param ['TensorBoard', 'ClearML', 'W&B']\n", "\n", - "if logger == 'Tensorboard':\n", + "if logger == 'TensorBoard':\n", " %load_ext tensorboard\n", " %tensorboard --logdir runs/train\n", "elif logger == 'ClearML':\n", @@ -1103,4 +1103,4 @@ "outputs": [] } ] -} \ No newline at end of file +} From 794f117f4bdd02171273d49da33c1e8a22037f76 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 13:33:13 +0200 Subject: [PATCH 500/661] Created using Colaboratory --- tutorial.ipynb | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 55e423d72833..040197bf8365 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -664,12 +664,10 @@ " %load_ext tensorboard\n", " %tensorboard --logdir runs/train\n", "elif logger == 'ClearML':\n", - " %pip install -q clearml\n", - " !clearml-init\n", + " %pip install -q clearml && clearml-init\n", "elif logger == 'W&B':\n", " %pip install -q wandb\n", - " import wandb\n", - " wandb.login()" + " import wandb; wandb.login()" ], "metadata": { "id": "i3oKtE4g-aNn" @@ -1103,4 +1101,4 @@ "outputs": [] } ] -} +} \ No newline at end of file From 1499526f5668f97832abf39c9e24e2acf3f98fdf Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 14:20:12 +0200 Subject: [PATCH 501/661] Created using Colaboratory --- tutorial.ipynb | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 040197bf8365..a8975424cb39 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -534,7 +534,7 @@ }, "source": [ "# Download COCO val\n", - "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip') # download COCO val (1GB - 5000 images)\n", + "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip') # download (780M - 5000 images)\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip # unzip" ], "execution_count": null, @@ -657,7 +657,7 @@ { "cell_type": "code", "source": [ - "#@title Select YOLOv5 🚀 logger\n", + "#@title Select YOLOv5 🚀 logger {run: 'auto'}\n", "logger = 'TensorBoard' #@param ['TensorBoard', 'ClearML', 'W&B']\n", "\n", "if logger == 'TensorBoard':\n", @@ -1077,7 +1077,7 @@ "cell_type": "code", "source": [ "# Validate on COCO test. Zip results.json and submit to eval server at https://competitions.codalab.org/competitions/20794\n", - "!bash data/scripts/get_coco.sh --test # download COCO test-dev2017 (7GB - 40,000 images, test 20,000)\n", + "!bash data/scripts/get_coco.sh --test # download COCO test-dev2017 (7G - 40,000 images, test 20,000)\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half --task test" ], "metadata": { From e6b4bf0bc26c06d54dd92eacef89decdc580a0f5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 14:21:48 +0200 Subject: [PATCH 502/661] Created using Colaboratory --- tutorial.ipynb | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index a8975424cb39..8753a2310b1d 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -1064,7 +1064,7 @@ "cell_type": "code", "source": [ "# Classification val\n", - "!bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images)\n", + "!bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G - 50000 images)\n", "!python classify/val.py --weights yolov5m-cls.pt --data ../datasets/imagenet --img 224 # validate" ], "metadata": { @@ -1077,7 +1077,7 @@ "cell_type": "code", "source": [ "# Validate on COCO test. Zip results.json and submit to eval server at https://competitions.codalab.org/competitions/20794\n", - "!bash data/scripts/get_coco.sh --test # download COCO test-dev2017 (7G - 40,000 images, test 20,000)\n", + "!bash data/scripts/get_coco.sh --test # download COCO test-dev2017 (7G - 40000 images, test 20000)\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half --task test" ], "metadata": { From 87e8deadd563982672a1c5104a68d1a67f0cf765 Mon Sep 17 00:00:00 2001 From: 0zppd <111682241+0zppd@users.noreply.github.com> Date: Sun, 21 Aug 2022 18:40:28 +0500 Subject: [PATCH 503/661] zero-mAP fix remove `torch.empty()` forward pass in `.train()` mode (#9068) * Fix Zero Map Issue Signed-off-by: 0zppd <111682241+0zppd@users.noreply.github.com> * Update __init__.py Signed-off-by: Glenn Jocher Signed-off-by: 0zppd <111682241+0zppd@users.noreply.github.com> Signed-off-by: Glenn Jocher Co-authored-by: Glenn Jocher --- utils/loggers/__init__.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index c5cdd92772f2..b9869df26a43 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -296,7 +296,7 @@ def log_tensorboard_graph(tb, model, imgsz=(640, 640)): try: p = next(model.parameters()) # for device, type imgsz = (imgsz, imgsz) if isinstance(imgsz, int) else imgsz # expand - im = torch.empty((1, 3, *imgsz)).to(p.device).type_as(p) # input image + im = torch.zeros((1, 3, *imgsz)).to(p.device).type_as(p) # input image (WARNING: must be zeros, not empty) with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress jit trace warning tb.add_graph(torch.jit.trace(de_parallel(model), im, strict=False), []) From 0b8639a40a9c73a9ee1556405fabfd2d46087299 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 15:50:02 +0200 Subject: [PATCH 504/661] Rename 'labels' to 'instances' (#9066) * Rename labels to instances * Rename labels to instances * align val --- train.py | 4 ++-- val.py | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/train.py b/train.py index 0bfcaffc16db..ac38d04dba90 100644 --- a/train.py +++ b/train.py @@ -271,7 +271,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio if RANK != -1: train_loader.sampler.set_epoch(epoch) pbar = enumerate(train_loader) - LOGGER.info(('\n' + '%10s' * 7) % ('Epoch', 'gpu_mem', 'box', 'obj', 'cls', 'labels', 'img_size')) + LOGGER.info(('\n' + '%11s' * 7) % ('Epoch', 'GPU_mem', 'box_loss', 'obj_loss', 'cls_loss', 'Instances', 'Size')) if RANK in {-1, 0}: pbar = tqdm(pbar, total=nb, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar optimizer.zero_grad() @@ -326,7 +326,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio if RANK in {-1, 0}: mloss = (mloss * i + loss_items) / (i + 1) # update mean losses mem = f'{torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0:.3g}G' # (GB) - pbar.set_description(('%10s' * 2 + '%10.4g' * 5) % + pbar.set_description(('%11s' * 2 + '%11.4g' * 5) % (f'{epoch}/{epochs - 1}', mem, *mloss, targets.shape[0], imgs.shape[-1])) callbacks.run('on_train_batch_end', ni, model, imgs, targets, paths, plots) if callbacks.stop_training: diff --git a/val.py b/val.py index fcaca889d7e2..f9557bba651d 100644 --- a/val.py +++ b/val.py @@ -186,7 +186,7 @@ def run( if isinstance(names, (list, tuple)): # old format names = dict(enumerate(names)) class_map = coco80_to_coco91_class() if is_coco else list(range(1000)) - s = ('%20s' + '%11s' * 6) % ('Class', 'Images', 'Labels', 'P', 'R', 'mAP@.5', 'mAP@.5:.95') + s = ('%22s' + '%11s' * 6) % ('Class', 'Images', 'Instances', 'P', 'R', 'mAP@.5', 'mAP@.5:.95') dt, p, r, f1, mp, mr, map50, map = (Profile(), Profile(), Profile()), 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 loss = torch.zeros(3, device=device) jdict, stats, ap, ap_class = [], [], [], [] @@ -270,7 +270,7 @@ def run( nt = np.bincount(stats[3].astype(int), minlength=nc) # number of targets per class # Print results - pf = '%20s' + '%11i' * 2 + '%11.3g' * 4 # print format + pf = '%22s' + '%11i' * 2 + '%11.3g' * 4 # print format LOGGER.info(pf % ('all', seen, nt.sum(), mp, mr, map50, map)) if nt.sum() == 0: LOGGER.warning(f'WARNING: no labels found in {task} set, can not compute metrics without labels ⚠️') From 8665d557c1caa66c190c1ec26b377eeae385d1d6 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 16:51:50 +0200 Subject: [PATCH 505/661] Threaded TensorBoard graph logging (#9070) * Log TensorBoard graph on pretrain_routine_end * fix --- train.py | 6 +++--- utils/loggers/__init__.py | 34 ++++++++++++++++++---------------- 2 files changed, 21 insertions(+), 19 deletions(-) diff --git a/train.py b/train.py index ac38d04dba90..e4c9b6ae6749 100644 --- a/train.py +++ b/train.py @@ -219,7 +219,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio check_anchors(dataset, model=model, thr=hyp['anchor_t'], imgsz=imgsz) # run AutoAnchor model.half().float() # pre-reduce anchor precision - callbacks.run('on_pretrain_routine_end', labels, names, plots) + callbacks.run('on_pretrain_routine_end', labels, names) # DDP mode if cuda and RANK != -1: @@ -328,7 +328,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio mem = f'{torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0:.3g}G' # (GB) pbar.set_description(('%11s' * 2 + '%11.4g' * 5) % (f'{epoch}/{epochs - 1}', mem, *mloss, targets.shape[0], imgs.shape[-1])) - callbacks.run('on_train_batch_end', ni, model, imgs, targets, paths, plots) + callbacks.run('on_train_batch_end', model, ni, imgs, targets, paths) if callbacks.stop_training: return # end batch ------------------------------------------------------------------------------------------------ @@ -420,7 +420,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio if is_coco: callbacks.run('on_fit_epoch_end', list(mloss) + list(results) + lr, epoch, best_fitness, fi) - callbacks.run('on_train_end', last, best, plots, epoch, results) + callbacks.run('on_train_end', last, best, epoch, results) torch.cuda.empty_cache() return results diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index b9869df26a43..98a123eee74d 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -49,6 +49,7 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, self.weights = weights self.opt = opt self.hyp = hyp + self.plots = not opt.noplots # plot results self.logger = logger # for printing results to console self.include = include self.keys = [ @@ -110,26 +111,26 @@ def on_train_start(self): # Callback runs on train start pass - def on_pretrain_routine_end(self, labels, names, plots): + def on_pretrain_routine_end(self, labels, names): # Callback runs on pre-train routine end - if plots: + if self.plots: plot_labels(labels, names, self.save_dir) - paths = self.save_dir.glob('*labels*.jpg') # training labels - if self.wandb: - self.wandb.log({"Labels": [wandb.Image(str(x), caption=x.name) for x in paths]}) - # if self.clearml: - # pass # ClearML saves these images automatically using hooks + paths = self.save_dir.glob('*labels*.jpg') # training labels + if self.wandb: + self.wandb.log({"Labels": [wandb.Image(str(x), caption=x.name) for x in paths]}) + # if self.clearml: + # pass # ClearML saves these images automatically using hooks - def on_train_batch_end(self, ni, model, imgs, targets, paths, plots): + def on_train_batch_end(self, model, ni, imgs, targets, paths): # Callback runs on train batch end # ni: number integrated batches (since train start) - if plots: - if ni == 0 and not self.opt.sync_bn and self.tb: - log_tensorboard_graph(self.tb, model, imgsz=list(imgs.shape[2:4])) + if self.plots: if ni < 3: f = self.save_dir / f'train_batch{ni}.jpg' # filename plot_images(imgs, targets, paths, f) - if (self.wandb or self.clearml) and ni == 10: + if ni == 0 and self.tb and not self.opt.sync_bn: + log_tensorboard_graph(self.tb, model, imgsz=(self.opt.imgsz, self.opt.imgsz)) + if ni == 10 and (self.wandb or self.clearml): files = sorted(self.save_dir.glob('train*.jpg')) if self.wandb: self.wandb.log({'Mosaics': [wandb.Image(str(f), caption=f.name) for f in files if f.exists()]}) @@ -197,9 +198,9 @@ def on_model_save(self, last, epoch, final_epoch, best_fitness, fi): model_name='Latest Model', auto_delete_file=False) - def on_train_end(self, last, best, plots, epoch, results): + def on_train_end(self, last, best, epoch, results): # Callback runs on training end, i.e. saving best model - if plots: + if self.plots: plot_results(file=self.save_dir / 'results.csv') # save results.png files = ['results.png', 'confusion_matrix.png', *(f'{x}_curve.png' for x in ('F1', 'PR', 'P', 'R'))] files = [(self.save_dir / f) for f in files if (self.save_dir / f).exists()] # filter @@ -291,6 +292,7 @@ def log_model(self, model_path, epoch=0, metadata={}): wandb.log_artifact(art) +@threaded def log_tensorboard_graph(tb, model, imgsz=(640, 640)): # Log model graph to TensorBoard try: @@ -300,5 +302,5 @@ def log_tensorboard_graph(tb, model, imgsz=(640, 640)): with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress jit trace warning tb.add_graph(torch.jit.trace(de_parallel(model), im, strict=False), []) - except Exception: - print('WARNING: TensorBoard graph visualization failure') + except Exception as e: + print(f'WARNING: TensorBoard graph visualization failure {e}') From 5373a28c1bcede65e513b7be0ab5a0d43125c90c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 17:01:05 +0200 Subject: [PATCH 506/661] Created using Colaboratory --- tutorial.ipynb | 451 ++++++++++++++++++++++++++----------------------- 1 file changed, 243 insertions(+), 208 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 8753a2310b1d..5b7b1f287d7e 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -17,110 +17,121 @@ "accelerator": "GPU", "widgets": { "application/vnd.jupyter.widget-state+json": { - "6d6b90ead2db49b3bdf624b6ba9b44e9": { + "da0946bcefd9414fa282977f7f609e36": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", - "model_module_version": "1.5.0", + "model_module_version": "2.0.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "1.5.0", + "_model_module_version": "2.0.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", - "_view_module_version": "1.5.0", + "_view_module_version": "2.0.0", "_view_name": "HBoxView", "box_style": "", "children": [ - "IPY_MODEL_cb77443edb9e43328a56aaa4413a0df3", - "IPY_MODEL_954c8b8699e143bf92be6bfc02fc52f6", - "IPY_MODEL_a64775946e13477f83d8bba6086385b9" + "IPY_MODEL_7838c0af44244ccc906c413cea0989d7", + "IPY_MODEL_309ea78b3e814198b4080beb878d5329", + "IPY_MODEL_b2d1d998e5db4ca1a36280902e1647c7" ], - "layout": "IPY_MODEL_1413611b7f4f4ef99e4f541f5ca35ed6" + "layout": "IPY_MODEL_e7d7f56c77884717ba122f1d603c0852", + "tabbable": null, + "tooltip": null } }, - "cb77443edb9e43328a56aaa4413a0df3": { + "7838c0af44244ccc906c413cea0989d7": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", - "model_module_version": "1.5.0", + "model_module_version": "2.0.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "1.5.0", + "_model_module_version": "2.0.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", - "_view_module_version": "1.5.0", + "_view_module_version": "2.0.0", "_view_name": "HTMLView", "description": "", - "description_tooltip": null, - "layout": "IPY_MODEL_00737f5558eb4fbd968172acb978e54a", + "description_allow_html": false, + "layout": "IPY_MODEL_abf60d6b8ea847f9bb358ae2b045458b", "placeholder": "​", - "style": "IPY_MODEL_f03e5ddfd1c04bedaf68ab02c3f6f0ea", + "style": "IPY_MODEL_379196a2761b4a29aca8ef088dc60c10", + "tabbable": null, + "tooltip": null, "value": "100%" } }, - "954c8b8699e143bf92be6bfc02fc52f6": { + "309ea78b3e814198b4080beb878d5329": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", - "model_module_version": "1.5.0", + "model_module_version": "2.0.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "1.5.0", + "_model_module_version": "2.0.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", - "_view_module_version": "1.5.0", + "_view_module_version": "2.0.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", - "description_tooltip": null, - "layout": "IPY_MODEL_6926db7e0035455f99e1dd4508c4b19c", + "description_allow_html": false, + "layout": "IPY_MODEL_52b546a356e54174a95049b30cb52c81", "max": 818322941, "min": 0, "orientation": "horizontal", - "style": "IPY_MODEL_a6a52c9f828b458e97ddf7a11ae9275f", + "style": "IPY_MODEL_0889e134327e4aa0a8719d03a0d6941b", + "tabbable": null, + "tooltip": null, "value": 818322941 } }, - "a64775946e13477f83d8bba6086385b9": { + "b2d1d998e5db4ca1a36280902e1647c7": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", - "model_module_version": "1.5.0", + "model_module_version": "2.0.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "1.5.0", + "_model_module_version": "2.0.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", - "_view_module_version": "1.5.0", + "_view_module_version": "2.0.0", "_view_name": "HTMLView", "description": "", - "description_tooltip": null, - "layout": "IPY_MODEL_c4c7dc45a1c24dc4b2c709e21271a37e", + "description_allow_html": false, + "layout": "IPY_MODEL_30f22a3e42d24f10ad9851f40a6703f3", "placeholder": "​", - "style": "IPY_MODEL_09c43ffe2c7e4bdc9489e83f9d82ab73", - "value": " 780M/780M [01:12<00:00, 23.8MB/s]" + "style": "IPY_MODEL_648b3512bb7d4ccca5d75af36c133e92", + "tabbable": null, + "tooltip": null, + "value": " 780M/780M [01:31<00:00, 12.3MB/s]" } }, - "1413611b7f4f4ef99e4f541f5ca35ed6": { + "e7d7f56c77884717ba122f1d603c0852": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", - "model_module_version": "1.2.0", + "model_module_version": "2.0.0", "state": { "_model_module": "@jupyter-widgets/base", - "_model_module_version": "1.2.0", + "_model_module_version": "2.0.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "1.2.0", + "_view_module_version": "2.0.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, - "border": null, + "border_bottom": null, + "border_left": null, + "border_right": null, + "border_top": null, "bottom": null, "display": null, "flex": null, @@ -148,8 +159,6 @@ "object_position": null, "order": null, "overflow": null, - "overflow_x": null, - "overflow_y": null, "padding": null, "right": null, "top": null, @@ -157,22 +166,25 @@ "width": null } }, - "00737f5558eb4fbd968172acb978e54a": { + "abf60d6b8ea847f9bb358ae2b045458b": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", - "model_module_version": "1.2.0", + "model_module_version": "2.0.0", "state": { "_model_module": "@jupyter-widgets/base", - "_model_module_version": "1.2.0", + "_model_module_version": "2.0.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "1.2.0", + "_view_module_version": "2.0.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, - "border": null, + "border_bottom": null, + "border_left": null, + "border_right": null, + "border_top": null, "bottom": null, "display": null, "flex": null, @@ -200,8 +212,6 @@ "object_position": null, "order": null, "overflow": null, - "overflow_x": null, - "overflow_y": null, "padding": null, "right": null, "top": null, @@ -209,37 +219,43 @@ "width": null } }, - "f03e5ddfd1c04bedaf68ab02c3f6f0ea": { + "379196a2761b4a29aca8ef088dc60c10": { "model_module": "@jupyter-widgets/controls", - "model_name": "DescriptionStyleModel", - "model_module_version": "1.5.0", + "model_name": "HTMLStyleModel", + "model_module_version": "2.0.0", "state": { "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "1.5.0", - "_model_name": "DescriptionStyleModel", + "_model_module_version": "2.0.0", + "_model_name": "HTMLStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "1.2.0", + "_view_module_version": "2.0.0", "_view_name": "StyleView", - "description_width": "" + "background": null, + "description_width": "", + "font_size": null, + "text_color": null } }, - "6926db7e0035455f99e1dd4508c4b19c": { + "52b546a356e54174a95049b30cb52c81": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", - "model_module_version": "1.2.0", + "model_module_version": "2.0.0", "state": { "_model_module": "@jupyter-widgets/base", - "_model_module_version": "1.2.0", + "_model_module_version": "2.0.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "1.2.0", + "_view_module_version": "2.0.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, - "border": null, + "border_bottom": null, + "border_left": null, + "border_right": null, + "border_top": null, "bottom": null, "display": null, "flex": null, @@ -267,8 +283,6 @@ "object_position": null, "order": null, "overflow": null, - "overflow_x": null, - "overflow_y": null, "padding": null, "right": null, "top": null, @@ -276,38 +290,41 @@ "width": null } }, - "a6a52c9f828b458e97ddf7a11ae9275f": { + "0889e134327e4aa0a8719d03a0d6941b": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", - "model_module_version": "1.5.0", + "model_module_version": "2.0.0", "state": { "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "1.5.0", + "_model_module_version": "2.0.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "1.2.0", + "_view_module_version": "2.0.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, - "c4c7dc45a1c24dc4b2c709e21271a37e": { + "30f22a3e42d24f10ad9851f40a6703f3": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", - "model_module_version": "1.2.0", + "model_module_version": "2.0.0", "state": { "_model_module": "@jupyter-widgets/base", - "_model_module_version": "1.2.0", + "_model_module_version": "2.0.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "1.2.0", + "_view_module_version": "2.0.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, - "border": null, + "border_bottom": null, + "border_left": null, + "border_right": null, + "border_top": null, "bottom": null, "display": null, "flex": null, @@ -335,8 +352,6 @@ "object_position": null, "order": null, "overflow": null, - "overflow_x": null, - "overflow_y": null, "padding": null, "right": null, "top": null, @@ -344,19 +359,22 @@ "width": null } }, - "09c43ffe2c7e4bdc9489e83f9d82ab73": { + "648b3512bb7d4ccca5d75af36c133e92": { "model_module": "@jupyter-widgets/controls", - "model_name": "DescriptionStyleModel", - "model_module_version": "1.5.0", + "model_name": "HTMLStyleModel", + "model_module_version": "2.0.0", "state": { "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "1.5.0", - "_model_name": "DescriptionStyleModel", + "_model_module_version": "2.0.0", + "_model_name": "HTMLStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "1.2.0", + "_view_module_version": "2.0.0", "_view_name": "StyleView", - "description_width": "" + "background": null, + "description_width": "", + "font_size": null, + "text_color": null } } } @@ -404,7 +422,7 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "508de90c-846e-495d-c7d6-50681af62a98" + "outputId": "4200fd6f-c6f5-4505-a4f9-a918f3ed1f86" }, "source": [ "!git clone https://github.com/ultralytics/yolov5 # clone\n", @@ -415,13 +433,13 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": null, + "execution_count": 1, "outputs": [ { "output_type": "stream", "name": "stderr", "text": [ - "YOLOv5 🚀 v6.2-15-g61adf01 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" + "YOLOv5 🚀 v6.2-41-g8665d55 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" ] }, { @@ -461,29 +479,29 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "93881540-331e-4890-cd38-4c2776933238" + "outputId": "1af15107-bcd1-4e8f-b5bd-0ee1a737e051" }, "source": [ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", "# display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": null, + "execution_count": 2, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mdetect: \u001b[0mweights=['yolov5s.pt'], source=data/images, data=data/coco128.yaml, imgsz=[640, 640], conf_thres=0.25, iou_thres=0.45, max_det=1000, device=, view_img=False, save_txt=False, save_conf=False, save_crop=False, nosave=False, classes=None, agnostic_nms=False, augment=False, visualize=False, update=False, project=runs/detect, name=exp, exist_ok=False, line_thickness=3, hide_labels=False, hide_conf=False, half=False, dnn=False\n", - "YOLOv5 🚀 v6.2-15-g61adf01 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-41-g8665d55 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s.pt to yolov5s.pt...\n", - "100% 14.1M/14.1M [00:00<00:00, 39.3MB/s]\n", + "100% 14.1M/14.1M [00:00<00:00, 41.7MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5s summary: 213 layers, 7225885 parameters, 0 gradients\n", - "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, 14.9ms\n", - "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, 22.0ms\n", - "Speed: 0.6ms pre-process, 18.4ms inference, 24.1ms NMS per image at shape (1, 3, 640, 640)\n", + "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, 14.5ms\n", + "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, 18.9ms\n", + "Speed: 0.5ms pre-process, 16.7ms inference, 21.4ms NMS per image at shape (1, 3, 640, 640)\n", "Results saved to \u001b[1mruns/detect/exp\u001b[0m\n" ] } @@ -515,29 +533,29 @@ "id": "WQPtK1QYVaD_", "colab": { "base_uri": "https://localhost:8080/", - "height": 49, + "height": 17, "referenced_widgets": [ - "6d6b90ead2db49b3bdf624b6ba9b44e9", - "cb77443edb9e43328a56aaa4413a0df3", - "954c8b8699e143bf92be6bfc02fc52f6", - "a64775946e13477f83d8bba6086385b9", - "1413611b7f4f4ef99e4f541f5ca35ed6", - "00737f5558eb4fbd968172acb978e54a", - "f03e5ddfd1c04bedaf68ab02c3f6f0ea", - "6926db7e0035455f99e1dd4508c4b19c", - "a6a52c9f828b458e97ddf7a11ae9275f", - "c4c7dc45a1c24dc4b2c709e21271a37e", - "09c43ffe2c7e4bdc9489e83f9d82ab73" + "da0946bcefd9414fa282977f7f609e36", + "7838c0af44244ccc906c413cea0989d7", + "309ea78b3e814198b4080beb878d5329", + "b2d1d998e5db4ca1a36280902e1647c7", + "e7d7f56c77884717ba122f1d603c0852", + "abf60d6b8ea847f9bb358ae2b045458b", + "379196a2761b4a29aca8ef088dc60c10", + "52b546a356e54174a95049b30cb52c81", + "0889e134327e4aa0a8719d03a0d6941b", + "30f22a3e42d24f10ad9851f40a6703f3", + "648b3512bb7d4ccca5d75af36c133e92" ] }, - "outputId": "ed2ca46e-a1a9-4a16-c449-859278d8aa18" + "outputId": "5f129105-eca5-4f33-fb1d-981255f814ad" }, "source": [ "# Download COCO val\n", "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip') # download (780M - 5000 images)\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip # unzip" ], - "execution_count": null, + "execution_count": 3, "outputs": [ { "output_type": "display_data", @@ -548,7 +566,24 @@ "application/vnd.jupyter.widget-view+json": { "version_major": 2, "version_minor": 0, - "model_id": "6d6b90ead2db49b3bdf624b6ba9b44e9" + "model_id": "da0946bcefd9414fa282977f7f609e36" + }, + "application/json": { + "n": 0, + "total": 818322941, + "elapsed": 0.020366430282592773, + "ncols": null, + "nrows": null, + "prefix": "", + "ascii": false, + "unit": "B", + "unit_scale": true, + "rate": null, + "bar_format": null, + "postfix": null, + "unit_divisor": 1024, + "initial": 0, + "colour": null } }, "metadata": {} @@ -562,48 +597,48 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "19a590ef-363e-424c-d9ce-78bbe0593cd5" + "outputId": "40d5d000-abee-46a0-c07d-1066e1662e01" }, "source": [ "# Validate YOLOv5x on COCO val\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" ], - "execution_count": null, + "execution_count": 4, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mval: \u001b[0mdata=/content/yolov5/data/coco.yaml, weights=['yolov5x.pt'], batch_size=32, imgsz=640, conf_thres=0.001, iou_thres=0.65, task=val, device=, workers=8, single_cls=False, augment=False, verbose=False, save_txt=False, save_hybrid=False, save_conf=False, save_json=True, project=runs/val, name=exp, exist_ok=False, half=True, dnn=False\n", - "YOLOv5 🚀 v6.2-15-g61adf01 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-41-g8665d55 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5x.pt to yolov5x.pt...\n", - "100% 166M/166M [00:06<00:00, 28.1MB/s]\n", + "100% 166M/166M [00:10<00:00, 16.6MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5x summary: 444 layers, 86705005 parameters, 0 gradients\n", "Downloading https://ultralytics.com/assets/Arial.ttf to /root/.config/Ultralytics/Arial.ttf...\n", - "100% 755k/755k [00:00<00:00, 47.3MB/s]\n", - "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 10756.32it/s]\n", + "100% 755k/755k [00:00<00:00, 1.39MB/s]\n", + "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 10506.48it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mNew cache created: /content/datasets/coco/val2017.cache\n", - " Class Images Labels P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:07<00:00, 2.33it/s]\n", - " all 5000 36335 0.743 0.625 0.683 0.504\n", - "Speed: 0.1ms pre-process, 4.6ms inference, 1.2ms NMS per image at shape (32, 3, 640, 640)\n", + " Class Images Instances P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:06<00:00, 2.36it/s]\n", + " all 5000 36335 0.743 0.625 0.683 0.504\n", + "Speed: 0.1ms pre-process, 4.6ms inference, 1.1ms NMS per image at shape (32, 3, 640, 640)\n", "\n", "Evaluating pycocotools mAP... saving runs/val/exp/yolov5x_predictions.json...\n", "loading annotations into memory...\n", - "Done (t=0.41s)\n", + "Done (t=0.38s)\n", "creating index...\n", "index created!\n", "Loading and preparing results...\n", - "DONE (t=5.64s)\n", + "DONE (t=5.49s)\n", "creating index...\n", "index created!\n", "Running per image evaluation...\n", "Evaluate annotation type *bbox*\n", - "DONE (t=76.80s).\n", + "DONE (t=72.10s).\n", "Accumulating evaluation results...\n", - "DONE (t=14.61s).\n", + "DONE (t=13.94s).\n", " Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.506\n", " Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.688\n", " Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.549\n", @@ -682,13 +717,13 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "47759d5e-34f0-4a6a-c714-ff533391cfff" + "outputId": "f0ce0354-7f50-4546-f3f9-672b4b522d59" }, "source": [ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": null, + "execution_count": 5, "outputs": [ { "output_type": "stream", @@ -696,7 +731,7 @@ "text": [ "\u001b[34m\u001b[1mtrain: \u001b[0mweights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch-low.yaml, epochs=3, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, noplots=False, evolve=None, bucket=, cache=ram, image_weights=False, device=, multi_scale=False, single_cls=False, optimizer=SGD, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, cos_lr=False, label_smoothing=0.0, patience=100, freeze=[0], save_period=-1, seed=0, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest\n", "\u001b[34m\u001b[1mgithub: \u001b[0mup to date with https://github.com/ultralytics/yolov5 ✅\n", - "YOLOv5 🚀 v6.2-15-g61adf01 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-41-g8665d55 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "\u001b[34m\u001b[1mhyperparameters: \u001b[0mlr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0\n", "\u001b[34m\u001b[1mWeights & Biases: \u001b[0mrun 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs in Weights & Biases\n", @@ -705,8 +740,8 @@ "\n", "Dataset not found ⚠️, missing paths ['/content/datasets/coco128/images/train2017']\n", "Downloading https://ultralytics.com/assets/coco128.zip to coco128.zip...\n", - "100% 6.66M/6.66M [00:00<00:00, 75.3MB/s]\n", - "Dataset download success ✅ (0.7s), saved to \u001b[1m/content/datasets\u001b[0m\n", + "100% 6.66M/6.66M [00:00<00:00, 76.7MB/s]\n", + "Dataset download success ✅ (0.5s), saved to \u001b[1m/content/datasets\u001b[0m\n", "\n", " from n params module arguments \n", " 0 -1 1 3520 models.common.Conv [3, 32, 6, 2, 2] \n", @@ -740,33 +775,33 @@ "\u001b[34m\u001b[1mAMP: \u001b[0mchecks passed ✅\n", "\u001b[34m\u001b[1moptimizer:\u001b[0m SGD(lr=0.01) with parameter groups 57 weight(decay=0.0), 60 weight(decay=0.0005), 60 bias\n", "\u001b[34m\u001b[1malbumentations: \u001b[0mBlur(p=0.01, blur_limit=(3, 7)), MedianBlur(p=0.01, blur_limit=(3, 7)), ToGray(p=0.01), CLAHE(p=0.01, clip_limit=(1, 4.0), tile_grid_size=(8, 8))\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017' images and labels...128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00<00:00, 7246.20it/s]\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017' images and labels...128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00<00:00, 7984.87it/s]\n", "\u001b[34m\u001b[1mtrain: \u001b[0mNew cache created: /content/datasets/coco128/labels/train2017.cache\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mCaching images (0.1GB ram): 100% 128/128 [00:00<00:00, 986.21it/s]\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mCaching images (0.1GB ram): 100% 128/128 [00:00<00:00, 1018.19it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco128/labels/train2017.cache' images and labels... 128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00 Date: Sun, 21 Aug 2022 17:07:56 +0200 Subject: [PATCH 507/661] De-thread TensorBoard graph logging (#9071) * De-thread TensorBoard graph logging Issues with Classification models Signed-off-by: Glenn Jocher * Update __init__.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/loggers/__init__.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 98a123eee74d..006125edbcd9 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -11,7 +11,7 @@ import torch from torch.utils.tensorboard import SummaryWriter -from utils.general import colorstr, cv2, threaded +from utils.general import colorstr, cv2 from utils.loggers.clearml.clearml_utils import ClearmlLogger from utils.loggers.wandb.wandb_utils import WandbLogger from utils.plots import plot_images, plot_labels, plot_results @@ -292,7 +292,6 @@ def log_model(self, model_path, epoch=0, metadata={}): wandb.log_artifact(art) -@threaded def log_tensorboard_graph(tb, model, imgsz=(640, 640)): # Log model graph to TensorBoard try: From 262187e95d304f80abf08abd850b7b5076f2a7a9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 21 Aug 2022 23:26:07 +0200 Subject: [PATCH 508/661] Two dimensional `size=(h,w)` AutoShape support (#9072) * Two dimensional `size=(h,w)` AutoShape support May resolve https://github.com/ultralytics/yolov5/issues/9039 Signed-off-by: Glenn Jocher * Update hubconf.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- hubconf.py | 10 +++++++--- models/common.py | 8 +++++--- 2 files changed, 12 insertions(+), 6 deletions(-) diff --git a/hubconf.py b/hubconf.py index 293f177dcbc1..0a7f917bd7d1 100644 --- a/hubconf.py +++ b/hubconf.py @@ -30,7 +30,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo from models.common import AutoShape, DetectMultiBackend from models.experimental import attempt_load - from models.yolo import DetectionModel + from models.yolo import ClassificationModel, DetectionModel from utils.downloads import attempt_download from utils.general import LOGGER, check_requirements, intersect_dicts, logging from utils.torch_utils import select_device @@ -45,8 +45,12 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo if pretrained and channels == 3 and classes == 80: try: model = DetectMultiBackend(path, device=device, fuse=autoshape) # detection model - if autoshape and isinstance(model.model, DetectionModel): - model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS + if autoshape: + if model.pt and isinstance(model.model, ClassificationModel): + LOGGER.warning('WARNING: YOLOv5 v6.2 ClassificationModel is not yet AutoShape compatible. ' + 'You must pass torch tensors in BCHW to this model, i.e. shape(1,3,224,224).') + else: + model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS except Exception: model = attempt_load(path, device=device, fuse=False) # arbitrary model else: diff --git a/models/common.py b/models/common.py index 44192e622bb5..d308244c4a44 100644 --- a/models/common.py +++ b/models/common.py @@ -589,7 +589,7 @@ def _apply(self, fn): @smart_inference_mode() def forward(self, ims, size=640, augment=False, profile=False): - # Inference from various sources. For height=640, width=1280, RGB images example inputs are: + # Inference from various sources. For size(height=640, width=1280), RGB images example inputs are: # file: ims = 'data/images/zidane.jpg' # str or PosixPath # URI: = 'https://ultralytics.com/images/zidane.jpg' # OpenCV: = cv2.imread('image.jpg')[:,:,::-1] # HWC BGR to RGB x(640,1280,3) @@ -600,6 +600,8 @@ def forward(self, ims, size=640, augment=False, profile=False): dt = (Profile(), Profile(), Profile()) with dt[0]: + if isinstance(size, int): # expand + size = (size, size) p = next(self.model.parameters()) if self.pt else torch.empty(1, device=self.model.device) # param autocast = self.amp and (p.device.type != 'cpu') # Automatic Mixed Precision (AMP) inference if isinstance(ims, torch.Tensor): # torch @@ -622,10 +624,10 @@ def forward(self, ims, size=640, augment=False, profile=False): im = im[..., :3] if im.ndim == 3 else cv2.cvtColor(im, cv2.COLOR_GRAY2BGR) # enforce 3ch input s = im.shape[:2] # HWC shape0.append(s) # image shape - g = (size / max(s)) # gain + g = max(size) / max(s) # gain shape1.append([y * g for y in s]) ims[i] = im if im.data.contiguous else np.ascontiguousarray(im) # update - shape1 = [make_divisible(x, self.stride) if self.pt else size for x in np.array(shape1).max(0)] # inf shape + shape1 = [make_divisible(x, self.stride) for x in np.array(shape1).max(0)] if self.pt else size # inf shape x = [letterbox(im, shape1, auto=False)[0] for im in ims] # pad x = np.ascontiguousarray(np.array(x).transpose((0, 3, 1, 2))) # stack and BHWC to BCHW x = torch.from_numpy(x).to(p.device).type_as(p) / 255 # uint8 to fp16/32 From 0abae780b356aa29332f7d50552e0ed88e38ee3a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 22 Aug 2022 00:04:30 +0200 Subject: [PATCH 509/661] Remove unused Timeout import (#9073) * Remove unused Timeout import Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/plots.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/plots.py b/utils/plots.py index 7e1de43aba1b..d35e2bdd168a 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -19,8 +19,8 @@ import torch from PIL import Image, ImageDraw, ImageFont -from utils.general import (CONFIG_DIR, FONT, LOGGER, Timeout, check_font, check_requirements, clip_coords, - increment_path, is_ascii, threaded, try_except, xywh2xyxy, xyxy2xywh) +from utils.general import (CONFIG_DIR, FONT, LOGGER, check_font, check_requirements, clip_coords, increment_path, + is_ascii, threaded, try_except, xywh2xyxy, xyxy2xywh) from utils.metrics import fitness # Settings From 06831aa9e905e0fa703958f6b3f3db443cf477f3 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 22 Aug 2022 01:06:29 +0200 Subject: [PATCH 510/661] Improved Usage example docstrings (#9075) * Updated Usage examples * Update detect.py Signed-off-by: Glenn Jocher * Update predict.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- classify/predict.py | 23 +++++++++++++++++++---- classify/train.py | 11 +++++++---- classify/val.py | 16 ++++++++++++++-- detect.py | 36 ++++++++++++++++++------------------ export.py | 22 +++++++++++----------- hubconf.py | 4 ++-- models/tf.py | 2 +- models/yolo.py | 2 +- train.py | 17 ++++++++++------- val.py | 24 ++++++++++++------------ 10 files changed, 95 insertions(+), 62 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 0bf99140b8e3..135470fd36ed 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -1,9 +1,24 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ -Run classification inference on file/dir/URL/glob - -Usage: - $ python classify/predict.py --weights yolov5s-cls.pt --source data/images/bus.jpg +Run YOLOv5 classification inference on images, videos, directories, and globs. + +Usage - sources: + $ python classify/predict.py --weights yolov5s.pt --source img.jpg # image + vid.mp4 # video + path/ # directory + 'path/*.jpg' # glob + +Usage - formats: + $ python classify/predict.py --weights yolov5s-cls.pt # PyTorch + yolov5s-cls.torchscript # TorchScript + yolov5s-cls.onnx # ONNX Runtime or OpenCV DNN with --dnn + yolov5s-cls.xml # OpenVINO + yolov5s-cls.engine # TensorRT + yolov5s-cls.mlmodel # CoreML (macOS-only) + yolov5s-cls_saved_model # TensorFlow SavedModel + yolov5s-cls.pb # TensorFlow GraphDef + yolov5s-cls.tflite # TensorFlow Lite + yolov5s-cls_edgetpu.tflite # TensorFlow Edge TPU """ import argparse diff --git a/classify/train.py b/classify/train.py index 8fe90c1b19eb..223367260bad 100644 --- a/classify/train.py +++ b/classify/train.py @@ -1,13 +1,16 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Train a YOLOv5 classifier model on a classification dataset -Datasets: --data mnist, fashion-mnist, cifar10, cifar100, imagenette, imagewoof, imagenet, or 'path/to/custom/dataset' -YOLOv5-cls models: --model yolov5n-cls.pt, yolov5s-cls.pt, yolov5m-cls.pt, yolov5l-cls.pt, yolov5x-cls.pt -Torchvision models: --model resnet50, efficientnet_b0, etc. See https://pytorch.org/vision/stable/models.html -Usage - Single-GPU and Multi-GPU DDP +Usage - Single-GPU training: $ python classify/train.py --model yolov5s-cls.pt --data imagenette160 --epochs 5 --img 128 + +Usage - Multi-GPU DDP training: $ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/train.py --model yolov5s-cls.pt --data imagenet --epochs 5 --img 224 --device 0,1,2,3 + +Datasets: --data mnist, fashion-mnist, cifar10, cifar100, imagenette, imagewoof, imagenet, or 'path/to/data' +YOLOv5-cls models: --model yolov5n-cls.pt, yolov5s-cls.pt, yolov5m-cls.pt, yolov5l-cls.pt, yolov5x-cls.pt +Torchvision models: --model resnet50, efficientnet_b0, etc. See https://pytorch.org/vision/stable/models.html """ import argparse diff --git a/classify/val.py b/classify/val.py index 2353737957d3..bf808bc21a84 100644 --- a/classify/val.py +++ b/classify/val.py @@ -1,10 +1,22 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ -Validate a classification model on a dataset +Validate a trained YOLOv5 classification model on a classification dataset Usage: $ bash data/scripts/get_imagenet.sh --val # download ImageNet val split (6.3G, 50000 images) - $ python classify/val.py --weights yolov5m-cls.pt --data ../datasets/imagenet --img 224 # validate + $ python classify/val.py --weights yolov5m-cls.pt --data ../datasets/imagenet --img 224 # validate ImageNet + +Usage - formats: + $ python classify/val.py --weights yolov5s-cls.pt # PyTorch + yolov5s-cls.torchscript # TorchScript + yolov5s-cls.onnx # ONNX Runtime or OpenCV DNN with --dnn + yolov5s-cls.xml # OpenVINO + yolov5s-cls.engine # TensorRT + yolov5s-cls.mlmodel # CoreML (macOS-only) + yolov5s-cls_saved_model # TensorFlow SavedModel + yolov5s-cls.pb # TensorFlow GraphDef + yolov5s-cls.tflite # TensorFlow Lite + yolov5s-cls_edgetpu.tflite # TensorFlow Edge TPU """ import argparse diff --git a/detect.py b/detect.py index 93ae0baccd13..541ad90e051d 100644 --- a/detect.py +++ b/detect.py @@ -1,27 +1,27 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ -Run inference on images, videos, directories, streams, etc. +Run YOLOv5 detection inference on images, videos, directories, streams, etc. Usage - sources: - $ python path/to/detect.py --weights yolov5s.pt --source 0 # webcam - img.jpg # image - vid.mp4 # video - path/ # directory - 'path/*.jpg' # glob - 'https://youtu.be/Zgi9g1ksQHc' # YouTube - 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP stream + $ python detect.py --weights yolov5s.pt --source 0 # webcam + img.jpg # image + vid.mp4 # video + path/ # directory + 'path/*.jpg' # glob + 'https://youtu.be/Zgi9g1ksQHc' # YouTube + 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP stream Usage - formats: - $ python path/to/detect.py --weights yolov5s.pt # PyTorch - yolov5s.torchscript # TorchScript - yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn - yolov5s.xml # OpenVINO - yolov5s.engine # TensorRT - yolov5s.mlmodel # CoreML (macOS-only) - yolov5s_saved_model # TensorFlow SavedModel - yolov5s.pb # TensorFlow GraphDef - yolov5s.tflite # TensorFlow Lite - yolov5s_edgetpu.tflite # TensorFlow Edge TPU + $ python detect.py --weights yolov5s.pt # PyTorch + yolov5s.torchscript # TorchScript + yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn + yolov5s.xml # OpenVINO + yolov5s.engine # TensorRT + yolov5s.mlmodel # CoreML (macOS-only) + yolov5s_saved_model # TensorFlow SavedModel + yolov5s.pb # TensorFlow GraphDef + yolov5s.tflite # TensorFlow Lite + yolov5s_edgetpu.tflite # TensorFlow Edge TPU """ import argparse diff --git a/export.py b/export.py index 166b5f406a20..7a746156b96d 100644 --- a/export.py +++ b/export.py @@ -21,19 +21,19 @@ $ pip install -r requirements.txt coremltools onnx onnx-simplifier onnxruntime-gpu openvino-dev tensorflow # GPU Usage: - $ python path/to/export.py --weights yolov5s.pt --include torchscript onnx openvino engine coreml tflite ... + $ python export.py --weights yolov5s.pt --include torchscript onnx openvino engine coreml tflite ... Inference: - $ python path/to/detect.py --weights yolov5s.pt # PyTorch - yolov5s.torchscript # TorchScript - yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn - yolov5s.xml # OpenVINO - yolov5s.engine # TensorRT - yolov5s.mlmodel # CoreML (macOS-only) - yolov5s_saved_model # TensorFlow SavedModel - yolov5s.pb # TensorFlow GraphDef - yolov5s.tflite # TensorFlow Lite - yolov5s_edgetpu.tflite # TensorFlow Edge TPU + $ python detect.py --weights yolov5s.pt # PyTorch + yolov5s.torchscript # TorchScript + yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn + yolov5s.xml # OpenVINO + yolov5s.engine # TensorRT + yolov5s.mlmodel # CoreML (macOS-only) + yolov5s_saved_model # TensorFlow SavedModel + yolov5s.pb # TensorFlow GraphDef + yolov5s.tflite # TensorFlow Lite + yolov5s_edgetpu.tflite # TensorFlow Edge TPU TensorFlow.js: $ cd .. && git clone https://github.com/zldrobit/tfjs-yolov5-example.git && cd tfjs-yolov5-example diff --git a/hubconf.py b/hubconf.py index 0a7f917bd7d1..33fc87930582 100644 --- a/hubconf.py +++ b/hubconf.py @@ -1,11 +1,11 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ -PyTorch Hub models https://pytorch.org/hub/ultralytics_yolov5/ +PyTorch Hub models https://pytorch.org/hub/ultralytics_yolov5 Usage: import torch model = torch.hub.load('ultralytics/yolov5', 'yolov5s') - model = torch.hub.load('ultralytics/yolov5:master', 'custom', 'path/to/yolov5s.onnx') # file from branch + model = torch.hub.load('ultralytics/yolov5:master', 'custom', 'path/to/yolov5s.onnx') # custom model from branch """ import torch diff --git a/models/tf.py b/models/tf.py index b0d98cc2a3a9..ecb0d4d79c78 100644 --- a/models/tf.py +++ b/models/tf.py @@ -7,7 +7,7 @@ $ python models/tf.py --weights yolov5s.pt Export: - $ python path/to/export.py --weights yolov5s.pt --include saved_model pb tflite tfjs + $ python export.py --weights yolov5s.pt --include saved_model pb tflite tfjs """ import argparse diff --git a/models/yolo.py b/models/yolo.py index 32a47e9591da..e154b72685b4 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -3,7 +3,7 @@ YOLO-specific modules Usage: - $ python path/to/models/yolo.py --cfg yolov5s.yaml + $ python models/yolo.py --cfg yolov5s.yaml """ import argparse diff --git a/train.py b/train.py index e4c9b6ae6749..0cd4a7f065a6 100644 --- a/train.py +++ b/train.py @@ -1,15 +1,18 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ Train a YOLOv5 model on a custom dataset. - Models and datasets download automatically from the latest YOLOv5 release. -Models: https://github.com/ultralytics/yolov5/tree/master/models -Datasets: https://github.com/ultralytics/yolov5/tree/master/data -Tutorial: https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data -Usage: - $ python path/to/train.py --data coco128.yaml --weights yolov5s.pt --img 640 # from pretrained (RECOMMENDED) - $ python path/to/train.py --data coco128.yaml --weights '' --cfg yolov5s.yaml --img 640 # from scratch +Usage - Single-GPU training: + $ python train.py --data coco128.yaml --weights yolov5s.pt --img 640 # from pretrained (recommended) + $ python train.py --data coco128.yaml --weights '' --cfg yolov5s.yaml --img 640 # from scratch + +Usage - Multi-GPU DDP training: + $ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 train.py --data coco128.yaml --weights yolov5s.pt --img 640 --device 0,1,2,3 + +Models: https://github.com/ultralytics/yolov5/tree/master/models +Datasets: https://github.com/ultralytics/yolov5/tree/master/data +Tutorial: https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data """ import argparse diff --git a/val.py b/val.py index f9557bba651d..58b9c9e1bec0 100644 --- a/val.py +++ b/val.py @@ -1,21 +1,21 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ -Validate a trained YOLOv5 model accuracy on a custom dataset +Validate a trained YOLOv5 detection model on a detection dataset Usage: - $ python path/to/val.py --weights yolov5s.pt --data coco128.yaml --img 640 + $ python val.py --weights yolov5s.pt --data coco128.yaml --img 640 Usage - formats: - $ python path/to/val.py --weights yolov5s.pt # PyTorch - yolov5s.torchscript # TorchScript - yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn - yolov5s.xml # OpenVINO - yolov5s.engine # TensorRT - yolov5s.mlmodel # CoreML (macOS-only) - yolov5s_saved_model # TensorFlow SavedModel - yolov5s.pb # TensorFlow GraphDef - yolov5s.tflite # TensorFlow Lite - yolov5s_edgetpu.tflite # TensorFlow Edge TPU + $ python val.py --weights yolov5s.pt # PyTorch + yolov5s.torchscript # TorchScript + yolov5s.onnx # ONNX Runtime or OpenCV DNN with --dnn + yolov5s.xml # OpenVINO + yolov5s.engine # TensorRT + yolov5s.mlmodel # CoreML (macOS-only) + yolov5s_saved_model # TensorFlow SavedModel + yolov5s.pb # TensorFlow GraphDef + yolov5s.tflite # TensorFlow Lite + yolov5s_edgetpu.tflite # TensorFlow Edge TPU """ import argparse From eab35f66f9104992a448fbd726c6c2dfdfdf240f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 22 Aug 2022 22:18:01 +0200 Subject: [PATCH 511/661] Install `torch` latest stable (#9092) Install torch 1.12.1 stable GPU assignment issues in 1.13 nightly that comes with image Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/docker/Dockerfile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index cf2c1c5cb3cb..4b9367cc27db 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -15,7 +15,7 @@ RUN apt update && apt install --no-install-recommends -y zip htop screen libgl1- # Install pip packages COPY requirements.txt . RUN python -m pip install --upgrade pip wheel -RUN pip uninstall -y Pillow torchtext # torch torchvision +RUN pip uninstall -y Pillow torchtext torch torchvision RUN pip install --no-cache -r requirements.txt albumentations wandb gsutil notebook Pillow>=9.1.0 \ 'opencv-python<4.6.0.66' \ --extra-index-url https://download.pytorch.org/whl/cu113 From d0fa0042bd7775b2dd191d66548f5d8b677bb756 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 23 Aug 2022 13:06:33 +0200 Subject: [PATCH 512/661] New `@try_export` decorator (#9096) * New export decorator * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * New export decorator * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Cleanup * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * rename fcn to func * rename to @try_export Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 633 +++++++++++++++++++++++------------------------ utils/general.py | 15 +- 2 files changed, 317 insertions(+), 331 deletions(-) diff --git a/export.py b/export.py index 7a746156b96d..1bb7ded8ab85 100644 --- a/export.py +++ b/export.py @@ -67,8 +67,8 @@ from models.experimental import attempt_load from models.yolo import Detect from utils.dataloaders import LoadImages -from utils.general import (LOGGER, check_dataset, check_img_size, check_requirements, check_version, check_yaml, - colorstr, file_size, print_args, url2file) +from utils.general import (LOGGER, Profile, check_dataset, check_img_size, check_requirements, check_version, + check_yaml, colorstr, file_size, get_default_args, print_args, url2file) from utils.torch_utils import select_device, smart_inference_mode @@ -89,200 +89,199 @@ def export_formats(): return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix', 'CPU', 'GPU']) +def try_export(inner_func): + # YOLOv5 export decorator, i..e @try_export + inner_args = get_default_args(inner_func) + + def outer_func(*args, **kwargs): + prefix = inner_args['prefix'] + try: + with Profile() as dt: + f, model = inner_func(*args, **kwargs) + LOGGER.info(f'{prefix} export success ✅ {dt.t:.1f}s, saved as {f} ({file_size(f):.1f} MB)') + return f, model + except Exception as e: + LOGGER.info(f'{prefix} export failure ❌ {dt.t:.1f}s: {e}') + return None, None + + return outer_func + + +@try_export def export_torchscript(model, im, file, optimize, prefix=colorstr('TorchScript:')): # YOLOv5 TorchScript model export - try: - LOGGER.info(f'\n{prefix} starting export with torch {torch.__version__}...') - f = file.with_suffix('.torchscript') - - ts = torch.jit.trace(model, im, strict=False) - d = {"shape": im.shape, "stride": int(max(model.stride)), "names": model.names} - extra_files = {'config.txt': json.dumps(d)} # torch._C.ExtraFilesMap() - if optimize: # https://pytorch.org/tutorials/recipes/mobile_interpreter.html - optimize_for_mobile(ts)._save_for_lite_interpreter(str(f), _extra_files=extra_files) - else: - ts.save(str(f), _extra_files=extra_files) + LOGGER.info(f'\n{prefix} starting export with torch {torch.__version__}...') + f = file.with_suffix('.torchscript') - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return f - except Exception as e: - LOGGER.info(f'{prefix} export failure: {e}') + ts = torch.jit.trace(model, im, strict=False) + d = {"shape": im.shape, "stride": int(max(model.stride)), "names": model.names} + extra_files = {'config.txt': json.dumps(d)} # torch._C.ExtraFilesMap() + if optimize: # https://pytorch.org/tutorials/recipes/mobile_interpreter.html + optimize_for_mobile(ts)._save_for_lite_interpreter(str(f), _extra_files=extra_files) + else: + ts.save(str(f), _extra_files=extra_files) + return f, None +@try_export def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorstr('ONNX:')): # YOLOv5 ONNX export - try: - check_requirements(('onnx',)) - import onnx - - LOGGER.info(f'\n{prefix} starting export with onnx {onnx.__version__}...') - f = file.with_suffix('.onnx') - - torch.onnx.export( - model.cpu() if dynamic else model, # --dynamic only compatible with cpu - im.cpu() if dynamic else im, - f, - verbose=False, - opset_version=opset, - training=torch.onnx.TrainingMode.TRAINING if train else torch.onnx.TrainingMode.EVAL, - do_constant_folding=not train, - input_names=['images'], - output_names=['output'], - dynamic_axes={ - 'images': { - 0: 'batch', - 2: 'height', - 3: 'width'}, # shape(1,3,640,640) - 'output': { - 0: 'batch', - 1: 'anchors'} # shape(1,25200,85) - } if dynamic else None) - - # Checks - model_onnx = onnx.load(f) # load onnx model - onnx.checker.check_model(model_onnx) # check onnx model - - # Metadata - d = {'stride': int(max(model.stride)), 'names': model.names} - for k, v in d.items(): - meta = model_onnx.metadata_props.add() - meta.key, meta.value = k, str(v) - onnx.save(model_onnx, f) - - # Simplify - if simplify: - try: - cuda = torch.cuda.is_available() - check_requirements(('onnxruntime-gpu' if cuda else 'onnxruntime', 'onnx-simplifier>=0.4.1')) - import onnxsim - - LOGGER.info(f'{prefix} simplifying with onnx-simplifier {onnxsim.__version__}...') - model_onnx, check = onnxsim.simplify(model_onnx) - assert check, 'assert check failed' - onnx.save(model_onnx, f) - except Exception as e: - LOGGER.info(f'{prefix} simplifier failure: {e}') - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return f - except Exception as e: - LOGGER.info(f'{prefix} export failure: {e}') + check_requirements(('onnx',)) + import onnx + + LOGGER.info(f'\n{prefix} starting export with onnx {onnx.__version__}...') + f = file.with_suffix('.onnx') + + torch.onnx.export( + model.cpu() if dynamic else model, # --dynamic only compatible with cpu + im.cpu() if dynamic else im, + f, + verbose=False, + opset_version=opset, + training=torch.onnx.TrainingMode.TRAINING if train else torch.onnx.TrainingMode.EVAL, + do_constant_folding=not train, + input_names=['images'], + output_names=['output'], + dynamic_axes={ + 'images': { + 0: 'batch', + 2: 'height', + 3: 'width'}, # shape(1,3,640,640) + 'output': { + 0: 'batch', + 1: 'anchors'} # shape(1,25200,85) + } if dynamic else None) + + # Checks + model_onnx = onnx.load(f) # load onnx model + onnx.checker.check_model(model_onnx) # check onnx model + + # Metadata + d = {'stride': int(max(model.stride)), 'names': model.names} + for k, v in d.items(): + meta = model_onnx.metadata_props.add() + meta.key, meta.value = k, str(v) + onnx.save(model_onnx, f) + + # Simplify + if simplify: + try: + cuda = torch.cuda.is_available() + check_requirements(('onnxruntime-gpu' if cuda else 'onnxruntime', 'onnx-simplifier>=0.4.1')) + import onnxsim + + LOGGER.info(f'{prefix} simplifying with onnx-simplifier {onnxsim.__version__}...') + model_onnx, check = onnxsim.simplify(model_onnx) + assert check, 'assert check failed' + onnx.save(model_onnx, f) + except Exception as e: + LOGGER.info(f'{prefix} simplifier failure: {e}') + return f, model_onnx +@try_export def export_openvino(model, file, half, prefix=colorstr('OpenVINO:')): # YOLOv5 OpenVINO export - try: - check_requirements(('openvino-dev',)) # requires openvino-dev: https://pypi.org/project/openvino-dev/ - import openvino.inference_engine as ie - - LOGGER.info(f'\n{prefix} starting export with openvino {ie.__version__}...') - f = str(file).replace('.pt', f'_openvino_model{os.sep}') + check_requirements(('openvino-dev',)) # requires openvino-dev: https://pypi.org/project/openvino-dev/ + import openvino.inference_engine as ie - cmd = f"mo --input_model {file.with_suffix('.onnx')} --output_dir {f} --data_type {'FP16' if half else 'FP32'}" - subprocess.check_output(cmd.split()) # export - with open(Path(f) / file.with_suffix('.yaml').name, 'w') as g: - yaml.dump({'stride': int(max(model.stride)), 'names': model.names}, g) # add metadata.yaml + LOGGER.info(f'\n{prefix} starting export with openvino {ie.__version__}...') + f = str(file).replace('.pt', f'_openvino_model{os.sep}') - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return f - except Exception as e: - LOGGER.info(f'\n{prefix} export failure: {e}') + cmd = f"mo --input_model {file.with_suffix('.onnx')} --output_dir {f} --data_type {'FP16' if half else 'FP32'}" + subprocess.check_output(cmd.split()) # export + with open(Path(f) / file.with_suffix('.yaml').name, 'w') as g: + yaml.dump({'stride': int(max(model.stride)), 'names': model.names}, g) # add metadata.yaml + return f, None +@try_export def export_coreml(model, im, file, int8, half, prefix=colorstr('CoreML:')): # YOLOv5 CoreML export - try: - check_requirements(('coremltools',)) - import coremltools as ct - - LOGGER.info(f'\n{prefix} starting export with coremltools {ct.__version__}...') - f = file.with_suffix('.mlmodel') - - ts = torch.jit.trace(model, im, strict=False) # TorchScript model - ct_model = ct.convert(ts, inputs=[ct.ImageType('image', shape=im.shape, scale=1 / 255, bias=[0, 0, 0])]) - bits, mode = (8, 'kmeans_lut') if int8 else (16, 'linear') if half else (32, None) - if bits < 32: - if platform.system() == 'Darwin': # quantization only supported on macOS - with warnings.catch_warnings(): - warnings.filterwarnings("ignore", category=DeprecationWarning) # suppress numpy==1.20 float warning - ct_model = ct.models.neural_network.quantization_utils.quantize_weights(ct_model, bits, mode) - else: - print(f'{prefix} quantization only supported on macOS, skipping...') - ct_model.save(f) - - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return ct_model, f - except Exception as e: - LOGGER.info(f'\n{prefix} export failure: {e}') - return None, None - - -def export_engine(model, im, file, train, half, dynamic, simplify, workspace=4, verbose=False): + check_requirements(('coremltools',)) + import coremltools as ct + + LOGGER.info(f'\n{prefix} starting export with coremltools {ct.__version__}...') + f = file.with_suffix('.mlmodel') + + ts = torch.jit.trace(model, im, strict=False) # TorchScript model + ct_model = ct.convert(ts, inputs=[ct.ImageType('image', shape=im.shape, scale=1 / 255, bias=[0, 0, 0])]) + bits, mode = (8, 'kmeans_lut') if int8 else (16, 'linear') if half else (32, None) + if bits < 32: + if platform.system() == 'Darwin': # quantization only supported on macOS + with warnings.catch_warnings(): + warnings.filterwarnings("ignore", category=DeprecationWarning) # suppress numpy==1.20 float warning + ct_model = ct.models.neural_network.quantization_utils.quantize_weights(ct_model, bits, mode) + else: + print(f'{prefix} quantization only supported on macOS, skipping...') + ct_model.save(f) + return f, ct_model + + +@try_export +def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose=False, prefix=colorstr('TensorRT:')): # YOLOv5 TensorRT export https://developer.nvidia.com/tensorrt - prefix = colorstr('TensorRT:') + assert im.device.type != 'cpu', 'export running on CPU but must be on GPU, i.e. `python export.py --device 0`' try: - assert im.device.type != 'cpu', 'export running on CPU but must be on GPU, i.e. `python export.py --device 0`' - try: - import tensorrt as trt - except Exception: - if platform.system() == 'Linux': - check_requirements(('nvidia-tensorrt',), cmds=('-U --index-url https://pypi.ngc.nvidia.com',)) - import tensorrt as trt - - if trt.__version__[0] == '7': # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012 - grid = model.model[-1].anchor_grid - model.model[-1].anchor_grid = [a[..., :1, :1, :] for a in grid] - export_onnx(model, im, file, 12, train, dynamic, simplify) # opset 12 - model.model[-1].anchor_grid = grid - else: # TensorRT >= 8 - check_version(trt.__version__, '8.0.0', hard=True) # require tensorrt>=8.0.0 - export_onnx(model, im, file, 13, train, dynamic, simplify) # opset 13 - onnx = file.with_suffix('.onnx') - - LOGGER.info(f'\n{prefix} starting export with TensorRT {trt.__version__}...') - assert onnx.exists(), f'failed to export ONNX file: {onnx}' - f = file.with_suffix('.engine') # TensorRT engine file - logger = trt.Logger(trt.Logger.INFO) - if verbose: - logger.min_severity = trt.Logger.Severity.VERBOSE - - builder = trt.Builder(logger) - config = builder.create_builder_config() - config.max_workspace_size = workspace * 1 << 30 - # config.set_memory_pool_limit(trt.MemoryPoolType.WORKSPACE, workspace << 30) # fix TRT 8.4 deprecation notice - - flag = (1 << int(trt.NetworkDefinitionCreationFlag.EXPLICIT_BATCH)) - network = builder.create_network(flag) - parser = trt.OnnxParser(network, logger) - if not parser.parse_from_file(str(onnx)): - raise RuntimeError(f'failed to load ONNX file: {onnx}') - - inputs = [network.get_input(i) for i in range(network.num_inputs)] - outputs = [network.get_output(i) for i in range(network.num_outputs)] - LOGGER.info(f'{prefix} Network Description:') + import tensorrt as trt + except Exception: + if platform.system() == 'Linux': + check_requirements(('nvidia-tensorrt',), cmds=('-U --index-url https://pypi.ngc.nvidia.com',)) + import tensorrt as trt + + if trt.__version__[0] == '7': # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012 + grid = model.model[-1].anchor_grid + model.model[-1].anchor_grid = [a[..., :1, :1, :] for a in grid] + export_onnx(model, im, file, 12, False, dynamic, simplify) # opset 12 + model.model[-1].anchor_grid = grid + else: # TensorRT >= 8 + check_version(trt.__version__, '8.0.0', hard=True) # require tensorrt>=8.0.0 + export_onnx(model, im, file, 13, False, dynamic, simplify) # opset 13 + onnx = file.with_suffix('.onnx') + + LOGGER.info(f'\n{prefix} starting export with TensorRT {trt.__version__}...') + assert onnx.exists(), f'failed to export ONNX file: {onnx}' + f = file.with_suffix('.engine') # TensorRT engine file + logger = trt.Logger(trt.Logger.INFO) + if verbose: + logger.min_severity = trt.Logger.Severity.VERBOSE + + builder = trt.Builder(logger) + config = builder.create_builder_config() + config.max_workspace_size = workspace * 1 << 30 + # config.set_memory_pool_limit(trt.MemoryPoolType.WORKSPACE, workspace << 30) # fix TRT 8.4 deprecation notice + + flag = (1 << int(trt.NetworkDefinitionCreationFlag.EXPLICIT_BATCH)) + network = builder.create_network(flag) + parser = trt.OnnxParser(network, logger) + if not parser.parse_from_file(str(onnx)): + raise RuntimeError(f'failed to load ONNX file: {onnx}') + + inputs = [network.get_input(i) for i in range(network.num_inputs)] + outputs = [network.get_output(i) for i in range(network.num_outputs)] + LOGGER.info(f'{prefix} Network Description:') + for inp in inputs: + LOGGER.info(f'{prefix}\tinput "{inp.name}" with shape {inp.shape} and dtype {inp.dtype}') + for out in outputs: + LOGGER.info(f'{prefix}\toutput "{out.name}" with shape {out.shape} and dtype {out.dtype}') + + if dynamic: + if im.shape[0] <= 1: + LOGGER.warning(f"{prefix}WARNING: --dynamic model requires maximum --batch-size argument") + profile = builder.create_optimization_profile() for inp in inputs: - LOGGER.info(f'{prefix}\tinput "{inp.name}" with shape {inp.shape} and dtype {inp.dtype}') - for out in outputs: - LOGGER.info(f'{prefix}\toutput "{out.name}" with shape {out.shape} and dtype {out.dtype}') - - if dynamic: - if im.shape[0] <= 1: - LOGGER.warning(f"{prefix}WARNING: --dynamic model requires maximum --batch-size argument") - profile = builder.create_optimization_profile() - for inp in inputs: - profile.set_shape(inp.name, (1, *im.shape[1:]), (max(1, im.shape[0] // 2), *im.shape[1:]), im.shape) - config.add_optimization_profile(profile) - - LOGGER.info(f'{prefix} building FP{16 if builder.platform_has_fast_fp16 and half else 32} engine in {f}') - if builder.platform_has_fast_fp16 and half: - config.set_flag(trt.BuilderFlag.FP16) - with builder.build_engine(network, config) as engine, open(f, 'wb') as t: - t.write(engine.serialize()) - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return f - except Exception as e: - LOGGER.info(f'\n{prefix} export failure: {e}') + profile.set_shape(inp.name, (1, *im.shape[1:]), (max(1, im.shape[0] // 2), *im.shape[1:]), im.shape) + config.add_optimization_profile(profile) + + LOGGER.info(f'{prefix} building FP{16 if builder.platform_has_fast_fp16 and half else 32} engine in {f}') + if builder.platform_has_fast_fp16 and half: + config.set_flag(trt.BuilderFlag.FP16) + with builder.build_engine(network, config) as engine, open(f, 'wb') as t: + t.write(engine.serialize()) + return f, None +@try_export def export_saved_model(model, im, file, @@ -296,162 +295,142 @@ def export_saved_model(model, keras=False, prefix=colorstr('TensorFlow SavedModel:')): # YOLOv5 TensorFlow SavedModel export - try: - import tensorflow as tf - from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2 - - from models.tf import TFDetect, TFModel - - LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') - f = str(file).replace('.pt', '_saved_model') - batch_size, ch, *imgsz = list(im.shape) # BCHW - - tf_model = TFModel(cfg=model.yaml, model=model, nc=model.nc, imgsz=imgsz) - im = tf.zeros((batch_size, *imgsz, ch)) # BHWC order for TensorFlow - _ = tf_model.predict(im, tf_nms, agnostic_nms, topk_per_class, topk_all, iou_thres, conf_thres) - inputs = tf.keras.Input(shape=(*imgsz, ch), batch_size=None if dynamic else batch_size) - outputs = tf_model.predict(inputs, tf_nms, agnostic_nms, topk_per_class, topk_all, iou_thres, conf_thres) - keras_model = tf.keras.Model(inputs=inputs, outputs=outputs) - keras_model.trainable = False - keras_model.summary() - if keras: - keras_model.save(f, save_format='tf') - else: - spec = tf.TensorSpec(keras_model.inputs[0].shape, keras_model.inputs[0].dtype) - m = tf.function(lambda x: keras_model(x)) # full model - m = m.get_concrete_function(spec) - frozen_func = convert_variables_to_constants_v2(m) - tfm = tf.Module() - tfm.__call__ = tf.function(lambda x: frozen_func(x)[:4] if tf_nms else frozen_func(x)[0], [spec]) - tfm.__call__(im) - tf.saved_model.save(tfm, - f, - options=tf.saved_model.SaveOptions(experimental_custom_gradients=False) - if check_version(tf.__version__, '2.6') else tf.saved_model.SaveOptions()) - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return keras_model, f - except Exception as e: - LOGGER.info(f'\n{prefix} export failure: {e}') - return None, None + import tensorflow as tf + from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2 + + from models.tf import TFModel + + LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') + f = str(file).replace('.pt', '_saved_model') + batch_size, ch, *imgsz = list(im.shape) # BCHW + + tf_model = TFModel(cfg=model.yaml, model=model, nc=model.nc, imgsz=imgsz) + im = tf.zeros((batch_size, *imgsz, ch)) # BHWC order for TensorFlow + _ = tf_model.predict(im, tf_nms, agnostic_nms, topk_per_class, topk_all, iou_thres, conf_thres) + inputs = tf.keras.Input(shape=(*imgsz, ch), batch_size=None if dynamic else batch_size) + outputs = tf_model.predict(inputs, tf_nms, agnostic_nms, topk_per_class, topk_all, iou_thres, conf_thres) + keras_model = tf.keras.Model(inputs=inputs, outputs=outputs) + keras_model.trainable = False + keras_model.summary() + if keras: + keras_model.save(f, save_format='tf') + else: + spec = tf.TensorSpec(keras_model.inputs[0].shape, keras_model.inputs[0].dtype) + m = tf.function(lambda x: keras_model(x)) # full model + m = m.get_concrete_function(spec) + frozen_func = convert_variables_to_constants_v2(m) + tfm = tf.Module() + tfm.__call__ = tf.function(lambda x: frozen_func(x)[:4] if tf_nms else frozen_func(x)[0], [spec]) + tfm.__call__(im) + tf.saved_model.save(tfm, + f, + options=tf.saved_model.SaveOptions(experimental_custom_gradients=False) if check_version( + tf.__version__, '2.6') else tf.saved_model.SaveOptions()) + return f, keras_model +@try_export def export_pb(keras_model, file, prefix=colorstr('TensorFlow GraphDef:')): # YOLOv5 TensorFlow GraphDef *.pb export https://github.com/leimao/Frozen_Graph_TensorFlow - try: - import tensorflow as tf - from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2 + import tensorflow as tf + from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2 - LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') - f = file.with_suffix('.pb') + LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') + f = file.with_suffix('.pb') - m = tf.function(lambda x: keras_model(x)) # full model - m = m.get_concrete_function(tf.TensorSpec(keras_model.inputs[0].shape, keras_model.inputs[0].dtype)) - frozen_func = convert_variables_to_constants_v2(m) - frozen_func.graph.as_graph_def() - tf.io.write_graph(graph_or_graph_def=frozen_func.graph, logdir=str(f.parent), name=f.name, as_text=False) - - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return f - except Exception as e: - LOGGER.info(f'\n{prefix} export failure: {e}') + m = tf.function(lambda x: keras_model(x)) # full model + m = m.get_concrete_function(tf.TensorSpec(keras_model.inputs[0].shape, keras_model.inputs[0].dtype)) + frozen_func = convert_variables_to_constants_v2(m) + frozen_func.graph.as_graph_def() + tf.io.write_graph(graph_or_graph_def=frozen_func.graph, logdir=str(f.parent), name=f.name, as_text=False) + return f, None +@try_export def export_tflite(keras_model, im, file, int8, data, nms, agnostic_nms, prefix=colorstr('TensorFlow Lite:')): # YOLOv5 TensorFlow Lite export - try: - import tensorflow as tf - - LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') - batch_size, ch, *imgsz = list(im.shape) # BCHW - f = str(file).replace('.pt', '-fp16.tflite') - - converter = tf.lite.TFLiteConverter.from_keras_model(keras_model) - converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS] - converter.target_spec.supported_types = [tf.float16] - converter.optimizations = [tf.lite.Optimize.DEFAULT] - if int8: - from models.tf import representative_dataset_gen - dataset = LoadImages(check_dataset(check_yaml(data))['train'], img_size=imgsz, auto=False) - converter.representative_dataset = lambda: representative_dataset_gen(dataset, ncalib=100) - converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8] - converter.target_spec.supported_types = [] - converter.inference_input_type = tf.uint8 # or tf.int8 - converter.inference_output_type = tf.uint8 # or tf.int8 - converter.experimental_new_quantizer = True - f = str(file).replace('.pt', '-int8.tflite') - if nms or agnostic_nms: - converter.target_spec.supported_ops.append(tf.lite.OpsSet.SELECT_TF_OPS) - - tflite_model = converter.convert() - open(f, "wb").write(tflite_model) - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return f - except Exception as e: - LOGGER.info(f'\n{prefix} export failure: {e}') - - + import tensorflow as tf + + LOGGER.info(f'\n{prefix} starting export with tensorflow {tf.__version__}...') + batch_size, ch, *imgsz = list(im.shape) # BCHW + f = str(file).replace('.pt', '-fp16.tflite') + + converter = tf.lite.TFLiteConverter.from_keras_model(keras_model) + converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS] + converter.target_spec.supported_types = [tf.float16] + converter.optimizations = [tf.lite.Optimize.DEFAULT] + if int8: + from models.tf import representative_dataset_gen + dataset = LoadImages(check_dataset(check_yaml(data))['train'], img_size=imgsz, auto=False) + converter.representative_dataset = lambda: representative_dataset_gen(dataset, ncalib=100) + converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8] + converter.target_spec.supported_types = [] + converter.inference_input_type = tf.uint8 # or tf.int8 + converter.inference_output_type = tf.uint8 # or tf.int8 + converter.experimental_new_quantizer = True + f = str(file).replace('.pt', '-int8.tflite') + if nms or agnostic_nms: + converter.target_spec.supported_ops.append(tf.lite.OpsSet.SELECT_TF_OPS) + + tflite_model = converter.convert() + open(f, "wb").write(tflite_model) + return f, None + + +@try_export def export_edgetpu(file, prefix=colorstr('Edge TPU:')): # YOLOv5 Edge TPU export https://coral.ai/docs/edgetpu/models-intro/ - try: - cmd = 'edgetpu_compiler --version' - help_url = 'https://coral.ai/docs/edgetpu/compiler/' - assert platform.system() == 'Linux', f'export only supported on Linux. See {help_url}' - if subprocess.run(f'{cmd} >/dev/null', shell=True).returncode != 0: - LOGGER.info(f'\n{prefix} export requires Edge TPU compiler. Attempting install from {help_url}') - sudo = subprocess.run('sudo --version >/dev/null', shell=True).returncode == 0 # sudo installed on system - for c in ( - 'curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -', - 'echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list', - 'sudo apt-get update', 'sudo apt-get install edgetpu-compiler'): - subprocess.run(c if sudo else c.replace('sudo ', ''), shell=True, check=True) - ver = subprocess.run(cmd, shell=True, capture_output=True, check=True).stdout.decode().split()[-1] - - LOGGER.info(f'\n{prefix} starting export with Edge TPU compiler {ver}...') - f = str(file).replace('.pt', '-int8_edgetpu.tflite') # Edge TPU model - f_tfl = str(file).replace('.pt', '-int8.tflite') # TFLite model - - cmd = f"edgetpu_compiler -s -d -k 10 --out_dir {file.parent} {f_tfl}" - subprocess.run(cmd.split(), check=True) - - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return f - except Exception as e: - LOGGER.info(f'\n{prefix} export failure: {e}') - - + cmd = 'edgetpu_compiler --version' + help_url = 'https://coral.ai/docs/edgetpu/compiler/' + assert platform.system() == 'Linux', f'export only supported on Linux. See {help_url}' + if subprocess.run(f'{cmd} >/dev/null', shell=True).returncode != 0: + LOGGER.info(f'\n{prefix} export requires Edge TPU compiler. Attempting install from {help_url}') + sudo = subprocess.run('sudo --version >/dev/null', shell=True).returncode == 0 # sudo installed on system + for c in ( + 'curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add -', + 'echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list', + 'sudo apt-get update', 'sudo apt-get install edgetpu-compiler'): + subprocess.run(c if sudo else c.replace('sudo ', ''), shell=True, check=True) + ver = subprocess.run(cmd, shell=True, capture_output=True, check=True).stdout.decode().split()[-1] + + LOGGER.info(f'\n{prefix} starting export with Edge TPU compiler {ver}...') + f = str(file).replace('.pt', '-int8_edgetpu.tflite') # Edge TPU model + f_tfl = str(file).replace('.pt', '-int8.tflite') # TFLite model + + cmd = f"edgetpu_compiler -s -d -k 10 --out_dir {file.parent} {f_tfl}" + subprocess.run(cmd.split(), check=True) + return f, None + + +@try_export def export_tfjs(file, prefix=colorstr('TensorFlow.js:')): # YOLOv5 TensorFlow.js export - try: - check_requirements(('tensorflowjs',)) - import re - - import tensorflowjs as tfjs - - LOGGER.info(f'\n{prefix} starting export with tensorflowjs {tfjs.__version__}...') - f = str(file).replace('.pt', '_web_model') # js dir - f_pb = file.with_suffix('.pb') # *.pb path - f_json = f'{f}/model.json' # *.json path - - cmd = f'tensorflowjs_converter --input_format=tf_frozen_model ' \ - f'--output_node_names=Identity,Identity_1,Identity_2,Identity_3 {f_pb} {f}' - subprocess.run(cmd.split()) - - json = Path(f_json).read_text() - with open(f_json, 'w') as j: # sort JSON Identity_* in ascending order - subst = re.sub( - r'{"outputs": {"Identity.?.?": {"name": "Identity.?.?"}, ' - r'"Identity.?.?": {"name": "Identity.?.?"}, ' - r'"Identity.?.?": {"name": "Identity.?.?"}, ' - r'"Identity.?.?": {"name": "Identity.?.?"}}}', r'{"outputs": {"Identity": {"name": "Identity"}, ' - r'"Identity_1": {"name": "Identity_1"}, ' - r'"Identity_2": {"name": "Identity_2"}, ' - r'"Identity_3": {"name": "Identity_3"}}}', json) - j.write(subst) - - LOGGER.info(f'{prefix} export success, saved as {f} ({file_size(f):.1f} MB)') - return f - except Exception as e: - LOGGER.info(f'\n{prefix} export failure: {e}') + check_requirements(('tensorflowjs',)) + import re + + import tensorflowjs as tfjs + + LOGGER.info(f'\n{prefix} starting export with tensorflowjs {tfjs.__version__}...') + f = str(file).replace('.pt', '_web_model') # js dir + f_pb = file.with_suffix('.pb') # *.pb path + f_json = f'{f}/model.json' # *.json path + + cmd = f'tensorflowjs_converter --input_format=tf_frozen_model ' \ + f'--output_node_names=Identity,Identity_1,Identity_2,Identity_3 {f_pb} {f}' + subprocess.run(cmd.split()) + + json = Path(f_json).read_text() + with open(f_json, 'w') as j: # sort JSON Identity_* in ascending order + subst = re.sub( + r'{"outputs": {"Identity.?.?": {"name": "Identity.?.?"}, ' + r'"Identity.?.?": {"name": "Identity.?.?"}, ' + r'"Identity.?.?": {"name": "Identity.?.?"}, ' + r'"Identity.?.?": {"name": "Identity.?.?"}}}', r'{"outputs": {"Identity": {"name": "Identity"}, ' + r'"Identity_1": {"name": "Identity_1"}, ' + r'"Identity_2": {"name": "Identity_2"}, ' + r'"Identity_3": {"name": "Identity_3"}}}', json) + j.write(subst) + return f, None @smart_inference_mode() @@ -524,22 +503,22 @@ def run( f = [''] * 10 # exported filenames warnings.filterwarnings(action='ignore', category=torch.jit.TracerWarning) # suppress TracerWarning if jit: - f[0] = export_torchscript(model, im, file, optimize) + f[0], _ = export_torchscript(model, im, file, optimize) if engine: # TensorRT required before ONNX - f[1] = export_engine(model, im, file, train, half, dynamic, simplify, workspace, verbose) + f[1], _ = export_engine(model, im, file, half, dynamic, simplify, workspace, verbose) if onnx or xml: # OpenVINO requires ONNX - f[2] = export_onnx(model, im, file, opset, train, dynamic, simplify) + f[2], _ = export_onnx(model, im, file, opset, train, dynamic, simplify) if xml: # OpenVINO - f[3] = export_openvino(model, file, half) + f[3], _ = export_openvino(model, file, half) if coreml: - _, f[4] = export_coreml(model, im, file, int8, half) + f[4], _ = export_coreml(model, im, file, int8, half) # TensorFlow Exports if any((saved_model, pb, tflite, edgetpu, tfjs)): if int8 or edgetpu: # TFLite --int8 bug https://github.com/ultralytics/yolov5/issues/5707 check_requirements(('flatbuffers==1.12',)) # required before `import tensorflow` assert not tflite or not tfjs, 'TFLite and TF.js models must be exported separately, please pass only one type.' - model, f[5] = export_saved_model(model.cpu(), + f[5], model = export_saved_model(model.cpu(), im, file, dynamic, @@ -551,19 +530,19 @@ def run( conf_thres=conf_thres, keras=keras) if pb or tfjs: # pb prerequisite to tfjs - f[6] = export_pb(model, file) + f[6], _ = export_pb(model, file) if tflite or edgetpu: - f[7] = export_tflite(model, im, file, int8=int8 or edgetpu, data=data, nms=nms, agnostic_nms=agnostic_nms) + f[7], _ = export_tflite(model, im, file, int8 or edgetpu, data=data, nms=nms, agnostic_nms=agnostic_nms) if edgetpu: - f[8] = export_edgetpu(file) + f[8], _ = export_edgetpu(file) if tfjs: - f[9] = export_tfjs(file) + f[9], _ = export_tfjs(file) # Finish f = [str(x) for x in f if x] # filter out '' and None if any(f): h = '--half' if half else '' # --half FP16 inference arg - LOGGER.info(f'\nExport complete ({time.time() - t:.2f}s)' + LOGGER.info(f'\nExport complete ({time.time() - t:.1f}s)' f"\nResults saved to {colorstr('bold', file.parent.resolve())}" f"\nDetect: python detect.py --weights {f[-1]} {h}" f"\nValidate: python val.py --weights {f[-1]} {h}" diff --git a/utils/general.py b/utils/general.py index 3bc6fbc22d57..d8c90f10ac8f 100755 --- a/utils/general.py +++ b/utils/general.py @@ -148,6 +148,7 @@ def __init__(self, t=0.0): def __enter__(self): self.start = self.time() + return self def __exit__(self, type, value, traceback): self.dt = self.time() - self.start # delta-time @@ -220,10 +221,10 @@ def methods(instance): return [f for f in dir(instance) if callable(getattr(instance, f)) and not f.startswith("__")] -def print_args(args: Optional[dict] = None, show_file=True, show_fcn=False): +def print_args(args: Optional[dict] = None, show_file=True, show_func=False): # Print function arguments (optional args dict) x = inspect.currentframe().f_back # previous frame - file, _, fcn, _, _ = inspect.getframeinfo(x) + file, _, func, _, _ = inspect.getframeinfo(x) if args is None: # get args automatically args, _, _, frm = inspect.getargvalues(x) args = {k: v for k, v in frm.items() if k in args} @@ -231,7 +232,7 @@ def print_args(args: Optional[dict] = None, show_file=True, show_fcn=False): file = Path(file).resolve().relative_to(ROOT).with_suffix('') except ValueError: file = Path(file).stem - s = (f'{file}: ' if show_file else '') + (f'{fcn}: ' if show_fcn else '') + s = (f'{file}: ' if show_file else '') + (f'{func}: ' if show_func else '') LOGGER.info(colorstr(s) + ', '.join(f'{k}={v}' for k, v in args.items())) @@ -255,7 +256,13 @@ def init_seeds(seed=0, deterministic=False): def intersect_dicts(da, db, exclude=()): # Dictionary intersection of matching keys and shapes, omitting 'exclude' keys, using da values - return {k: v for k, v in da.items() if k in db and not any(x in k for x in exclude) and v.shape == db[k].shape} + return {k: v for k, v in da.items() if k in db and all(x not in k for x in exclude) and v.shape == db[k].shape} + + +def get_default_args(func): + # Get func() default arguments + signature = inspect.signature(func) + return {k: v.default for k, v in signature.parameters.items() if v.default is not inspect.Parameter.empty} def get_latest_run(search_dir='.'): From 48e56d3c9bede445d49e8f2af458d70955032e91 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 23 Aug 2022 14:37:46 +0200 Subject: [PATCH 513/661] Add optional `transforms` argument to LoadStreams() (#9105) * Add optional `transforms` argument to LoadStreams() Prepare for streaming classification support Signed-off-by: Glenn Jocher * Cleanup Signed-off-by: Glenn Jocher * fix * batch size > 1 fix Signed-off-by: Glenn Jocher --- utils/dataloaders.py | 54 ++++++++++++++++++++------------------------ 1 file changed, 25 insertions(+), 29 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index e73b20a58915..675c2898e7d7 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -251,7 +251,7 @@ def __next__(self): s = f'image {self.count}/{self.nf} {path}: ' if self.transforms: - im = self.transforms(cv2.cvtColor(im0, cv2.COLOR_BGR2RGB)) # classify transforms + im = self.transforms(cv2.cvtColor(im0, cv2.COLOR_BGR2RGB)) # transforms else: im = letterbox(im0, self.img_size, stride=self.stride, auto=self.auto)[0] # padded resize im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB @@ -289,22 +289,20 @@ def __next__(self): raise StopIteration # Read frame - ret_val, img0 = self.cap.read() - img0 = cv2.flip(img0, 1) # flip left-right + ret_val, im0 = self.cap.read() + im0 = cv2.flip(im0, 1) # flip left-right # Print assert ret_val, f'Camera Error {self.pipe}' img_path = 'webcam.jpg' s = f'webcam {self.count}: ' - # Padded resize - img = letterbox(img0, self.img_size, stride=self.stride)[0] + # Process + im = letterbox(im0, self.img_size, stride=self.stride)[0] # resize + im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB + im = np.ascontiguousarray(im) # contiguous - # Convert - img = img.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB - img = np.ascontiguousarray(img) - - return img_path, img, img0, None, s + return img_path, im, im0, None, s def __len__(self): return 0 @@ -312,7 +310,7 @@ def __len__(self): class LoadStreams: # YOLOv5 streamloader, i.e. `python detect.py --source 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP streams` - def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True): + def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True, transforms=None): self.mode = 'stream' self.img_size = img_size self.stride = stride @@ -326,7 +324,6 @@ def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True): n = len(sources) self.imgs, self.fps, self.frames, self.threads = [None] * n, [0] * n, [0] * n, [None] * n self.sources = [clean_str(x) for x in sources] # clean source names for later - self.auto = auto for i, s in enumerate(sources): # index, source # Start thread to read frames from video stream st = f'{i + 1}/{n}: {s}... ' @@ -353,8 +350,10 @@ def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True): LOGGER.info('') # newline # check for common shapes - s = np.stack([letterbox(x, self.img_size, stride=self.stride, auto=self.auto)[0].shape for x in self.imgs]) + s = np.stack([letterbox(x, img_size, stride=stride, auto=auto)[0].shape for x in self.imgs]) self.rect = np.unique(s, axis=0).shape[0] == 1 # rect inference if all shapes equal + self.auto = auto and self.rect + self.transforms = transforms # optional if not self.rect: LOGGER.warning('WARNING: Stream shapes differ. For optimal performance supply similarly-shaped streams.') @@ -385,18 +384,15 @@ def __next__(self): cv2.destroyAllWindows() raise StopIteration - # Letterbox - img0 = self.imgs.copy() - img = [letterbox(x, self.img_size, stride=self.stride, auto=self.rect and self.auto)[0] for x in img0] - - # Stack - img = np.stack(img, 0) - - # Convert - img = img[..., ::-1].transpose((0, 3, 1, 2)) # BGR to RGB, BHWC to BCHW - img = np.ascontiguousarray(img) + im0 = self.imgs.copy() + if self.transforms: + im = np.stack([self.transforms(cv2.cvtColor(x, cv2.COLOR_BGR2RGB)) for x in im0]) # transforms + else: + im = np.stack([letterbox(x, self.img_size, stride=self.stride, auto=self.auto)[0] for x in im0]) # resize + im = im[..., ::-1].transpose((0, 3, 1, 2)) # BGR to RGB, BHWC to BCHW + im = np.ascontiguousarray(im) # contiguous - return self.sources, img, img0, None, '' + return self.sources, im, im0, None, '' def __len__(self): return len(self.sources) # 1E12 frames = 32 streams at 30 FPS for 30 years @@ -836,7 +832,7 @@ def collate_fn(batch): @staticmethod def collate_fn4(batch): - img, label, path, shapes = zip(*batch) # transposed + im, label, path, shapes = zip(*batch) # transposed n = len(shapes) // 4 im4, label4, path4, shapes4 = [], [], path[:n], shapes[:n] @@ -846,13 +842,13 @@ def collate_fn4(batch): for i in range(n): # zidane torch.zeros(16,3,720,1280) # BCHW i *= 4 if random.random() < 0.5: - im = F.interpolate(img[i].unsqueeze(0).float(), scale_factor=2.0, mode='bilinear', - align_corners=False)[0].type(img[i].type()) + im1 = F.interpolate(im[i].unsqueeze(0).float(), scale_factor=2.0, mode='bilinear', + align_corners=False)[0].type(im[i].type()) lb = label[i] else: - im = torch.cat((torch.cat((img[i], img[i + 1]), 1), torch.cat((img[i + 2], img[i + 3]), 1)), 2) + im1 = torch.cat((torch.cat((im[i], im[i + 1]), 1), torch.cat((im[i + 2], im[i + 3]), 1)), 2) lb = torch.cat((label[i], label[i + 1] + ho, label[i + 2] + wo, label[i + 3] + ho + wo), 0) * s - im4.append(im) + im4.append(im1) label4.append(lb) for i, lb in enumerate(label4): From 51c9f9229731021f55a9ceb9f9504abfc979a54b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 23 Aug 2022 17:54:51 +0200 Subject: [PATCH 514/661] Streaming Classification support (#9106) * Streaming Classification support * Streaming Classification support * Streaming Classification support --- classify/predict.py | 168 +++++++++++++++++++++++++++++++---------- detect.py | 2 +- utils/augmentations.py | 1 + 3 files changed, 131 insertions(+), 40 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 135470fd36ed..b430c0645f21 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -1,12 +1,15 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ -Run YOLOv5 classification inference on images, videos, directories, and globs. +Run YOLOv5 classification inference on images, videos, directories, globs, YouTube, webcam, streams, etc. Usage - sources: - $ python classify/predict.py --weights yolov5s.pt --source img.jpg # image - vid.mp4 # video - path/ # directory - 'path/*.jpg' # glob + $ python classify/predict.py --weights yolov5s-cls.pt --source 0 # webcam + img.jpg # image + vid.mp4 # video + path/ # directory + 'path/*.jpg' # glob + 'https://youtu.be/Zgi9g1ksQHc' # YouTube + 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP stream Usage - formats: $ python classify/predict.py --weights yolov5s-cls.pt # PyTorch @@ -23,9 +26,11 @@ import argparse import os +import platform import sys from pathlib import Path +import torch.backends.cudnn as cudnn import torch.nn.functional as F FILE = Path(__file__).resolve() @@ -36,45 +41,70 @@ from models.common import DetectMultiBackend from utils.augmentations import classify_transforms -from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages -from utils.general import LOGGER, Profile, check_file, check_requirements, colorstr, increment_path, print_args +from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadStreams +from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, + increment_path, print_args, strip_optimizer) +from utils.plots import Annotator from utils.torch_utils import select_device, smart_inference_mode @smart_inference_mode() def run( weights=ROOT / 'yolov5s-cls.pt', # model.pt path(s) - source=ROOT / 'data/images', # file/dir/URL/glob - imgsz=224, # inference size + source=ROOT / 'data/images', # file/dir/URL/glob, 0 for webcam + data=ROOT / 'data/coco128.yaml', # dataset.yaml path + imgsz=(224, 224), # inference size (height, width) device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu + view_img=False, # show results + save_txt=False, # save results to *.txt + nosave=False, # do not save images/videos + augment=False, # augmented inference + visualize=False, # visualize features + update=False, # update all models + project=ROOT / 'runs/predict-cls', # save results to project/name + name='exp', # save results to project/name + exist_ok=False, # existing project/name ok, do not increment half=False, # use FP16 half-precision inference dnn=False, # use OpenCV DNN for ONNX inference - project=ROOT / 'runs/predict-cls', # save to project/name - name='exp', # save to project/name - exist_ok=False, # existing project/name ok, do not increment ): source = str(source) + save_img = not nosave and not source.endswith('.txt') # save inference images is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS) is_url = source.lower().startswith(('rtsp://', 'rtmp://', 'http://', 'https://')) + webcam = source.isnumeric() or source.endswith('.txt') or (is_url and not is_file) if is_url and is_file: source = check_file(source) # download - dt = Profile(), Profile(), Profile() - device = select_device(device) - # Directories save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run - save_dir.mkdir(parents=True, exist_ok=True) # make dir + (save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir # Load model - model = DetectMultiBackend(weights, device=device, dnn=dnn, fp16=half) - model.warmup(imgsz=(1, 3, imgsz, imgsz)) # warmup - dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz)) - for seen, (path, im, im0s, vid_cap, s) in enumerate(dataset): - # Image + device = select_device(device) + model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data, fp16=half) + stride, names, pt = model.stride, model.names, model.pt + imgsz = check_img_size(imgsz, s=stride) # check image size + + # Dataloader + if webcam: + view_img = check_imshow() + cudnn.benchmark = True # set True to speed up constant image size inference + dataset = LoadStreams(source, img_size=imgsz, transforms=classify_transforms(imgsz[0])) + bs = len(dataset) # batch_size + else: + dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz[0])) + bs = 1 # batch_size + vid_path, vid_writer = [None] * bs, [None] * bs + + # Run inference + model.warmup(imgsz=(1 if pt else bs, 3, *imgsz)) # warmup + seen, windows, dt = 0, [], (Profile(), Profile(), Profile()) + for path, im, im0s, vid_cap, s in dataset: with dt[0]: - im = im.unsqueeze(0).to(device) - im = im.half() if model.fp16 else im.float() + im = im.to(device) + im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 + if len(im.shape) == 3: + im = im[None] # expand for batch dim # Inference with dt[1]: @@ -82,33 +112,93 @@ def run( # Post-process with dt[2]: - p = F.softmax(results, dim=1) # probabilities - i = p.argsort(1, descending=True)[:, :5].squeeze().tolist() # top 5 indices - # if save: - # imshow_cls(im, f=save_dir / Path(path).name, verbose=True) - LOGGER.info( - f"{s}{imgsz}x{imgsz} {', '.join(f'{model.names[j]} {p[0, j]:.2f}' for j in i)}, {dt[1].dt * 1E3:.1f}ms") + pred = F.softmax(results, dim=1) # probabilities + + # Process predictions + for i, prob in enumerate(pred): # per image + seen += 1 + if webcam: # batch_size >= 1 + p, im0 = path[i], im0s[i].copy() + s += f'{i}: ' + else: + p, im0 = path, im0s.copy() + + p = Path(p) # to Path + save_path = str(save_dir / p.name) # im.jpg + s += '%gx%g ' % im.shape[2:] # print string + annotator = Annotator(im0, example=str(names), pil=True) + + # Print results + top5i = prob.argsort(0, descending=True)[:5].tolist() # top 5 indices + s += f"{', '.join(f'{names[j]} {prob[j]:.2f}' for j in top5i)}, " + + # Write results + if save_img or view_img: # Add bbox to image + text = '\n'.join(f'{prob[j]:.2f} {names[j]}' for j in top5i) + annotator.text((64, 64), text, txt_color=(255, 255, 255)) + + # Stream results + im0 = annotator.result() + if view_img: + if platform.system() == 'Linux' and p not in windows: + windows.append(p) + cv2.namedWindow(str(p), cv2.WINDOW_NORMAL | cv2.WINDOW_KEEPRATIO) # allow window resize (Linux) + cv2.resizeWindow(str(p), im0.shape[1], im0.shape[0]) + cv2.imshow(str(p), im0) + cv2.waitKey(1) # 1 millisecond + + # Save results (image with detections) + if save_img: + if dataset.mode == 'image': + cv2.imwrite(save_path, im0) + else: # 'video' or 'stream' + if vid_path[i] != save_path: # new video + vid_path[i] = save_path + if isinstance(vid_writer[i], cv2.VideoWriter): + vid_writer[i].release() # release previous video writer + if vid_cap: # video + fps = vid_cap.get(cv2.CAP_PROP_FPS) + w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH)) + h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT)) + else: # stream + fps, w, h = 30, im0.shape[1], im0.shape[0] + save_path = str(Path(save_path).with_suffix('.mp4')) # force *.mp4 suffix on results videos + vid_writer[i] = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h)) + vid_writer[i].write(im0) + + # Print time (inference-only) + LOGGER.info(f"{s}{dt[1].dt * 1E3:.1f}ms") # Print results - t = tuple(x.t / (seen + 1) * 1E3 for x in dt) # speeds per image - shape = (1, 3, imgsz, imgsz) - LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms post-process per image at shape {shape}' % t) - LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") - return p + t = tuple(x.t / seen * 1E3 for x in dt) # speeds per image + LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {(1, 3, *imgsz)}' % t) + if save_txt or save_img: + s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else '' + LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}{s}") + if update: + strip_optimizer(weights[0]) # update model (to fix SourceChangeWarning) def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s-cls.pt', help='model path(s)') - parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob') - parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=224, help='train, val image size (pixels)') + parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob, 0 for webcam') + parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='(optional) dataset.yaml path') + parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[224], help='inference size h,w') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--view-img', action='store_true', help='show results') + parser.add_argument('--save-txt', action='store_true', help='save results to *.txt') + parser.add_argument('--nosave', action='store_true', help='do not save images/videos') + parser.add_argument('--augment', action='store_true', help='augmented inference') + parser.add_argument('--visualize', action='store_true', help='visualize features') + parser.add_argument('--update', action='store_true', help='update all models') + parser.add_argument('--project', default=ROOT / 'runs/predict-cls', help='save results to project/name') + parser.add_argument('--name', default='exp', help='save results to project/name') + parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') - parser.add_argument('--project', default=ROOT / 'runs/predict-cls', help='save to project/name') - parser.add_argument('--name', default='exp', help='save to project/name') - parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') opt = parser.parse_args() + opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand print_args(vars(opt)) return opt diff --git a/detect.py b/detect.py index 541ad90e051d..60a821b59a03 100644 --- a/detect.py +++ b/detect.py @@ -1,6 +1,6 @@ # YOLOv5 🚀 by Ultralytics, GPL-3.0 license """ -Run YOLOv5 detection inference on images, videos, directories, streams, etc. +Run YOLOv5 detection inference on images, videos, directories, globs, YouTube, webcam, streams, etc. Usage - sources: $ python detect.py --weights yolov5s.pt --source 0 # webcam diff --git a/utils/augmentations.py b/utils/augmentations.py index a55fefa68a76..c8499b3fc8ae 100644 --- a/utils/augmentations.py +++ b/utils/augmentations.py @@ -344,4 +344,5 @@ def classify_albumentations(augment=True, def classify_transforms(size=224): # Transforms to apply if albumentations not installed + assert isinstance(size, int), f'ERROR: classify_transforms size {size} must be integer, not (list, tuple)' return T.Compose([T.ToTensor(), T.Resize(size), T.CenterCrop(size), T.Normalize(IMAGENET_MEAN, IMAGENET_STD)]) From e6f54c5b32340278474e922d456fa3eb7f74599d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 23 Aug 2022 23:54:05 +0200 Subject: [PATCH 515/661] Fix numpy to torch cls streaming bug (#9112) * Fix numpy to torch cls streaming bug Resolves https://github.com/ultralytics/yolov5/issues/9111 Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- classify/predict.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/classify/predict.py b/classify/predict.py index b430c0645f21..b33b5bcc9933 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -30,6 +30,7 @@ import sys from pathlib import Path +import torch import torch.backends.cudnn as cudnn import torch.nn.functional as F @@ -101,7 +102,7 @@ def run( seen, windows, dt = 0, [], (Profile(), Profile(), Profile()) for path, im, im0s, vid_cap, s in dataset: with dt[0]: - im = im.to(device) + im = torch.Tensor(im).to(device) im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 if len(im.shape) == 3: im = im[None] # expand for batch dim From f8816f58b7f4bf018ec0fdf546430295e5719205 Mon Sep 17 00:00:00 2001 From: Ayush Chaurasia Date: Wed, 24 Aug 2022 15:45:37 +0530 Subject: [PATCH 516/661] Infer Loggers project name (#9117) * smart project name inference * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update __init__.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- utils/loggers/__init__.py | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 006125edbcd9..59d4b566836a 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -252,7 +252,7 @@ def __init__(self, opt, console_logger, include=('tb', 'wandb')): self.tb = SummaryWriter(str(self.save_dir)) if wandb and 'wandb' in self.include: - self.wandb = wandb.init(project="YOLOv5-Classifier" if opt.project == "runs/train" else opt.project, + self.wandb = wandb.init(project=web_project_name(str(opt.project)), name=None if opt.name == "exp" else opt.name, config=opt) else: @@ -303,3 +303,11 @@ def log_tensorboard_graph(tb, model, imgsz=(640, 640)): tb.add_graph(torch.jit.trace(de_parallel(model), im, strict=False), []) except Exception as e: print(f'WARNING: TensorBoard graph visualization failure {e}') + + +def web_project_name(project): + # Convert local project name to web project name + if not project.startswith('runs/train'): + return project + suffix = '-Classify' if project.endswith('-cls') else '-Segment' if project.endswith('-seg') else '' + return f'YOLOv5{suffix}' From f0e5a608f50ac647827bede88fded7908c7edeab Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 24 Aug 2022 12:31:50 +0200 Subject: [PATCH 517/661] Add CSV logging to GenericLogger (#9128) Enable CSV logging for Classify training. Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/loggers/__init__.py | 21 +++++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 59d4b566836a..880039b1914c 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -242,9 +242,10 @@ class GenericLogger: def __init__(self, opt, console_logger, include=('tb', 'wandb')): # init default loggers - self.save_dir = opt.save_dir + self.save_dir = Path(opt.save_dir) self.include = include self.console_logger = console_logger + self.csv = self.save_dir / 'results.csv' # CSV logger if 'tb' in self.include: prefix = colorstr('TensorBoard: ') self.console_logger.info( @@ -258,14 +259,21 @@ def __init__(self, opt, console_logger, include=('tb', 'wandb')): else: self.wandb = None - def log_metrics(self, metrics_dict, epoch): + def log_metrics(self, metrics, epoch): # Log metrics dictionary to all loggers + if self.csv: + keys, vals = list(metrics.keys()), list(metrics.values()) + n = len(metrics) + 1 # number of cols + s = '' if self.csv.exists() else (('%23s,' * n % tuple(['epoch'] + keys)).rstrip(',') + '\n') # header + with open(self.csv, 'a') as f: + f.write(s + ('%23.5g,' * n % tuple([epoch] + vals)).rstrip(',') + '\n') + if self.tb: - for k, v in metrics_dict.items(): + for k, v in metrics.items(): self.tb.add_scalar(k, v, epoch) if self.wandb: - self.wandb.log(metrics_dict, step=epoch) + self.wandb.log(metrics, step=epoch) def log_images(self, files, name='Images', epoch=0): # Log images to all loggers @@ -291,6 +299,11 @@ def log_model(self, model_path, epoch=0, metadata={}): art.add_file(str(model_path)) wandb.log_artifact(art) + def update_params(self, params): + # Update the paramters logged + if self.wandb: + wandb.run.config.update(params, allow_val_change=True) + def log_tensorboard_graph(tb, model, imgsz=(640, 640)): # Log model graph to TensorBoard From d07ddc69960ed71111457cbe41ab25ded1ab3155 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 25 Aug 2022 14:34:26 +0200 Subject: [PATCH 518/661] New TryExcept decorator (#9154) * New TryExcept decorator * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/__init__.py | 27 ++++++++++++++++++ utils/general.py | 27 ++---------------- utils/metrics.py | 73 ++++++++++++++++++++++++----------------------- utils/plots.py | 5 ++-- 4 files changed, 71 insertions(+), 61 deletions(-) diff --git a/utils/__init__.py b/utils/__init__.py index a63c473a4340..7466a486caf4 100644 --- a/utils/__init__.py +++ b/utils/__init__.py @@ -3,6 +3,33 @@ utils/initialization """ +import contextlib +import threading + + +class TryExcept(contextlib.ContextDecorator): + # YOLOv5 TryExcept class. Usage: @TryExcept() decorator or 'with TryExcept():' context manager + def __init__(self, msg='default message here'): + self.msg = msg + + def __enter__(self): + pass + + def __exit__(self, exc_type, value, traceback): + if value: + print(f'{self.msg}: {value}') + return True + + +def threaded(func): + # Multi-threads a target function and returns thread. Usage: @threaded decorator + def wrapper(*args, **kwargs): + thread = threading.Thread(target=func, args=args, kwargs=kwargs, daemon=True) + thread.start() + return thread + + return wrapper + def notebook_init(verbose=True): # Check system software and hardware diff --git a/utils/general.py b/utils/general.py index d8c90f10ac8f..91b13f84a6c4 100755 --- a/utils/general.py +++ b/utils/general.py @@ -15,7 +15,6 @@ import shutil import signal import sys -import threading import time import urllib from datetime import datetime @@ -34,6 +33,7 @@ import torchvision import yaml +from utils import TryExcept from utils.downloads import gsutil_getsize from utils.metrics import box_iou, fitness @@ -195,27 +195,6 @@ def __exit__(self, exc_type, exc_val, exc_tb): os.chdir(self.cwd) -def try_except(func): - # try-except function. Usage: @try_except decorator - def handler(*args, **kwargs): - try: - func(*args, **kwargs) - except Exception as e: - print(e) - - return handler - - -def threaded(func): - # Multi-threads a target function and returns thread. Usage: @threaded decorator - def wrapper(*args, **kwargs): - thread = threading.Thread(target=func, args=args, kwargs=kwargs, daemon=True) - thread.start() - return thread - - return wrapper - - def methods(instance): # Get class/instance methods return [f for f in dir(instance) if callable(getattr(instance, f)) and not f.startswith("__")] @@ -319,7 +298,7 @@ def git_describe(path=ROOT): # path must be a directory return '' -@try_except +@TryExcept() @WorkingDirectory(ROOT) def check_git_status(repo='ultralytics/yolov5'): # YOLOv5 status check, recommend 'git pull' if code is out of date @@ -364,7 +343,7 @@ def check_version(current='0.0.0', minimum='0.0.0', name='version ', pinned=Fals return result -@try_except +@TryExcept() def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), install=True, cmds=()): # Check installed dependencies meet YOLOv5 requirements (pass *.txt file or list of packages) prefix = colorstr('red', 'bold', 'requirements:') diff --git a/utils/metrics.py b/utils/metrics.py index 8fa3c7e217c7..de1bf05b326b 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -11,6 +11,8 @@ import numpy as np import torch +from utils import TryExcept, threaded + def fitness(x): # Model fitness as a weighted combination of metrics @@ -184,36 +186,35 @@ def tp_fp(self): # fn = self.matrix.sum(0) - tp # false negatives (missed detections) return tp[:-1], fp[:-1] # remove background class + @TryExcept('WARNING: ConfusionMatrix plot failure') def plot(self, normalize=True, save_dir='', names=()): - try: - import seaborn as sn - - array = self.matrix / ((self.matrix.sum(0).reshape(1, -1) + 1E-9) if normalize else 1) # normalize columns - array[array < 0.005] = np.nan # don't annotate (would appear as 0.00) - - fig = plt.figure(figsize=(12, 9), tight_layout=True) - nc, nn = self.nc, len(names) # number of classes, names - sn.set(font_scale=1.0 if nc < 50 else 0.8) # for label size - labels = (0 < nn < 99) and (nn == nc) # apply names to ticklabels - with warnings.catch_warnings(): - warnings.simplefilter('ignore') # suppress empty matrix RuntimeWarning: All-NaN slice encountered - sn.heatmap(array, - annot=nc < 30, - annot_kws={ - "size": 8}, - cmap='Blues', - fmt='.2f', - square=True, - vmin=0.0, - xticklabels=names + ['background FP'] if labels else "auto", - yticklabels=names + ['background FN'] if labels else "auto").set_facecolor((1, 1, 1)) - fig.axes[0].set_xlabel('True') - fig.axes[0].set_ylabel('Predicted') - plt.title('Confusion Matrix') - fig.savefig(Path(save_dir) / 'confusion_matrix.png', dpi=250) - plt.close() - except Exception as e: - print(f'WARNING: ConfusionMatrix plot failure: {e}') + import seaborn as sn + + array = self.matrix / ((self.matrix.sum(0).reshape(1, -1) + 1E-9) if normalize else 1) # normalize columns + array[array < 0.005] = np.nan # don't annotate (would appear as 0.00) + + fig, ax = plt.subplots(1, 1, figsize=(12, 9), tight_layout=True) + nc, nn = self.nc, len(names) # number of classes, names + sn.set(font_scale=1.0 if nc < 50 else 0.8) # for label size + labels = (0 < nn < 99) and (nn == nc) # apply names to ticklabels + with warnings.catch_warnings(): + warnings.simplefilter('ignore') # suppress empty matrix RuntimeWarning: All-NaN slice encountered + sn.heatmap(array, + ax=ax, + annot=nc < 30, + annot_kws={ + "size": 8}, + cmap='Blues', + fmt='.2f', + square=True, + vmin=0.0, + xticklabels=names + ['background FP'] if labels else "auto", + yticklabels=names + ['background FN'] if labels else "auto").set_facecolor((1, 1, 1)) + ax.set_ylabel('True') + ax.set_ylabel('Predicted') + ax.set_title('Confusion Matrix') + fig.savefig(Path(save_dir) / 'confusion_matrix.png', dpi=250) + plt.close(fig) def print(self): for i in range(self.nc + 1): @@ -320,6 +321,7 @@ def wh_iou(wh1, wh2, eps=1e-7): # Plots ---------------------------------------------------------------------------------------------------------------- +@threaded def plot_pr_curve(px, py, ap, save_dir=Path('pr_curve.png'), names=()): # Precision-recall curve fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True) @@ -336,12 +338,13 @@ def plot_pr_curve(px, py, ap, save_dir=Path('pr_curve.png'), names=()): ax.set_ylabel('Precision') ax.set_xlim(0, 1) ax.set_ylim(0, 1) - plt.legend(bbox_to_anchor=(1.04, 1), loc="upper left") - plt.title('Precision-Recall Curve') + ax.legend(bbox_to_anchor=(1.04, 1), loc="upper left") + ax.set_title('Precision-Recall Curve') fig.savefig(save_dir, dpi=250) - plt.close() + plt.close(fig) +@threaded def plot_mc_curve(px, py, save_dir=Path('mc_curve.png'), names=(), xlabel='Confidence', ylabel='Metric'): # Metric-confidence curve fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True) @@ -358,7 +361,7 @@ def plot_mc_curve(px, py, save_dir=Path('mc_curve.png'), names=(), xlabel='Confi ax.set_ylabel(ylabel) ax.set_xlim(0, 1) ax.set_ylim(0, 1) - plt.legend(bbox_to_anchor=(1.04, 1), loc="upper left") - plt.title(f'{ylabel}-Confidence Curve') + ax.legend(bbox_to_anchor=(1.04, 1), loc="upper left") + ax.set_title(f'{ylabel}-Confidence Curve') fig.savefig(save_dir, dpi=250) - plt.close() + plt.close(fig) diff --git a/utils/plots.py b/utils/plots.py index d35e2bdd168a..2aa163268336 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -19,8 +19,9 @@ import torch from PIL import Image, ImageDraw, ImageFont +from utils import TryExcept, threaded from utils.general import (CONFIG_DIR, FONT, LOGGER, check_font, check_requirements, clip_coords, increment_path, - is_ascii, threaded, try_except, xywh2xyxy, xyxy2xywh) + is_ascii, xywh2xyxy, xyxy2xywh) from utils.metrics import fitness # Settings @@ -339,7 +340,7 @@ def plot_val_study(file='', dir='', x=None): # from utils.plots import *; plot_ plt.savefig(f, dpi=300) -@try_except # known issue https://github.com/ultralytics/yolov5/issues/5395 +@TryExcept() # known issue https://github.com/ultralytics/yolov5/issues/5395 def plot_labels(labels, names=(), save_dir=Path('')): # plot dataset labels LOGGER.info(f"Plotting labels to {save_dir / 'labels.jpg'}... ") From 729dc169baeab2eb55b79ef0c29e3174306c8a0e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 25 Aug 2022 15:04:27 +0200 Subject: [PATCH 519/661] Fixed Classify offsets (#9155) --- classify/predict.py | 2 +- utils/plots.py | 10 ++++++---- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index b33b5bcc9933..937704d0f080 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -136,7 +136,7 @@ def run( # Write results if save_img or view_img: # Add bbox to image text = '\n'.join(f'{prob[j]:.2f} {names[j]}' for j in top5i) - annotator.text((64, 64), text, txt_color=(255, 255, 255)) + annotator.text((32, 32), text, txt_color=(255, 255, 255)) # Stream results im0 = annotator.result() diff --git a/utils/plots.py b/utils/plots.py index 2aa163268336..0f322b6b5844 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -117,10 +117,12 @@ def rectangle(self, xy, fill=None, outline=None, width=1): # Add rectangle to image (PIL-only) self.draw.rectangle(xy, fill, outline, width) - def text(self, xy, text, txt_color=(255, 255, 255)): + def text(self, xy, text, txt_color=(255, 255, 255), anchor='top'): # Add text to image (PIL-only) - w, h = self.font.getsize(text) # text width, height - self.draw.text((xy[0], xy[1] - h + 1), text, fill=txt_color, font=self.font) + if anchor == 'bottom': # start y from font bottom + w, h = self.font.getsize(text) # text width, height + xy[1] += 1 - h + self.draw.text(xy, text, fill=txt_color, font=self.font) def result(self): # Return annotated image as array @@ -222,7 +224,7 @@ def plot_images(images, targets, paths=None, fname='images.jpg', names=None, max x, y = int(w * (i // ns)), int(h * (i % ns)) # block origin annotator.rectangle([x, y, x + w, y + h], None, (255, 255, 255), width=2) # borders if paths: - annotator.text((x + 5, y + 5 + h), text=Path(paths[i]).name[:40], txt_color=(220, 220, 220)) # filenames + annotator.text((x + 5, y + 5), text=Path(paths[i]).name[:40], txt_color=(220, 220, 220)) # filenames if len(targets) > 0: ti = targets[targets[:, 0] == i] # image targets boxes = xywh2xyxy(ti[:, 2:6]).T From 30e674b14d6bb4e13ceea84a5ef67d08e6dd2f7d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 25 Aug 2022 15:06:20 +0200 Subject: [PATCH 520/661] New YOLOv5 v6.2 splash images (#9142) * New YOLOv5 v6.2 splash images @AyushExel @AlanDimmer Signed-off-by: Glenn Jocher * Created using Colaboratory * Update README.md Signed-off-by: Glenn Jocher * Update README.md Signed-off-by: Glenn Jocher * Update README.md Signed-off-by: Glenn Jocher * Update README_cn.md Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/README_cn.md | 98 +++++++++++++++++++++--------------------- README.md | 100 +++++++++++++++++++++---------------------- tutorial.ipynb | 42 +++++++++--------- 3 files changed, 119 insertions(+), 121 deletions(-) diff --git a/.github/README_cn.md b/.github/README_cn.md index 46aafd86ec9b..bb62714f003f 100644 --- a/.github/README_cn.md +++ b/.github/README_cn.md @@ -1,55 +1,55 @@
-

- - -

-
- -[English](../README.md) | 简体中文 -
- CI CPU testing - YOLOv5 Citation - Docker Pulls -
- Open In Colab - Open In Kaggle - Join Forum -
- -
-

-YOLOv5🚀是一个在COCO数据集上预训练的物体检测架构和模型系列，它代表了Ultralytics对未来视觉AI方法的公开研究，其中包含了在数千小时的研究和开发中所获得的经验和最佳实践。 -

- -
- - - - - - - - - - - - - - - - - - - - +

+ + +

+ +   + + +

+ + [English](../README.md) | 简体中文 +
+
+ CI CPU testing + YOLOv5 Citation + Docker Pulls +
+ Open In Colab + Open In Kaggle + Join Forum +
+ +
+

+ YOLOv5🚀是一个在COCO数据集上预训练的物体检测架构和模型系列，它代表了Ultralytics对未来视觉AI方法的公开研究，其中包含了在数千小时的研究和开发中所获得的经验和最佳实践。 +

+ +
+ + + + + + + + + + + + + + + + + + + + +
- - -
##
文件
diff --git a/README.md b/README.md index 89e4f1199cde..1d6b4e153d82 100644 --- a/README.md +++ b/README.md @@ -1,56 +1,56 @@
-

- - -

- -English | [简体中文](.github/README_cn.md) -
-
- CI CPU testing - YOLOv5 Citation - Docker Pulls -
- Open In Colab - Open In Kaggle - Join Forum -
- -
-

-YOLOv5 🚀 is a family of object detection architectures and models pretrained on the COCO dataset, and represents Ultralytics - open-source research into future vision AI methods, incorporating lessons learned and best practices evolved over thousands of hours of research and development. -

- -
- - - - - - - - - - - - - - - - - - - - +

+ + +

+ +   + + +

+ + English | [简体中文](.github/README_cn.md) +
+
+ CI CPU testing + YOLOv5 Citation + Docker Pulls +
+ Open In Colab + Open In Kaggle + Join Forum +
+ +
+

+ YOLOv5 🚀 is a family of object detection architectures and models pretrained on the COCO dataset, and represents Ultralytics + open-source research into future vision AI methods, incorporating lessons learned and best practices evolved over thousands of hours of research and development. +

+ +
+ + + + + + + + + + + + + + + + + + + + +
- - -
##
Documentation
diff --git a/tutorial.ipynb b/tutorial.ipynb index 5b7b1f287d7e..3af5517c9623 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -7,8 +7,7 @@ "provenance": [], "collapsed_sections": [], "machine_shape": "hm", - "toc_visible": true, - "include_colab_link": true + "toc_visible": true }, "kernelspec": { "name": "python3", @@ -381,27 +380,26 @@ } }, "cells": [ - { - "cell_type": "markdown", - "metadata": { - "id": "view-in-github", - "colab_type": "text" - }, - "source": [ - "\"Open" - ] - }, { "cell_type": "markdown", "metadata": { "id": "t6MPjfT5NrKQ" }, "source": [ - "\n", - "\n", + "
\n", + "\n", + " \n", + " \n", + "\n", "\n", - "This is the **official YOLOv5 🚀 notebook** by **Ultralytics**, and is freely available for redistribution under the [GPL-3.0 license](https://choosealicense.com/licenses/gpl-3.0/). \n", - "For more information please visit https://github.com/ultralytics/yolov5 and https://ultralytics.com. Thank you!" + "
\n", + " \"Open\n", + " \"Open\n", + "
\n", + "\n", + "This YOLOv5 🚀 notebook by Ultralytics presents simple train, validate and predict examples to help start your AI adventure.
See GitHub for community support or contact us for professional support.\n", + "\n", + "
" ] }, { @@ -412,7 +410,7 @@ "source": [ "# Setup\n", "\n", - "Clone repo, install dependencies and check PyTorch and GPU." + "Clone GitHub [repository](https://github.com/ultralytics/yolov5), install [dependencies](https://github.com/ultralytics/yolov5/blob/master/requirements.txt) and check PyTorch and GPU." ] }, { @@ -433,7 +431,7 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": 1, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -485,7 +483,7 @@ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", "# display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": 2, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -555,7 +553,7 @@ "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip') # download (780M - 5000 images)\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip # unzip" ], - "execution_count": 3, + "execution_count": null, "outputs": [ { "output_type": "display_data", @@ -603,7 +601,7 @@ "# Validate YOLOv5x on COCO val\n", "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" ], - "execution_count": 4, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -723,7 +721,7 @@ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": 5, + "execution_count": null, "outputs": [ { "output_type": "stream", From f2b8f3fe3a3ae2b601706e5bea9f25265eb2fcd9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 25 Aug 2022 22:17:28 +0200 Subject: [PATCH 521/661] Created using Colaboratory --- tutorial.ipynb | 474 +++++++++++++++++++++++-------------------------- 1 file changed, 218 insertions(+), 256 deletions(-) diff --git a/tutorial.ipynb b/tutorial.ipynb index 3af5517c9623..12840063b1f1 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -16,121 +16,110 @@ "accelerator": "GPU", "widgets": { "application/vnd.jupyter.widget-state+json": { - "da0946bcefd9414fa282977f7f609e36": { + "9b8caa3522fc4cbab31e13b5dfc7808d": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", - "model_module_version": "2.0.0", + "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "2.0.0", + "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", - "_view_module_version": "2.0.0", + "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ - "IPY_MODEL_7838c0af44244ccc906c413cea0989d7", - "IPY_MODEL_309ea78b3e814198b4080beb878d5329", - "IPY_MODEL_b2d1d998e5db4ca1a36280902e1647c7" + "IPY_MODEL_574140e4c4bc48c9a171541a02cd0211", + "IPY_MODEL_35e03ce5090346c9ae602891470fc555", + "IPY_MODEL_c942c208e72d46568b476bb0f2d75496" ], - "layout": "IPY_MODEL_e7d7f56c77884717ba122f1d603c0852", - "tabbable": null, - "tooltip": null + "layout": "IPY_MODEL_65881db1db8a4e9c930fab9172d45143" } }, - "7838c0af44244ccc906c413cea0989d7": { + "574140e4c4bc48c9a171541a02cd0211": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", - "model_module_version": "2.0.0", + "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "2.0.0", + "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", - "_view_module_version": "2.0.0", + "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", - "description_allow_html": false, - "layout": "IPY_MODEL_abf60d6b8ea847f9bb358ae2b045458b", + "description_tooltip": null, + "layout": "IPY_MODEL_60b913d755b34d638478e30705a2dde1", "placeholder": "​", - "style": "IPY_MODEL_379196a2761b4a29aca8ef088dc60c10", - "tabbable": null, - "tooltip": null, + "style": "IPY_MODEL_0856bea36ec148b68522ff9c9eb258d8", "value": "100%" } }, - "309ea78b3e814198b4080beb878d5329": { + "35e03ce5090346c9ae602891470fc555": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", - "model_module_version": "2.0.0", + "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "2.0.0", + "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", - "_view_module_version": "2.0.0", + "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", - "description_allow_html": false, - "layout": "IPY_MODEL_52b546a356e54174a95049b30cb52c81", + "description_tooltip": null, + "layout": "IPY_MODEL_76879f6f2aa54637a7a07faeea2bd684", "max": 818322941, "min": 0, "orientation": "horizontal", - "style": "IPY_MODEL_0889e134327e4aa0a8719d03a0d6941b", - "tabbable": null, - "tooltip": null, + "style": "IPY_MODEL_0ace3934ec6f4d36a1b3a9e086390926", "value": 818322941 } }, - "b2d1d998e5db4ca1a36280902e1647c7": { + "c942c208e72d46568b476bb0f2d75496": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", - "model_module_version": "2.0.0", + "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "2.0.0", + "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", - "_view_module_version": "2.0.0", + "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", - "description_allow_html": false, - "layout": "IPY_MODEL_30f22a3e42d24f10ad9851f40a6703f3", + "description_tooltip": null, + "layout": "IPY_MODEL_d6b7a2243e0c4beca714d99dceec23d6", "placeholder": "​", - "style": "IPY_MODEL_648b3512bb7d4ccca5d75af36c133e92", - "tabbable": null, - "tooltip": null, - "value": " 780M/780M [01:31<00:00, 12.3MB/s]" + "style": "IPY_MODEL_5966ba6e6f114d8c9d8d1d6b1bd4f4c7", + "value": " 780M/780M [02:19<00:00, 6.24MB/s]" } }, - "e7d7f56c77884717ba122f1d603c0852": { + "65881db1db8a4e9c930fab9172d45143": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", - "model_module_version": "2.0.0", + "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", - "_model_module_version": "2.0.0", + "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "2.0.0", + "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, - "border_bottom": null, - "border_left": null, - "border_right": null, - "border_top": null, + "border": null, "bottom": null, "display": null, "flex": null, @@ -158,6 +147,8 @@ "object_position": null, "order": null, "overflow": null, + "overflow_x": null, + "overflow_y": null, "padding": null, "right": null, "top": null, @@ -165,25 +156,22 @@ "width": null } }, - "abf60d6b8ea847f9bb358ae2b045458b": { + "60b913d755b34d638478e30705a2dde1": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", - "model_module_version": "2.0.0", + "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", - "_model_module_version": "2.0.0", + "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "2.0.0", + "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, - "border_bottom": null, - "border_left": null, - "border_right": null, - "border_top": null, + "border": null, "bottom": null, "display": null, "flex": null, @@ -211,6 +199,8 @@ "object_position": null, "order": null, "overflow": null, + "overflow_x": null, + "overflow_y": null, "padding": null, "right": null, "top": null, @@ -218,43 +208,37 @@ "width": null } }, - "379196a2761b4a29aca8ef088dc60c10": { + "0856bea36ec148b68522ff9c9eb258d8": { "model_module": "@jupyter-widgets/controls", - "model_name": "HTMLStyleModel", - "model_module_version": "2.0.0", + "model_name": "DescriptionStyleModel", + "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "2.0.0", - "_model_name": "HTMLStyleModel", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "2.0.0", + "_view_module_version": "1.2.0", "_view_name": "StyleView", - "background": null, - "description_width": "", - "font_size": null, - "text_color": null + "description_width": "" } }, - "52b546a356e54174a95049b30cb52c81": { + "76879f6f2aa54637a7a07faeea2bd684": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", - "model_module_version": "2.0.0", + "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", - "_model_module_version": "2.0.0", + "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "2.0.0", + "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, - "border_bottom": null, - "border_left": null, - "border_right": null, - "border_top": null, + "border": null, "bottom": null, "display": null, "flex": null, @@ -282,6 +266,8 @@ "object_position": null, "order": null, "overflow": null, + "overflow_x": null, + "overflow_y": null, "padding": null, "right": null, "top": null, @@ -289,41 +275,38 @@ "width": null } }, - "0889e134327e4aa0a8719d03a0d6941b": { + "0ace3934ec6f4d36a1b3a9e086390926": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", - "model_module_version": "2.0.0", + "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "2.0.0", + "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "2.0.0", + "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, - "30f22a3e42d24f10ad9851f40a6703f3": { + "d6b7a2243e0c4beca714d99dceec23d6": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", - "model_module_version": "2.0.0", + "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", - "_model_module_version": "2.0.0", + "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "2.0.0", + "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, - "border_bottom": null, - "border_left": null, - "border_right": null, - "border_top": null, + "border": null, "bottom": null, "display": null, "flex": null, @@ -351,6 +334,8 @@ "object_position": null, "order": null, "overflow": null, + "overflow_x": null, + "overflow_y": null, "padding": null, "right": null, "top": null, @@ -358,22 +343,19 @@ "width": null } }, - "648b3512bb7d4ccca5d75af36c133e92": { + "5966ba6e6f114d8c9d8d1d6b1bd4f4c7": { "model_module": "@jupyter-widgets/controls", - "model_name": "HTMLStyleModel", - "model_module_version": "2.0.0", + "model_name": "DescriptionStyleModel", + "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", - "_model_module_version": "2.0.0", - "_model_name": "HTMLStyleModel", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", - "_view_module_version": "2.0.0", + "_view_module_version": "1.2.0", "_view_name": "StyleView", - "background": null, - "description_width": "", - "font_size": null, - "text_color": null + "description_width": "" } } } @@ -420,7 +402,7 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "4200fd6f-c6f5-4505-a4f9-a918f3ed1f86" + "outputId": "0f9ee467-cea4-48e8-9050-7a76ae1b6141" }, "source": [ "!git clone https://github.com/ultralytics/yolov5 # clone\n", @@ -431,13 +413,13 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": null, + "execution_count": 1, "outputs": [ { "output_type": "stream", "name": "stderr", "text": [ - "YOLOv5 🚀 v6.2-41-g8665d55 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" + "YOLOv5 🚀 v6.2-56-g30e674b Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n" ] }, { @@ -477,29 +459,29 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "1af15107-bcd1-4e8f-b5bd-0ee1a737e051" + "outputId": "60647b99-e8d4-402c-f444-331bf6746da4" }, "source": [ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", "# display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": null, + "execution_count": 2, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[34m\u001b[1mdetect: \u001b[0mweights=['yolov5s.pt'], source=data/images, data=data/coco128.yaml, imgsz=[640, 640], conf_thres=0.25, iou_thres=0.45, max_det=1000, device=, view_img=False, save_txt=False, save_conf=False, save_crop=False, nosave=False, classes=None, agnostic_nms=False, augment=False, visualize=False, update=False, project=runs/detect, name=exp, exist_ok=False, line_thickness=3, hide_labels=False, hide_conf=False, half=False, dnn=False\n", - "YOLOv5 🚀 v6.2-41-g8665d55 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-56-g30e674b Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s.pt to yolov5s.pt...\n", - "100% 14.1M/14.1M [00:00<00:00, 41.7MB/s]\n", + "100% 14.1M/14.1M [00:00<00:00, 27.8MB/s]\n", "\n", "Fusing layers... \n", "YOLOv5s summary: 213 layers, 7225885 parameters, 0 gradients\n", - "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, 14.5ms\n", - "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, 18.9ms\n", - "Speed: 0.5ms pre-process, 16.7ms inference, 21.4ms NMS per image at shape (1, 3, 640, 640)\n", + "image 1/2 /content/yolov5/data/images/bus.jpg: 640x480 4 persons, 1 bus, 14.8ms\n", + "image 2/2 /content/yolov5/data/images/zidane.jpg: 384x640 2 persons, 2 ties, 20.1ms\n", + "Speed: 0.6ms pre-process, 17.4ms inference, 21.6ms NMS per image at shape (1, 3, 640, 640)\n", "Results saved to \u001b[1mruns/detect/exp\u001b[0m\n" ] } @@ -531,29 +513,29 @@ "id": "WQPtK1QYVaD_", "colab": { "base_uri": "https://localhost:8080/", - "height": 17, + "height": 49, "referenced_widgets": [ - "da0946bcefd9414fa282977f7f609e36", - "7838c0af44244ccc906c413cea0989d7", - "309ea78b3e814198b4080beb878d5329", - "b2d1d998e5db4ca1a36280902e1647c7", - "e7d7f56c77884717ba122f1d603c0852", - "abf60d6b8ea847f9bb358ae2b045458b", - "379196a2761b4a29aca8ef088dc60c10", - "52b546a356e54174a95049b30cb52c81", - "0889e134327e4aa0a8719d03a0d6941b", - "30f22a3e42d24f10ad9851f40a6703f3", - "648b3512bb7d4ccca5d75af36c133e92" + "9b8caa3522fc4cbab31e13b5dfc7808d", + "574140e4c4bc48c9a171541a02cd0211", + "35e03ce5090346c9ae602891470fc555", + "c942c208e72d46568b476bb0f2d75496", + "65881db1db8a4e9c930fab9172d45143", + "60b913d755b34d638478e30705a2dde1", + "0856bea36ec148b68522ff9c9eb258d8", + "76879f6f2aa54637a7a07faeea2bd684", + "0ace3934ec6f4d36a1b3a9e086390926", + "d6b7a2243e0c4beca714d99dceec23d6", + "5966ba6e6f114d8c9d8d1d6b1bd4f4c7" ] }, - "outputId": "5f129105-eca5-4f33-fb1d-981255f814ad" + "outputId": "102dabed-bc31-42fe-9133-d9ce28a2c01e" }, "source": [ "# Download COCO val\n", "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip') # download (780M - 5000 images)\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip # unzip" ], - "execution_count": null, + "execution_count": 3, "outputs": [ { "output_type": "display_data", @@ -564,24 +546,7 @@ "application/vnd.jupyter.widget-view+json": { "version_major": 2, "version_minor": 0, - "model_id": "da0946bcefd9414fa282977f7f609e36" - }, - "application/json": { - "n": 0, - "total": 818322941, - "elapsed": 0.020366430282592773, - "ncols": null, - "nrows": null, - "prefix": "", - "ascii": false, - "unit": "B", - "unit_scale": true, - "rate": null, - "bar_format": null, - "postfix": null, - "unit_divisor": 1024, - "initial": 0, - "colour": null + "model_id": "9b8caa3522fc4cbab31e13b5dfc7808d" } }, "metadata": {} @@ -595,60 +560,57 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "40d5d000-abee-46a0-c07d-1066e1662e01" + "outputId": "daf60b1b-b098-4657-c863-584f4c9cf078" }, "source": [ - "# Validate YOLOv5x on COCO val\n", - "!python val.py --weights yolov5x.pt --data coco.yaml --img 640 --iou 0.65 --half" + "# Validate YOLOv5s on COCO val\n", + "!python val.py --weights yolov5s.pt --data coco.yaml --img 640 --half" ], - "execution_count": null, + "execution_count": 4, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ - "\u001b[34m\u001b[1mval: \u001b[0mdata=/content/yolov5/data/coco.yaml, weights=['yolov5x.pt'], batch_size=32, imgsz=640, conf_thres=0.001, iou_thres=0.65, task=val, device=, workers=8, single_cls=False, augment=False, verbose=False, save_txt=False, save_hybrid=False, save_conf=False, save_json=True, project=runs/val, name=exp, exist_ok=False, half=True, dnn=False\n", - "YOLOv5 🚀 v6.2-41-g8665d55 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", - "\n", - "Downloading https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5x.pt to yolov5x.pt...\n", - "100% 166M/166M [00:10<00:00, 16.6MB/s]\n", + "\u001b[34m\u001b[1mval: \u001b[0mdata=/content/yolov5/data/coco.yaml, weights=['yolov5s.pt'], batch_size=32, imgsz=640, conf_thres=0.001, iou_thres=0.6, task=val, device=, workers=8, single_cls=False, augment=False, verbose=False, save_txt=False, save_hybrid=False, save_conf=False, save_json=True, project=runs/val, name=exp, exist_ok=False, half=True, dnn=False\n", + "YOLOv5 🚀 v6.2-56-g30e674b Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "Fusing layers... \n", - "YOLOv5x summary: 444 layers, 86705005 parameters, 0 gradients\n", + "YOLOv5s summary: 213 layers, 7225885 parameters, 0 gradients\n", "Downloading https://ultralytics.com/assets/Arial.ttf to /root/.config/Ultralytics/Arial.ttf...\n", - "100% 755k/755k [00:00<00:00, 1.39MB/s]\n", - "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 10506.48it/s]\n", + "100% 755k/755k [00:00<00:00, 52.7MB/s]\n", + "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco/val2017' images and labels...4952 found, 48 missing, 0 empty, 0 corrupt: 100% 5000/5000 [00:00<00:00, 10509.20it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mNew cache created: /content/datasets/coco/val2017.cache\n", - " Class Images Instances P R mAP@.5 mAP@.5:.95: 100% 157/157 [01:06<00:00, 2.36it/s]\n", - " all 5000 36335 0.743 0.625 0.683 0.504\n", - "Speed: 0.1ms pre-process, 4.6ms inference, 1.1ms NMS per image at shape (32, 3, 640, 640)\n", + " Class Images Instances P R mAP@.5 mAP@.5:.95: 100% 157/157 [00:50<00:00, 3.10it/s]\n", + " all 5000 36335 0.67 0.521 0.566 0.371\n", + "Speed: 0.1ms pre-process, 1.0ms inference, 1.5ms NMS per image at shape (32, 3, 640, 640)\n", "\n", - "Evaluating pycocotools mAP... saving runs/val/exp/yolov5x_predictions.json...\n", + "Evaluating pycocotools mAP... saving runs/val/exp/yolov5s_predictions.json...\n", "loading annotations into memory...\n", - "Done (t=0.38s)\n", + "Done (t=0.81s)\n", "creating index...\n", "index created!\n", "Loading and preparing results...\n", - "DONE (t=5.49s)\n", + "DONE (t=5.62s)\n", "creating index...\n", "index created!\n", "Running per image evaluation...\n", "Evaluate annotation type *bbox*\n", - "DONE (t=72.10s).\n", + "DONE (t=77.03s).\n", "Accumulating evaluation results...\n", - "DONE (t=13.94s).\n", - " Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.506\n", - " Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.688\n", - " Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.549\n", - " Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.340\n", - " Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.558\n", - " Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.651\n", - " Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.382\n", - " Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.631\n", - " Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.684\n", - " Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.528\n", - " Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.737\n", - " Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.833\n", + "DONE (t=14.63s).\n", + " Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.374\n", + " Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.572\n", + " Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.402\n", + " Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.211\n", + " Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.423\n", + " Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.489\n", + " Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.311\n", + " Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.516\n", + " Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.566\n", + " Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.378\n", + " Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.625\n", + " Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.724\n", "Results saved to \u001b[1mruns/val/exp\u001b[0m\n" ] } @@ -715,13 +677,13 @@ "colab": { "base_uri": "https://localhost:8080/" }, - "outputId": "f0ce0354-7f50-4546-f3f9-672b4b522d59" + "outputId": "baa6d4be-3379-4aab-844a-d5a5396c0e49" }, "source": [ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": null, + "execution_count": 5, "outputs": [ { "output_type": "stream", @@ -729,7 +691,7 @@ "text": [ "\u001b[34m\u001b[1mtrain: \u001b[0mweights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch-low.yaml, epochs=3, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, noplots=False, evolve=None, bucket=, cache=ram, image_weights=False, device=, multi_scale=False, single_cls=False, optimizer=SGD, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, cos_lr=False, label_smoothing=0.0, patience=100, freeze=[0], save_period=-1, seed=0, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest\n", "\u001b[34m\u001b[1mgithub: \u001b[0mup to date with https://github.com/ultralytics/yolov5 ✅\n", - "YOLOv5 🚀 v6.2-41-g8665d55 Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", + "YOLOv5 🚀 v6.2-56-g30e674b Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "\u001b[34m\u001b[1mhyperparameters: \u001b[0mlr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0\n", "\u001b[34m\u001b[1mWeights & Biases: \u001b[0mrun 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs in Weights & Biases\n", @@ -738,8 +700,8 @@ "\n", "Dataset not found ⚠️, missing paths ['/content/datasets/coco128/images/train2017']\n", "Downloading https://ultralytics.com/assets/coco128.zip to coco128.zip...\n", - "100% 6.66M/6.66M [00:00<00:00, 76.7MB/s]\n", - "Dataset download success ✅ (0.5s), saved to \u001b[1m/content/datasets\u001b[0m\n", + "100% 6.66M/6.66M [00:00<00:00, 41.1MB/s]\n", + "Dataset download success ✅ (0.8s), saved to \u001b[1m/content/datasets\u001b[0m\n", "\n", " from n params module arguments \n", " 0 -1 1 3520 models.common.Conv [3, 32, 6, 2, 2] \n", @@ -773,11 +735,11 @@ "\u001b[34m\u001b[1mAMP: \u001b[0mchecks passed ✅\n", "\u001b[34m\u001b[1moptimizer:\u001b[0m SGD(lr=0.01) with parameter groups 57 weight(decay=0.0), 60 weight(decay=0.0005), 60 bias\n", "\u001b[34m\u001b[1malbumentations: \u001b[0mBlur(p=0.01, blur_limit=(3, 7)), MedianBlur(p=0.01, blur_limit=(3, 7)), ToGray(p=0.01), CLAHE(p=0.01, clip_limit=(1, 4.0), tile_grid_size=(8, 8))\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017' images and labels...128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00<00:00, 7984.87it/s]\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mScanning '/content/datasets/coco128/labels/train2017' images and labels...128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00<00:00, 9659.25it/s]\n", "\u001b[34m\u001b[1mtrain: \u001b[0mNew cache created: /content/datasets/coco128/labels/train2017.cache\n", - "\u001b[34m\u001b[1mtrain: \u001b[0mCaching images (0.1GB ram): 100% 128/128 [00:00<00:00, 1018.19it/s]\n", + "\u001b[34m\u001b[1mtrain: \u001b[0mCaching images (0.1GB ram): 100% 128/128 [00:00<00:00, 951.31it/s]\n", "\u001b[34m\u001b[1mval: \u001b[0mScanning '/content/datasets/coco128/labels/train2017.cache' images and labels... 128 found, 0 missing, 2 empty, 0 corrupt: 100% 128/128 [00:00 Date: Fri, 26 Aug 2022 14:34:28 +0200 Subject: [PATCH 522/661] Rename onnx_dynamic -> dynamic (#9168) --- export.py | 2 +- models/yolo.py | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/export.py b/export.py index 1bb7ded8ab85..0f26e63e9adc 100644 --- a/export.py +++ b/export.py @@ -489,7 +489,7 @@ def run( for k, m in model.named_modules(): if isinstance(m, Detect): m.inplace = inplace - m.onnx_dynamic = dynamic + m.dynamic = dynamic m.export = True for _ in range(2): diff --git a/models/yolo.py b/models/yolo.py index e154b72685b4..7a7308312a14 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -37,7 +37,7 @@ class Detect(nn.Module): stride = None # strides computed during build - onnx_dynamic = False # ONNX export parameter + dynamic = False # force grid reconstruction export = False # export mode def __init__(self, nc=80, anchors=(), ch=(), inplace=True): # detection layer @@ -60,7 +60,7 @@ def forward(self, x): x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous() if not self.training: # inference - if self.onnx_dynamic or self.grid[i].shape[2:4] != x[i].shape[2:4]: + if self.dynamic or self.grid[i].shape[2:4] != x[i].shape[2:4]: self.grid[i], self.anchor_grid[i] = self._make_grid(nx, ny, i) y = x[i].sigmoid() From 5d3d051c9b6bb25c45d254ceabab669c758ed72b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 26 Aug 2022 15:29:31 +0200 Subject: [PATCH 523/661] Inline `_make_grid()` meshgrid (#9170) * Inline _make_grid() meshgrid Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- models/yolo.py | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index 7a7308312a14..fa05fcf9a8d9 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -81,10 +81,7 @@ def _make_grid(self, nx=20, ny=20, i=0, torch_1_10=check_version(torch.__version t = self.anchors[i].dtype shape = 1, self.na, ny, nx, 2 # grid shape y, x = torch.arange(ny, device=d, dtype=t), torch.arange(nx, device=d, dtype=t) - if torch_1_10: # torch>=1.10.0 meshgrid workaround for torch>=0.7 compatibility - yv, xv = torch.meshgrid(y, x, indexing='ij') - else: - yv, xv = torch.meshgrid(y, x) + yv, xv = torch.meshgrid(y, x, indexing='ij') if torch_1_10 else torch.meshgrid(y, x) # torch>=0.7 compatibility grid = torch.stack((xv, yv), 2).expand(shape) - 0.5 # add grid offset, i.e. y = 2.0 * x - 0.5 anchor_grid = (self.anchors[i] * self.stride[i]).view((1, self.na, 1, 1, 2)).expand(shape) return grid, anchor_grid From cff9717d730710ad0f5e858ca54cb19731e6a6b5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 26 Aug 2022 20:06:26 +0200 Subject: [PATCH 524/661] Comment EMA assert (#9173) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/torch_utils.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 5fbe8bbf10f6..abf0bbc19a98 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -422,7 +422,7 @@ def update(self, model): if v.dtype.is_floating_point: # true for FP16 and FP32 v *= d v += (1 - d) * msd[k].detach() - assert v.dtype == msd[k].detach().dtype == torch.float32, f'EMA {v.dtype} and model {msd[k]} must both be FP32' + # assert v.dtype == msd[k].dtype == torch.float32, f'{k}: EMA {v.dtype} and model {msd[k].dtype} must be FP32' def update_attr(self, model, include=(), exclude=('process_group', 'reducer')): # Update EMA attributes From ffbce3858ae3d0d1d0978a5927daa2d4f94e55b6 Mon Sep 17 00:00:00 2001 From: HighMans <42877729+HighMans@users.noreply.github.com> Date: Fri, 26 Aug 2022 19:39:11 -0400 Subject: [PATCH 525/661] Fix confidence threshold for ClearML debug images (#9174) * Fix confidence threshold The confidence is converted to a percentage on line 144, but it is being compared to a default conf_threshold value of a decimal value instead of percent value. Signed-off-by: HighMans <42877729+HighMans@users.noreply.github.com> * Revert "Fix confidence threshold" This reverts commit f84a09967f83d70626ca8dfe7625dce60fb0102e. * Fix confidence comparison Fix the confidence percentage is being compared to a decimal value. Signed-off-by: HighMans <42877729+HighMans@users.noreply.github.com> --- utils/loggers/clearml/clearml_utils.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/utils/loggers/clearml/clearml_utils.py b/utils/loggers/clearml/clearml_utils.py index 52320c090ddd..1e136907367d 100644 --- a/utils/loggers/clearml/clearml_utils.py +++ b/utils/loggers/clearml/clearml_utils.py @@ -141,10 +141,10 @@ def log_image_with_boxes(self, image_path, boxes, class_names, image, conf_thres color = colors(i) class_name = class_names[int(class_nr)] - confidence = round(float(conf) * 100, 2) - label = f"{class_name}: {confidence}%" + confidence_percentage = round(float(conf) * 100, 2) + label = f"{class_name}: {confidence_percentage}%" - if confidence > conf_threshold: + if conf > conf_threshold: annotator.rectangle(box.cpu().numpy(), outline=color) annotator.box_label(box.cpu().numpy(), label=label, color=color) From f58fe6b6c12f1b0d25d95ab07a6656b87ac31b25 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 28 Aug 2022 21:36:05 +0200 Subject: [PATCH 526/661] Update Dockerfile-cpu (#9184) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/docker/Dockerfile-cpu | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/utils/docker/Dockerfile-cpu b/utils/docker/Dockerfile-cpu index d61dfeffe22c..5dc75d83c20f 100644 --- a/utils/docker/Dockerfile-cpu +++ b/utils/docker/Dockerfile-cpu @@ -18,7 +18,8 @@ RUN apt install --no-install-recommends -y python3-pip git zip curl htop libgl1- COPY requirements.txt . RUN python3 -m pip install --upgrade pip wheel RUN pip install --no-cache -r requirements.txt albumentations gsutil notebook \ - coremltools onnx onnx-simplifier onnxruntime openvino-dev tensorflow-cpu tensorflowjs \ + coremltools onnx onnx-simplifier onnxruntime tensorflow-cpu tensorflowjs \ + # openvino-dev \ --extra-index-url https://download.pytorch.org/whl/cpu # Create working directory From 985e000d813c739fe6e4c05b8df6f80f40ca3c7a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 28 Aug 2022 21:48:58 +0200 Subject: [PATCH 527/661] Update Dockerfile-cpu to libpython3-dev (#9185) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/docker/Dockerfile-cpu | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/docker/Dockerfile-cpu b/utils/docker/Dockerfile-cpu index 5dc75d83c20f..d6fac645dba1 100644 --- a/utils/docker/Dockerfile-cpu +++ b/utils/docker/Dockerfile-cpu @@ -11,7 +11,7 @@ ADD https://ultralytics.com/assets/Arial.ttf https://ultralytics.com/assets/Aria # Install linux packages RUN apt update RUN DEBIAN_FRONTEND=noninteractive TZ=Etc/UTC apt install -y tzdata -RUN apt install --no-install-recommends -y python3-pip git zip curl htop libgl1-mesa-glx libglib2.0-0 libpython3.8-dev +RUN apt install --no-install-recommends -y python3-pip git zip curl htop libgl1-mesa-glx libglib2.0-0 libpython3-dev # RUN alias python=python3 # Install pip packages From 53711bacea004389a603697e02c5aa8f7cd4b78e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 28 Aug 2022 22:14:21 +0200 Subject: [PATCH 528/661] Update Dockerfile-arm64 to libpython3-dev (#9187) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/docker/Dockerfile-arm64 | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/utils/docker/Dockerfile-arm64 b/utils/docker/Dockerfile-arm64 index fe92c8d56146..6e8ff77545c5 100644 --- a/utils/docker/Dockerfile-arm64 +++ b/utils/docker/Dockerfile-arm64 @@ -11,8 +11,7 @@ ADD https://ultralytics.com/assets/Arial.ttf https://ultralytics.com/assets/Aria # Install linux packages RUN apt update RUN DEBIAN_FRONTEND=noninteractive TZ=Etc/UTC apt install -y tzdata -RUN apt install --no-install-recommends -y python3-pip git zip curl htop gcc \ - libgl1-mesa-glx libglib2.0-0 libpython3.8-dev +RUN apt install --no-install-recommends -y python3-pip git zip curl htop gcc libgl1-mesa-glx libglib2.0-0 libpython3-dev # RUN alias python=python3 # Install pip packages From 13530402f8b960544aed45db4f71d7056a3ffdfc Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 29 Aug 2022 01:51:41 +0200 Subject: [PATCH 529/661] Fix AutoAnchor MPS bug (#9188) Resolves https://github.com/ultralytics/yolov5/issues/8862 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/autoanchor.py | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/utils/autoanchor.py b/utils/autoanchor.py index f2222203e24c..ac17c6cafc90 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -10,6 +10,7 @@ import yaml from tqdm import tqdm +from utils import TryExcept from utils.general import LOGGER, colorstr PREFIX = colorstr('AutoAnchor: ') @@ -25,6 +26,7 @@ def check_anchor_order(m): m.anchors[:] = m.anchors.flip(0) +@TryExcept(f'{PREFIX}ERROR:') def check_anchors(dataset, model, thr=4.0, imgsz=640): # Check anchor fit to data, recompute if necessary m = model.module.model[-1] if hasattr(model, 'module') else model.model[-1] # Detect() @@ -49,10 +51,7 @@ def metric(k): # compute metric else: LOGGER.info(f'{s}Anchors are a poor fit to dataset ⚠️, attempting to improve...') na = m.anchors.numel() // 2 # number of anchors - try: - anchors = kmean_anchors(dataset, n=na, img_size=imgsz, thr=thr, gen=1000, verbose=False) - except Exception as e: - LOGGER.info(f'{PREFIX}ERROR: {e}') + anchors = kmean_anchors(dataset, n=na, img_size=imgsz, thr=thr, gen=1000, verbose=False) new_bpr = metric(anchors)[0] if new_bpr > bpr: # replace anchors anchors = torch.tensor(anchors, device=m.anchors.device).type_as(m.anchors) @@ -124,7 +123,7 @@ def print_results(k, verbose=True): i = (wh0 < 3.0).any(1).sum() if i: LOGGER.info(f'{PREFIX}WARNING: Extremely small objects found: {i} of {len(wh0)} labels are < 3 pixels in size') - wh = wh0[(wh0 >= 2.0).any(1)] # filter > 2 pixels + wh = wh0[(wh0 >= 2.0).any(1)].astype(np.float32) # filter > 2 pixels # wh = wh * (npr.rand(wh.shape[0], 1) * 0.9 + 0.1) # multiply by random scale 0-1 # Kmeans init @@ -167,4 +166,4 @@ def print_results(k, verbose=True): if verbose: print_results(k, verbose) - return print_results(k) + return print_results(k).astype(np.float32) From e57275a2d8713ec6b6fe88fd341d24c6c6e2d29d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 29 Aug 2022 02:00:42 +0200 Subject: [PATCH 530/661] Skip AMP check on MPS (#9189) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/general.py b/utils/general.py index 91b13f84a6c4..842f28c60886 100755 --- a/utils/general.py +++ b/utils/general.py @@ -535,8 +535,8 @@ def amp_allclose(model, im): prefix = colorstr('AMP: ') device = next(model.parameters()).device # get model device - if device.type == 'cpu': - return False # AMP disabled on CPU + if device.type in ('cpu', 'mps'): + return False # AMP only used on CUDA devices f = ROOT / 'data' / 'images' / 'bus.jpg' # image to check im = f if f.exists() else 'https://ultralytics.com/images/bus.jpg' if check_online() else np.ones((640, 640, 3)) try: From cf5d9cbc33ed6849801311765d0c90cdce8ddfd9 Mon Sep 17 00:00:00 2001 From: HighMans <42877729+HighMans@users.noreply.github.com> Date: Mon, 29 Aug 2022 08:58:55 -0400 Subject: [PATCH 531/661] ClearML's set_report_period's time is defined in minutes not seconds. (#9186) * ClearML's set_report_period's time is defined in minutes not seconds. https://clear.ml/docs/latest/docs/references/sdk/hpo_optimization_hyperparameteroptimizer/#set_report_period set_report_period function takes in time in terms of minutes, not seconds. Signed-off-by: HighMans <42877729+HighMans@users.noreply.github.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: HighMans <42877729+HighMans@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- utils/loggers/clearml/hpo.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/loggers/clearml/hpo.py b/utils/loggers/clearml/hpo.py index 96c2c544c84c..ee518b0fbfc8 100644 --- a/utils/loggers/clearml/hpo.py +++ b/utils/loggers/clearml/hpo.py @@ -69,7 +69,7 @@ ) # report every 10 seconds, this is way too often, but we are testing here -optimizer.set_report_period(10) +optimizer.set_report_period(10 / 60) # You can also use the line below instead to run all the optimizer tasks locally, without using queues or agent # an_optimizer.start_locally(job_complete_callback=job_complete_callback) # set the time limit for the optimization process (2 hours) From f65081c4360887ead430e44ee3eb23566a8145ef Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 29 Aug 2022 15:00:55 +0200 Subject: [PATCH 532/661] Add `check_git_status(..., branch='master')` argument (#9199) Add check_git_status(branch='master') argument Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/utils/general.py b/utils/general.py index 842f28c60886..ac9b0cefd7c8 100755 --- a/utils/general.py +++ b/utils/general.py @@ -300,7 +300,7 @@ def git_describe(path=ROOT): # path must be a directory @TryExcept() @WorkingDirectory(ROOT) -def check_git_status(repo='ultralytics/yolov5'): +def check_git_status(repo='ultralytics/yolov5', branch='master'): # YOLOv5 status check, recommend 'git pull' if code is out of date url = f'https://github.com/{repo}' msg = f', for updates see {url}' @@ -316,10 +316,10 @@ def check_git_status(repo='ultralytics/yolov5'): remote = 'ultralytics' check_output(f'git remote add {remote} {url}', shell=True) check_output(f'git fetch {remote}', shell=True, timeout=5) # git fetch - branch = check_output('git rev-parse --abbrev-ref HEAD', shell=True).decode().strip() # checked out - n = int(check_output(f'git rev-list {branch}..{remote}/master --count', shell=True)) # commits behind + local_branch = check_output('git rev-parse --abbrev-ref HEAD', shell=True).decode().strip() # checked out + n = int(check_output(f'git rev-list {local_branch}..{remote}/{branch} --count', shell=True)) # commits behind if n > 0: - pull = 'git pull' if remote == 'origin' else f'git pull {remote} master' + pull = 'git pull' if remote == 'origin' else f'git pull {remote} {branch}' s += f"⚠️ YOLOv5 is out of date by {n} commit{'s' * (n > 1)}. Use `{pull}` or `git clone {url}` to update." else: s += f'up to date with {url} ✅' From fef1913d288a170a19df33493e241b593de99e41 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 29 Aug 2022 15:12:15 +0200 Subject: [PATCH 533/661] `check_font()` on notebook init (#9200) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/__init__.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/utils/__init__.py b/utils/__init__.py index 7466a486caf4..8bdffd47b3b2 100644 --- a/utils/__init__.py +++ b/utils/__init__.py @@ -38,10 +38,12 @@ def notebook_init(verbose=True): import os import shutil - from utils.general import check_requirements, emojis, is_colab + from utils.general import check_font, check_requirements, emojis, is_colab from utils.torch_utils import select_device # imports check_requirements(('psutil', 'IPython')) + check_font() + import psutil from IPython import display # to display images and clear console output From bd5fd78411115f8f9ebed4c95a26f0d3da316ac5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 29 Aug 2022 17:13:55 +0200 Subject: [PATCH 534/661] Comment `protobuf` in requirements.txt (#9207) The low package version is causing conflicts among other dependencies, commenting it causes no ill effects in CI so this should be fine. Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 10620566ca66..44fe1ce697b7 100644 --- a/requirements.txt +++ b/requirements.txt @@ -12,7 +12,7 @@ scipy>=1.4.1 torch>=1.7.0 torchvision>=0.8.1 tqdm>=4.64.0 -protobuf<=3.20.1 # https://github.com/ultralytics/yolov5/issues/8012 +# protobuf<=3.20.1 # https://github.com/ultralytics/yolov5/issues/8012 # Logging ------------------------------------- tensorboard>=2.4.1 From da22e01a6818199d9222a13e58aa145b0477c342 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 29 Aug 2022 18:10:19 +0200 Subject: [PATCH 535/661] `check_font()` fstring update (#9208) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index ac9b0cefd7c8..3e42e887283c 100755 --- a/utils/general.py +++ b/utils/general.py @@ -456,7 +456,7 @@ def check_font(font=FONT, progress=False): font = Path(font) file = CONFIG_DIR / font.name if not font.exists() and not file.exists(): - url = "https://ultralytics.com/assets/" + font.name + url = f'https://ultralytics.com/assets/{font.name}' LOGGER.info(f'Downloading {url} to {file}...') torch.hub.download_url_to_file(url, str(file), progress=progress) From 3c64d891043643cede117c8e54e30e35aecf2e56 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 29 Aug 2022 20:06:00 +0200 Subject: [PATCH 536/661] AutoBatch protect from extreme batch sizes (#9209) If < 1 or > 1024 set output to default batch size 16. May partially address https://github.com/ultralytics/yolov5/issues/9156 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/autobatch.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/autobatch.py b/utils/autobatch.py index 8d12e46f0f09..01152055196d 100644 --- a/utils/autobatch.py +++ b/utils/autobatch.py @@ -60,8 +60,8 @@ def autobatch(model, imgsz=640, fraction=0.9, batch_size=16): i = results.index(None) # first fail index if b >= batch_sizes[i]: # y intercept above failure point b = batch_sizes[max(i - 1, 0)] # select prior safe point - if b < 1: # zero or negative batch size - b = 16 + if b < 1 or b > 1024: # b outside of safe range + b = batch_size LOGGER.warning(f'{prefix}WARNING: ⚠️ CUDA anomaly detected, recommend restart environment and retry command.') fraction = np.polyval(p, b) / t # actual fraction predicted From 91a81d48fa4e34dbdbaf0e45a1f841c11216aab5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 29 Aug 2022 20:41:54 +0200 Subject: [PATCH 537/661] Default AutoBatch 0.8 fraction (#9212) --- hubconf.py | 2 +- utils/autobatch.py | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/hubconf.py b/hubconf.py index 33fc87930582..bffe2d588b4f 100644 --- a/hubconf.py +++ b/hubconf.py @@ -47,7 +47,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo model = DetectMultiBackend(path, device=device, fuse=autoshape) # detection model if autoshape: if model.pt and isinstance(model.model, ClassificationModel): - LOGGER.warning('WARNING: YOLOv5 v6.2 ClassificationModel is not yet AutoShape compatible. ' + LOGGER.warning('WARNING: ⚠️ YOLOv5 v6.2 ClassificationModel is not yet AutoShape compatible. ' 'You must pass torch tensors in BCHW to this model, i.e. shape(1,3,224,224).') else: model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS diff --git a/utils/autobatch.py b/utils/autobatch.py index 01152055196d..641b055b9fe3 100644 --- a/utils/autobatch.py +++ b/utils/autobatch.py @@ -18,7 +18,7 @@ def check_train_batch_size(model, imgsz=640, amp=True): return autobatch(deepcopy(model).train(), imgsz) # compute optimal batch size -def autobatch(model, imgsz=640, fraction=0.9, batch_size=16): +def autobatch(model, imgsz=640, fraction=0.8, batch_size=16): # Automatically estimate best batch size to use `fraction` of available CUDA memory # Usage: # import torch From f37ac8d611c0972851831fdf534cdb2b7f126cff Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 30 Aug 2022 11:36:38 +0200 Subject: [PATCH 538/661] Delete rebase.yml (#9202) * Delete rebase.yml No longer required with new built-in GitHub PR merge master feature Signed-off-by: Glenn Jocher * Update CONTRIBUTING.md Signed-off-by: Glenn Jocher * Update greetings.yml Signed-off-by: Glenn Jocher * Update CONTRIBUTING.md Signed-off-by: Glenn Jocher * cleanup Signed-off-by: Glenn Jocher --- .github/workflows/greetings.yml | 14 ++++---------- .github/workflows/rebase.yml | 21 --------------------- CONTRIBUTING.md | 23 +++++++++-------------- 3 files changed, 13 insertions(+), 45 deletions(-) delete mode 100644 .github/workflows/rebase.yml diff --git a/.github/workflows/greetings.yml b/.github/workflows/greetings.yml index d5dad7a25559..91bf190eb727 100644 --- a/.github/workflows/greetings.yml +++ b/.github/workflows/greetings.yml @@ -17,16 +17,10 @@ jobs: repo-token: ${{ secrets.GITHUB_TOKEN }} pr-message: | 👋 Hello @${{ github.actor }}, thank you for submitting a YOLOv5 🚀 PR! To allow your work to be integrated as seamlessly as possible, we advise you to: - - ✅ Verify your PR is **up-to-date with upstream/master.** If your PR is behind upstream/master an automatic [GitHub Actions](https://github.com/ultralytics/yolov5/blob/master/.github/workflows/rebase.yml) merge may be attempted by writing /rebase in a new comment, or by running the following code, replacing 'feature' with the name of your local branch: - ```bash - git remote add upstream https://github.com/ultralytics/yolov5.git - git fetch upstream - # git checkout feature # <--- replace 'feature' with local branch name - git merge upstream/master - git push -u origin -f - ``` - - ✅ Verify all Continuous Integration (CI) **checks are passing**. - - ✅ Reduce changes to the absolute **minimum** required for your bug fix or feature addition. _"It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is."_ -Bruce Lee + + - ✅ Verify your PR is **up-to-date** with `ultralytics/yolov5` `master` branch. If your PR is behind you can update your code by clicking the 'Update branch' button or by running `git pull` and `git merge master` locally. + - ✅ Verify all YOLOv5 Continuous Integration (CI) **checks are passing**. + - ✅ Reduce changes to the absolute **minimum** required for your bug fix or feature addition. _"It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is."_ — Bruce Lee issue-message: | 👋 Hello @${{ github.actor }}, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ [Tutorials](https://github.com/ultralytics/yolov5/wiki#tutorials) to get started, where you can find quickstart guides for simple tasks like [Custom Data Training](https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data) all the way to advanced concepts like [Hyperparameter Evolution](https://github.com/ultralytics/yolov5/issues/607). diff --git a/.github/workflows/rebase.yml b/.github/workflows/rebase.yml deleted file mode 100644 index a4dc9e5092fd..000000000000 --- a/.github/workflows/rebase.yml +++ /dev/null @@ -1,21 +0,0 @@ -# https://github.com/marketplace/actions/automatic-rebase - -name: Automatic Rebase -on: - issue_comment: - types: [created] -jobs: - rebase: - name: Rebase - if: github.event.issue.pull_request != '' && contains(github.event.comment.body, '/rebase') - runs-on: ubuntu-latest - steps: - - name: Checkout the latest code - uses: actions/checkout@v3 - with: - token: ${{ secrets.ACTIONS_TOKEN }} - fetch-depth: 0 # otherwise, you will fail to push refs to dest repo - - name: Automatic Rebase - uses: cirrus-actions/rebase@1.7 - env: - GITHUB_TOKEN: ${{ secrets.ACTIONS_TOKEN }} diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 13b9b73b50cc..7498f8995d40 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -45,20 +45,15 @@ changes** button. All done, your PR is now submitted to YOLOv5 for review and ap To allow your work to be integrated as seamlessly as possible, we advise you to: -- ✅ Verify your PR is **up-to-date with upstream/master.** If your PR is behind upstream/master an - automatic [GitHub Actions](https://github.com/ultralytics/yolov5/blob/master/.github/workflows/rebase.yml) merge may - be attempted by writing /rebase in a new comment, or by running the following code, replacing 'feature' with the name - of your local branch: - -```bash -git remote add upstream https://github.com/ultralytics/yolov5.git -git fetch upstream -# git checkout feature # <--- replace 'feature' with local branch name -git merge upstream/master -git push -u origin -f -``` - -- ✅ Verify all Continuous Integration (CI) **checks are passing**. +- ✅ Verify your PR is **up-to-date** with `ultralytics/yolov5` `master` branch. If your PR is behind you can update + your code by clicking the 'Update branch' button or by running `git pull` and `git merge master` locally. + +

Screenshot 2022-08-29 at 22 47 15

+ +- ✅ Verify all YOLOv5 Continuous Integration (CI) **checks are passing**. + +

Screenshot 2022-08-29 at 22 47 03

+ - ✅ Reduce changes to the absolute **minimum** required for your bug fix or feature addition. _"It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is."_ — Bruce Lee From 5fb267f3e5dc86675d508e1b08d20fc0e2e84003 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 30 Aug 2022 13:40:51 +0200 Subject: [PATCH 539/661] Duplicate segment verification fix (#9225) Solution by @Laughing-q to resolve duplicate segment verification bug in https://github.com/ultralytics/yolov5/pull/9052#issuecomment-1231426638 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/dataloaders.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 675c2898e7d7..f027307ccb94 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -963,7 +963,7 @@ def verify_image_label(args): if len(i) < nl: # duplicate row check lb = lb[i] # remove duplicates if segments: - segments = segments[i] + segments = [segments[x] for x in i] msg = f'{prefix}WARNING: {im_file}: {nl - len(i)} duplicate labels removed' else: ne = 1 # label empty From 6e7a7ae7edee8f66d7ce5617f9f75724bb7d6992 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 30 Aug 2022 15:17:58 +0200 Subject: [PATCH 540/661] New `LetterBox(size)` `CenterCrop(size)`, `ToTensor()` transforms (#9213) * New LetterBox transform YOLOv5 LetterBox class for image preprocessing, i.e. T.Compose([T.ToTensor(), LetterBox(size)]) Signed-off-by: Glenn Jocher * Update augmentations.py Signed-off-by: Glenn Jocher * Update augmentations.py Signed-off-by: Glenn Jocher * Update augmentations.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * cleanup * cleanup * cleanup * cleanup * cleanup Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/augmentations.py | 50 +++++++++++++++++++++++++++++++++++++++++- utils/dataloaders.py | 22 +++++++++---------- 2 files changed, 60 insertions(+), 12 deletions(-) diff --git a/utils/augmentations.py b/utils/augmentations.py index c8499b3fc8ae..a5587351f75b 100644 --- a/utils/augmentations.py +++ b/utils/augmentations.py @@ -8,6 +8,7 @@ import cv2 import numpy as np +import torch import torchvision.transforms as T import torchvision.transforms.functional as TF @@ -345,4 +346,51 @@ def classify_albumentations(augment=True, def classify_transforms(size=224): # Transforms to apply if albumentations not installed assert isinstance(size, int), f'ERROR: classify_transforms size {size} must be integer, not (list, tuple)' - return T.Compose([T.ToTensor(), T.Resize(size), T.CenterCrop(size), T.Normalize(IMAGENET_MEAN, IMAGENET_STD)]) + # T.Compose([T.ToTensor(), T.Resize(size), T.CenterCrop(size), T.Normalize(IMAGENET_MEAN, IMAGENET_STD)]) + return T.Compose([CenterCrop(size), ToTensor(), T.Normalize(IMAGENET_MEAN, IMAGENET_STD)]) + + +class LetterBox: + # YOLOv5 LetterBox class for image preprocessing, i.e. T.Compose([LetterBox(size), ToTensor()]) + def __init__(self, size=(640, 640), auto=False, stride=32): + super().__init__() + self.h, self.w = (size, size) if isinstance(size, int) else size + self.auto = auto # pass max size integer, automatically solve for short side using stride + self.stride = stride # used with auto + + def __call__(self, im): # im = np.array HWC + imh, imw = im.shape[:2] + r = min(self.h / imh, self.w / imw) # ratio of new/old + h, w = round(imh * r), round(imw * r) # resized image + hs, ws = (math.ceil(x / self.stride) * self.stride for x in (h, w)) if self.auto else self.h, self.w + top, left = round((hs - h) / 2 - 0.1), round((ws - w) / 2 - 0.1) + im_out = np.full((self.h, self.w, 3), 114, dtype=im.dtype) + im_out[top:top + h, left:left + w] = cv2.resize(im, (w, h), interpolation=cv2.INTER_LINEAR) + return im_out + + +class CenterCrop: + # YOLOv5 CenterCrop class for image preprocessing, i.e. T.Compose([CenterCrop(size), ToTensor()]) + def __init__(self, size=640): + super().__init__() + self.h, self.w = (size, size) if isinstance(size, int) else size + + def __call__(self, im): # im = np.array HWC + imh, imw = im.shape[:2] + m = min(imh, imw) # min dimension + top, left = (imh - m) // 2, (imw - m) // 2 + return cv2.resize(im[top:top + m, left:left + m], (self.w, self.h), interpolation=cv2.INTER_LINEAR) + + +class ToTensor: + # YOLOv5 ToTensor class for image preprocessing, i.e. T.Compose([LetterBox(size), ToTensor()]) + def __init__(self, half=False): + super().__init__() + self.half = half + + def __call__(self, im): # im = np.array HWC in BGR order + im = np.ascontiguousarray(im.transpose((2, 0, 1))[::-1]) # HWC to CHW -> BGR to RGB -> contiguous + im = torch.from_numpy(im) # to torch + im = im.half() if self.half else im.float() # uint8 to fp16/32 + im /= 255.0 # 0-255 to 0.0-1.0 + return im diff --git a/utils/dataloaders.py b/utils/dataloaders.py index f027307ccb94..d4ab592bbea7 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -251,7 +251,7 @@ def __next__(self): s = f'image {self.count}/{self.nf} {path}: ' if self.transforms: - im = self.transforms(cv2.cvtColor(im0, cv2.COLOR_BGR2RGB)) # transforms + im = self.transforms(im0) # transforms else: im = letterbox(im0, self.img_size, stride=self.stride, auto=self.auto)[0] # padded resize im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB @@ -386,7 +386,7 @@ def __next__(self): im0 = self.imgs.copy() if self.transforms: - im = np.stack([self.transforms(cv2.cvtColor(x, cv2.COLOR_BGR2RGB)) for x in im0]) # transforms + im = np.stack([self.transforms(x) for x in im0]) # transforms else: im = np.stack([letterbox(x, self.img_size, stride=self.stride, auto=self.auto)[0] for x in im0]) # resize im = im[..., ::-1].transpose((0, 3, 1, 2)) # BGR to RGB, BHWC to BCHW @@ -1113,18 +1113,18 @@ def __init__(self, root, augment, imgsz, cache=False): def __getitem__(self, i): f, j, fn, im = self.samples[i] # filename, index, filename.with_suffix('.npy'), image + if self.cache_ram and im is None: + im = self.samples[i][3] = cv2.imread(f) + elif self.cache_disk: + if not fn.exists(): # load npy + np.save(fn.as_posix(), cv2.imread(f)) + im = np.load(fn) + else: # read image + im = cv2.imread(f) # BGR if self.album_transforms: - if self.cache_ram and im is None: - im = self.samples[i][3] = cv2.imread(f) - elif self.cache_disk: - if not fn.exists(): # load npy - np.save(fn.as_posix(), cv2.imread(f)) - im = np.load(fn) - else: # read image - im = cv2.imread(f) # BGR sample = self.album_transforms(image=cv2.cvtColor(im, cv2.COLOR_BGR2RGB))["image"] else: - sample = self.torch_transforms(self.loader(f)) + sample = self.torch_transforms(im) return sample, j From 4a37381ee8f9b650dde21fe352a94ff932c5b08d Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 30 Aug 2022 16:18:01 +0200 Subject: [PATCH 541/661] Add ClassificationModel TF export assert (#9226) * Add ClassificationModel TF export assert Export to TF not yet supported, warning alerts users. Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/export.py b/export.py index 0f26e63e9adc..4d0144af9efb 100644 --- a/export.py +++ b/export.py @@ -65,7 +65,7 @@ ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.experimental import attempt_load -from models.yolo import Detect +from models.yolo import ClassificationModel, Detect from utils.dataloaders import LoadImages from utils.general import (LOGGER, Profile, check_dataset, check_img_size, check_requirements, check_version, check_yaml, colorstr, file_size, get_default_args, print_args, url2file) @@ -518,6 +518,7 @@ def run( if int8 or edgetpu: # TFLite --int8 bug https://github.com/ultralytics/yolov5/issues/5707 check_requirements(('flatbuffers==1.12',)) # required before `import tensorflow` assert not tflite or not tfjs, 'TFLite and TF.js models must be exported separately, please pass only one type.' + assert not isinstance(model, ClassificationModel), 'ClassificationModel export to TF formats not yet supported.' f[5], model = export_saved_model(model.cpu(), im, file, From 5f1000a499dc8de5f9083412796324ebe091ba10 Mon Sep 17 00:00:00 2001 From: Yannick Merkli Date: Tue, 30 Aug 2022 21:57:36 +0200 Subject: [PATCH 542/661] Remove usage of `pathlib.Path.unlink(missing_ok=...)` (#9227) remove usage of pathlib.Path.unlink(missing_ok=...) Co-authored-by: Yannick Merkli --- utils/dataloaders.py | 4 +++- utils/downloads.py | 18 ++++++++++++------ 2 files changed, 15 insertions(+), 7 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index d4ab592bbea7..c61068ea316f 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -917,7 +917,9 @@ def autosplit(path=DATASETS_DIR / 'coco128/images', weights=(0.9, 0.1, 0.0), ann indices = random.choices([0, 1, 2], weights=weights, k=n) # assign each image to a split txt = ['autosplit_train.txt', 'autosplit_val.txt', 'autosplit_test.txt'] # 3 txt files - [(path.parent / x).unlink(missing_ok=True) for x in txt] # remove existing + for x in txt: + if (path.parent / x).exists(): + (path.parent / x).unlink() # remove existing print(f'Autosplitting images from {path}' + ', using *.txt labeled images only' * annotated_only) for i, img in tqdm(zip(indices, files), total=n): diff --git a/utils/downloads.py b/utils/downloads.py index 69887a579966..b56fc28c3bde 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -44,12 +44,14 @@ def safe_download(file, url, url2=None, min_bytes=1E0, error_msg=''): torch.hub.download_url_to_file(url, str(file), progress=LOGGER.level <= logging.INFO) assert file.exists() and file.stat().st_size > min_bytes, assert_msg # check except Exception as e: # url2 - file.unlink(missing_ok=True) # remove partial downloads + if file.exists(): + file.unlink() # remove partial downloads LOGGER.info(f'ERROR: {e}\nRe-attempting {url2 or url} to {file}...') os.system(f"curl -# -L '{url2 or url}' -o '{file}' --retry 3 -C -") # curl download, retry and resume on fail finally: if not file.exists() or file.stat().st_size < min_bytes: # check - file.unlink(missing_ok=True) # remove partial downloads + if file.exists(): + file.unlink() # remove partial downloads LOGGER.info(f"ERROR: {assert_msg}\n{error_msg}") LOGGER.info('') @@ -112,8 +114,10 @@ def gdrive_download(id='16TiPfZj7htmTyhntwcZyEEAejOUxuT6m', file='tmp.zip'): file = Path(file) cookie = Path('cookie') # gdrive cookie print(f'Downloading https://drive.google.com/uc?export=download&id={id} as {file}... ', end='') - file.unlink(missing_ok=True) # remove existing file - cookie.unlink(missing_ok=True) # remove existing cookie + if file.exists(): + file.unlink() # remove existing file + if cookie.exists(): + cookie.unlink() # remove existing cookie # Attempt file download out = "NUL" if platform.system() == "Windows" else "/dev/null" @@ -123,11 +127,13 @@ def gdrive_download(id='16TiPfZj7htmTyhntwcZyEEAejOUxuT6m', file='tmp.zip'): else: # small file s = f'curl -s -L -o {file} "drive.google.com/uc?export=download&id={id}"' r = os.system(s) # execute, capture return - cookie.unlink(missing_ok=True) # remove existing cookie + if cookie.exists(): + cookie.unlink() # remove existing cookie # Error check if r != 0: - file.unlink(missing_ok=True) # remove partial + if file.exists(): + file.unlink() # remove partial print('Download error ') # raise Exception('Download error') return r From 79e181a83badd31c5013fffa0b80b55ff090c761 Mon Sep 17 00:00:00 2001 From: spacewalk01 Date: Thu, 1 Sep 2022 00:31:13 +0900 Subject: [PATCH 543/661] Add support for *`.pfm` images (#9230) add support for pfm image --- utils/dataloaders.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index c61068ea316f..84215925284e 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -36,7 +36,7 @@ # Parameters HELP_URL = 'See https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data' -IMG_FORMATS = 'bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp' # include image suffixes +IMG_FORMATS = 'bmp', 'dng', 'jpeg', 'jpg', 'mpo', 'png', 'tif', 'tiff', 'webp', 'pfm' # include image suffixes VID_FORMATS = 'asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'ts', 'wmv' # include video suffixes BAR_FORMAT = '{l_bar}{bar:10}{r_bar}{bar:-10b}' # tqdm bar format LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html From 55b009616b4701f73311d1272cc87057d84a93e6 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 31 Aug 2022 18:53:46 +0200 Subject: [PATCH 544/661] Python check warning emoji (#9238) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index 3e42e887283c..bc978ea221f3 100755 --- a/utils/general.py +++ b/utils/general.py @@ -335,7 +335,7 @@ def check_version(current='0.0.0', minimum='0.0.0', name='version ', pinned=Fals # Check version vs. required version current, minimum = (pkg.parse_version(x) for x in (current, minimum)) result = (current == minimum) if pinned else (current >= minimum) # bool - s = f'{name}{minimum} required by YOLOv5, but {name}{current} is currently installed' # string + s = f'WARNING: ⚠️ {name}{minimum} is required by YOLOv5, but {name}{current} is currently installed' # string if hard: assert result, s # assert min requirements met if verbose and not result: From 223c59dbe07357a0bf760ea49cef6e1d7b66df91 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 1 Sep 2022 12:13:53 +0200 Subject: [PATCH 545/661] Add `url_getsize()` function (#9247) * Add `url_getsize()` function Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update downloads.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/downloads.py | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/utils/downloads.py b/utils/downloads.py index b56fc28c3bde..dd2698f995a4 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -33,6 +33,12 @@ def gsutil_getsize(url=''): return eval(s.split(' ')[0]) if len(s) else 0 # bytes +def url_getsize(url='https://ultralytics.com/images/bus.jpg'): + # Return downloadable file size in bytes + response = requests.head(url, allow_redirects=True) + return int(response.headers.get('content-length', -1)) + + def safe_download(file, url, url2=None, min_bytes=1E0, error_msg=''): # Attempts to download file from url or url2, checks and removes incomplete downloads < min_bytes from utils.general import LOGGER From c91d1db7161f4cffe70535378b81faf3ff4549b4 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 1 Sep 2022 14:30:21 +0200 Subject: [PATCH 546/661] Update dataloaders.py (#9250) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/dataloaders.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 84215925284e..a4e6c0cfef18 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -40,6 +40,7 @@ VID_FORMATS = 'asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'ts', 'wmv' # include video suffixes BAR_FORMAT = '{l_bar}{bar:10}{r_bar}{bar:-10b}' # tqdm bar format LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html +PIN_MEMORY = str(os.getenv('PIN_MEMORY', True)).lower() == 'true' # global pin_memory for dataloaders # Get orientation exif tag for orientation in ExifTags.TAGS.keys(): @@ -83,7 +84,7 @@ def exif_transpose(image): 5: Image.TRANSPOSE, 6: Image.ROTATE_270, 7: Image.TRANSVERSE, - 8: Image.ROTATE_90,}.get(orientation) + 8: Image.ROTATE_90}.get(orientation) if method is not None: image = image.transpose(method) del exif[0x0112] @@ -144,7 +145,7 @@ def create_dataloader(path, shuffle=shuffle and sampler is None, num_workers=nw, sampler=sampler, - pin_memory=True, + pin_memory=PIN_MEMORY, collate_fn=LoadImagesAndLabels.collate_fn4 if quad else LoadImagesAndLabels.collate_fn, worker_init_fn=seed_worker, generator=generator), dataset @@ -1152,6 +1153,6 @@ def create_classification_dataloader(path, shuffle=shuffle and sampler is None, num_workers=nw, sampler=sampler, - pin_memory=True, + pin_memory=PIN_MEMORY, worker_init_fn=seed_worker, generator=generator) # or DataLoader(persistent_workers=True) From 2d082a07bd28952159bf534c8728865ba577a449 Mon Sep 17 00:00:00 2001 From: Ayush Chaurasia Date: Thu, 1 Sep 2022 22:47:36 +0530 Subject: [PATCH 547/661] Refactor Loggers : Move code outside train.py (#9241) * update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * update * update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- train.py | 11 +++++------ utils/loggers/__init__.py | 11 +++++++++++ 2 files changed, 16 insertions(+), 6 deletions(-) diff --git a/train.py b/train.py index 0cd4a7f065a6..29293aa612cf 100644 --- a/train.py +++ b/train.py @@ -91,17 +91,16 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio data_dict = None if RANK in {-1, 0}: loggers = Loggers(save_dir, weights, opt, hyp, LOGGER) # loggers instance - if loggers.clearml: - data_dict = loggers.clearml.data_dict # None if no ClearML dataset or filled in by ClearML - if loggers.wandb: - data_dict = loggers.wandb.data_dict - if resume: - weights, epochs, hyp, batch_size = opt.weights, opt.epochs, opt.hyp, opt.batch_size # Register actions for k in methods(loggers): callbacks.register_action(k, callback=getattr(loggers, k)) + # Process custom dataset artifact link + data_dict = loggers.remote_dataset + if resume: # If resuming runs from remote artifact + weights, epochs, hyp, batch_size = opt.weights, opt.epochs, opt.hyp, opt.batch_size + # Config plots = not evolve and not opt.noplots # create plots cuda = device.type != 'cpu' diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 880039b1914c..1aa8427f9127 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -107,6 +107,17 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, else: self.clearml = None + @property + def remote_dataset(self): + # Get data_dict if custom dataset artifact link is provided + data_dict = None + if self.clearml: + data_dict = self.clearml.data_dict + if self.wandb: + data_dict = self.wandb.data_dict + + return data_dict + def on_train_start(self): # Callback runs on train start pass From ea98199041088a378b4f13316ba96afc637dfb83 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 1 Sep 2022 19:36:27 +0200 Subject: [PATCH 548/661] Update general.py (#9252) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index bc978ea221f3..ba6d9e165901 100755 --- a/utils/general.py +++ b/utils/general.py @@ -337,7 +337,7 @@ def check_version(current='0.0.0', minimum='0.0.0', name='version ', pinned=Fals result = (current == minimum) if pinned else (current >= minimum) # bool s = f'WARNING: ⚠️ {name}{minimum} is required by YOLOv5, but {name}{current} is currently installed' # string if hard: - assert result, s # assert min requirements met + assert result, emojis(s) # assert min requirements met if verbose and not result: LOGGER.warning(s) return result From 9da6d0f9f5bc37fa386b7b82d2a963f94650949a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 1 Sep 2022 22:30:26 +0200 Subject: [PATCH 549/661] Add LoadImages._cv2_rotate() (#9249) Optional manual rotation code per iPhone rotation issue in https://github.com/ultralytics/yolov5/issues/8493 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/dataloaders.py | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index a4e6c0cfef18..5f86f83786db 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -213,7 +213,7 @@ def __init__(self, path, img_size=640, stride=32, auto=True, transforms=None): self.auto = auto self.transforms = transforms # optional if any(videos): - self.new_video(videos[0]) # new video + self._new_video(videos[0]) # new video else: self.cap = None assert self.nf > 0, f'No images or videos found in {p}. ' \ @@ -238,10 +238,11 @@ def __next__(self): if self.count == self.nf: # last video raise StopIteration path = self.files[self.count] - self.new_video(path) + self._new_video(path) ret_val, im0 = self.cap.read() self.frame += 1 + # im0 = self._cv2_rotate(im0) # for use if cv2 auto rotation is False s = f'video {self.count + 1}/{self.nf} ({self.frame}/{self.frames}) {path}: ' else: @@ -260,10 +261,23 @@ def __next__(self): return path, im, im0, self.cap, s - def new_video(self, path): + def _new_video(self, path): + # Create a new video capture object self.frame = 0 self.cap = cv2.VideoCapture(path) self.frames = int(self.cap.get(cv2.CAP_PROP_FRAME_COUNT)) + self.orientation = int(self.cap.get(cv2.CAP_PROP_ORIENTATION_META)) # rotation degrees + # self.cap.set(cv2.CAP_PROP_ORIENTATION_AUTO, 0) # disable https://github.com/ultralytics/yolov5/issues/8493 + + def _cv2_rotate(self, im): + # Rotate a cv2 video manually + if self.orientation == 0: + return cv2.rotate(im, cv2.ROTATE_90_CLOCKWISE) + elif self.orientation == 180: + return cv2.rotate(im, cv2.ROTATE_90_COUNTERCLOCKWISE) + elif self.orientation == 90: + return cv2.rotate(im, cv2.ROTATE_180) + return im def __len__(self): return self.nf # number of files From ffdb58b0e07d964eb2d148a6814d22a4a26d47cc Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 2 Sep 2022 14:12:10 +0200 Subject: [PATCH 550/661] Move `cudnn.benchmarks(True)` to LoadStreams (#9258) * Move cudnn.benchmarks(True) to LoadStreams * Update dataloaders.py Signed-off-by: Glenn Jocher * Move cudnn.benchmarks(True) to LoadStreams Signed-off-by: Glenn Jocher --- classify/predict.py | 2 -- detect.py | 2 -- utils/dataloaders.py | 54 ++++---------------------------------------- 3 files changed, 4 insertions(+), 54 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 937704d0f080..76115c75029f 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -31,7 +31,6 @@ from pathlib import Path import torch -import torch.backends.cudnn as cudnn import torch.nn.functional as F FILE = Path(__file__).resolve() @@ -89,7 +88,6 @@ def run( # Dataloader if webcam: view_img = check_imshow() - cudnn.benchmark = True # set True to speed up constant image size inference dataset = LoadStreams(source, img_size=imgsz, transforms=classify_transforms(imgsz[0])) bs = len(dataset) # batch_size else: diff --git a/detect.py b/detect.py index 60a821b59a03..cf75d0f11c92 100644 --- a/detect.py +++ b/detect.py @@ -31,7 +31,6 @@ from pathlib import Path import torch -import torch.backends.cudnn as cudnn FILE = Path(__file__).resolve() ROOT = FILE.parents[0] # YOLOv5 root directory @@ -97,7 +96,6 @@ def run( # Dataloader if webcam: view_img = check_imshow() - cudnn.benchmark = True # set True to speed up constant image size inference dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt) bs = len(dataset) # batch_size else: diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 5f86f83786db..38ae3399ce26 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -283,62 +283,17 @@ def __len__(self): return self.nf # number of files -class LoadWebcam: # for inference - # YOLOv5 local webcam dataloader, i.e. `python detect.py --source 0` - def __init__(self, pipe='0', img_size=640, stride=32): - self.img_size = img_size - self.stride = stride - self.pipe = eval(pipe) if pipe.isnumeric() else pipe - self.cap = cv2.VideoCapture(self.pipe) # video capture object - self.cap.set(cv2.CAP_PROP_BUFFERSIZE, 3) # set buffer size - - def __iter__(self): - self.count = -1 - return self - - def __next__(self): - self.count += 1 - if cv2.waitKey(1) == ord('q'): # q to quit - self.cap.release() - cv2.destroyAllWindows() - raise StopIteration - - # Read frame - ret_val, im0 = self.cap.read() - im0 = cv2.flip(im0, 1) # flip left-right - - # Print - assert ret_val, f'Camera Error {self.pipe}' - img_path = 'webcam.jpg' - s = f'webcam {self.count}: ' - - # Process - im = letterbox(im0, self.img_size, stride=self.stride)[0] # resize - im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB - im = np.ascontiguousarray(im) # contiguous - - return img_path, im, im0, None, s - - def __len__(self): - return 0 - - class LoadStreams: # YOLOv5 streamloader, i.e. `python detect.py --source 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP streams` def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True, transforms=None): + torch.backends.cudnn.benchmark = True # faster for fixed-size inference self.mode = 'stream' self.img_size = img_size self.stride = stride - - if os.path.isfile(sources): - with open(sources) as f: - sources = [x.strip() for x in f.read().strip().splitlines() if len(x.strip())] - else: - sources = [sources] - + sources = Path(sources).read_text().rsplit() if Path(sources).is_file() else [sources] n = len(sources) - self.imgs, self.fps, self.frames, self.threads = [None] * n, [0] * n, [0] * n, [None] * n self.sources = [clean_str(x) for x in sources] # clean source names for later + self.imgs, self.fps, self.frames, self.threads = [None] * n, [0] * n, [0] * n, [None] * n for i, s in enumerate(sources): # index, source # Start thread to read frames from video stream st = f'{i + 1}/{n}: {s}... ' @@ -377,8 +332,7 @@ def update(self, i, cap, stream): n, f, read = 0, self.frames[i], 1 # frame number, frame array, inference every 'read' frame while cap.isOpened() and n < f: n += 1 - # _, self.imgs[index] = cap.read() - cap.grab() + cap.grab() # .read() = .grab() followed by .retrieve() if n % read == 0: success, im = cap.retrieve() if success: From 5d4787baabea694369ad95c7d762139eb9f04e56 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 2 Sep 2022 16:05:23 +0200 Subject: [PATCH 551/661] `cudnn.benchmark = True` on Seed 0 (#9259) * `cudnn.benchmark = True` on Seed 0 Signed-off-by: Glenn Jocher * Update general.py Signed-off-by: Glenn Jocher * Update general.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 15 ++++++--------- 1 file changed, 6 insertions(+), 9 deletions(-) diff --git a/utils/general.py b/utils/general.py index ba6d9e165901..25a1a1456009 100755 --- a/utils/general.py +++ b/utils/general.py @@ -217,20 +217,17 @@ def print_args(args: Optional[dict] = None, show_file=True, show_func=False): def init_seeds(seed=0, deterministic=False): # Initialize random number generator (RNG) seeds https://pytorch.org/docs/stable/notes/randomness.html - # cudnn seed 0 settings are slower and more reproducible, else faster and less reproducible - import torch.backends.cudnn as cudnn - - if deterministic and check_version(torch.__version__, '1.12.0'): # https://github.com/ultralytics/yolov5/pull/8213 - torch.use_deterministic_algorithms(True) - os.environ['CUBLAS_WORKSPACE_CONFIG'] = ':4096:8' - os.environ['PYTHONHASHSEED'] = str(seed) - random.seed(seed) np.random.seed(seed) torch.manual_seed(seed) - cudnn.benchmark, cudnn.deterministic = (False, True) if seed == 0 else (True, False) torch.cuda.manual_seed(seed) torch.cuda.manual_seed_all(seed) # for Multi-GPU, exception safe + torch.backends.cudnn.benchmark = True # for faster training + if deterministic and check_version(torch.__version__, '1.12.0'): # https://github.com/ultralytics/yolov5/pull/8213 + torch.use_deterministic_algorithms(True) + torch.backends.cudnn.deterministic = True + os.environ['CUBLAS_WORKSPACE_CONFIG'] = ':4096:8' + os.environ['PYTHONHASHSEED'] = str(seed) def intersect_dicts(da, db, exclude=()): From 15e82d296720d4be344bf42a34d60ffd57b3eb28 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 2 Sep 2022 16:24:30 +0200 Subject: [PATCH 552/661] Update `TryExcept(msg='...')`` (#9261) --- utils/__init__.py | 4 ++-- utils/autoanchor.py | 2 +- utils/metrics.py | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/utils/__init__.py b/utils/__init__.py index 8bdffd47b3b2..46225c2208ce 100644 --- a/utils/__init__.py +++ b/utils/__init__.py @@ -9,7 +9,7 @@ class TryExcept(contextlib.ContextDecorator): # YOLOv5 TryExcept class. Usage: @TryExcept() decorator or 'with TryExcept():' context manager - def __init__(self, msg='default message here'): + def __init__(self, msg=''): self.msg = msg def __enter__(self): @@ -17,7 +17,7 @@ def __enter__(self): def __exit__(self, exc_type, value, traceback): if value: - print(f'{self.msg}: {value}') + print(f'{self.msg}{value}') return True diff --git a/utils/autoanchor.py b/utils/autoanchor.py index ac17c6cafc90..0b49ab3319c0 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -26,7 +26,7 @@ def check_anchor_order(m): m.anchors[:] = m.anchors.flip(0) -@TryExcept(f'{PREFIX}ERROR:') +@TryExcept(f'{PREFIX}ERROR: ') def check_anchors(dataset, model, thr=4.0, imgsz=640): # Check anchor fit to data, recompute if necessary m = model.module.model[-1] if hasattr(model, 'module') else model.model[-1] # Detect() diff --git a/utils/metrics.py b/utils/metrics.py index de1bf05b326b..ee7d33982cfc 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -186,7 +186,7 @@ def tp_fp(self): # fn = self.matrix.sum(0) - tp # false negatives (missed detections) return tp[:-1], fp[:-1] # remove background class - @TryExcept('WARNING: ConfusionMatrix plot failure') + @TryExcept('WARNING: ConfusionMatrix plot failure: ') def plot(self, normalize=True, save_dir='', names=()): import seaborn as sn From 5cb9fe612a215e0b7f6d99bf39e91cc52ab13c53 Mon Sep 17 00:00:00 2001 From: Victor Sonck Date: Sat, 3 Sep 2022 20:49:25 +0200 Subject: [PATCH 553/661] Make sure best.pt model file is preserved ClearML (#9265) * Make sure best.pt model file is preserved ClearML * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/loggers/__init__.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 1aa8427f9127..3aee35844f52 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -233,7 +233,9 @@ def on_train_end(self, last, best, epoch, results): self.wandb.finish_run() if self.clearml and not self.opt.evolve: - self.clearml.task.update_output_model(model_path=str(best if best.exists() else last), name='Best Model') + self.clearml.task.update_output_model(model_path=str(best if best.exists() else last), + name='Best Model', + auto_delete_file=False) def on_params_update(self, params: dict): # Update hyperparams or configs of the experiment From 63ecce60eab055bd5fec3223ee2b8d8a3d099349 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 4 Sep 2022 01:33:38 +0200 Subject: [PATCH 554/661] DetectMultiBackend improvements (#9269) * Update common.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- models/common.py | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/models/common.py b/models/common.py index d308244c4a44..2e5d5a198e33 100644 --- a/models/common.py +++ b/models/common.py @@ -354,6 +354,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, import onnxruntime providers = ['CUDAExecutionProvider', 'CPUExecutionProvider'] if cuda else ['CPUExecutionProvider'] session = onnxruntime.InferenceSession(w, providers=providers) + output_names = [x.name for x in session.get_outputs()] meta = session.get_modelmeta().custom_metadata_map # metadata if 'stride' in meta: stride, names = int(meta['stride']), eval(meta['names']) @@ -372,9 +373,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, batch_size = batch_dim.get_length() executable_network = ie.compile_model(network, device_name="CPU") # device_name="MYRIAD" for Intel NCS2 output_layer = next(iter(executable_network.outputs)) - meta = Path(w).with_suffix('.yaml') - if meta.exists(): - stride, names = self._load_metadata(meta) # load metadata + stride, names = self._load_metadata(Path(w).with_suffix('.yaml')) # load metadata elif engine: # TensorRT LOGGER.info(f'Loading {w} for TensorRT inference...') import tensorrt as trt # https://developer.nvidia.com/nvidia-tensorrt-download @@ -476,7 +475,7 @@ def forward(self, im, augment=False, visualize=False, val=False): y = self.net.forward() elif self.onnx: # ONNX Runtime im = im.cpu().numpy() # torch to numpy - y = self.session.run([self.session.get_outputs()[0].name], {self.session.get_inputs()[0].name: im})[0] + y = self.session.run(self.output_names, {self.session.get_inputs()[0].name: im})[0] elif self.xml: # OpenVINO im = im.cpu().numpy() # FP32 y = self.executable_network([im])[self.output_layer] @@ -524,7 +523,7 @@ def forward(self, im, augment=False, visualize=False, val=False): y[..., :4] *= [w, h, w, h] # xywh normalized to pixels if isinstance(y, np.ndarray): - y = torch.tensor(y, device=self.device) + y = torch.from_numpy(y).to(self.device) return (y, []) if val else y def warmup(self, imgsz=(1, 3, 640, 640)): @@ -548,10 +547,12 @@ def _model_type(p='path/to/model.pt'): return pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs @staticmethod - def _load_metadata(f='path/to/meta.yaml'): + def _load_metadata(f=Path('path/to/meta.yaml')): # Load metadata from meta.yaml if it exists - d = yaml_load(f) - return d['stride'], d['names'] # assign stride, names + if f.exists(): + d = yaml_load(f) + return d['stride'], d['names'] # assign stride, names + return None, None class AutoShape(nn.Module): From 96c3c7f71d6af51819c270e2752603665680ced7 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 4 Sep 2022 14:01:43 +0200 Subject: [PATCH 555/661] Update DetectMultiBackend for tuple outputs (#9274) Update --- models/common.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/models/common.py b/models/common.py index 2e5d5a198e33..5c82b18f102c 100644 --- a/models/common.py +++ b/models/common.py @@ -465,17 +465,15 @@ def forward(self, im, augment=False, visualize=False, val=False): if self.pt: # PyTorch y = self.model(im, augment=augment, visualize=visualize) if augment or visualize else self.model(im) - if isinstance(y, tuple): - y = y[0] elif self.jit: # TorchScript - y = self.model(im)[0] + y = self.model(im) elif self.dnn: # ONNX OpenCV DNN im = im.cpu().numpy() # torch to numpy self.net.setInput(im) y = self.net.forward() elif self.onnx: # ONNX Runtime im = im.cpu().numpy() # torch to numpy - y = self.session.run(self.output_names, {self.session.get_inputs()[0].name: im})[0] + y = self.session.run(self.output_names, {self.session.get_inputs()[0].name: im}) elif self.xml: # OpenVINO im = im.cpu().numpy() # FP32 y = self.executable_network([im])[self.output_layer] @@ -522,6 +520,8 @@ def forward(self, im, augment=False, visualize=False, val=False): y = (y.astype(np.float32) - zero_point) * scale # re-scale y[..., :4] *= [w, h, w, h] # xywh normalized to pixels + if isinstance(y, (list, tuple)): + y = y[0] if isinstance(y, np.ndarray): y = torch.from_numpy(y).to(self.device) return (y, []) if val else y From 7aa263c5f2f526472435babf86ddd33eed1dbd78 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 4 Sep 2022 15:39:57 +0200 Subject: [PATCH 556/661] Update DetectMultiBackend for tuple outputs 2 (#9275) * Update DetectMultiBackend for tuple outputs 2 Signed-off-by: Glenn Jocher * Update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update * Update * Update Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- models/common.py | 12 +++++++----- utils/general.py | 3 +++ val.py | 4 ++-- 3 files changed, 12 insertions(+), 7 deletions(-) diff --git a/models/common.py b/models/common.py index 5c82b18f102c..7ac3a4a29672 100644 --- a/models/common.py +++ b/models/common.py @@ -457,7 +457,7 @@ def wrap_frozen_graph(gd, inputs, outputs): self.__dict__.update(locals()) # assign all variables to self - def forward(self, im, augment=False, visualize=False, val=False): + def forward(self, im, augment=False, visualize=False): # YOLOv5 MultiBackend inference b, ch, h, w = im.shape # batch, channel, height, width if self.fp16 and im.dtype != torch.float16: @@ -521,10 +521,12 @@ def forward(self, im, augment=False, visualize=False, val=False): y[..., :4] *= [w, h, w, h] # xywh normalized to pixels if isinstance(y, (list, tuple)): - y = y[0] - if isinstance(y, np.ndarray): - y = torch.from_numpy(y).to(self.device) - return (y, []) if val else y + return self.from_numpy(y[0]) if len(y) == 1 else [self.from_numpy(x) for x in y] + else: + return self.from_numpy(y) + + def from_numpy(self, x): + return torch.from_numpy(x).to(self.device) if isinstance(x, np.ndarray) else x def warmup(self, imgsz=(1, 3, 640, 640)): # Warmup model by running inference once diff --git a/utils/general.py b/utils/general.py index 25a1a1456009..cae63fd9dd21 100755 --- a/utils/general.py +++ b/utils/general.py @@ -813,6 +813,9 @@ def non_max_suppression(prediction, list of detections, on (n,6) tensor per image [xyxy, conf, cls] """ + if isinstance(prediction, (list, tuple)): # YOLOv5 model in validation model, output = (inference_out, loss_out) + prediction = prediction[0] # select only inference output + bs = prediction.shape[0] # batch size nc = prediction.shape[2] - 5 # number of classes xc = prediction[..., 4] > conf_thres # candidates diff --git a/val.py b/val.py index 58b9c9e1bec0..5427ee7b3619 100644 --- a/val.py +++ b/val.py @@ -204,11 +204,11 @@ def run( # Inference with dt[1]: - out, train_out = model(im) if training else model(im, augment=augment, val=True) # inference, loss outputs + out, train_out = model(im) if compute_loss else (model(im, augment=augment), None) # Loss if compute_loss: - loss += compute_loss([x.float() for x in train_out], targets)[1] # box, obj, cls + loss += compute_loss(train_out, targets)[1] # box, obj, cls # NMS targets[:, 2:] *= torch.tensor((width, height, width, height), device=device) # to pixels From e45d335bbc4a891a2a9f49311f4448e252d3d88f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 4 Sep 2022 16:35:16 +0200 Subject: [PATCH 557/661] Update benchmarks CI with `--hard-fail` min metric floor (#9276) * Update benchmarks CI with min metric floor Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher * Update benchmarks.py Signed-off-by: Glenn Jocher * Update benchmarks.py Signed-off-by: Glenn Jocher * Update benchmarks.py Signed-off-by: Glenn Jocher * Update benchmarks.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/workflows/ci-testing.yml | 2 +- utils/benchmarks.py | 8 ++++++-- 2 files changed, 7 insertions(+), 3 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 4ef930c61233..6fb277676959 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -39,7 +39,7 @@ jobs: pip list - name: Run benchmarks run: | - python utils/benchmarks.py --weights ${{ matrix.model }}.pt --img 320 --hard-fail + python utils/benchmarks.py --weights ${{ matrix.model }}.pt --img 320 --hard-fail 0.29 Tests: timeout-minutes: 60 diff --git a/utils/benchmarks.py b/utils/benchmarks.py index d412653c866f..d5f4c1d61fbe 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -92,10 +92,14 @@ def run( LOGGER.info('\n') parse_opt() notebook_init() # print system info - c = ['Format', 'Size (MB)', 'mAP@0.5:0.95', 'Inference time (ms)'] if map else ['Format', 'Export', '', ''] + c = ['Format', 'Size (MB)', 'mAP50-95', 'Inference time (ms)'] if map else ['Format', 'Export', '', ''] py = pd.DataFrame(y, columns=c) LOGGER.info(f'\nBenchmarks complete ({time.time() - t:.2f}s)') LOGGER.info(str(py if map else py.iloc[:, :2])) + if hard_fail and isinstance(hard_fail, str): + metrics = py['mAP50-95'].array # values to compare to floor + floor = eval(hard_fail) # minimum metric floor to pass, i.e. = 0.29 mAP for YOLOv5n + assert all(x > floor for x in metrics if pd.notna(x)), f'HARD FAIL: mAP50-95 < floor {floor}' return py @@ -141,7 +145,7 @@ def parse_opt(): parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') parser.add_argument('--test', action='store_true', help='test exports only') parser.add_argument('--pt-only', action='store_true', help='test PyTorch only') - parser.add_argument('--hard-fail', action='store_true', help='throw error on benchmark failure') + parser.add_argument('--hard-fail', nargs='?', const=True, default=False, help='Exception on error or < min metric') opt = parser.parse_args() opt.data = check_yaml(opt.data) # check YAML print_args(vars(opt)) From 1aea74cddbc78e7f79dac07090cb157dfc24dbcc Mon Sep 17 00:00:00 2001 From: VELC Date: Sun, 4 Sep 2022 17:15:53 +0200 Subject: [PATCH 558/661] Add new `--vid-stride` inference parameter for videos (#9256) * fps feature/skip frame added * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * predict.py updates * Update dataloaders.py Signed-off-by: Glenn Jocher * Update dataloaders.py Signed-off-by: Glenn Jocher * remove unused attribute Signed-off-by: Glenn Jocher * Cleanup Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update predict.py Signed-off-by: Glenn Jocher * Update detect.py Signed-off-by: Glenn Jocher * Update dataloaders.py Signed-off-by: Glenn Jocher * Rename skip_frame to vid_stride * cleanup * cleanup2 Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- classify/predict.py | 6 ++++-- detect.py | 6 ++++-- utils/dataloaders.py | 15 +++++++++------ 3 files changed, 17 insertions(+), 10 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 76115c75029f..701b5b1ac92d 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -66,6 +66,7 @@ def run( exist_ok=False, # existing project/name ok, do not increment half=False, # use FP16 half-precision inference dnn=False, # use OpenCV DNN for ONNX inference + vid_stride=1, # video frame-rate stride ): source = str(source) save_img = not nosave and not source.endswith('.txt') # save inference images @@ -88,10 +89,10 @@ def run( # Dataloader if webcam: view_img = check_imshow() - dataset = LoadStreams(source, img_size=imgsz, transforms=classify_transforms(imgsz[0])) + dataset = LoadStreams(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) bs = len(dataset) # batch_size else: - dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz[0])) + dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) bs = 1 # batch_size vid_path, vid_writer = [None] * bs, [None] * bs @@ -196,6 +197,7 @@ def parse_opt(): parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') + parser.add_argument('--vid-stride', type=int, default=1, help='video frame-rate stride') opt = parser.parse_args() opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand print_args(vars(opt)) diff --git a/detect.py b/detect.py index cf75d0f11c92..69a1bf13aac6 100644 --- a/detect.py +++ b/detect.py @@ -74,6 +74,7 @@ def run( hide_conf=False, # hide confidences half=False, # use FP16 half-precision inference dnn=False, # use OpenCV DNN for ONNX inference + vid_stride=1, # video frame-rate stride ): source = str(source) save_img = not nosave and not source.endswith('.txt') # save inference images @@ -96,10 +97,10 @@ def run( # Dataloader if webcam: view_img = check_imshow() - dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt) + dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) bs = len(dataset) # batch_size else: - dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt) + dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) bs = 1 # batch_size vid_path, vid_writer = [None] * bs, [None] * bs @@ -236,6 +237,7 @@ def parse_opt(): parser.add_argument('--hide-conf', default=False, action='store_true', help='hide confidences') parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') + parser.add_argument('--vid-stride', type=int, default=1, help='video frame-rate stride') opt = parser.parse_args() opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand print_args(vars(opt)) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 38ae3399ce26..c1ad1f1a4b83 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -187,7 +187,7 @@ def __iter__(self): class LoadImages: # YOLOv5 image/video dataloader, i.e. `python detect.py --source image.jpg/vid.mp4` - def __init__(self, path, img_size=640, stride=32, auto=True, transforms=None): + def __init__(self, path, img_size=640, stride=32, auto=True, transforms=None, vid_stride=1): files = [] for p in sorted(path) if isinstance(path, (list, tuple)) else [path]: p = str(Path(p).resolve()) @@ -212,6 +212,7 @@ def __init__(self, path, img_size=640, stride=32, auto=True, transforms=None): self.mode = 'image' self.auto = auto self.transforms = transforms # optional + self.vid_stride = vid_stride # video frame-rate stride if any(videos): self._new_video(videos[0]) # new video else: @@ -232,6 +233,7 @@ def __next__(self): # Read video self.mode = 'video' ret_val, im0 = self.cap.read() + self.cap.set(cv2.CAP_PROP_POS_FRAMES, self.vid_stride * (self.frame + 1)) # read at vid_stride while not ret_val: self.count += 1 self.cap.release() @@ -242,7 +244,7 @@ def __next__(self): ret_val, im0 = self.cap.read() self.frame += 1 - # im0 = self._cv2_rotate(im0) # for use if cv2 auto rotation is False + # im0 = self._cv2_rotate(im0) # for use if cv2 autorotation is False s = f'video {self.count + 1}/{self.nf} ({self.frame}/{self.frames}) {path}: ' else: @@ -265,7 +267,7 @@ def _new_video(self, path): # Create a new video capture object self.frame = 0 self.cap = cv2.VideoCapture(path) - self.frames = int(self.cap.get(cv2.CAP_PROP_FRAME_COUNT)) + self.frames = int(self.cap.get(cv2.CAP_PROP_FRAME_COUNT) / self.vid_stride) self.orientation = int(self.cap.get(cv2.CAP_PROP_ORIENTATION_META)) # rotation degrees # self.cap.set(cv2.CAP_PROP_ORIENTATION_AUTO, 0) # disable https://github.com/ultralytics/yolov5/issues/8493 @@ -285,11 +287,12 @@ def __len__(self): class LoadStreams: # YOLOv5 streamloader, i.e. `python detect.py --source 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP streams` - def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True, transforms=None): + def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True, transforms=None, vid_stride=1): torch.backends.cudnn.benchmark = True # faster for fixed-size inference self.mode = 'stream' self.img_size = img_size self.stride = stride + self.vid_stride = vid_stride # video frame-rate stride sources = Path(sources).read_text().rsplit() if Path(sources).is_file() else [sources] n = len(sources) self.sources = [clean_str(x) for x in sources] # clean source names for later @@ -329,11 +332,11 @@ def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True, tr def update(self, i, cap, stream): # Read stream `i` frames in daemon thread - n, f, read = 0, self.frames[i], 1 # frame number, frame array, inference every 'read' frame + n, f = 0, self.frames[i] # frame number, frame array while cap.isOpened() and n < f: n += 1 cap.grab() # .read() = .grab() followed by .retrieve() - if n % read == 0: + if n % self.vid_stride == 0: success, im = cap.retrieve() if success: self.imgs[i] = im From 32794c130bc0de9cbd1fe34819b7032138bbd81d Mon Sep 17 00:00:00 2001 From: "pre-commit-ci[bot]" <66853113+pre-commit-ci[bot]@users.noreply.github.com> Date: Tue, 6 Sep 2022 19:00:26 +0300 Subject: [PATCH 559/661] [pre-commit.ci] pre-commit suggestions (#9295) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit updates: - [github.com/executablebooks/mdformat: 0.7.14 → 0.7.16](https://github.com/executablebooks/mdformat/compare/0.7.14...0.7.16) - [github.com/asottile/yesqa: v1.3.0 → v1.4.0](https://github.com/asottile/yesqa/compare/v1.3.0...v1.4.0) - [github.com/PyCQA/flake8: 5.0.2 → 5.0.4](https://github.com/PyCQA/flake8/compare/5.0.2...5.0.4) Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- .pre-commit-config.yaml | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 43aca019feb1..ba8005535397 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -43,7 +43,7 @@ repos: name: YAPF formatting - repo: https://github.com/executablebooks/mdformat - rev: 0.7.14 + rev: 0.7.16 hooks: - id: mdformat name: MD formatting @@ -53,12 +53,12 @@ repos: exclude: "README.md|README_cn.md" - repo: https://github.com/asottile/yesqa - rev: v1.3.0 + rev: v1.4.0 hooks: - id: yesqa - repo: https://github.com/PyCQA/flake8 - rev: 5.0.2 + rev: 5.0.4 hooks: - id: flake8 name: PEP8 From 5a134e06530a8c24fdb9774c2c4ab0b513b08260 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 7 Sep 2022 10:11:30 +0300 Subject: [PATCH 560/661] Replace deprecated `np.int` with `int` (#9307) Per ``` /content/yolov5/utils/dataloaders.py:458: DeprecationWarning: `np.int` is a deprecated alias for the builtin `int`. To silence this warning, use `int` by itself. Doing this will not modify any behavior and is safe. When replacing `np.int`, you may wish to use e.g. `np.int64` or `np.int32` to specify the precision. If you wish to review your current use, check the release note link for additional information. Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations ``` Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/dataloaders.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index c1ad1f1a4b83..d8ef11fd94b4 100755 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -455,7 +455,7 @@ def __init__(self, self.im_files = list(cache.keys()) # update self.label_files = img2label_paths(cache.keys()) # update n = len(shapes) # number of images - bi = np.floor(np.arange(n) / batch_size).astype(np.int) # batch index + bi = np.floor(np.arange(n) / batch_size).astype(int) # batch index nb = bi[-1] + 1 # number of batches self.batch = bi # batch index of image self.n = n @@ -497,7 +497,7 @@ def __init__(self, elif mini > 1: shapes[i] = [1, 1 / mini] - self.batch_shapes = np.ceil(np.array(shapes) * img_size / stride + pad).astype(np.int) * stride + self.batch_shapes = np.ceil(np.array(shapes) * img_size / stride + pad).astype(int) * stride # Cache images into RAM/disk for faster training (WARNING: large datasets may exceed system resources) self.ims = [None] * n @@ -867,7 +867,7 @@ def extract_boxes(path=DATASETS_DIR / 'coco128'): # from utils.dataloaders impo b = x[1:] * [w, h, w, h] # box # b[2:] = b[2:].max() # rectangle to square b[2:] = b[2:] * 1.2 + 3 # pad - b = xywh2xyxy(b.reshape(-1, 4)).ravel().astype(np.int) + b = xywh2xyxy(b.reshape(-1, 4)).ravel().astype(int) b[[0, 2]] = np.clip(b[[0, 2]], 0, w) # clip boxes outside of image b[[1, 3]] = np.clip(b[[1, 3]], 0, h) From 903b239f1338e7ad8b12dd8e4a3c53f4f362e07f Mon Sep 17 00:00:00 2001 From: Dhruv Nair Date: Wed, 7 Sep 2022 11:28:46 -0400 Subject: [PATCH 561/661] Comet Logging and Visualization Integration (#9232) * add comet to logger interface * add comet logger * add support for updated parameters * clean up offline logger creation * update callback args for comet logger * add comet optimizer * add optimizer config * add comet README * update tutorial notebook with Comet section * add option to log class level metrics * add support for class level metrics and confusion matrix * handle errors when adding files to artifacts * fix typo * clean resume workflow * updates for HPO * update comet README * fix typo in comet README * update code snippets in comet README * update comet links in tutorial * updated links * change optimizer batch size param and update comet README image * update comet section in tutorial * use prexisting cmd line flags to configure logger * update artifact upload/download flow * remove come remove comet logger specific cmd line args * move downloading weights into comet logger code * remove extra argparse * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * change checkpoint logging flow to follow offline logger * update resume flow * add comet logger to remote dataset property * update cmd line args in hpo * set types for integer/float env variables * update README * fix typo in README * default to always logging model predictions * Update tutorial.ipynb * Update train.py * Add Comet to Integrations table * Update README.md * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: Ayush Chaurasia Co-authored-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- README.md | 59 +-- train.py | 17 +- tutorial.ipynb | 39 +- utils/loggers/__init__.py | 77 +++- utils/loggers/comet/README.md | 256 +++++++++++ utils/loggers/comet/__init__.py | 496 ++++++++++++++++++++++ utils/loggers/comet/comet_utils.py | 150 +++++++ utils/loggers/comet/hpo.py | 118 +++++ utils/loggers/comet/optimizer_config.json | 209 +++++++++ val.py | 4 +- 10 files changed, 1376 insertions(+), 49 deletions(-) create mode 100644 utils/loggers/comet/README.md create mode 100644 utils/loggers/comet/__init__.py create mode 100644 utils/loggers/comet/comet_utils.py create mode 100644 utils/loggers/comet/hpo.py create mode 100644 utils/loggers/comet/optimizer_config.json diff --git a/README.md b/README.md index 1d6b4e153d82..7763d174f92b 100644 --- a/README.md +++ b/README.md @@ -160,46 +160,31 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12
-##
Environments
- -Get started in seconds with our verified environments. Click each icon below for details. - -
- - - - - - - - - - - - - - -
##
Integrations
+ +
+ + + - + - + - +
-|Deci ⭐ NEW|ClearML ⭐ NEW|Roboflow|Weights & Biases -|:-:|:-:|:-:|:-:| -|Automatically compile and quantize YOLOv5 for better inference performance in one click at [Deci](https://bit.ly/yolov5-deci-platform)|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) +|Comet ⭐ NEW|Deci ⭐ NEW|ClearML ⭐ NEW|Roboflow|Weights & Biases +|:-:|:-:|:-:|:-:|:-:| +|Visualize model metrics and predictions and upload models and datasets in realtime with [Comet](https://www.comet.com/site/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration)|Automatically compile and quantize YOLOv5 for better inference performance in one click at [Deci](https://bit.ly/yolov5-deci-platform)|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) ##
Why YOLOv5
@@ -323,6 +308,28 @@ python export.py --weights yolov5s-cls.pt resnet50.pt efficientnet_b0.pt --inclu
+##
Environments
+ +Get started in seconds with our verified environments. Click each icon below for details. + +
+ + + + + + + + + + + + + + +
+ + ##
Contribute
We love your input! We want to make contributing to YOLOv5 as easy and transparent as possible. Please see our [Contributing Guide](CONTRIBUTING.md) to get started, and fill out the [YOLOv5 Survey](https://ultralytics.com/survey?utm_source=github&utm_medium=social&utm_campaign=Survey) to send us feedback on your experiences. Thank you to all our contributors! diff --git a/train.py b/train.py index 29293aa612cf..e16c17c499f0 100644 --- a/train.py +++ b/train.py @@ -52,6 +52,7 @@ init_seeds, intersect_dicts, labels_to_class_weights, labels_to_image_weights, methods, one_cycle, print_args, print_mutation, strip_optimizer, yaml_save) from utils.loggers import Loggers +from utils.loggers.comet.comet_utils import check_comet_resume from utils.loggers.wandb.wandb_utils import check_wandb_resume from utils.loss import ComputeLoss from utils.metrics import fitness @@ -330,7 +331,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio mem = f'{torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0:.3g}G' # (GB) pbar.set_description(('%11s' * 2 + '%11.4g' * 5) % (f'{epoch}/{epochs - 1}', mem, *mloss, targets.shape[0], imgs.shape[-1])) - callbacks.run('on_train_batch_end', model, ni, imgs, targets, paths) + callbacks.run('on_train_batch_end', model, ni, imgs, targets, paths, list(mloss)) if callbacks.stop_training: return # end batch ------------------------------------------------------------------------------------------------ @@ -465,11 +466,11 @@ def parse_opt(known=False): parser.add_argument('--seed', type=int, default=0, help='Global training seed') parser.add_argument('--local_rank', type=int, default=-1, help='Automatic DDP Multi-GPU argument, do not modify') - # Weights & Biases arguments - parser.add_argument('--entity', default=None, help='W&B: Entity') - parser.add_argument('--upload_dataset', nargs='?', const=True, default=False, help='W&B: Upload data, "val" option') - parser.add_argument('--bbox_interval', type=int, default=-1, help='W&B: Set bounding-box image logging interval') - parser.add_argument('--artifact_alias', type=str, default='latest', help='W&B: Version of dataset artifact to use') + # Logger arguments + parser.add_argument('--entity', default=None, help='Entity') + parser.add_argument('--upload_dataset', nargs='?', const=True, default=False, help='Upload data, "val" option') + parser.add_argument('--bbox_interval', type=int, default=-1, help='Set bounding-box image logging interval') + parser.add_argument('--artifact_alias', type=str, default='latest', help='Version of dataset artifact to use') return parser.parse_known_args()[0] if known else parser.parse_args() @@ -481,8 +482,8 @@ def main(opt, callbacks=Callbacks()): check_git_status() check_requirements() - # Resume - if opt.resume and not (check_wandb_resume(opt) or opt.evolve): # resume from specified or most recent last.pt + # Resume (from specified or most recent last.pt) + if opt.resume and not check_wandb_resume(opt) and not check_comet_resume(opt) or opt.evolve: last = Path(check_file(opt.resume) if isinstance(opt.resume, str) else get_latest_run()) opt_yaml = last.parent.parent / 'opt.yaml' # train options yaml opt_data = opt.data # original dataset diff --git a/tutorial.ipynb b/tutorial.ipynb index 12840063b1f1..957437b2be6d 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -413,7 +413,7 @@ "import utils\n", "display = utils.notebook_init() # checks" ], - "execution_count": 1, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -465,7 +465,7 @@ "!python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data/images\n", "# display.Image(filename='runs/detect/exp/zidane.jpg', width=600)" ], - "execution_count": 2, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -535,7 +535,7 @@ "torch.hub.download_url_to_file('https://ultralytics.com/assets/coco2017val.zip', 'tmp.zip') # download (780M - 5000 images)\n", "!unzip -q tmp.zip -d ../datasets && rm tmp.zip # unzip" ], - "execution_count": 3, + "execution_count": null, "outputs": [ { "output_type": "display_data", @@ -566,7 +566,7 @@ "# Validate YOLOv5s on COCO val\n", "!python val.py --weights yolov5s.pt --data coco.yaml --img 640 --half" ], - "execution_count": 4, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -653,11 +653,14 @@ "cell_type": "code", "source": [ "#@title Select YOLOv5 🚀 logger {run: 'auto'}\n", - "logger = 'TensorBoard' #@param ['TensorBoard', 'ClearML', 'W&B']\n", + "logger = 'TensorBoard' #@param ['TensorBoard', 'Comet', 'ClearML', 'W&B']\n", "\n", "if logger == 'TensorBoard':\n", " %load_ext tensorboard\n", " %tensorboard --logdir runs/train\n", + "elif logger == 'Comet':\n", + " %pip install -q comet_ml\n", + " import comet_ml; comet_ml.init()\n", "elif logger == 'ClearML':\n", " %pip install -q clearml && clearml-init\n", "elif logger == 'W&B':\n", @@ -683,7 +686,7 @@ "# Train YOLOv5s on COCO128 for 3 epochs\n", "!python train.py --img 640 --batch 16 --epochs 3 --data coco128.yaml --weights yolov5s.pt --cache" ], - "execution_count": 5, + "execution_count": null, "outputs": [ { "output_type": "stream", @@ -857,6 +860,28 @@ "# 4. Visualize" ] }, + { + "cell_type": "markdown", + "source": [ + "## Comet Logging and Visualization 🌟 NEW\n", + "[Comet](https://www.comet.com/site/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration) is now fully integrated with YOLOv5. Track and visualize model metrics in real time, save your hyperparameters, datasets, and model checkpoints, and visualize your model predictions with [Comet Custom Panels](https://www.comet.com/docs/v2/guides/comet-dashboard/code-panels/about-panels/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration)! Comet makes sure you never lose track of your work and makes it easy to share results and collaborate across teams of all sizes! \n", + "\n", + "Getting started is easy:\n", + "```shell\n", + "pip install comet_ml # 1. install\n", + "export COMET_API_KEY= # 2. paste API key\n", + "python train.py --img 640 --epochs 3 --data coco128.yaml --weights yolov5s.pt # 3. train\n", + "```\n", + "\n", + "To learn more about all of the supported Comet features for this integration, check out the [Comet Tutorial](https://github.com/ultralytics/yolov5/tree/master/utils/loggers/comet). If you'd like to learn more about Comet, head over to our [documentation](https://www.comet.com/docs/v2/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration). Get started by trying out the Comet Colab Notebook:\n", + "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1RG0WOQyxlDlo5Km8GogJpIEJlg_5lyYO?usp=sharing)\n", + "\n", + "\"yolo-ui\"" + ], + "metadata": { + "id": "nWOsI5wJR1o3" + } + }, { "cell_type": "markdown", "source": [ @@ -1096,4 +1121,4 @@ "outputs": [] } ] -} \ No newline at end of file +} diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 3aee35844f52..f29debb76907 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -17,7 +17,7 @@ from utils.plots import plot_images, plot_labels, plot_results from utils.torch_utils import de_parallel -LOGGERS = ('csv', 'tb', 'wandb', 'clearml') # *.csv, TensorBoard, Weights & Biases, ClearML +LOGGERS = ('csv', 'tb', 'wandb', 'clearml', 'comet') # *.csv, TensorBoard, Weights & Biases, ClearML RANK = int(os.getenv('RANK', -1)) try: @@ -41,6 +41,18 @@ except (ImportError, AssertionError): clearml = None +try: + if RANK not in [0, -1]: + comet_ml = None + else: + import comet_ml + + assert hasattr(comet_ml, '__version__') # verify package import not local dir + from utils.loggers.comet import CometLogger + +except (ModuleNotFoundError, ImportError, AssertionError): + comet_ml = None + class Loggers(): # YOLOv5 Loggers class @@ -80,7 +92,10 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, prefix = colorstr('ClearML: ') s = f"{prefix}run 'pip install clearml' to automatically track, visualize and remotely train YOLOv5 🚀 in ClearML" self.logger.info(s) - + if not comet_ml: + prefix = colorstr('Comet: ') + s = f"{prefix}run 'pip install comet_ml' to automatically track and visualize YOLOv5 🚀 runs in Comet" + self.logger.info(s) # TensorBoard s = self.save_dir if 'tb' in self.include and not self.opt.evolve: @@ -107,6 +122,18 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, else: self.clearml = None + # Comet + if comet_ml and 'comet' in self.include: + if isinstance(self.opt.resume, str) and self.opt.resume.startswith("comet://"): + run_id = self.opt.resume.split("/")[-1] + self.comet_logger = CometLogger(self.opt, self.hyp, run_id=run_id) + + else: + self.comet_logger = CometLogger(self.opt, self.hyp) + + else: + self.comet_logger = None + @property def remote_dataset(self): # Get data_dict if custom dataset artifact link is provided @@ -115,12 +142,18 @@ def remote_dataset(self): data_dict = self.clearml.data_dict if self.wandb: data_dict = self.wandb.data_dict + if self.comet_logger: + data_dict = self.comet_logger.data_dict return data_dict def on_train_start(self): - # Callback runs on train start - pass + if self.comet_logger: + self.comet_logger.on_train_start() + + def on_pretrain_routine_start(self): + if self.comet_logger: + self.comet_logger.on_pretrain_routine_start() def on_pretrain_routine_end(self, labels, names): # Callback runs on pre-train routine end @@ -131,8 +164,11 @@ def on_pretrain_routine_end(self, labels, names): self.wandb.log({"Labels": [wandb.Image(str(x), caption=x.name) for x in paths]}) # if self.clearml: # pass # ClearML saves these images automatically using hooks + if self.comet_logger: + self.comet_logger.on_pretrain_routine_end(paths) - def on_train_batch_end(self, model, ni, imgs, targets, paths): + def on_train_batch_end(self, model, ni, imgs, targets, paths, vals): + log_dict = dict(zip(self.keys[0:3], vals)) # Callback runs on train batch end # ni: number integrated batches (since train start) if self.plots: @@ -148,11 +184,21 @@ def on_train_batch_end(self, model, ni, imgs, targets, paths): if self.clearml: self.clearml.log_debug_samples(files, title='Mosaics') + if self.comet_logger: + self.comet_logger.on_train_batch_end(log_dict, step=ni) + def on_train_epoch_end(self, epoch): # Callback runs on train epoch end if self.wandb: self.wandb.current_epoch = epoch + 1 + if self.comet_logger: + self.comet_logger.on_train_epoch_end(epoch) + + def on_val_start(self): + if self.comet_logger: + self.comet_logger.on_val_start() + def on_val_image_end(self, pred, predn, path, names, im): # Callback runs on val image end if self.wandb: @@ -160,7 +206,11 @@ def on_val_image_end(self, pred, predn, path, names, im): if self.clearml: self.clearml.log_image_with_boxes(path, pred, names, im) - def on_val_end(self): + def on_val_batch_end(self, batch_i, im, targets, paths, shapes, out): + if self.comet_logger: + self.comet_logger.on_val_batch_end(batch_i, im, targets, paths, shapes, out) + + def on_val_end(self, nt, tp, fp, p, r, f1, ap, ap50, ap_class, confusion_matrix): # Callback runs on val end if self.wandb or self.clearml: files = sorted(self.save_dir.glob('val*.jpg')) @@ -169,6 +219,9 @@ def on_val_end(self): if self.clearml: self.clearml.log_debug_samples(files, title='Validation') + if self.comet_logger: + self.comet_logger.on_val_end(nt, tp, fp, p, r, f1, ap, ap50, ap_class, confusion_matrix) + def on_fit_epoch_end(self, vals, epoch, best_fitness, fi): # Callback runs at the end of each fit (train+val) epoch x = dict(zip(self.keys, vals)) @@ -199,6 +252,9 @@ def on_fit_epoch_end(self, vals, epoch, best_fitness, fi): self.clearml.current_epoch_logged_images = set() # reset epoch image limit self.clearml.current_epoch += 1 + if self.comet_logger: + self.comet_logger.on_fit_epoch_end(x, epoch=epoch) + def on_model_save(self, last, epoch, final_epoch, best_fitness, fi): # Callback runs on model save event if (epoch + 1) % self.opt.save_period == 0 and not final_epoch and self.opt.save_period != -1: @@ -209,6 +265,9 @@ def on_model_save(self, last, epoch, final_epoch, best_fitness, fi): model_name='Latest Model', auto_delete_file=False) + if self.comet_logger: + self.comet_logger.on_model_save(last, epoch, final_epoch, best_fitness, fi) + def on_train_end(self, last, best, epoch, results): # Callback runs on training end, i.e. saving best model if self.plots: @@ -237,10 +296,16 @@ def on_train_end(self, last, best, epoch, results): name='Best Model', auto_delete_file=False) + if self.comet_logger: + final_results = dict(zip(self.keys[3:10], results)) + self.comet_logger.on_train_end(files, self.save_dir, last, best, epoch, final_results) + def on_params_update(self, params: dict): # Update hyperparams or configs of the experiment if self.wandb: self.wandb.wandb_run.config.update(params, allow_val_change=True) + if self.comet_logger: + self.comet_logger.on_params_update(params) class GenericLogger: diff --git a/utils/loggers/comet/README.md b/utils/loggers/comet/README.md new file mode 100644 index 000000000000..7b0b8e0e2f09 --- /dev/null +++ b/utils/loggers/comet/README.md @@ -0,0 +1,256 @@ + + +# YOLOv5 with Comet + +This guide will cover how to use YOLOv5 with [Comet](https://www.comet.com/site/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration) + +# About Comet + +Comet builds tools that help data scientists, engineers, and team leaders accelerate and optimize machine learning and deep learning models. + +Track and visualize model metrics in real time, save your hyperparameters, datasets, and model checkpoints, and visualize your model predictions with [Comet Custom Panels](https://www.comet.com/examples/comet-example-yolov5?shareable=YcwMiJaZSXfcEXpGOHDD12vA1&ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration)! +Comet makes sure you never lose track of your work and makes it easy to share results and collaborate across teams of all sizes! + +# Getting Started + +## Install Comet + +```shell +pip install comet_ml +``` + +## Configure Comet Credentials + +There are two ways to configure Comet with YOLOv5. + +You can either set your credentials through enviroment variables + +**Environment Variables** + +```shell +export COMET_API_KEY= +export COMET_PROJECT_NAME= # This will default to 'yolov5' +``` + +Or create a `.comet.config` file in your working directory and set your credentials there. + +**Comet Configuration File** + +``` +[comet] +api_key= +project_name= # This will default to 'yolov5' +``` + +## Run the Training Script + +```shell +# Train YOLOv5s on COCO128 for 5 epochs +python train.py --img 640 --batch 16 --epochs 5 --data coco128.yaml --weights yolov5s.pt +``` + +That's it! Comet will automatically log your hyperparameters, command line arguments, training and valiation metrics. You can visualize and analyze your runs in the Comet UI + +yolo-ui + +# Try out an Example! +Check out an example of a [completed run here](https://www.comet.com/examples/comet-example-yolov5/a0e29e0e9b984e4a822db2a62d0cb357?experiment-tab=chart&showOutliers=true&smoothing=0&transformY=smoothing&xAxis=step&ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration) + +Or better yet, try it out yourself in this Colab Notebook + +[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1RG0WOQyxlDlo5Km8GogJpIEJlg_5lyYO?usp=sharing) + +# Log automatically + +By default, Comet will log the following items + +## Metrics +- Box Loss, Object Loss, Classification Loss for the training and validation data +- mAP_0.5, mAP_0.5:0.95 metrics for the validation data. +- Precision and Recall for the validation data + +## Parameters + +- Model Hyperparameters +- All parameters passed through the command line options + +## Visualizations + +- Confusion Matrix of the model predictions on the validation data +- Plots for the PR and F1 curves across all classes +- Correlogram of the Class Labels + +# Configure Comet Logging + +Comet can be configured to log additional data either through command line flags passed to the training script +or through environment variables. + +```shell +export COMET_MODE=online # Set whether to run Comet in 'online' or 'offline' mode. Defaults to online +export COMET_MODEL_NAME= #Set the name for the saved model. Defaults to yolov5 +export COMET_LOG_CONFUSION_MATRIX=false # Set to disable logging a Comet Confusion Matrix. Defaults to true +export COMET_MAX_IMAGE_UPLOADS= # Controls how many total image predictions to log to Comet. Defaults to 100. +export COMET_LOG_PER_CLASS_METRICS=true # Set to log evaluation metrics for each detected class at the end of training. Defaults to false +export COMET_DEFAULT_CHECKPOINT_FILENAME= # Set this if you would like to resume training from a different checkpoint. Defaults to 'last.pt' +export COMET_LOG_BATCH_LEVEL_METRICS=true # Set this if you would like to log training metrics at the batch level. Defaults to false. +export COMET_LOG_PREDICTIONS=true # Set this to false to disable logging model predictions +``` + +## Logging Checkpoints with Comet + +Logging Models to Comet is disabled by default. To enable it, pass the `save-period` argument to the training script. This will save the +logged checkpoints to Comet based on the interval value provided by `save-period` + +```shell +python train.py \ +--img 640 \ +--batch 16 \ +--epochs 5 \ +--data coco128.yaml \ +--weights yolov5s.pt \ +--save-period 1 +``` + +## Logging Model Predictions + +By default, model predictions (images, ground truth labels and bounding boxes) will be logged to Comet. + +You can control the frequency of logged predictions and the associated images by passing the `bbox_interval` command line argument. Predictions can be visualized using Comet's Object Detection Custom Panel. This frequency corresponds to every Nth batch of data per epoch. In the example below, we are logging every 2nd batch of data for each epoch. + +**Note:** The YOLOv5 validation dataloader will default to a batch size of 32, so you will have to set the logging frequency accordingly. + +Here is an [example project using the Panel](https://www.comet.com/examples/comet-example-yolov5?shareable=YcwMiJaZSXfcEXpGOHDD12vA1&ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration) + + +```shell +python train.py \ +--img 640 \ +--batch 16 \ +--epochs 5 \ +--data coco128.yaml \ +--weights yolov5s.pt \ +--bbox_interval 2 +``` + +### Controlling the number of Prediction Images logged to Comet + +When logging predictions from YOLOv5, Comet will log the images associated with each set of predictions. By default a maximum of 100 validation images are logged. You can increase or decrease this number using the `COMET_MAX_IMAGE_UPLOADS` environment variable. + +```shell +env COMET_MAX_IMAGE_UPLOADS=200 python train.py \ +--img 640 \ +--batch 16 \ +--epochs 5 \ +--data coco128.yaml \ +--weights yolov5s.pt \ +--bbox_interval 1 +``` + +### Logging Class Level Metrics + +Use the `COMET_LOG_PER_CLASS_METRICS` environment variable to log mAP, precision, recall, f1 for each class. + +```shell +env COMET_LOG_PER_CLASS_METRICS=true python train.py \ +--img 640 \ +--batch 16 \ +--epochs 5 \ +--data coco128.yaml \ +--weights yolov5s.pt +``` + +## Uploading a Dataset to Comet Artifacts + +If you would like to store your data using [Comet Artifacts](https://www.comet.com/docs/v2/guides/data-management/using-artifacts/#learn-more?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration), you can do so using the `upload_dataset` flag. + +The dataset be organized in the way described in the [YOLOv5 documentation](https://docs.ultralytics.com/tutorials/train-custom-datasets/#3-organize-directories). The dataset config `yaml` file must follow the same format as that of the `coco128.yaml` file. + +```shell +python train.py \ +--img 640 \ +--batch 16 \ +--epochs 5 \ +--data coco128.yaml \ +--weights yolov5s.pt \ +--upload_dataset +``` + +You can find the uploaded dataset in the Artifacts tab in your Comet Workspace +artifact-1 + +You can preview the data directly in the Comet UI. +artifact-2 + +Artifacts are versioned and also support adding metadata about the dataset. Comet will automatically log the metadata from your dataset `yaml` file +artifact-3 + +### Using a saved Artifact + +If you would like to use a dataset from Comet Artifacts, set the `path` variable in your dataset `yaml` file to point to the following Artifact resource URL. + +``` +# contents of artifact.yaml file +path: "comet:///:" +``` +Then pass this file to your training script in the following way + +```shell +python train.py \ +--img 640 \ +--batch 16 \ +--epochs 5 \ +--data artifact.yaml \ +--weights yolov5s.pt +``` + +Artifacts also allow you to track the lineage of data as it flows through your Experimentation workflow. Here you can see a graph that shows you all the experiments that have used your uploaded dataset. +artifact-4 + +## Resuming a Training Run + +If your training run is interrupted for any reason, e.g. disrupted internet connection, you can resume the run using the `resume` flag and the Comet Run Path. + +The Run Path has the following format `comet:////`. + +This will restore the run to its state before the interruption, which includes restoring the model from a checkpoint, restoring all hyperparameters and training arguments and downloading Comet dataset Artifacts if they were used in the original run. The resumed run will continue logging to the existing Experiment in the Comet UI + +```shell +python train.py \ +--resume "comet://" +``` + +## Hyperparameter Search with the Comet Optimizer + +YOLOv5 is also integrated with Comet's Optimizer, making is simple to visualie hyperparameter sweeps in the Comet UI. + +### Configuring an Optimizer Sweep + +To configure the Comet Optimizer, you will have to create a JSON file with the information about the sweep. An example file has been provided in `utils/loggers/comet/optimizer_config.json` + +```shell +python utils/loggers/comet/hpo.py \ + --comet_optimizer_config "utils/loggers/comet/optimizer_config.json" +``` + +The `hpo.py` script accepts the same arguments as `train.py`. If you wish to pass additional arguments to your sweep simply add them after +the script. + +```shell +python utils/loggers/comet/hpo.py \ + --comet_optimizer_config "utils/loggers/comet/optimizer_config.json" \ + --save-period 1 \ + --bbox_interval 1 +``` + +### Running a Sweep in Parallel + +```shell +comet optimizer -j utils/loggers/comet/hpo.py \ + utils/loggers/comet/optimizer_config.json" +``` + +### Visualizing Results + +Comet provides a number of ways to visualize the results of your sweep. Take a look at a [project with a completed sweep here](https://www.comet.com/examples/comet-example-yolov5/view/PrlArHGuuhDTKC1UuBmTtOSXD/panels?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration) + +hyperparameter-yolo \ No newline at end of file diff --git a/utils/loggers/comet/__init__.py b/utils/loggers/comet/__init__.py new file mode 100644 index 000000000000..b168687dd7b2 --- /dev/null +++ b/utils/loggers/comet/__init__.py @@ -0,0 +1,496 @@ +import glob +import json +import logging +import os +import sys +from pathlib import Path + +logger = logging.getLogger(__name__) + +FILE = Path(__file__).resolve() +ROOT = FILE.parents[3] # YOLOv5 root directory +if str(ROOT) not in sys.path: + sys.path.append(str(ROOT)) # add ROOT to PATH + +try: + import comet_ml + + # Project Configuration + config = comet_ml.config.get_config() + COMET_PROJECT_NAME = config.get_string(os.getenv("COMET_PROJECT_NAME"), "comet.project_name", default="yolov5") +except (ModuleNotFoundError, ImportError): + comet_ml = None + COMET_PROJECT_NAME = None + +import torch +import torchvision.transforms as T +import yaml + +from utils.dataloaders import img2label_paths +from utils.general import check_dataset, scale_coords, xywh2xyxy +from utils.metrics import box_iou + +COMET_PREFIX = "comet://" + +COMET_MODE = os.getenv("COMET_MODE", "online") + +# Model Saving Settings +COMET_MODEL_NAME = os.getenv("COMET_MODEL_NAME", "yolov5") + +# Dataset Artifact Settings +COMET_UPLOAD_DATASET = os.getenv("COMET_UPLOAD_DATASET", "false").lower() == "true" + +# Evaluation Settings +COMET_LOG_CONFUSION_MATRIX = os.getenv("COMET_LOG_CONFUSION_MATRIX", "true").lower() == "true" +COMET_LOG_PREDICTIONS = os.getenv("COMET_LOG_PREDICTIONS", "true").lower() == "true" +COMET_MAX_IMAGE_UPLOADS = int(os.getenv("COMET_MAX_IMAGE_UPLOADS", 100)) + +# Confusion Matrix Settings +CONF_THRES = float(os.getenv("CONF_THRES", 0.001)) +IOU_THRES = float(os.getenv("IOU_THRES", 0.6)) + +# Batch Logging Settings +COMET_LOG_BATCH_METRICS = os.getenv("COMET_LOG_BATCH_METRICS", "false").lower() == "true" +COMET_BATCH_LOGGING_INTERVAL = os.getenv("COMET_BATCH_LOGGING_INTERVAL", 1) +COMET_PREDICTION_LOGGING_INTERVAL = os.getenv("COMET_PREDICTION_LOGGING_INTERVAL", 1) +COMET_LOG_PER_CLASS_METRICS = os.getenv("COMET_LOG_PER_CLASS_METRICS", "false").lower() == "true" + +RANK = int(os.getenv("RANK", -1)) + +to_pil = T.ToPILImage() + + +class CometLogger: + """Log metrics, parameters, source code, models and much more + with Comet + """ + + def __init__(self, opt, hyp, run_id=None, job_type="Training", **experiment_kwargs) -> None: + self.job_type = job_type + self.opt = opt + self.hyp = hyp + + # Comet Flags + self.comet_mode = COMET_MODE + + self.save_model = opt.save_period > -1 + self.model_name = COMET_MODEL_NAME + + # Batch Logging Settings + self.log_batch_metrics = COMET_LOG_BATCH_METRICS + self.comet_log_batch_interval = COMET_BATCH_LOGGING_INTERVAL + + # Dataset Artifact Settings + self.upload_dataset = self.opt.upload_dataset if self.opt.upload_dataset else COMET_UPLOAD_DATASET + self.resume = self.opt.resume + + # Default parameters to pass to Experiment objects + self.default_experiment_kwargs = { + "log_code": False, + "log_env_gpu": True, + "log_env_cpu": True, + "project_name": COMET_PROJECT_NAME,} + self.default_experiment_kwargs.update(experiment_kwargs) + self.experiment = self._get_experiment(self.comet_mode, run_id) + + self.data_dict = self.check_dataset(self.opt.data) + self.class_names = self.data_dict["names"] + self.num_classes = self.data_dict["nc"] + + self.logged_images_count = 0 + self.max_images = COMET_MAX_IMAGE_UPLOADS + + if run_id is None: + self.experiment.log_other("Created from", "YOLOv5") + if not isinstance(self.experiment, comet_ml.OfflineExperiment): + workspace, project_name, experiment_id = self.experiment.url.split("/")[-3:] + self.experiment.log_other( + "Run Path", + f"{workspace}/{project_name}/{experiment_id}", + ) + self.log_parameters(vars(opt)) + self.log_parameters(self.opt.hyp) + self.log_asset_data( + self.opt.hyp, + name="hyperparameters.json", + metadata={"type": "hyp-config-file"}, + ) + self.log_asset( + f"{self.opt.save_dir}/opt.yaml", + metadata={"type": "opt-config-file"}, + ) + + self.comet_log_confusion_matrix = COMET_LOG_CONFUSION_MATRIX + + if hasattr(self.opt, "conf_thres"): + self.conf_thres = self.opt.conf_thres + else: + self.conf_thres = CONF_THRES + if hasattr(self.opt, "iou_thres"): + self.iou_thres = self.opt.iou_thres + else: + self.iou_thres = IOU_THRES + + self.comet_log_predictions = COMET_LOG_PREDICTIONS + if self.opt.bbox_interval == -1: + self.comet_log_prediction_interval = self.opt.epochs // 10 if self.opt.epochs < 10 else 1 + else: + self.comet_log_prediction_interval = self.opt.bbox_interval + + if self.comet_log_predictions: + self.metadata_dict = {} + + self.comet_log_per_class_metrics = COMET_LOG_PER_CLASS_METRICS + + self.experiment.log_others({ + "comet_mode": COMET_MODE, + "comet_max_image_uploads": COMET_MAX_IMAGE_UPLOADS, + "comet_log_per_class_metrics": COMET_LOG_PER_CLASS_METRICS, + "comet_log_batch_metrics": COMET_LOG_BATCH_METRICS, + "comet_log_confusion_matrix": COMET_LOG_CONFUSION_MATRIX, + "comet_model_name": COMET_MODEL_NAME,}) + + # Check if running the Experiment with the Comet Optimizer + if hasattr(self.opt, "comet_optimizer_id"): + self.experiment.log_other("optimizer_id", self.opt.comet_optimizer_id) + self.experiment.log_other("optimizer_objective", self.opt.comet_optimizer_objective) + self.experiment.log_other("optimizer_metric", self.opt.comet_optimizer_metric) + self.experiment.log_other("optimizer_parameters", json.dumps(self.hyp)) + + def _get_experiment(self, mode, experiment_id=None): + if mode == "offline": + if experiment_id is not None: + return comet_ml.ExistingOfflineExperiment( + previous_experiment=experiment_id, + **self.default_experiment_kwargs, + ) + + return comet_ml.OfflineExperiment(**self.default_experiment_kwargs,) + + else: + try: + if experiment_id is not None: + return comet_ml.ExistingExperiment( + previous_experiment=experiment_id, + **self.default_experiment_kwargs, + ) + + return comet_ml.Experiment(**self.default_experiment_kwargs) + + except ValueError: + logger.warning("COMET WARNING: " + "Comet credentials have not been set. " + "Comet will default to offline logging. " + "Please set your credentials to enable online logging.") + return self._get_experiment("offline", experiment_id) + + return + + def log_metrics(self, log_dict, **kwargs): + self.experiment.log_metrics(log_dict, **kwargs) + + def log_parameters(self, log_dict, **kwargs): + self.experiment.log_parameters(log_dict, **kwargs) + + def log_asset(self, asset_path, **kwargs): + self.experiment.log_asset(asset_path, **kwargs) + + def log_asset_data(self, asset, **kwargs): + self.experiment.log_asset_data(asset, **kwargs) + + def log_image(self, img, **kwargs): + self.experiment.log_image(img, **kwargs) + + def log_model(self, path, opt, epoch, fitness_score, best_model=False): + if not self.save_model: + return + + model_metadata = { + "fitness_score": fitness_score[-1], + "epochs_trained": epoch + 1, + "save_period": opt.save_period, + "total_epochs": opt.epochs,} + + model_files = glob.glob(f"{path}/*.pt") + for model_path in model_files: + name = Path(model_path).name + + self.experiment.log_model( + self.model_name, + file_or_folder=model_path, + file_name=name, + metadata=model_metadata, + overwrite=True, + ) + + def check_dataset(self, data_file): + with open(data_file) as f: + data_config = yaml.safe_load(f) + + if data_config['path'].startswith(COMET_PREFIX): + path = data_config['path'].replace(COMET_PREFIX, "") + data_dict = self.download_dataset_artifact(path) + + return data_dict + + self.log_asset(self.opt.data, metadata={"type": "data-config-file"}) + + return check_dataset(data_file) + + def log_predictions(self, image, labelsn, path, shape, predn): + if self.logged_images_count >= self.max_images: + return + detections = predn[predn[:, 4] > self.conf_thres] + iou = box_iou(labelsn[:, 1:], detections[:, :4]) + mask, _ = torch.where(iou > self.iou_thres) + if len(mask) == 0: + return + + filtered_detections = detections[mask] + filtered_labels = labelsn[mask] + + processed_image = (image * 255).to(torch.uint8) + + image_id = path.split("/")[-1].split(".")[0] + image_name = f"{image_id}_curr_epoch_{self.experiment.curr_epoch}" + self.log_image(to_pil(processed_image), name=image_name) + + metadata = [] + for cls, *xyxy in filtered_labels.tolist(): + metadata.append({ + "label": f"{self.class_names[int(cls)]}-gt", + "score": 100, + "box": { + "x": xyxy[0], + "y": xyxy[1], + "x2": xyxy[2], + "y2": xyxy[3]},}) + for *xyxy, conf, cls in filtered_detections.tolist(): + metadata.append({ + "label": f"{self.class_names[int(cls)]}", + "score": conf * 100, + "box": { + "x": xyxy[0], + "y": xyxy[1], + "x2": xyxy[2], + "y2": xyxy[3]},}) + + self.metadata_dict[image_name] = metadata + self.logged_images_count += 1 + + return + + def preprocess_prediction(self, image, labels, shape, pred): + nl, _ = labels.shape[0], pred.shape[0] + + # Predictions + if self.opt.single_cls: + pred[:, 5] = 0 + + predn = pred.clone() + scale_coords(image.shape[1:], predn[:, :4], shape[0], shape[1]) + + labelsn = None + if nl: + tbox = xywh2xyxy(labels[:, 1:5]) # target boxes + scale_coords(image.shape[1:], tbox, shape[0], shape[1]) # native-space labels + labelsn = torch.cat((labels[:, 0:1], tbox), 1) # native-space labels + scale_coords(image.shape[1:], predn[:, :4], shape[0], shape[1]) # native-space pred + + return predn, labelsn + + def add_assets_to_artifact(self, artifact, path, asset_path, split): + img_paths = sorted(glob.glob(f"{asset_path}/*")) + label_paths = img2label_paths(img_paths) + + for image_file, label_file in zip(img_paths, label_paths): + image_logical_path, label_logical_path = map(lambda x: os.path.relpath(x, path), [image_file, label_file]) + + try: + artifact.add(image_file, logical_path=image_logical_path, metadata={"split": split}) + artifact.add(label_file, logical_path=label_logical_path, metadata={"split": split}) + except ValueError as e: + logger.error('COMET ERROR: Error adding file to Artifact. Skipping file.') + logger.error(f"COMET ERROR: {e}") + continue + + return artifact + + def upload_dataset_artifact(self): + dataset_name = self.data_dict.get("dataset_name", "yolov5-dataset") + path = str((ROOT / Path(self.data_dict["path"])).resolve()) + + metadata = self.data_dict.copy() + for key in ["train", "val", "test"]: + split_path = metadata.get(key) + if split_path is not None: + metadata[key] = split_path.replace(path, "") + + artifact = comet_ml.Artifact(name=dataset_name, artifact_type="dataset", metadata=metadata) + for key in metadata.keys(): + if key in ["train", "val", "test"]: + if isinstance(self.upload_dataset, str) and (key != self.upload_dataset): + continue + + asset_path = self.data_dict.get(key) + if asset_path is not None: + artifact = self.add_assets_to_artifact(artifact, path, asset_path, key) + + self.experiment.log_artifact(artifact) + + return + + def download_dataset_artifact(self, artifact_path): + logged_artifact = self.experiment.get_artifact(artifact_path) + artifact_save_dir = str(Path(self.opt.save_dir) / logged_artifact.name) + logged_artifact.download(artifact_save_dir) + + metadata = logged_artifact.metadata + data_dict = metadata.copy() + data_dict["path"] = artifact_save_dir + data_dict["names"] = {int(k): v for k, v in metadata.get("names").items()} + + data_dict = self.update_data_paths(data_dict) + return data_dict + + def update_data_paths(self, data_dict): + path = data_dict.get("path", "") + + for split in ["train", "val", "test"]: + if data_dict.get(split): + split_path = data_dict.get(split) + data_dict[split] = (f"{path}/{split_path}" if isinstance(split, str) else [ + f"{path}/{x}" for x in split_path]) + + return data_dict + + def on_pretrain_routine_end(self, paths): + if self.opt.resume: + return + + for path in paths: + self.log_asset(str(path)) + + if self.upload_dataset: + if not self.resume: + self.upload_dataset_artifact() + + return + + def on_train_start(self): + self.log_parameters(self.hyp) + + def on_train_epoch_start(self): + return + + def on_train_epoch_end(self, epoch): + self.experiment.curr_epoch = epoch + + return + + def on_train_batch_start(self): + return + + def on_train_batch_end(self, log_dict, step): + self.experiment.curr_step = step + if self.log_batch_metrics and (step % self.comet_log_batch_interval == 0): + self.log_metrics(log_dict, step=step) + + return + + def on_train_end(self, files, save_dir, last, best, epoch, results): + if self.comet_log_predictions: + curr_epoch = self.experiment.curr_epoch + self.experiment.log_asset_data(self.metadata_dict, "image-metadata.json", epoch=curr_epoch) + + for f in files: + self.log_asset(f, metadata={"epoch": epoch}) + self.log_asset(f"{save_dir}/results.csv", metadata={"epoch": epoch}) + + if not self.opt.evolve: + model_path = str(best if best.exists() else last) + name = Path(model_path).name + if self.save_model: + self.experiment.log_model( + self.model_name, + file_or_folder=model_path, + file_name=name, + overwrite=True, + ) + + # Check if running Experiment with Comet Optimizer + if hasattr(self.opt, 'comet_optimizer_id'): + metric = results.get(self.opt.comet_optimizer_metric) + self.experiment.log_other('optimizer_metric_value', metric) + + self.finish_run() + + def on_val_start(self): + return + + def on_val_batch_start(self): + return + + def on_val_batch_end(self, batch_i, images, targets, paths, shapes, outputs): + if not (self.comet_log_predictions and ((batch_i + 1) % self.comet_log_prediction_interval == 0)): + return + + for si, pred in enumerate(outputs): + if len(pred) == 0: + continue + + image = images[si] + labels = targets[targets[:, 0] == si, 1:] + shape = shapes[si] + path = paths[si] + predn, labelsn = self.preprocess_prediction(image, labels, shape, pred) + if labelsn is not None: + self.log_predictions(image, labelsn, path, shape, predn) + + return + + def on_val_end(self, nt, tp, fp, p, r, f1, ap, ap50, ap_class, confusion_matrix): + if self.comet_log_per_class_metrics: + if self.num_classes > 1: + for i, c in enumerate(ap_class): + class_name = self.class_names[c] + self.experiment.log_metrics( + { + 'mAP@.5': ap50[i], + 'mAP@.5:.95': ap[i], + 'precision': p[i], + 'recall': r[i], + 'f1': f1[i], + 'true_positives': tp[i], + 'false_positives': fp[i], + 'support': nt[c]}, + prefix=class_name) + + if self.comet_log_confusion_matrix: + epoch = self.experiment.curr_epoch + class_names = list(self.class_names.values()) + class_names.append("background") + num_classes = len(class_names) + + self.experiment.log_confusion_matrix( + matrix=confusion_matrix.matrix, + max_categories=num_classes, + labels=class_names, + epoch=epoch, + column_label='Actual Category', + row_label='Predicted Category', + file_name=f"confusion-matrix-epoch-{epoch}.json", + ) + + def on_fit_epoch_end(self, result, epoch): + self.log_metrics(result, epoch=epoch) + + def on_model_save(self, last, epoch, final_epoch, best_fitness, fi): + if ((epoch + 1) % self.opt.save_period == 0 and not final_epoch) and self.opt.save_period != -1: + self.log_model(last.parent, self.opt, epoch, fi, best_model=best_fitness == fi) + + def on_params_update(self, params): + self.log_parameters(params) + + def finish_run(self): + self.experiment.end() diff --git a/utils/loggers/comet/comet_utils.py b/utils/loggers/comet/comet_utils.py new file mode 100644 index 000000000000..3cbd45156b57 --- /dev/null +++ b/utils/loggers/comet/comet_utils.py @@ -0,0 +1,150 @@ +import logging +import os +from urllib.parse import urlparse + +try: + import comet_ml +except (ModuleNotFoundError, ImportError): + comet_ml = None + +import yaml + +logger = logging.getLogger(__name__) + +COMET_PREFIX = "comet://" +COMET_MODEL_NAME = os.getenv("COMET_MODEL_NAME", "yolov5") +COMET_DEFAULT_CHECKPOINT_FILENAME = os.getenv("COMET_DEFAULT_CHECKPOINT_FILENAME", "last.pt") + + +def download_model_checkpoint(opt, experiment): + model_dir = f"{opt.project}/{experiment.name}" + os.makedirs(model_dir, exist_ok=True) + + model_name = COMET_MODEL_NAME + model_asset_list = experiment.get_model_asset_list(model_name) + + if len(model_asset_list) == 0: + logger.error(f"COMET ERROR: No checkpoints found for model name : {model_name}") + return + + model_asset_list = sorted( + model_asset_list, + key=lambda x: x["step"], + reverse=True, + ) + logged_checkpoint_map = {asset["fileName"]: asset["assetId"] for asset in model_asset_list} + + resource_url = urlparse(opt.weights) + checkpoint_filename = resource_url.query + + if checkpoint_filename: + asset_id = logged_checkpoint_map.get(checkpoint_filename) + else: + asset_id = logged_checkpoint_map.get(COMET_DEFAULT_CHECKPOINT_FILENAME) + checkpoint_filename = COMET_DEFAULT_CHECKPOINT_FILENAME + + if asset_id is None: + logger.error(f"COMET ERROR: Checkpoint {checkpoint_filename} not found in the given Experiment") + return + + try: + logger.info(f"COMET INFO: Downloading checkpoint {checkpoint_filename}") + asset_filename = checkpoint_filename + + model_binary = experiment.get_asset(asset_id, return_type="binary", stream=False) + model_download_path = f"{model_dir}/{asset_filename}" + with open(model_download_path, "wb") as f: + f.write(model_binary) + + opt.weights = model_download_path + + except Exception as e: + logger.warning("COMET WARNING: Unable to download checkpoint from Comet") + logger.exception(e) + + +def set_opt_parameters(opt, experiment): + """Update the opts Namespace with parameters + from Comet's ExistingExperiment when resuming a run + + Args: + opt (argparse.Namespace): Namespace of command line options + experiment (comet_ml.APIExperiment): Comet API Experiment object + """ + asset_list = experiment.get_asset_list() + resume_string = opt.resume + + for asset in asset_list: + if asset["fileName"] == "opt.yaml": + asset_id = asset["assetId"] + asset_binary = experiment.get_asset(asset_id, return_type="binary", stream=False) + opt_dict = yaml.safe_load(asset_binary) + for key, value in opt_dict.items(): + setattr(opt, key, value) + opt.resume = resume_string + + # Save hyperparameters to YAML file + # Necessary to pass checks in training script + save_dir = f"{opt.project}/{experiment.name}" + os.makedirs(save_dir, exist_ok=True) + + hyp_yaml_path = f"{save_dir}/hyp.yaml" + with open(hyp_yaml_path, "w") as f: + yaml.dump(opt.hyp, f) + opt.hyp = hyp_yaml_path + + +def check_comet_weights(opt): + """Downloads model weights from Comet and updates the + weights path to point to saved weights location + + Args: + opt (argparse.Namespace): Command Line arguments passed + to YOLOv5 training script + + Returns: + None/bool: Return True if weights are successfully downloaded + else return None + """ + if comet_ml is None: + return + + if isinstance(opt.weights, str): + if opt.weights.startswith(COMET_PREFIX): + api = comet_ml.API() + resource = urlparse(opt.weights) + experiment_path = f"{resource.netloc}{resource.path}" + experiment = api.get(experiment_path) + download_model_checkpoint(opt, experiment) + return True + + return None + + +def check_comet_resume(opt): + """Restores run parameters to its original state based on the model checkpoint + and logged Experiment parameters. + + Args: + opt (argparse.Namespace): Command Line arguments passed + to YOLOv5 training script + + Returns: + None/bool: Return True if the run is restored successfully + else return None + """ + if comet_ml is None: + return + + if isinstance(opt.resume, str): + if opt.resume.startswith(COMET_PREFIX): + api = comet_ml.API() + resource = urlparse(opt.resume) + experiment_path = f"{resource.netloc}{resource.path}" + experiment = api.get(experiment_path) + set_opt_parameters(opt, experiment) + download_model_checkpoint(opt, experiment) + + return True + + return None diff --git a/utils/loggers/comet/hpo.py b/utils/loggers/comet/hpo.py new file mode 100644 index 000000000000..eab4df9978cf --- /dev/null +++ b/utils/loggers/comet/hpo.py @@ -0,0 +1,118 @@ +import argparse +import json +import logging +import os +import sys +from pathlib import Path + +import comet_ml + +logger = logging.getLogger(__name__) + +FILE = Path(__file__).resolve() +ROOT = FILE.parents[3] # YOLOv5 root directory +if str(ROOT) not in sys.path: + sys.path.append(str(ROOT)) # add ROOT to PATH + +from train import parse_opt, train +from utils.callbacks import Callbacks +from utils.general import increment_path +from utils.torch_utils import select_device + +# Project Configuration +config = comet_ml.config.get_config() +COMET_PROJECT_NAME = config.get_string(os.getenv("COMET_PROJECT_NAME"), "comet.project_name", default="yolov5") + + +def get_args(known=False): + parser = argparse.ArgumentParser() + parser.add_argument('--weights', type=str, default=ROOT / 'yolov5s.pt', help='initial weights path') + parser.add_argument('--cfg', type=str, default='', help='model.yaml path') + parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path') + parser.add_argument('--hyp', type=str, default=ROOT / 'data/hyps/hyp.scratch-low.yaml', help='hyperparameters path') + parser.add_argument('--epochs', type=int, default=300, help='total training epochs') + parser.add_argument('--batch-size', type=int, default=16, help='total batch size for all GPUs, -1 for autobatch') + parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='train, val image size (pixels)') + parser.add_argument('--rect', action='store_true', help='rectangular training') + parser.add_argument('--resume', nargs='?', const=True, default=False, help='resume most recent training') + parser.add_argument('--nosave', action='store_true', help='only save final checkpoint') + parser.add_argument('--noval', action='store_true', help='only validate final epoch') + parser.add_argument('--noautoanchor', action='store_true', help='disable AutoAnchor') + parser.add_argument('--noplots', action='store_true', help='save no plot files') + parser.add_argument('--evolve', type=int, nargs='?', const=300, help='evolve hyperparameters for x generations') + parser.add_argument('--bucket', type=str, default='', help='gsutil bucket') + parser.add_argument('--cache', type=str, nargs='?', const='ram', help='--cache images in "ram" (default) or "disk"') + parser.add_argument('--image-weights', action='store_true', help='use weighted image selection for training') + parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--multi-scale', action='store_true', help='vary img-size +/- 50%%') + parser.add_argument('--single-cls', action='store_true', help='train multi-class data as single-class') + parser.add_argument('--optimizer', type=str, choices=['SGD', 'Adam', 'AdamW'], default='SGD', help='optimizer') + parser.add_argument('--sync-bn', action='store_true', help='use SyncBatchNorm, only available in DDP mode') + parser.add_argument('--workers', type=int, default=8, help='max dataloader workers (per RANK in DDP mode)') + parser.add_argument('--project', default=ROOT / 'runs/train', help='save to project/name') + parser.add_argument('--name', default='exp', help='save to project/name') + parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') + parser.add_argument('--quad', action='store_true', help='quad dataloader') + parser.add_argument('--cos-lr', action='store_true', help='cosine LR scheduler') + parser.add_argument('--label-smoothing', type=float, default=0.0, help='Label smoothing epsilon') + parser.add_argument('--patience', type=int, default=100, help='EarlyStopping patience (epochs without improvement)') + parser.add_argument('--freeze', nargs='+', type=int, default=[0], help='Freeze layers: backbone=10, first3=0 1 2') + parser.add_argument('--save-period', type=int, default=-1, help='Save checkpoint every x epochs (disabled if < 1)') + parser.add_argument('--seed', type=int, default=0, help='Global training seed') + parser.add_argument('--local_rank', type=int, default=-1, help='Automatic DDP Multi-GPU argument, do not modify') + + # Weights & Biases arguments + parser.add_argument('--entity', default=None, help='W&B: Entity') + parser.add_argument('--upload_dataset', nargs='?', const=True, default=False, help='W&B: Upload data, "val" option') + parser.add_argument('--bbox_interval', type=int, default=-1, help='W&B: Set bounding-box image logging interval') + parser.add_argument('--artifact_alias', type=str, default='latest', help='W&B: Version of dataset artifact to use') + + # Comet Arguments + parser.add_argument("--comet_optimizer_config", type=str, help="Comet: Path to a Comet Optimizer Config File.") + parser.add_argument("--comet_optimizer_id", type=str, help="Comet: ID of the Comet Optimizer sweep.") + parser.add_argument("--comet_optimizer_objective", type=str, help="Comet: Set to 'minimize' or 'maximize'.") + parser.add_argument("--comet_optimizer_metric", type=str, help="Comet: Metric to Optimize.") + parser.add_argument("--comet_optimizer_workers", + type=int, + default=1, + help="Comet: Number of Parallel Workers to use with the Comet Optimizer.") + + return parser.parse_known_args()[0] if known else parser.parse_args() + + +def run(parameters, opt): + hyp_dict = {k: v for k, v in parameters.items() if k not in ["epochs", "batch_size"]} + + opt.save_dir = str(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok or opt.evolve)) + opt.batch_size = parameters.get("batch_size") + opt.epochs = parameters.get("epochs") + + device = select_device(opt.device, batch_size=opt.batch_size) + train(hyp_dict, opt, device, callbacks=Callbacks()) + + +if __name__ == "__main__": + opt = get_args(known=True) + + opt.weights = str(opt.weights) + opt.cfg = str(opt.cfg) + opt.data = str(opt.data) + opt.project = str(opt.project) + + optimizer_id = os.getenv("COMET_OPTIMIZER_ID") + if optimizer_id is None: + with open(opt.comet_optimizer_config) as f: + optimizer_config = json.load(f) + optimizer = comet_ml.Optimizer(optimizer_config) + else: + optimizer = comet_ml.Optimizer(optimizer_id) + + opt.comet_optimizer_id = optimizer.id + status = optimizer.status() + + opt.comet_optimizer_objective = status["spec"]["objective"] + opt.comet_optimizer_metric = status["spec"]["metric"] + + logger.info("COMET INFO: Starting Hyperparameter Sweep") + for parameter in optimizer.get_parameters(): + run(parameter["parameters"], opt) diff --git a/utils/loggers/comet/optimizer_config.json b/utils/loggers/comet/optimizer_config.json new file mode 100644 index 000000000000..83ddddab6f20 --- /dev/null +++ b/utils/loggers/comet/optimizer_config.json @@ -0,0 +1,209 @@ +{ + "algorithm": "random", + "parameters": { + "anchor_t": { + "type": "discrete", + "values": [ + 2, + 8 + ] + }, + "batch_size": { + "type": "discrete", + "values": [ + 16, + 32, + 64 + ] + }, + "box": { + "type": "discrete", + "values": [ + 0.02, + 0.2 + ] + }, + "cls": { + "type": "discrete", + "values": [ + 0.2 + ] + }, + "cls_pw": { + "type": "discrete", + "values": [ + 0.5 + ] + }, + "copy_paste": { + "type": "discrete", + "values": [ + 1 + ] + }, + "degrees": { + "type": "discrete", + "values": [ + 0, + 45 + ] + }, + "epochs": { + "type": "discrete", + "values": [ + 5 + ] + }, + "fl_gamma": { + "type": "discrete", + "values": [ + 0 + ] + }, + "fliplr": { + "type": "discrete", + "values": [ + 0 + ] + }, + "flipud": { + "type": "discrete", + "values": [ + 0 + ] + }, + "hsv_h": { + "type": "discrete", + "values": [ + 0 + ] + }, + "hsv_s": { + "type": "discrete", + "values": [ + 0 + ] + }, + "hsv_v": { + "type": "discrete", + "values": [ + 0 + ] + }, + "iou_t": { + "type": "discrete", + "values": [ + 0.7 + ] + }, + "lr0": { + "type": "discrete", + "values": [ + 1e-05, + 0.1 + ] + }, + "lrf": { + "type": "discrete", + "values": [ + 0.01, + 1 + ] + }, + "mixup": { + "type": "discrete", + "values": [ + 1 + ] + }, + "momentum": { + "type": "discrete", + "values": [ + 0.6 + ] + }, + "mosaic": { + "type": "discrete", + "values": [ + 0 + ] + }, + "obj": { + "type": "discrete", + "values": [ + 0.2 + ] + }, + "obj_pw": { + "type": "discrete", + "values": [ + 0.5 + ] + }, + "optimizer": { + "type": "categorical", + "values": [ + "SGD", + "Adam", + "AdamW" + ] + }, + "perspective": { + "type": "discrete", + "values": [ + 0 + ] + }, + "scale": { + "type": "discrete", + "values": [ + 0 + ] + }, + "shear": { + "type": "discrete", + "values": [ + 0 + ] + }, + "translate": { + "type": "discrete", + "values": [ + 0 + ] + }, + "warmup_bias_lr": { + "type": "discrete", + "values": [ + 0, + 0.2 + ] + }, + "warmup_epochs": { + "type": "discrete", + "values": [ + 5 + ] + }, + "warmup_momentum": { + "type": "discrete", + "values": [ + 0, + 0.95 + ] + }, + "weight_decay": { + "type": "discrete", + "values": [ + 0, + 0.001 + ] + } + }, + "spec": { + "maxCombo": 0, + "metric": "metrics/mAP_0.5", + "objective": "maximize" + }, + "trials": 1 +} diff --git a/val.py b/val.py index 5427ee7b3619..665d92f9286d 100644 --- a/val.py +++ b/val.py @@ -259,7 +259,7 @@ def run( plot_images(im, targets, paths, save_dir / f'val_batch{batch_i}_labels.jpg', names) # labels plot_images(im, output_to_target(out), paths, save_dir / f'val_batch{batch_i}_pred.jpg', names) # pred - callbacks.run('on_val_batch_end') + callbacks.run('on_val_batch_end', batch_i, im, targets, paths, shapes, out) # Compute metrics stats = [torch.cat(x, 0).cpu().numpy() for x in zip(*stats)] # to numpy @@ -289,7 +289,7 @@ def run( # Plots if plots: confusion_matrix.plot(save_dir=save_dir, names=list(names.values())) - callbacks.run('on_val_end') + callbacks.run('on_val_end', nt, tp, fp, p, r, f1, ap, ap50, ap_class, confusion_matrix) # Save JSON if save_json and len(jdict): From 5f075eedf221852aab85b4d2b5d98289e13077b4 Mon Sep 17 00:00:00 2001 From: Dhruv Nair Date: Thu, 8 Sep 2022 11:17:14 -0400 Subject: [PATCH 562/661] Comet changes (#9328) * add link to comte tutorial from main README * fix prediction interval bug --- README.md | 1 + utils/loggers/comet/__init__.py | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index 7763d174f92b..da8bf1dad862 100644 --- a/README.md +++ b/README.md @@ -157,6 +157,7 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 - [Roboflow for Datasets, Labeling, and Active Learning](https://github.com/ultralytics/yolov5/issues/4975)  🌟 NEW - [ClearML Logging](https://github.com/ultralytics/yolov5/tree/master/utils/loggers/clearml) 🌟 NEW - [Deci Platform](https://github.com/ultralytics/yolov5/wiki/Deci-Platform) 🌟 NEW +- [Comet Logging](https://github.com/ultralytics/yolov5/tree/master/utils/loggers/comet) 🌟 NEW
diff --git a/utils/loggers/comet/__init__.py b/utils/loggers/comet/__init__.py index b168687dd7b2..4ee86dd70d6e 100644 --- a/utils/loggers/comet/__init__.py +++ b/utils/loggers/comet/__init__.py @@ -133,7 +133,7 @@ def __init__(self, opt, hyp, run_id=None, job_type="Training", **experiment_kwar self.comet_log_predictions = COMET_LOG_PREDICTIONS if self.opt.bbox_interval == -1: - self.comet_log_prediction_interval = self.opt.epochs // 10 if self.opt.epochs < 10 else 1 + self.comet_log_prediction_interval = 1 if self.opt.epochs < 10 else self.opt.epochs // 10 else: self.comet_log_prediction_interval = self.opt.bbox_interval From 3cd66b1c3863a8524c6cc564029c29ac783f7310 Mon Sep 17 00:00:00 2001 From: robinned <78896580+robinned@users.noreply.github.com> Date: Thu, 8 Sep 2022 12:00:54 -0700 Subject: [PATCH 563/661] Train.py line 486 typo fix (#9330) fixed issue Signed-off-by: robinned <78896580+robinned@users.noreply.github.com> Signed-off-by: robinned <78896580+robinned@users.noreply.github.com> Co-authored-by: Ayush Chaurasia --- train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train.py b/train.py index e16c17c499f0..4eff6e5d645a 100644 --- a/train.py +++ b/train.py @@ -483,7 +483,7 @@ def main(opt, callbacks=Callbacks()): check_requirements() # Resume (from specified or most recent last.pt) - if opt.resume and not check_wandb_resume(opt) and not check_comet_resume(opt) or opt.evolve: + if opt.resume and not check_wandb_resume(opt) and not check_comet_resume(opt) and not opt.evolve: last = Path(check_file(opt.resume) if isinstance(opt.resume, str) else get_latest_run()) opt_yaml = last.parent.parent / 'opt.yaml' # train options yaml opt_data = opt.data # original dataset From 8aa196ce08007aa1033b0e42931c247e1e491321 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E9=BB=84=E4=B8=8D=E7=9B=88?= <33193090+YellowAndGreen@users.noreply.github.com> Date: Sat, 10 Sep 2022 05:16:07 +0800 Subject: [PATCH 564/661] Add dilated conv support (#9347) * added dilate conv support * added dilate conv support * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update common.py * Update common.py Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- models/common.py | 16 +++++++++------- utils/torch_utils.py | 1 + 2 files changed, 10 insertions(+), 7 deletions(-) diff --git a/models/common.py b/models/common.py index 7ac3a4a29672..c30c8ee94777 100644 --- a/models/common.py +++ b/models/common.py @@ -28,18 +28,20 @@ from utils.torch_utils import copy_attr, smart_inference_mode -def autopad(k, p=None): # kernel, padding - # Pad to 'same' +def autopad(k, p=None, d=1): # kernel, padding, dilation + # Pad to 'same' shape outputs + if d > 1: + k = d * (k - 1) + 1 if isinstance(k, int) else [d * (x - 1) + 1 for x in k] # actual kernel-size if p is None: p = k // 2 if isinstance(k, int) else [x // 2 for x in k] # auto-pad return p class Conv(nn.Module): - # Standard convolution - def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups + # Standard convolution with args(ch_in, ch_out, kernel, stride, padding, groups, dilation, activation) + def __init__(self, c1, c2, k=1, s=1, p=None, g=1, d=1, act=True): super().__init__() - self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g, bias=False) + self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p, d), groups=g, dilation=d, bias=False) self.bn = nn.BatchNorm2d(c2) self.act = nn.SiLU() if act is True else (act if isinstance(act, nn.Module) else nn.Identity()) @@ -51,13 +53,13 @@ def forward_fuse(self, x): class DWConv(Conv): - # Depth-wise convolution class + # Depth-wise convolution def __init__(self, c1, c2, k=1, s=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups super().__init__(c1, c2, k, s, g=math.gcd(c1, c2), act=act) class DWConvTranspose2d(nn.ConvTranspose2d): - # Depth-wise transpose convolution class + # Depth-wise transpose convolution def __init__(self, c1, c2, k=1, s=1, p1=0, p2=0): # ch_in, ch_out, kernel, stride, padding, padding_out super().__init__(c1, c2, k, s, p1, p2, groups=math.gcd(c1, c2)) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index abf0bbc19a98..8a3366ca3e27 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -251,6 +251,7 @@ def fuse_conv_and_bn(conv, bn): kernel_size=conv.kernel_size, stride=conv.stride, padding=conv.padding, + dilation=conv.dilation, groups=conv.groups, bias=True).requires_grad_(False).to(conv.weight.device) From 24bf9cceb406a7e380bdb9e100417318615a78a1 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 10 Sep 2022 11:11:56 +0300 Subject: [PATCH 565/661] Update `check_requirements()` single install (#9353) * Update `check_requirements()` single install Faster install and better conflict resolution with single installation Signed-off-by: Glenn Jocher * Update * Update * Update Signed-off-by: Glenn Jocher --- export.py | 12 ++++++------ models/common.py | 4 ++-- utils/general.py | 48 +++++++++++++++++++++++------------------------- val.py | 4 ++-- 4 files changed, 33 insertions(+), 35 deletions(-) diff --git a/export.py b/export.py index 4d0144af9efb..8fed4d3e3661 100644 --- a/export.py +++ b/export.py @@ -126,7 +126,7 @@ def export_torchscript(model, im, file, optimize, prefix=colorstr('TorchScript:' @try_export def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorstr('ONNX:')): # YOLOv5 ONNX export - check_requirements(('onnx',)) + check_requirements('onnx') import onnx LOGGER.info(f'\n{prefix} starting export with onnx {onnx.__version__}...') @@ -182,7 +182,7 @@ def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorst @try_export def export_openvino(model, file, half, prefix=colorstr('OpenVINO:')): # YOLOv5 OpenVINO export - check_requirements(('openvino-dev',)) # requires openvino-dev: https://pypi.org/project/openvino-dev/ + check_requirements('openvino-dev') # requires openvino-dev: https://pypi.org/project/openvino-dev/ import openvino.inference_engine as ie LOGGER.info(f'\n{prefix} starting export with openvino {ie.__version__}...') @@ -198,7 +198,7 @@ def export_openvino(model, file, half, prefix=colorstr('OpenVINO:')): @try_export def export_coreml(model, im, file, int8, half, prefix=colorstr('CoreML:')): # YOLOv5 CoreML export - check_requirements(('coremltools',)) + check_requirements('coremltools') import coremltools as ct LOGGER.info(f'\n{prefix} starting export with coremltools {ct.__version__}...') @@ -226,7 +226,7 @@ def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose import tensorrt as trt except Exception: if platform.system() == 'Linux': - check_requirements(('nvidia-tensorrt',), cmds=('-U --index-url https://pypi.ngc.nvidia.com',)) + check_requirements('nvidia-tensorrt', cmds=['-U --index-url https://pypi.ngc.nvidia.com']) import tensorrt as trt if trt.__version__[0] == '7': # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012 @@ -405,7 +405,7 @@ def export_edgetpu(file, prefix=colorstr('Edge TPU:')): @try_export def export_tfjs(file, prefix=colorstr('TensorFlow.js:')): # YOLOv5 TensorFlow.js export - check_requirements(('tensorflowjs',)) + check_requirements('tensorflowjs') import re import tensorflowjs as tfjs @@ -516,7 +516,7 @@ def run( # TensorFlow Exports if any((saved_model, pb, tflite, edgetpu, tfjs)): if int8 or edgetpu: # TFLite --int8 bug https://github.com/ultralytics/yolov5/issues/5707 - check_requirements(('flatbuffers==1.12',)) # required before `import tensorflow` + check_requirements('flatbuffers==1.12') # required before `import tensorflow` assert not tflite or not tfjs, 'TFLite and TF.js models must be exported separately, please pass only one type.' assert not isinstance(model, ClassificationModel), 'ClassificationModel export to TF formats not yet supported.' f[5], model = export_saved_model(model.cpu(), diff --git a/models/common.py b/models/common.py index c30c8ee94777..0e01b60e81e5 100644 --- a/models/common.py +++ b/models/common.py @@ -347,7 +347,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, stride, names = int(d['stride']), d['names'] elif dnn: # ONNX OpenCV DNN LOGGER.info(f'Loading {w} for ONNX OpenCV DNN inference...') - check_requirements(('opencv-python>=4.5.4',)) + check_requirements('opencv-python>=4.5.4') net = cv2.dnn.readNetFromONNX(w) elif onnx: # ONNX Runtime LOGGER.info(f'Loading {w} for ONNX Runtime inference...') @@ -362,7 +362,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, stride, names = int(meta['stride']), eval(meta['names']) elif xml: # OpenVINO LOGGER.info(f'Loading {w} for OpenVINO inference...') - check_requirements(('openvino',)) # requires openvino-dev: https://pypi.org/project/openvino-dev/ + check_requirements('openvino') # requires openvino-dev: https://pypi.org/project/openvino-dev/ from openvino.runtime import Core, Layout, get_batch ie = Core() if not Path(w).is_file(): # if not *.xml diff --git a/utils/general.py b/utils/general.py index cae63fd9dd21..629df32ebc54 100755 --- a/utils/general.py +++ b/utils/general.py @@ -342,39 +342,37 @@ def check_version(current='0.0.0', minimum='0.0.0', name='version ', pinned=Fals @TryExcept() def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), install=True, cmds=()): - # Check installed dependencies meet YOLOv5 requirements (pass *.txt file or list of packages) + # Check installed dependencies meet YOLOv5 requirements (pass *.txt file or list of packages or single package str) prefix = colorstr('red', 'bold', 'requirements:') check_python() # check python version - if isinstance(requirements, (str, Path)): # requirements.txt file - file = Path(requirements) + if isinstance(requirements, Path): # requirements.txt file + file = requirements assert file.exists(), f"{prefix} {file.resolve()} not found, check failed." with file.open() as f: requirements = [f'{x.name}{x.specifier}' for x in pkg.parse_requirements(f) if x.name not in exclude] - else: # list or tuple of packages - requirements = [x for x in requirements if x not in exclude] + elif isinstance(requirements, str): + requirements = [requirements] - n = 0 # number of packages updates - for i, r in enumerate(requirements): + s = '' + n = 0 + for r in requirements: try: pkg.require(r) - except Exception: # DistributionNotFound or VersionConflict if requirements not met - s = f"{prefix} {r} not found and is required by YOLOv5" - if install and AUTOINSTALL: # check environment variable - LOGGER.info(f"{s}, attempting auto-update...") - try: - assert check_online(), f"'pip install {r}' skipped (offline)" - LOGGER.info(check_output(f'pip install "{r}" {cmds[i] if cmds else ""}', shell=True).decode()) - n += 1 - except Exception as e: - LOGGER.warning(f'{prefix} {e}') - else: - LOGGER.info(f'{s}. Please install and rerun your command.') - - if n: # if packages updated - source = file.resolve() if 'file' in locals() else requirements - s = f"{prefix} {n} package{'s' * (n > 1)} updated per {source}\n" \ - f"{prefix} ⚠️ {colorstr('bold', 'Restart runtime or rerun command for updates to take effect')}\n" - LOGGER.info(s) + except (pkg.VersionConflict, pkg.DistributionNotFound): # exception if requirements not met + s += f'"{r}" ' + n += 1 + + if s and install and AUTOINSTALL: # check environment variable + LOGGER.info(f"{prefix} YOLOv5 requirements {s}not found, attempting AutoUpdate...") + try: + assert check_online(), "AutoUpdate skipped (offline)" + LOGGER.info(check_output(f'pip install {s} {" ".join(cmds) if cmds else ""}', shell=True).decode()) + source = file.resolve() if 'file' in locals() else requirements + s = f"{prefix} {n} package{'s' * (n > 1)} updated per {source}\n" \ + f"{prefix} ⚠️ {colorstr('bold', 'Restart runtime or rerun command for updates to take effect')}\n" + LOGGER.info(s) + except Exception as e: + LOGGER.warning(f'{prefix} {e}') def check_img_size(imgsz, s=32, floor=0): diff --git a/val.py b/val.py index 665d92f9286d..fed5e21577e5 100644 --- a/val.py +++ b/val.py @@ -301,7 +301,7 @@ def run( json.dump(jdict, f) try: # https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocoEvalDemo.ipynb - check_requirements(['pycocotools']) + check_requirements('pycocotools') from pycocotools.coco import COCO from pycocotools.cocoeval import COCOeval @@ -360,7 +360,7 @@ def parse_opt(): def main(opt): - check_requirements(requirements=ROOT / 'requirements.txt', exclude=('tensorboard', 'thop')) + check_requirements(exclude=('tensorboard', 'thop')) if opt.task in ('train', 'val', 'test'): # run normally if opt.conf_thres > 0.001: # https://github.com/ultralytics/yolov5/issues/1466 From e9ddc5b5274be1d795a28542159d7c9293efccea Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 10 Sep 2022 12:00:16 +0300 Subject: [PATCH 566/661] Update `check_requirements(args, cmds='')` (#9355) * Update `check_requirements(args, cmds='')` * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 2 +- utils/general.py | 10 +++++----- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/export.py b/export.py index 8fed4d3e3661..cdf5dcddd07a 100644 --- a/export.py +++ b/export.py @@ -226,7 +226,7 @@ def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose import tensorrt as trt except Exception: if platform.system() == 'Linux': - check_requirements('nvidia-tensorrt', cmds=['-U --index-url https://pypi.ngc.nvidia.com']) + check_requirements('nvidia-tensorrt', cmds='-U --index-url https://pypi.ngc.nvidia.com') import tensorrt as trt if trt.__version__[0] == '7': # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012 diff --git a/utils/general.py b/utils/general.py index 629df32ebc54..187d2c6b2d4a 100755 --- a/utils/general.py +++ b/utils/general.py @@ -341,13 +341,13 @@ def check_version(current='0.0.0', minimum='0.0.0', name='version ', pinned=Fals @TryExcept() -def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), install=True, cmds=()): +def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), install=True, cmds=''): # Check installed dependencies meet YOLOv5 requirements (pass *.txt file or list of packages or single package str) prefix = colorstr('red', 'bold', 'requirements:') check_python() # check python version if isinstance(requirements, Path): # requirements.txt file - file = requirements - assert file.exists(), f"{prefix} {file.resolve()} not found, check failed." + file = requirements.resolve() + assert file.exists(), f"{prefix} {file} not found, check failed." with file.open() as f: requirements = [f'{x.name}{x.specifier}' for x in pkg.parse_requirements(f) if x.name not in exclude] elif isinstance(requirements, str): @@ -366,8 +366,8 @@ def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), insta LOGGER.info(f"{prefix} YOLOv5 requirements {s}not found, attempting AutoUpdate...") try: assert check_online(), "AutoUpdate skipped (offline)" - LOGGER.info(check_output(f'pip install {s} {" ".join(cmds) if cmds else ""}', shell=True).decode()) - source = file.resolve() if 'file' in locals() else requirements + LOGGER.info(check_output(f'pip install {s} {cmds}', shell=True).decode()) + source = file if 'file' in locals() else requirements s = f"{prefix} {n} package{'s' * (n > 1)} updated per {source}\n" \ f"{prefix} ⚠️ {colorstr('bold', 'Restart runtime or rerun command for updates to take effect')}\n" LOGGER.info(s) From 57ef676af2358d70bd5902059531655789135510 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 10 Sep 2022 12:14:31 +0300 Subject: [PATCH 567/661] Update `check_requirements()` multiple string (#9356) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index 187d2c6b2d4a..33232efac9fd 100755 --- a/utils/general.py +++ b/utils/general.py @@ -363,7 +363,7 @@ def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), insta n += 1 if s and install and AUTOINSTALL: # check environment variable - LOGGER.info(f"{prefix} YOLOv5 requirements {s}not found, attempting AutoUpdate...") + LOGGER.info(f"{prefix} YOLOv5 requirement{'s' * (n > 1)} {s}not found, attempting AutoUpdate...") try: assert check_online(), "AutoUpdate skipped (offline)" LOGGER.info(check_output(f'pip install {s} {cmds}', shell=True).decode()) From e3e5122f82b0d1f24c11a90b2377fbb5a1673274 Mon Sep 17 00:00:00 2001 From: Katteria <39751846+kisaragychihaya@users.noreply.github.com> Date: Sat, 10 Sep 2022 17:20:46 +0800 Subject: [PATCH 568/661] Add PaddlePaddle export and inference (#9240) * Add PaddlePaddle Model Export Test on Yolov5 DockerEnviroment with paddlepaddle-gpu v2.2 Signed-off-by: Katteria <39751846+kisaragychihaya@users.noreply.github.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Cleanup Paddle Export Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update common.py Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update export.py Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher * Use PyTorch2Paddle Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Paddle no longer requires ONNX Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update benchmarks.py Signed-off-by: Glenn Jocher * Add inference code of PaddlePaddle Signed-off-by: Katteria <39751846+kisaragychihaya@users.noreply.github.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update common.py Signed-off-by: Katteria <39751846+kisaragychihaya@users.noreply.github.com> * Update common.py Signed-off-by: Glenn Jocher * Add paddlepaddle-gpu install if cuda Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update common.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher Signed-off-by: Katteria <39751846+kisaragychihaya@users.noreply.github.com> Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- export.py | 72 +++++++++++++++++----------- models/common.py | 114 ++++++++++++++++++++++++++------------------ utils/benchmarks.py | 2 +- 3 files changed, 112 insertions(+), 76 deletions(-) diff --git a/export.py b/export.py index cdf5dcddd07a..262b11a1a268 100644 --- a/export.py +++ b/export.py @@ -15,6 +15,7 @@ TensorFlow Lite | `tflite` | yolov5s.tflite TensorFlow Edge TPU | `edgetpu` | yolov5s_edgetpu.tflite TensorFlow.js | `tfjs` | yolov5s_web_model/ +PaddlePaddle | `paddle` | yolov5s_paddle_model/ Requirements: $ pip install -r requirements.txt coremltools onnx onnx-simplifier onnxruntime openvino-dev tensorflow-cpu # CPU @@ -54,7 +55,6 @@ import pandas as pd import torch -import yaml from torch.utils.mobile_optimizer import optimize_for_mobile FILE = Path(__file__).resolve() @@ -68,7 +68,7 @@ from models.yolo import ClassificationModel, Detect from utils.dataloaders import LoadImages from utils.general import (LOGGER, Profile, check_dataset, check_img_size, check_requirements, check_version, - check_yaml, colorstr, file_size, get_default_args, print_args, url2file) + check_yaml, colorstr, file_size, get_default_args, print_args, url2file, yaml_save) from utils.torch_utils import select_device, smart_inference_mode @@ -85,7 +85,8 @@ def export_formats(): ['TensorFlow GraphDef', 'pb', '.pb', True, True], ['TensorFlow Lite', 'tflite', '.tflite', True, False], ['TensorFlow Edge TPU', 'edgetpu', '_edgetpu.tflite', False, False], - ['TensorFlow.js', 'tfjs', '_web_model', False, False],] + ['TensorFlow.js', 'tfjs', '_web_model', False, False], + ['PaddlePaddle', 'paddle', '_paddle_model', True, True],] return pd.DataFrame(x, columns=['Format', 'Argument', 'Suffix', 'CPU', 'GPU']) @@ -180,7 +181,7 @@ def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorst @try_export -def export_openvino(model, file, half, prefix=colorstr('OpenVINO:')): +def export_openvino(file, metadata, half, prefix=colorstr('OpenVINO:')): # YOLOv5 OpenVINO export check_requirements('openvino-dev') # requires openvino-dev: https://pypi.org/project/openvino-dev/ import openvino.inference_engine as ie @@ -189,9 +190,23 @@ def export_openvino(model, file, half, prefix=colorstr('OpenVINO:')): f = str(file).replace('.pt', f'_openvino_model{os.sep}') cmd = f"mo --input_model {file.with_suffix('.onnx')} --output_dir {f} --data_type {'FP16' if half else 'FP32'}" - subprocess.check_output(cmd.split()) # export - with open(Path(f) / file.with_suffix('.yaml').name, 'w') as g: - yaml.dump({'stride': int(max(model.stride)), 'names': model.names}, g) # add metadata.yaml + subprocess.run(cmd.split(), check=True, env=os.environ) # export + yaml_save(Path(f) / file.with_suffix('.yaml').name, metadata) # add metadata.yaml + return f, None + + +@try_export +def export_paddle(model, im, file, metadata, prefix=colorstr('PaddlePaddle:')): + # YOLOv5 Paddle export + check_requirements(('paddlepaddle', 'x2paddle')) + import x2paddle + from x2paddle.convert import pytorch2paddle + + LOGGER.info(f'\n{prefix} starting export with X2Paddle {x2paddle.__version__}...') + f = str(file).replace('.pt', f'_paddle_model{os.sep}') + + pytorch2paddle(module=model, save_dir=f, jit_type='trace', input_examples=[im]) # export + yaml_save(Path(f) / file.with_suffix('.yaml').name, metadata) # add metadata.yaml return f, None @@ -464,7 +479,7 @@ def run( fmts = tuple(export_formats()['Argument'][1:]) # --include arguments flags = [x in include for x in fmts] assert sum(flags) == len(include), f'ERROR: Invalid --include {include}, valid --include arguments are {fmts}' - jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs = flags # export booleans + jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle = flags # export booleans file = Path(url2file(weights) if str(weights).startswith(('http:/', 'https:/')) else weights) # PyTorch weights # Load PyTorch model @@ -497,47 +512,48 @@ def run( if half and not coreml: im, model = im.half(), model.half() # to FP16 shape = tuple((y[0] if isinstance(y, tuple) else y).shape) # model output shape + metadata = {'stride': int(max(model.stride)), 'names': model.names} # model metadata LOGGER.info(f"\n{colorstr('PyTorch:')} starting from {file} with output shape {shape} ({file_size(file):.1f} MB)") # Exports - f = [''] * 10 # exported filenames + f = [''] * len(fmts) # exported filenames warnings.filterwarnings(action='ignore', category=torch.jit.TracerWarning) # suppress TracerWarning - if jit: + if jit: # TorchScript f[0], _ = export_torchscript(model, im, file, optimize) if engine: # TensorRT required before ONNX f[1], _ = export_engine(model, im, file, half, dynamic, simplify, workspace, verbose) if onnx or xml: # OpenVINO requires ONNX f[2], _ = export_onnx(model, im, file, opset, train, dynamic, simplify) if xml: # OpenVINO - f[3], _ = export_openvino(model, file, half) - if coreml: + f[3], _ = export_openvino(file, metadata, half) + if coreml: # CoreML f[4], _ = export_coreml(model, im, file, int8, half) - - # TensorFlow Exports - if any((saved_model, pb, tflite, edgetpu, tfjs)): + if any((saved_model, pb, tflite, edgetpu, tfjs)): # TensorFlow formats if int8 or edgetpu: # TFLite --int8 bug https://github.com/ultralytics/yolov5/issues/5707 check_requirements('flatbuffers==1.12') # required before `import tensorflow` assert not tflite or not tfjs, 'TFLite and TF.js models must be exported separately, please pass only one type.' assert not isinstance(model, ClassificationModel), 'ClassificationModel export to TF formats not yet supported.' - f[5], model = export_saved_model(model.cpu(), - im, - file, - dynamic, - tf_nms=nms or agnostic_nms or tfjs, - agnostic_nms=agnostic_nms or tfjs, - topk_per_class=topk_per_class, - topk_all=topk_all, - iou_thres=iou_thres, - conf_thres=conf_thres, - keras=keras) + f[5], s_model = export_saved_model(model.cpu(), + im, + file, + dynamic, + tf_nms=nms or agnostic_nms or tfjs, + agnostic_nms=agnostic_nms or tfjs, + topk_per_class=topk_per_class, + topk_all=topk_all, + iou_thres=iou_thres, + conf_thres=conf_thres, + keras=keras) if pb or tfjs: # pb prerequisite to tfjs - f[6], _ = export_pb(model, file) + f[6], _ = export_pb(s_model, file) if tflite or edgetpu: - f[7], _ = export_tflite(model, im, file, int8 or edgetpu, data=data, nms=nms, agnostic_nms=agnostic_nms) + f[7], _ = export_tflite(s_model, im, file, int8 or edgetpu, data=data, nms=nms, agnostic_nms=agnostic_nms) if edgetpu: f[8], _ = export_edgetpu(file) if tfjs: f[9], _ = export_tfjs(file) + if paddle: # PaddlePaddle + f[10], _ = export_paddle(model, im, file, metadata) # Finish f = [str(x) for x in f if x] # filter out '' and None diff --git a/models/common.py b/models/common.py index 0e01b60e81e5..396b5de0b505 100644 --- a/models/common.py +++ b/models/common.py @@ -320,14 +320,16 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, # TensorFlow GraphDef: *.pb # TensorFlow Lite: *.tflite # TensorFlow Edge TPU: *_edgetpu.tflite + # PaddlePaddle: *_paddle_model from models.experimental import attempt_download, attempt_load # scoped to avoid circular import super().__init__() w = str(weights[0] if isinstance(weights, list) else weights) - pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs = self._model_type(w) # get backend + pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle = self._model_type(w) # type w = attempt_download(w) # download if not local fp16 &= pt or jit or onnx or engine # FP16 stride = 32 # default stride + cuda = torch.cuda.is_available() and device.type != 'cpu' # use CUDA if pt: # PyTorch model = attempt_load(weights if isinstance(weights, list) else w, device=device, inplace=True, fuse=fuse) @@ -351,7 +353,6 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, net = cv2.dnn.readNetFromONNX(w) elif onnx: # ONNX Runtime LOGGER.info(f'Loading {w} for ONNX Runtime inference...') - cuda = torch.cuda.is_available() and device.type != 'cpu' check_requirements(('onnx', 'onnxruntime-gpu' if cuda else 'onnxruntime')) import onnxruntime providers = ['CUDAExecutionProvider', 'CPUExecutionProvider'] if cuda else ['CPUExecutionProvider'] @@ -408,48 +409,60 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, LOGGER.info(f'Loading {w} for CoreML inference...') import coremltools as ct model = ct.models.MLModel(w) - else: # TensorFlow (SavedModel, GraphDef, Lite, Edge TPU) - if saved_model: # SavedModel - LOGGER.info(f'Loading {w} for TensorFlow SavedModel inference...') - import tensorflow as tf - keras = False # assume TF1 saved_model - model = tf.keras.models.load_model(w) if keras else tf.saved_model.load(w) - elif pb: # GraphDef https://www.tensorflow.org/guide/migrate#a_graphpb_or_graphpbtxt - LOGGER.info(f'Loading {w} for TensorFlow GraphDef inference...') + elif saved_model: # TF SavedModel + LOGGER.info(f'Loading {w} for TensorFlow SavedModel inference...') + import tensorflow as tf + keras = False # assume TF1 saved_model + model = tf.keras.models.load_model(w) if keras else tf.saved_model.load(w) + elif pb: # GraphDef https://www.tensorflow.org/guide/migrate#a_graphpb_or_graphpbtxt + LOGGER.info(f'Loading {w} for TensorFlow GraphDef inference...') + import tensorflow as tf + + def wrap_frozen_graph(gd, inputs, outputs): + x = tf.compat.v1.wrap_function(lambda: tf.compat.v1.import_graph_def(gd, name=""), []) # wrapped + ge = x.graph.as_graph_element + return x.prune(tf.nest.map_structure(ge, inputs), tf.nest.map_structure(ge, outputs)) + + gd = tf.Graph().as_graph_def() # TF GraphDef + with open(w, 'rb') as f: + gd.ParseFromString(f.read()) + frozen_func = wrap_frozen_graph(gd, inputs="x:0", outputs="Identity:0") + elif tflite or edgetpu: # https://www.tensorflow.org/lite/guide/python#install_tensorflow_lite_for_python + try: # https://coral.ai/docs/edgetpu/tflite-python/#update-existing-tf-lite-code-for-the-edge-tpu + from tflite_runtime.interpreter import Interpreter, load_delegate + except ImportError: import tensorflow as tf - - def wrap_frozen_graph(gd, inputs, outputs): - x = tf.compat.v1.wrap_function(lambda: tf.compat.v1.import_graph_def(gd, name=""), []) # wrapped - ge = x.graph.as_graph_element - return x.prune(tf.nest.map_structure(ge, inputs), tf.nest.map_structure(ge, outputs)) - - gd = tf.Graph().as_graph_def() # graph_def - with open(w, 'rb') as f: - gd.ParseFromString(f.read()) - frozen_func = wrap_frozen_graph(gd, inputs="x:0", outputs="Identity:0") - elif tflite or edgetpu: # https://www.tensorflow.org/lite/guide/python#install_tensorflow_lite_for_python - try: # https://coral.ai/docs/edgetpu/tflite-python/#update-existing-tf-lite-code-for-the-edge-tpu - from tflite_runtime.interpreter import Interpreter, load_delegate - except ImportError: - import tensorflow as tf - Interpreter, load_delegate = tf.lite.Interpreter, tf.lite.experimental.load_delegate, - if edgetpu: # Edge TPU https://coral.ai/software/#edgetpu-runtime - LOGGER.info(f'Loading {w} for TensorFlow Lite Edge TPU inference...') - delegate = { - 'Linux': 'libedgetpu.so.1', - 'Darwin': 'libedgetpu.1.dylib', - 'Windows': 'edgetpu.dll'}[platform.system()] - interpreter = Interpreter(model_path=w, experimental_delegates=[load_delegate(delegate)]) - else: # Lite - LOGGER.info(f'Loading {w} for TensorFlow Lite inference...') - interpreter = Interpreter(model_path=w) # load TFLite model - interpreter.allocate_tensors() # allocate - input_details = interpreter.get_input_details() # inputs - output_details = interpreter.get_output_details() # outputs - elif tfjs: - raise NotImplementedError('ERROR: YOLOv5 TF.js inference is not supported') - else: - raise NotImplementedError(f'ERROR: {w} is not a supported format') + Interpreter, load_delegate = tf.lite.Interpreter, tf.lite.experimental.load_delegate, + if edgetpu: # TF Edge TPU https://coral.ai/software/#edgetpu-runtime + LOGGER.info(f'Loading {w} for TensorFlow Lite Edge TPU inference...') + delegate = { + 'Linux': 'libedgetpu.so.1', + 'Darwin': 'libedgetpu.1.dylib', + 'Windows': 'edgetpu.dll'}[platform.system()] + interpreter = Interpreter(model_path=w, experimental_delegates=[load_delegate(delegate)]) + else: # TFLite + LOGGER.info(f'Loading {w} for TensorFlow Lite inference...') + interpreter = Interpreter(model_path=w) # load TFLite model + interpreter.allocate_tensors() # allocate + input_details = interpreter.get_input_details() # inputs + output_details = interpreter.get_output_details() # outputs + elif tfjs: # TF.js + raise NotImplementedError('ERROR: YOLOv5 TF.js inference is not supported') + elif paddle: # PaddlePaddle + LOGGER.info(f'Loading {w} for PaddlePaddle inference...') + check_requirements('paddlepaddle-gpu' if cuda else 'paddlepaddle') + import paddle.inference as pdi + if not Path(w).is_file(): # if not *.pdmodel + w = next(Path(w).rglob('*.pdmodel')) # get *.xml file from *_openvino_model dir + weights = Path(w).with_suffix('.pdiparams') + config = pdi.Config(str(w), str(weights)) + if cuda: + config.enable_use_gpu(memory_pool_init_size_mb=2048, device_id=0) + predictor = pdi.create_predictor(config) + input_names = predictor.get_input_names() + input_handle = predictor.get_input_handle(input_names[0]) + else: + raise NotImplementedError(f'ERROR: {w} is not a supported format') # class names if 'names' not in locals(): @@ -502,6 +515,13 @@ def forward(self, im, augment=False, visualize=False): else: k = 'var_' + str(sorted(int(k.replace('var_', '')) for k in y)[-1]) # output key y = y[k] # output + elif self.paddle: # PaddlePaddle + im = im.cpu().numpy().astype("float32") + self.input_handle.copy_from_cpu(im) + self.predictor.run() + output_names = self.predictor.get_output_names() + output_handle = self.predictor.get_output_handle(output_names[0]) + y = output_handle.copy_to_cpu() else: # TensorFlow (SavedModel, GraphDef, Lite, Edge TPU) im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) if self.saved_model: # SavedModel @@ -542,13 +562,13 @@ def warmup(self, imgsz=(1, 3, 640, 640)): def _model_type(p='path/to/model.pt'): # Return model type from model path, i.e. path='path/to/model.onnx' -> type=onnx from export import export_formats - suffixes = list(export_formats().Suffix) + ['.xml'] # export suffixes - check_suffix(p, suffixes) # checks + sf = list(export_formats().Suffix) + ['.xml'] # export suffixes + check_suffix(p, sf) # checks p = Path(p).name # eliminate trailing separators - pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, xml2 = (s in p for s in suffixes) + pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle, xml2 = (s in p for s in sf) xml |= xml2 # *_openvino_model or *.xml tflite &= not edgetpu # *.tflite - return pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs + return pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle @staticmethod def _load_metadata(f=Path('path/to/meta.yaml')): diff --git a/utils/benchmarks.py b/utils/benchmarks.py index d5f4c1d61fbe..9d5c7f2965d5 100644 --- a/utils/benchmarks.py +++ b/utils/benchmarks.py @@ -61,7 +61,7 @@ def run( device = select_device(device) for i, (name, f, suffix, cpu, gpu) in export.export_formats().iterrows(): # index, (name, file, suffix, CPU, GPU) try: - assert i not in (9, 10), 'inference not supported' # Edge TPU and TF.js are unsupported + assert i not in (9, 10, 11), 'inference not supported' # Edge TPU, TF.js and Paddle are unsupported assert i != 5 or platform.system() == 'Darwin', 'inference only supported on macOS>=10.13' # CoreML if 'cpu' in device.type: assert cpu, 'inference not supported on CPU' From 4e8504abd9c1a7287dfcf9f96dfa04f061086cca Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 10 Sep 2022 13:25:01 +0300 Subject: [PATCH 569/661] PaddlePaddle Usage examples (#9358) --- classify/predict.py | 1 + classify/val.py | 1 + detect.py | 1 + export.py | 1 + models/common.py | 2 +- val.py | 1 + 6 files changed, 6 insertions(+), 1 deletion(-) diff --git a/classify/predict.py b/classify/predict.py index 701b5b1ac92d..878cf48b6fef 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -22,6 +22,7 @@ yolov5s-cls.pb # TensorFlow GraphDef yolov5s-cls.tflite # TensorFlow Lite yolov5s-cls_edgetpu.tflite # TensorFlow Edge TPU + yolov5s-cls_paddle_model # PaddlePaddle """ import argparse diff --git a/classify/val.py b/classify/val.py index bf808bc21a84..3c16ec8092d8 100644 --- a/classify/val.py +++ b/classify/val.py @@ -17,6 +17,7 @@ yolov5s-cls.pb # TensorFlow GraphDef yolov5s-cls.tflite # TensorFlow Lite yolov5s-cls_edgetpu.tflite # TensorFlow Edge TPU + yolov5s-cls_paddle_model # PaddlePaddle """ import argparse diff --git a/detect.py b/detect.py index 69a1bf13aac6..a69606a3dff9 100644 --- a/detect.py +++ b/detect.py @@ -22,6 +22,7 @@ yolov5s.pb # TensorFlow GraphDef yolov5s.tflite # TensorFlow Lite yolov5s_edgetpu.tflite # TensorFlow Edge TPU + yolov5s_paddle_model # PaddlePaddle """ import argparse diff --git a/export.py b/export.py index 262b11a1a268..9d33024a9ca4 100644 --- a/export.py +++ b/export.py @@ -35,6 +35,7 @@ yolov5s.pb # TensorFlow GraphDef yolov5s.tflite # TensorFlow Lite yolov5s_edgetpu.tflite # TensorFlow Edge TPU + yolov5s_paddle_model # PaddlePaddle TensorFlow.js: $ cd .. && git clone https://github.com/zldrobit/tfjs-yolov5-example.git && cd tfjs-yolov5-example diff --git a/models/common.py b/models/common.py index 396b5de0b505..c601aacc885c 100644 --- a/models/common.py +++ b/models/common.py @@ -312,7 +312,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, # PyTorch: weights = *.pt # TorchScript: *.torchscript # ONNX Runtime: *.onnx - # ONNX OpenCV DNN: *.onnx with --dnn + # ONNX OpenCV DNN: *.onnx --dnn # OpenVINO: *.xml # CoreML: *.mlmodel # TensorRT: *.engine diff --git a/val.py b/val.py index fed5e21577e5..4b0bdddae3b1 100644 --- a/val.py +++ b/val.py @@ -16,6 +16,7 @@ yolov5s.pb # TensorFlow GraphDef yolov5s.tflite # TensorFlow Lite yolov5s_edgetpu.tflite # TensorFlow Edge TPU + yolov5s_paddle_model # PaddlePaddle """ import argparse From 2b5c9a83ec4953c68159a924b338a646554a4490 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 10 Sep 2022 22:24:46 +0300 Subject: [PATCH 570/661] labels.jpg names fix (#9361) Partially resolves https://github.com/ultralytics/yolov5/issues/9360 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/plots.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/plots.py b/utils/plots.py index 0f322b6b5844..0530d0abdf48 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -364,7 +364,7 @@ def plot_labels(labels, names=(), save_dir=Path('')): ax[0].set_ylabel('instances') if 0 < len(names) < 30: ax[0].set_xticks(range(len(names))) - ax[0].set_xticklabels(names, rotation=90, fontsize=10) + ax[0].set_xticklabels(list(names.values()), rotation=90, fontsize=10) else: ax[0].set_xlabel('classes') sn.histplot(x, x='x', y='y', ax=ax[2], bins=50, pmax=0.9) From cafdd189397992cf93ec0ad6b76929c60ff09a17 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 10 Sep 2022 22:58:24 +0300 Subject: [PATCH 571/661] Exclude `ipython` from hubconf.py `check_requirements()` (#9362) Exclude ipython from hubconf.py `check_requirements()` Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- hubconf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hubconf.py b/hubconf.py index bffe2d588b4f..2f05565629a5 100644 --- a/hubconf.py +++ b/hubconf.py @@ -37,7 +37,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo if not verbose: LOGGER.setLevel(logging.WARNING) - check_requirements(exclude=('tensorboard', 'thop', 'opencv-python')) + check_requirements(exclude=('ipython', 'opencv-python', 'tensorboard', 'thop')) name = Path(name) path = name.with_suffix('.pt') if name.suffix == '' and not name.is_dir() else name # checkpoint path try: From 23d0456b08cac22f783d63292cc7c2bf87a19a60 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 10 Sep 2022 23:55:18 +0300 Subject: [PATCH 572/661] `torch.jit.trace()` fix (#9363) * Update common.py Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/workflows/ci-testing.yml | 3 +++ models/common.py | 1 + 2 files changed, 4 insertions(+) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 6fb277676959..a83f997cbfc2 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -119,9 +119,12 @@ jobs: python export.py --weights $m.pt --img 64 --include torchscript # export python - < Date: Sun, 11 Sep 2022 13:56:51 +0300 Subject: [PATCH 573/661] AMP Check fix (#9367) Resolves https://github.com/ultralytics/yolov5/issues/9365 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index 33232efac9fd..f5fb2c93a3d5 100755 --- a/utils/general.py +++ b/utils/general.py @@ -17,6 +17,7 @@ import sys import time import urllib +from copy import deepcopy from datetime import datetime from itertools import repeat from multiprocessing.pool import ThreadPool @@ -535,7 +536,7 @@ def amp_allclose(model, im): f = ROOT / 'data' / 'images' / 'bus.jpg' # image to check im = f if f.exists() else 'https://ultralytics.com/images/bus.jpg' if check_online() else np.ones((640, 640, 3)) try: - assert amp_allclose(model, im) or amp_allclose(DetectMultiBackend('yolov5n.pt', device), im) + assert amp_allclose(deepcopy(model), im) or amp_allclose(DetectMultiBackend('yolov5n.pt', device), im) LOGGER.info(f'{prefix}checks passed ✅') return True except Exception: From a4ed9888938a090631ca4dba5be6363f8b66575c Mon Sep 17 00:00:00 2001 From: Jiacong Fang Date: Wed, 14 Sep 2022 05:50:23 +0800 Subject: [PATCH 574/661] Remove duplicate line in setup.cfg (#9380) --- setup.cfg | 1 - 1 file changed, 1 deletion(-) diff --git a/setup.cfg b/setup.cfg index 020a75740e97..f12995da3e8e 100644 --- a/setup.cfg +++ b/setup.cfg @@ -34,7 +34,6 @@ ignore = F401 # module imported but unused W504 # line break after binary operator E127 # continuation line over-indented for visual indent - W504 # line break after binary operator E231 # missing whitespace after ‘,’, ‘;’, or ‘:’ E501 # line too long F403 # ‘from module import *’ used; unable to detect undefined names From 1323b4805319ca18e4ffd8f93f3e855b87093ad4 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 15 Sep 2022 19:05:10 +0200 Subject: [PATCH 575/661] Remove `.train()` mode exports (#9429) * Remove `.train()` mode exports No common use cases. Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/export.py b/export.py index 9d33024a9ca4..1b25f3f8221b 100644 --- a/export.py +++ b/export.py @@ -126,7 +126,7 @@ def export_torchscript(model, im, file, optimize, prefix=colorstr('TorchScript:' @try_export -def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorstr('ONNX:')): +def export_onnx(model, im, file, opset, dynamic, simplify, prefix=colorstr('ONNX:')): # YOLOv5 ONNX export check_requirements('onnx') import onnx @@ -140,8 +140,7 @@ def export_onnx(model, im, file, opset, train, dynamic, simplify, prefix=colorst f, verbose=False, opset_version=opset, - training=torch.onnx.TrainingMode.TRAINING if train else torch.onnx.TrainingMode.EVAL, - do_constant_folding=not train, + do_constant_folding=True, input_names=['images'], output_names=['output'], dynamic_axes={ @@ -459,7 +458,6 @@ def run( include=('torchscript', 'onnx'), # include formats half=False, # FP16 half-precision export inplace=False, # set YOLOv5 Detect() inplace=True - train=False, # model.train() mode keras=False, # use Keras optimize=False, # TorchScript: optimize for mobile int8=False, # CoreML/TF INT8 quantization @@ -501,7 +499,7 @@ def run( im = torch.zeros(batch_size, 3, *imgsz).to(device) # image size(1,3,320,192) BCHW iDetection # Update model - model.train() if train else model.eval() # training mode = no Detect() layer grid construction + model.eval() for k, m in model.named_modules(): if isinstance(m, Detect): m.inplace = inplace @@ -524,7 +522,7 @@ def run( if engine: # TensorRT required before ONNX f[1], _ = export_engine(model, im, file, half, dynamic, simplify, workspace, verbose) if onnx or xml: # OpenVINO requires ONNX - f[2], _ = export_onnx(model, im, file, opset, train, dynamic, simplify) + f[2], _ = export_onnx(model, im, file, opset, dynamic, simplify) if xml: # OpenVINO f[3], _ = export_openvino(file, metadata, half) if coreml: # CoreML @@ -578,7 +576,6 @@ def parse_opt(): parser.add_argument('--device', default='cpu', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--half', action='store_true', help='FP16 half-precision export') parser.add_argument('--inplace', action='store_true', help='set YOLOv5 Detect() inplace=True') - parser.add_argument('--train', action='store_true', help='model.train() mode') parser.add_argument('--keras', action='store_true', help='TF: use Keras') parser.add_argument('--optimize', action='store_true', help='TorchScript: optimize for mobile') parser.add_argument('--int8', action='store_true', help='CoreML/TF INT8 quantization') From 36cb05b7b211d4c5d99586dd49d3195de16e4485 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 15 Sep 2022 23:28:33 +0200 Subject: [PATCH 576/661] Continue on Docker arm64 failure (#9430) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/workflows/docker.yml | 1 + 1 file changed, 1 insertion(+) diff --git a/.github/workflows/docker.yml b/.github/workflows/docker.yml index c89d0ada3219..67ef565474a4 100644 --- a/.github/workflows/docker.yml +++ b/.github/workflows/docker.yml @@ -30,6 +30,7 @@ jobs: - name: Build and push arm64 image uses: docker/build-push-action@v3 + continue-on-error: true with: context: . platforms: linux/arm64 From 65afaa78beaa3d68d457e9c49109dc6327003962 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 15 Sep 2022 23:53:36 +0200 Subject: [PATCH 577/661] Continue on Docker failure (all backends) (#9432) Continue on Docker failure (all) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/workflows/docker.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/docker.yml b/.github/workflows/docker.yml index 67ef565474a4..f9eec3bd839e 100644 --- a/.github/workflows/docker.yml +++ b/.github/workflows/docker.yml @@ -12,6 +12,7 @@ jobs: if: github.repository == 'ultralytics/yolov5' name: Push Docker image to Docker Hub runs-on: ubuntu-latest + continue-on-error: true steps: - name: Checkout repo uses: actions/checkout@v3 @@ -30,7 +31,6 @@ jobs: - name: Build and push arm64 image uses: docker/build-push-action@v3 - continue-on-error: true with: context: . platforms: linux/arm64 From abea53ea5b7d4eba6b58535d31e17336912d0d1f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 16 Sep 2022 00:10:10 +0200 Subject: [PATCH 578/661] Continue on Docker fail (all backends) fix (#9433) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/workflows/docker.yml | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/.github/workflows/docker.yml b/.github/workflows/docker.yml index f9eec3bd839e..1d0bd30b22cb 100644 --- a/.github/workflows/docker.yml +++ b/.github/workflows/docker.yml @@ -12,7 +12,6 @@ jobs: if: github.repository == 'ultralytics/yolov5' name: Push Docker image to Docker Hub runs-on: ubuntu-latest - continue-on-error: true steps: - name: Checkout repo uses: actions/checkout@v3 @@ -31,6 +30,7 @@ jobs: - name: Build and push arm64 image uses: docker/build-push-action@v3 + continue-on-error: true with: context: . platforms: linux/arm64 @@ -40,6 +40,7 @@ jobs: - name: Build and push CPU image uses: docker/build-push-action@v3 + continue-on-error: true with: context: . file: utils/docker/Dockerfile-cpu @@ -48,6 +49,7 @@ jobs: - name: Build and push GPU image uses: docker/build-push-action@v3 + continue-on-error: true with: context: . file: utils/docker/Dockerfile From f9869f7ffdbce757f260d28a6b799c5fa50263ee Mon Sep 17 00:00:00 2001 From: Ayush Chaurasia Date: Fri, 16 Sep 2022 03:42:46 +0530 Subject: [PATCH 579/661] YOLOv5 segmentation model support (#9052) * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix duplicate plots.py * Fix check_font() * # torch.use_deterministic_algorithms(True) * update doc detect->predict * Resolve precommit for segment/train and segment/val * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Resolve precommit for utils/segment * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Resolve precommit min_wh * Resolve precommit utils/segment/plots * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Resolve precommit utils/segment/general * Align NMS-seg closer to NMS * restore deterministic init_seeds code * remove easydict dependency * update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * restore output_to_target mask * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * update * cleanup * Remove unused ImageFont import * Unified NMS * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * DetectMultiBackend compatibility * segment/predict.py update * update plot colors * fix bbox shifted * sort bbox by confidence * enable overlap by default * Merge detect/segment output_to_target() function * Start segmentation CI * fix plots * Update ci-testing.yml * fix training whitespace * optimize process mask functions (can we merge both?) * Update predict/detect * Update plot_images * Update plot_images_and_masks * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix * Add train to CI * fix precommit * fix precommit CI * fix precommit pycocotools * fix val float issues * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix masks float float issues * suppress errors * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix no-predictions plotting bug * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Add CSV Logger * fix val len(plot_masks) * speed up evaluation * fix process_mask * fix plots * update segment/utils build_targets * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * optimize utils/segment/general crop() * optimize utils/segment/general crop() 2 * minor updates * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * torch.where revert * downsample only if different shape * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * loss cleanup * loss cleanup 2 * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * loss cleanup 3 * update project names * Rename -seg yamls from _underscore to -dash * prepare for yolov5n-seg.pt * precommit space fix * add coco128-seg.yaml * update coco128-seg comments * cleanup val.py * Major val.py cleanup * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * precommit fix * precommit fix * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * optional pycocotools * remove CI pip install pycocotools (auto-installed now) * seg yaml fix * optimize mask_iou() and masks_iou() * threaded fix * Major train.py update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Major segments/val/process_batch() update * yolov5/val updates from segment * process_batch numpy/tensor fix * opt-in to pycocotools with --save-json * threaded pycocotools ops for 2x speed increase * Avoid permute contiguous if possible * Add max_det=300 argument to both val.py and segment/val.py * fix onnx_dynamic * speed up pycocotools ops * faster process_mask(upsample=True) for predict * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * eliminate permutations for process_mask(upsample=True) * eliminate permute-contiguous in crop(), use native dimension order * cleanup comment * Add Proto() module * fix class count * fix anchor order * broadcast mask_gti in loss for speed * Cleanup seg loss * faster indexing * faster indexing fix * faster indexing fix2 * revert faster indexing * fix validation plotting * Loss cleanup and mxyxy simplification * Loss cleanup and mxyxy simplification 2 * revert validation plotting * replace missing tanh * Eliminate last permutation * delete unneeded .float() * Remove MaskIOULoss and crop(if HWC) * Final v6.3 SegmentationModel architecture updates * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Add support for TF export * remove debugger trace * add call * update * update * Merge master * Merge master * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update dataloaders.py * Restore CI * Update dataloaders.py * Fix TF/TFLite export for segmentation model * Merge master * Cleanup predict.py mask plotting * cleanup scale_masks() * rename scale_masks to scale_image * cleanup/optimize plot_masks * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Add Annotator.masks() * Annotator.masks() fix * Update plots.py * Annotator mask optimization * Rename crop() to crop_mask() * Do not crop in predict.py * crop always * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Merge master * Add vid-stride from master PR * Update seg model outputs * Update seg model outputs * Add segmentation benchmarks * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Add segmentation benchmarks * Add segmentation benchmarks * Add segmentation benchmarks * Fix DetectMultiBackend for OpenVINO * update Annotator.masks * fix val plot * revert val plot * clean up * revert pil * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix CI error * fix predict log * remove upsample * update interpolate * fix validation plot logging * Annotator.masks() cleanup * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Remove segmentation_model definition * Restore 0.99999 decimals Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher Co-authored-by: Laughing-q <1185102784@qq.com> Co-authored-by: Jiacong Fang --- .github/workflows/ci-testing.yml | 22 +- utils/benchmarks.py => benchmarks.py | 20 +- data/coco128-seg.yaml | 101 ++++ detect.py | 4 +- models/common.py | 18 +- models/segment/yolov5l-seg.yaml | 48 ++ models/segment/yolov5m-seg.yaml | 48 ++ models/segment/yolov5n-seg.yaml | 48 ++ models/segment/yolov5s-seg.yaml | 48 ++ models/segment/yolov5x-seg.yaml | 48 ++ models/tf.py | 36 +- models/yolo.py | 58 ++- segment/predict.py | 266 +++++++++++ segment/train.py | 676 +++++++++++++++++++++++++++ segment/val.py | 471 +++++++++++++++++++ utils/dataloaders.py | 1 + utils/general.py | 45 +- utils/metrics.py | 10 +- utils/plots.py | 71 ++- utils/segment/__init__.py | 0 utils/segment/augmentations.py | 104 +++++ utils/segment/dataloaders.py | 330 +++++++++++++ utils/segment/general.py | 120 +++++ utils/segment/loss.py | 186 ++++++++ utils/segment/metrics.py | 210 +++++++++ utils/segment/plots.py | 143 ++++++ val.py | 30 +- 27 files changed, 3091 insertions(+), 71 deletions(-) rename utils/benchmarks.py => benchmarks.py (87%) create mode 100644 data/coco128-seg.yaml create mode 100644 models/segment/yolov5l-seg.yaml create mode 100644 models/segment/yolov5m-seg.yaml create mode 100644 models/segment/yolov5n-seg.yaml create mode 100644 models/segment/yolov5s-seg.yaml create mode 100644 models/segment/yolov5x-seg.yaml create mode 100644 segment/predict.py create mode 100644 segment/train.py create mode 100644 segment/val.py mode change 100755 => 100644 utils/dataloaders.py mode change 100755 => 100644 utils/general.py create mode 100644 utils/segment/__init__.py create mode 100644 utils/segment/augmentations.py create mode 100644 utils/segment/dataloaders.py create mode 100644 utils/segment/general.py create mode 100644 utils/segment/loss.py create mode 100644 utils/segment/metrics.py create mode 100644 utils/segment/plots.py diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index a83f997cbfc2..537ba96e7225 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -15,6 +15,7 @@ jobs: Benchmarks: runs-on: ${{ matrix.os }} strategy: + fail-fast: false matrix: os: [ ubuntu-latest ] python-version: [ '3.9' ] # requires python<=3.9 @@ -37,9 +38,12 @@ jobs: python --version pip --version pip list - - name: Run benchmarks + - name: Benchmark DetectionModel + run: | + python benchmarks.py --data coco128.yaml --weights ${{ matrix.model }}.pt --img 320 --hard-fail 0.29 + - name: Benchmark SegmentationModel run: | - python utils/benchmarks.py --weights ${{ matrix.model }}.pt --img 320 --hard-fail 0.29 + python benchmarks.py --data coco128-seg.yaml --weights ${{ matrix.model }}-seg.pt --img 320 Tests: timeout-minutes: 60 @@ -126,6 +130,20 @@ jobs: model(im) # warmup, build grids for trace torch.jit.trace(model, [im]) EOF + - name: Test segmentation + shell: bash # for Windows compatibility + run: | + m=${{ matrix.model }}-seg # official weights + b=runs/train-seg/exp/weights/best # best.pt checkpoint + python segment/train.py --imgsz 64 --batch 32 --weights $m.pt --cfg $m.yaml --epochs 1 --device cpu # train + python segment/train.py --imgsz 64 --batch 32 --weights '' --cfg $m.yaml --epochs 1 --device cpu # train + for d in cpu; do # devices + for w in $m $b; do # weights + python segment/val.py --imgsz 64 --batch 32 --weights $w.pt --device $d # val + python segment/predict.py --imgsz 64 --weights $w.pt --device $d # predict + python export.py --weights $w.pt --img 64 --include torchscript --device $d # export + done + done - name: Test classification shell: bash # for Windows compatibility run: | diff --git a/utils/benchmarks.py b/benchmarks.py similarity index 87% rename from utils/benchmarks.py rename to benchmarks.py index 9d5c7f2965d5..58e083c95d55 100644 --- a/utils/benchmarks.py +++ b/benchmarks.py @@ -34,16 +34,19 @@ import pandas as pd FILE = Path(__file__).resolve() -ROOT = FILE.parents[1] # YOLOv5 root directory +ROOT = FILE.parents[0] # YOLOv5 root directory if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH # ROOT = ROOT.relative_to(Path.cwd()) # relative import export -import val +from models.experimental import attempt_load +from models.yolo import SegmentationModel +from segment.val import run as val_seg from utils import notebook_init from utils.general import LOGGER, check_yaml, file_size, print_args from utils.torch_utils import select_device +from val import run as val_det def run( @@ -59,6 +62,7 @@ def run( ): y, t = [], time.time() device = select_device(device) + model_type = type(attempt_load(weights, fuse=False)) # DetectionModel, SegmentationModel, etc. for i, (name, f, suffix, cpu, gpu) in export.export_formats().iterrows(): # index, (name, file, suffix, CPU, GPU) try: assert i not in (9, 10, 11), 'inference not supported' # Edge TPU, TF.js and Paddle are unsupported @@ -76,10 +80,14 @@ def run( assert suffix in str(w), 'export failed' # Validate - result = val.run(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) - metrics = result[0] # metrics (mp, mr, map50, map, *losses(box, obj, cls)) - speeds = result[2] # times (preprocess, inference, postprocess) - y.append([name, round(file_size(w), 1), round(metrics[3], 4), round(speeds[1], 2)]) # MB, mAP, t_inference + if model_type == SegmentationModel: + result = val_seg(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) + metric = result[0][7] # (box(p, r, map50, map), mask(p, r, map50, map), *loss(box, obj, cls)) + else: # DetectionModel: + result = val_det(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) + metric = result[0][3] # (p, r, map50, map, *loss(box, obj, cls)) + speed = result[2][1] # times (preprocess, inference, postprocess) + y.append([name, round(file_size(w), 1), round(metric, 4), round(speed, 2)]) # MB, mAP, t_inference except Exception as e: if hard_fail: assert type(e) is AssertionError, f'Benchmark --hard-fail for {name}: {e}' diff --git a/data/coco128-seg.yaml b/data/coco128-seg.yaml new file mode 100644 index 000000000000..5e81910cc456 --- /dev/null +++ b/data/coco128-seg.yaml @@ -0,0 +1,101 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +# COCO128-seg dataset https://www.kaggle.com/ultralytics/coco128 (first 128 images from COCO train2017) by Ultralytics +# Example usage: python train.py --data coco128.yaml +# parent +# ├── yolov5 +# └── datasets +# └── coco128-seg ← downloads here (7 MB) + + +# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..] +path: ../datasets/coco128-seg # dataset root dir +train: images/train2017 # train images (relative to 'path') 128 images +val: images/train2017 # val images (relative to 'path') 128 images +test: # test images (optional) + +# Classes +names: + 0: person + 1: bicycle + 2: car + 3: motorcycle + 4: airplane + 5: bus + 6: train + 7: truck + 8: boat + 9: traffic light + 10: fire hydrant + 11: stop sign + 12: parking meter + 13: bench + 14: bird + 15: cat + 16: dog + 17: horse + 18: sheep + 19: cow + 20: elephant + 21: bear + 22: zebra + 23: giraffe + 24: backpack + 25: umbrella + 26: handbag + 27: tie + 28: suitcase + 29: frisbee + 30: skis + 31: snowboard + 32: sports ball + 33: kite + 34: baseball bat + 35: baseball glove + 36: skateboard + 37: surfboard + 38: tennis racket + 39: bottle + 40: wine glass + 41: cup + 42: fork + 43: knife + 44: spoon + 45: bowl + 46: banana + 47: apple + 48: sandwich + 49: orange + 50: broccoli + 51: carrot + 52: hot dog + 53: pizza + 54: donut + 55: cake + 56: chair + 57: couch + 58: potted plant + 59: bed + 60: dining table + 61: toilet + 62: tv + 63: laptop + 64: mouse + 65: remote + 66: keyboard + 67: cell phone + 68: microwave + 69: oven + 70: toaster + 71: sink + 72: refrigerator + 73: book + 74: clock + 75: vase + 76: scissors + 77: teddy bear + 78: hair drier + 79: toothbrush + + +# Download script/URL (optional) +download: https://ultralytics.com/assets/coco128-seg.zip diff --git a/detect.py b/detect.py index a69606a3dff9..310d169281bf 100644 --- a/detect.py +++ b/detect.py @@ -149,8 +149,8 @@ def run( det[:, :4] = scale_coords(im.shape[2:], det[:, :4], im0.shape).round() # Print results - for c in det[:, -1].unique(): - n = (det[:, -1] == c).sum() # detections per class + for c in det[:, 5].unique(): + n = (det[:, 5] == c).sum() # detections per class s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string # Write results diff --git a/models/common.py b/models/common.py index 8b7dbbfa95fe..0d90ff4f8827 100644 --- a/models/common.py +++ b/models/common.py @@ -375,7 +375,6 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, if batch_dim.is_static: batch_size = batch_dim.get_length() executable_network = ie.compile_model(network, device_name="CPU") # device_name="MYRIAD" for Intel NCS2 - output_layer = next(iter(executable_network.outputs)) stride, names = self._load_metadata(Path(w).with_suffix('.yaml')) # load metadata elif engine: # TensorRT LOGGER.info(f'Loading {w} for TensorRT inference...') @@ -491,7 +490,7 @@ def forward(self, im, augment=False, visualize=False): y = self.session.run(self.output_names, {self.session.get_inputs()[0].name: im}) elif self.xml: # OpenVINO im = im.cpu().numpy() # FP32 - y = self.executable_network([im])[self.output_layer] + y = list(self.executable_network([im]).values()) elif self.engine: # TensorRT if self.dynamic and im.shape != self.bindings['images'].shape: i_in, i_out = (self.model.get_binding_index(x) for x in ('images', 'output')) @@ -786,8 +785,21 @@ def __str__(self): return '' +class Proto(nn.Module): + # YOLOv5 mask Proto module for segmentation models + def __init__(self, c1, c_=256, c2=32): # ch_in, number of protos, number of masks + super().__init__() + self.cv1 = Conv(c1, c_, k=3) + self.upsample = nn.Upsample(scale_factor=2, mode='nearest') + self.cv2 = Conv(c_, c_, k=3) + self.cv3 = Conv(c_, c2) + + def forward(self, x): + return self.cv3(self.cv2(self.upsample(self.cv1(x)))) + + class Classify(nn.Module): - # Classification head, i.e. x(b,c1,20,20) to x(b,c2) + # YOLOv5 classification head, i.e. x(b,c1,20,20) to x(b,c2) def __init__(self, c1, c2, k=1, s=1, p=None, g=1): # ch_in, ch_out, kernel, stride, padding, groups super().__init__() c_ = 1280 # efficientnet_b0 size diff --git a/models/segment/yolov5l-seg.yaml b/models/segment/yolov5l-seg.yaml new file mode 100644 index 000000000000..4782de11dd2d --- /dev/null +++ b/models/segment/yolov5l-seg.yaml @@ -0,0 +1,48 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license + +# Parameters +nc: 80 # number of classes +depth_multiple: 1.0 # model depth multiple +width_multiple: 1.0 # layer channel multiple +anchors: + - [10,13, 16,30, 33,23] # P3/8 + - [30,61, 62,45, 59,119] # P4/16 + - [116,90, 156,198, 373,326] # P5/32 + +# YOLOv5 v6.0 backbone +backbone: + # [from, number, module, args] + [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 + [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 + [-1, 3, C3, [128]], + [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 + [-1, 6, C3, [256]], + [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 + [-1, 9, C3, [512]], + [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 + [-1, 3, C3, [1024]], + [-1, 1, SPPF, [1024, 5]], # 9 + ] + +# YOLOv5 v6.0 head +head: + [[-1, 1, Conv, [512, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 6], 1, Concat, [1]], # cat backbone P4 + [-1, 3, C3, [512, False]], # 13 + + [-1, 1, Conv, [256, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 4], 1, Concat, [1]], # cat backbone P3 + [-1, 3, C3, [256, False]], # 17 (P3/8-small) + + [-1, 1, Conv, [256, 3, 2]], + [[-1, 14], 1, Concat, [1]], # cat head P4 + [-1, 3, C3, [512, False]], # 20 (P4/16-medium) + + [-1, 1, Conv, [512, 3, 2]], + [[-1, 10], 1, Concat, [1]], # cat head P5 + [-1, 3, C3, [1024, False]], # 23 (P5/32-large) + + [[17, 20, 23], 1, Segment, [nc, anchors, 32, 256]], # Detect(P3, P4, P5) + ] diff --git a/models/segment/yolov5m-seg.yaml b/models/segment/yolov5m-seg.yaml new file mode 100644 index 000000000000..f73d1992ac19 --- /dev/null +++ b/models/segment/yolov5m-seg.yaml @@ -0,0 +1,48 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license + +# Parameters +nc: 80 # number of classes +depth_multiple: 0.67 # model depth multiple +width_multiple: 0.75 # layer channel multiple +anchors: + - [10,13, 16,30, 33,23] # P3/8 + - [30,61, 62,45, 59,119] # P4/16 + - [116,90, 156,198, 373,326] # P5/32 + +# YOLOv5 v6.0 backbone +backbone: + # [from, number, module, args] + [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 + [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 + [-1, 3, C3, [128]], + [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 + [-1, 6, C3, [256]], + [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 + [-1, 9, C3, [512]], + [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 + [-1, 3, C3, [1024]], + [-1, 1, SPPF, [1024, 5]], # 9 + ] + +# YOLOv5 v6.0 head +head: + [[-1, 1, Conv, [512, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 6], 1, Concat, [1]], # cat backbone P4 + [-1, 3, C3, [512, False]], # 13 + + [-1, 1, Conv, [256, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 4], 1, Concat, [1]], # cat backbone P3 + [-1, 3, C3, [256, False]], # 17 (P3/8-small) + + [-1, 1, Conv, [256, 3, 2]], + [[-1, 14], 1, Concat, [1]], # cat head P4 + [-1, 3, C3, [512, False]], # 20 (P4/16-medium) + + [-1, 1, Conv, [512, 3, 2]], + [[-1, 10], 1, Concat, [1]], # cat head P5 + [-1, 3, C3, [1024, False]], # 23 (P5/32-large) + + [[17, 20, 23], 1, Segment, [nc, anchors, 32, 256]], # Detect(P3, P4, P5) + ] \ No newline at end of file diff --git a/models/segment/yolov5n-seg.yaml b/models/segment/yolov5n-seg.yaml new file mode 100644 index 000000000000..c28225ab4a50 --- /dev/null +++ b/models/segment/yolov5n-seg.yaml @@ -0,0 +1,48 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license + +# Parameters +nc: 80 # number of classes +depth_multiple: 0.33 # model depth multiple +width_multiple: 0.25 # layer channel multiple +anchors: + - [10,13, 16,30, 33,23] # P3/8 + - [30,61, 62,45, 59,119] # P4/16 + - [116,90, 156,198, 373,326] # P5/32 + +# YOLOv5 v6.0 backbone +backbone: + # [from, number, module, args] + [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 + [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 + [-1, 3, C3, [128]], + [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 + [-1, 6, C3, [256]], + [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 + [-1, 9, C3, [512]], + [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 + [-1, 3, C3, [1024]], + [-1, 1, SPPF, [1024, 5]], # 9 + ] + +# YOLOv5 v6.0 head +head: + [[-1, 1, Conv, [512, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 6], 1, Concat, [1]], # cat backbone P4 + [-1, 3, C3, [512, False]], # 13 + + [-1, 1, Conv, [256, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 4], 1, Concat, [1]], # cat backbone P3 + [-1, 3, C3, [256, False]], # 17 (P3/8-small) + + [-1, 1, Conv, [256, 3, 2]], + [[-1, 14], 1, Concat, [1]], # cat head P4 + [-1, 3, C3, [512, False]], # 20 (P4/16-medium) + + [-1, 1, Conv, [512, 3, 2]], + [[-1, 10], 1, Concat, [1]], # cat head P5 + [-1, 3, C3, [1024, False]], # 23 (P5/32-large) + + [[17, 20, 23], 1, Segment, [nc, anchors, 32, 256]], # Detect(P3, P4, P5) + ] diff --git a/models/segment/yolov5s-seg.yaml b/models/segment/yolov5s-seg.yaml new file mode 100644 index 000000000000..7cbdb36b425c --- /dev/null +++ b/models/segment/yolov5s-seg.yaml @@ -0,0 +1,48 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license + +# Parameters +nc: 80 # number of classes +depth_multiple: 0.33 # model depth multiple +width_multiple: 0.5 # layer channel multiple +anchors: + - [10,13, 16,30, 33,23] # P3/8 + - [30,61, 62,45, 59,119] # P4/16 + - [116,90, 156,198, 373,326] # P5/32 + +# YOLOv5 v6.0 backbone +backbone: + # [from, number, module, args] + [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 + [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 + [-1, 3, C3, [128]], + [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 + [-1, 6, C3, [256]], + [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 + [-1, 9, C3, [512]], + [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 + [-1, 3, C3, [1024]], + [-1, 1, SPPF, [1024, 5]], # 9 + ] + +# YOLOv5 v6.0 head +head: + [[-1, 1, Conv, [512, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 6], 1, Concat, [1]], # cat backbone P4 + [-1, 3, C3, [512, False]], # 13 + + [-1, 1, Conv, [256, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 4], 1, Concat, [1]], # cat backbone P3 + [-1, 3, C3, [256, False]], # 17 (P3/8-small) + + [-1, 1, Conv, [256, 3, 2]], + [[-1, 14], 1, Concat, [1]], # cat head P4 + [-1, 3, C3, [512, False]], # 20 (P4/16-medium) + + [-1, 1, Conv, [512, 3, 2]], + [[-1, 10], 1, Concat, [1]], # cat head P5 + [-1, 3, C3, [1024, False]], # 23 (P5/32-large) + + [[17, 20, 23], 1, Segment, [nc, anchors, 32, 256]], # Detect(P3, P4, P5) + ] \ No newline at end of file diff --git a/models/segment/yolov5x-seg.yaml b/models/segment/yolov5x-seg.yaml new file mode 100644 index 000000000000..5d0c4524a99c --- /dev/null +++ b/models/segment/yolov5x-seg.yaml @@ -0,0 +1,48 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license + +# Parameters +nc: 80 # number of classes +depth_multiple: 1.33 # model depth multiple +width_multiple: 1.25 # layer channel multiple +anchors: + - [10,13, 16,30, 33,23] # P3/8 + - [30,61, 62,45, 59,119] # P4/16 + - [116,90, 156,198, 373,326] # P5/32 + +# YOLOv5 v6.0 backbone +backbone: + # [from, number, module, args] + [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 + [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 + [-1, 3, C3, [128]], + [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 + [-1, 6, C3, [256]], + [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 + [-1, 9, C3, [512]], + [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 + [-1, 3, C3, [1024]], + [-1, 1, SPPF, [1024, 5]], # 9 + ] + +# YOLOv5 v6.0 head +head: + [[-1, 1, Conv, [512, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 6], 1, Concat, [1]], # cat backbone P4 + [-1, 3, C3, [512, False]], # 13 + + [-1, 1, Conv, [256, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 4], 1, Concat, [1]], # cat backbone P3 + [-1, 3, C3, [256, False]], # 17 (P3/8-small) + + [-1, 1, Conv, [256, 3, 2]], + [[-1, 14], 1, Concat, [1]], # cat head P4 + [-1, 3, C3, [512, False]], # 20 (P4/16-medium) + + [-1, 1, Conv, [512, 3, 2]], + [[-1, 10], 1, Concat, [1]], # cat head P5 + [-1, 3, C3, [1024, False]], # 23 (P5/32-large) + + [[17, 20, 23], 1, Segment, [nc, anchors, 32, 256]], # Detect(P3, P4, P5) + ] diff --git a/models/tf.py b/models/tf.py index ecb0d4d79c78..8cce147059d3 100644 --- a/models/tf.py +++ b/models/tf.py @@ -30,7 +30,7 @@ from models.common import (C3, SPP, SPPF, Bottleneck, BottleneckCSP, C3x, Concat, Conv, CrossConv, DWConv, DWConvTranspose2d, Focus, autopad) from models.experimental import MixConv2d, attempt_load -from models.yolo import Detect +from models.yolo import Detect, Segment from utils.activations import SiLU from utils.general import LOGGER, make_divisible, print_args @@ -320,6 +320,36 @@ def _make_grid(nx=20, ny=20): return tf.cast(tf.reshape(tf.stack([xv, yv], 2), [1, 1, ny * nx, 2]), dtype=tf.float32) +class TFSegment(TFDetect): + # YOLOv5 Segment head for segmentation models + def __init__(self, nc=80, anchors=(), nm=32, npr=256, ch=(), imgsz=(640, 640), w=None): + super().__init__(nc, anchors, ch, imgsz, w) + self.nm = nm # number of masks + self.npr = npr # number of protos + self.no = 5 + nc + self.nm # number of outputs per anchor + self.m = [TFConv2d(x, self.no * self.na, 1, w=w.m[i]) for i, x in enumerate(ch)] # output conv + self.proto = TFProto(ch[0], self.npr, self.nm, w=w.proto) # protos + self.detect = TFDetect.call + + def call(self, x): + p = self.proto(x[0]) + x = self.detect(self, x) + return (x, p) if self.training else ((x[0], p),) + + +class TFProto(keras.layers.Layer): + + def __init__(self, c1, c_=256, c2=32, w=None): + super().__init__() + self.cv1 = TFConv(c1, c_, k=3, w=w.cv1) + self.upsample = TFUpsample(None, scale_factor=2, mode='nearest') + self.cv2 = TFConv(c_, c_, k=3, w=w.cv2) + self.cv3 = TFConv(c_, c2, w=w.cv3) + + def call(self, inputs): + return self.cv3(self.cv2(self.upsample(self.cv1(inputs)))) + + class TFUpsample(keras.layers.Layer): # TF version of torch.nn.Upsample() def __init__(self, size, scale_factor, mode, w=None): # warning: all arguments needed including 'w' @@ -377,10 +407,12 @@ def parse_model(d, ch, model, imgsz): # model_dict, input_channels(3) args = [ch[f]] elif m is Concat: c2 = sum(ch[-1 if x == -1 else x + 1] for x in f) - elif m is Detect: + elif m in [Detect, Segment]: args.append([ch[x + 1] for x in f]) if isinstance(args[1], int): # number of anchors args[1] = [list(range(args[1] * 2))] * len(f) + if m is Segment: + args[3] = make_divisible(args[3] * gw, 8) args.append(imgsz) else: c2 = ch[f] diff --git a/models/yolo.py b/models/yolo.py index fa05fcf9a8d9..a0702a7c0257 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -36,6 +36,7 @@ class Detect(nn.Module): + # YOLOv5 Detect head for detection models stride = None # strides computed during build dynamic = False # force grid reconstruction export = False # export mode @@ -63,15 +64,16 @@ def forward(self, x): if self.dynamic or self.grid[i].shape[2:4] != x[i].shape[2:4]: self.grid[i], self.anchor_grid[i] = self._make_grid(nx, ny, i) - y = x[i].sigmoid() + y = x[i].clone() + y[..., :5 + self.nc].sigmoid_() if self.inplace: y[..., 0:2] = (y[..., 0:2] * 2 + self.grid[i]) * self.stride[i] # xy y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953 - xy, wh, conf = y.split((2, 2, self.nc + 1), 4) # y.tensor_split((2, 4, 5), 4) # torch 1.8.0 + xy, wh, etc = y.split((2, 2, self.no - 4), 4) # tensor_split((2, 4, 5), 4) if torch 1.8.0 xy = (xy * 2 + self.grid[i]) * self.stride[i] # xy wh = (wh * 2) ** 2 * self.anchor_grid[i] # wh - y = torch.cat((xy, wh, conf), 4) + y = torch.cat((xy, wh, etc), 4) z.append(y.view(bs, -1, self.no)) return x if self.training else (torch.cat(z, 1),) if self.export else (torch.cat(z, 1), x) @@ -87,6 +89,23 @@ def _make_grid(self, nx=20, ny=20, i=0, torch_1_10=check_version(torch.__version return grid, anchor_grid +class Segment(Detect): + # YOLOv5 Segment head for segmentation models + def __init__(self, nc=80, anchors=(), nm=32, npr=256, ch=(), inplace=True): + super().__init__(nc, anchors, ch, inplace) + self.nm = nm # number of masks + self.npr = npr # number of protos + self.no = 5 + nc + self.nm # number of outputs per anchor + self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv + self.proto = Proto(ch[0], self.npr, self.nm) # protos + self.detect = Detect.forward + + def forward(self, x): + p = self.proto(x[0]) + x = self.detect(self, x) + return (x, p) if self.training else (x[0], p) if self.export else (x[0], p, x[1]) + + class BaseModel(nn.Module): # YOLOv5 base model def forward(self, x, profile=False, visualize=False): @@ -135,7 +154,7 @@ def _apply(self, fn): # Apply to(), cpu(), cuda(), half() to model tensors that are not parameters or registered buffers self = super()._apply(fn) m = self.model[-1] # Detect() - if isinstance(m, Detect): + if isinstance(m, (Detect, Segment)): m.stride = fn(m.stride) m.grid = list(map(fn, m.grid)) if isinstance(m.anchor_grid, list): @@ -169,11 +188,12 @@ def __init__(self, cfg='yolov5s.yaml', ch=3, nc=None, anchors=None): # model, i # Build strides, anchors m = self.model[-1] # Detect() - if isinstance(m, Detect): + if isinstance(m, (Detect, Segment)): s = 256 # 2x min stride m.inplace = self.inplace - m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.empty(1, ch, s, s))]) # forward - check_anchor_order(m) # must be in pixel-space (not grid-space) + forward = lambda x: self.forward(x)[0] if isinstance(m, Segment) else self.forward(x) + m.stride = torch.tensor([s / x.shape[-2] for x in forward(torch.zeros(1, ch, s, s))]) # forward + check_anchor_order(m) m.anchors /= m.stride.view(-1, 1, 1) self.stride = m.stride self._initialize_biases() # only run once @@ -235,15 +255,21 @@ def _initialize_biases(self, cf=None): # initialize biases into Detect(), cf is # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1. m = self.model[-1] # Detect() module for mi, s in zip(m.m, m.stride): # from - b = mi.bias.view(m.na, -1).detach() # conv.bias(255) to (3,85) - b[:, 4] += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image) - b[:, 5:] += math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum()) # cls + b = mi.bias.view(m.na, -1) # conv.bias(255) to (3,85) + b.data[:, 4] += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image) + b.data[:, 5:5 + m.nc] += math.log(0.6 / (m.nc - 0.99999)) if cf is None else torch.log(cf / cf.sum()) # cls mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True) Model = DetectionModel # retain YOLOv5 'Model' class for backwards compatibility +class SegmentationModel(DetectionModel): + # YOLOv5 segmentation model + def __init__(self, cfg='yolov5s-seg.yaml', ch=3, nc=None, anchors=None): + super().__init__(cfg, ch, nc, anchors) + + class ClassificationModel(BaseModel): # YOLOv5 classification model def __init__(self, cfg=None, model=None, nc=1000, cutoff=10): # yaml, model, number of classes, cutoff index @@ -284,24 +310,28 @@ def parse_model(d, ch): # model_dict, input_channels(3) args[j] = eval(a) if isinstance(a, str) else a # eval strings n = n_ = max(round(n * gd), 1) if n > 1 else n # depth gain - if m in (Conv, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv, - BottleneckCSP, C3, C3TR, C3SPP, C3Ghost, nn.ConvTranspose2d, DWConvTranspose2d, C3x): + if m in { + Conv, GhostConv, Bottleneck, GhostBottleneck, SPP, SPPF, DWConv, MixConv2d, Focus, CrossConv, + BottleneckCSP, C3, C3TR, C3SPP, C3Ghost, nn.ConvTranspose2d, DWConvTranspose2d, C3x}: c1, c2 = ch[f], args[0] if c2 != no: # if not output c2 = make_divisible(c2 * gw, 8) args = [c1, c2, *args[1:]] - if m in [BottleneckCSP, C3, C3TR, C3Ghost, C3x]: + if m in {BottleneckCSP, C3, C3TR, C3Ghost, C3x}: args.insert(2, n) # number of repeats n = 1 elif m is nn.BatchNorm2d: args = [ch[f]] elif m is Concat: c2 = sum(ch[x] for x in f) - elif m is Detect: + # TODO: channel, gw, gd + elif m in {Detect, Segment}: args.append([ch[x] for x in f]) if isinstance(args[1], int): # number of anchors args[1] = [list(range(args[1] * 2))] * len(f) + if m is Segment: + args[3] = make_divisible(args[3] * gw, 8) elif m is Contract: c2 = ch[f] * args[0] ** 2 elif m is Expand: diff --git a/segment/predict.py b/segment/predict.py new file mode 100644 index 000000000000..ba4cf2905255 --- /dev/null +++ b/segment/predict.py @@ -0,0 +1,266 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Run YOLOv5 segmentation inference on images, videos, directories, streams, etc. + +Usage - sources: + $ python segment/predict.py --weights yolov5s-seg.pt --source 0 # webcam + img.jpg # image + vid.mp4 # video + path/ # directory + 'path/*.jpg' # glob + 'https://youtu.be/Zgi9g1ksQHc' # YouTube + 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP stream + +Usage - formats: + $ python segment/predict.py --weights yolov5s-seg.pt # PyTorch + yolov5s-seg.torchscript # TorchScript + yolov5s-seg.onnx # ONNX Runtime or OpenCV DNN with --dnn + yolov5s-seg.xml # OpenVINO + yolov5s-seg.engine # TensorRT + yolov5s-seg.mlmodel # CoreML (macOS-only) + yolov5s-seg_saved_model # TensorFlow SavedModel + yolov5s-seg.pb # TensorFlow GraphDef + yolov5s-seg.tflite # TensorFlow Lite + yolov5s-seg_edgetpu.tflite # TensorFlow Edge TPU + yolov5s-seg_paddle_model # PaddlePaddle +""" + +import argparse +import os +import platform +import sys +from pathlib import Path + +import torch + +FILE = Path(__file__).resolve() +ROOT = FILE.parents[1] # YOLOv5 root directory +if str(ROOT) not in sys.path: + sys.path.append(str(ROOT)) # add ROOT to PATH +ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative + +from models.common import DetectMultiBackend +from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadStreams +from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, + increment_path, non_max_suppression, print_args, scale_coords, strip_optimizer, xyxy2xywh) +from utils.plots import Annotator, colors, save_one_box +from utils.segment.general import process_mask +from utils.torch_utils import select_device, smart_inference_mode + + +@smart_inference_mode() +def run( + weights=ROOT / 'yolov5s-seg.pt', # model.pt path(s) + source=ROOT / 'data/images', # file/dir/URL/glob, 0 for webcam + data=ROOT / 'data/coco128.yaml', # dataset.yaml path + imgsz=(640, 640), # inference size (height, width) + conf_thres=0.25, # confidence threshold + iou_thres=0.45, # NMS IOU threshold + max_det=1000, # maximum detections per image + device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu + view_img=False, # show results + save_txt=False, # save results to *.txt + save_conf=False, # save confidences in --save-txt labels + save_crop=False, # save cropped prediction boxes + nosave=False, # do not save images/videos + classes=None, # filter by class: --class 0, or --class 0 2 3 + agnostic_nms=False, # class-agnostic NMS + augment=False, # augmented inference + visualize=False, # visualize features + update=False, # update all models + project=ROOT / 'runs/predict-seg', # save results to project/name + name='exp', # save results to project/name + exist_ok=False, # existing project/name ok, do not increment + line_thickness=3, # bounding box thickness (pixels) + hide_labels=False, # hide labels + hide_conf=False, # hide confidences + half=False, # use FP16 half-precision inference + dnn=False, # use OpenCV DNN for ONNX inference + vid_stride=1, # video frame-rate stride + retina_masks=False, +): + source = str(source) + save_img = not nosave and not source.endswith('.txt') # save inference images + is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS) + is_url = source.lower().startswith(('rtsp://', 'rtmp://', 'http://', 'https://')) + webcam = source.isnumeric() or source.endswith('.txt') or (is_url and not is_file) + if is_url and is_file: + source = check_file(source) # download + + # Directories + save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run + (save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir + + # Load model + device = select_device(device) + model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data, fp16=half) + stride, names, pt = model.stride, model.names, model.pt + imgsz = check_img_size(imgsz, s=stride) # check image size + + # Dataloader + if webcam: + view_img = check_imshow() + dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) + bs = len(dataset) # batch_size + else: + dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) + bs = 1 # batch_size + vid_path, vid_writer = [None] * bs, [None] * bs + + # Run inference + model.warmup(imgsz=(1 if pt else bs, 3, *imgsz)) # warmup + seen, windows, dt = 0, [], (Profile(), Profile(), Profile()) + for path, im, im0s, vid_cap, s in dataset: + with dt[0]: + im = torch.from_numpy(im).to(device) + im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 + im /= 255 # 0 - 255 to 0.0 - 1.0 + if len(im.shape) == 3: + im = im[None] # expand for batch dim + + # Inference + with dt[1]: + visualize = increment_path(save_dir / Path(path).stem, mkdir=True) if visualize else False + pred, proto = model(im, augment=augment, visualize=visualize)[:2] + + # NMS + with dt[2]: + pred = non_max_suppression(pred, conf_thres, iou_thres, classes, agnostic_nms, max_det=max_det, nm=32) + + # Second-stage classifier (optional) + # pred = utils.general.apply_classifier(pred, classifier_model, im, im0s) + + # Process predictions + for i, det in enumerate(pred): # per image + seen += 1 + if webcam: # batch_size >= 1 + p, im0, frame = path[i], im0s[i].copy(), dataset.count + s += f'{i}: ' + else: + p, im0, frame = path, im0s.copy(), getattr(dataset, 'frame', 0) + + p = Path(p) # to Path + save_path = str(save_dir / p.name) # im.jpg + txt_path = str(save_dir / 'labels' / p.stem) + ('' if dataset.mode == 'image' else f'_{frame}') # im.txt + s += '%gx%g ' % im.shape[2:] # print string + gn = torch.tensor(im0.shape)[[1, 0, 1, 0]] # normalization gain whwh + imc = im0.copy() if save_crop else im0 # for save_crop + annotator = Annotator(im0, line_width=line_thickness, example=str(names)) + if len(det): + masks = process_mask(proto[i], det[:, 6:], det[:, :4], im.shape[2:], upsample=True) # HWC + + # Rescale boxes from img_size to im0 size + det[:, :4] = scale_coords(im.shape[2:], det[:, :4], im0.shape).round() + + # Print results + for c in det[:, 5].unique(): + n = (det[:, 5] == c).sum() # detections per class + s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string + + # Mask plotting + annotator.masks(masks, + colors=[colors(x, True) for x in det[:, 5]], + im_gpu=None if retina_masks else im[i]) + + # Write results + for *xyxy, conf, cls in reversed(det[:, :6]): + if save_txt: # Write to file + xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh + line = (cls, *xywh, conf) if save_conf else (cls, *xywh) # label format + with open(f'{txt_path}.txt', 'a') as f: + f.write(('%g ' * len(line)).rstrip() % line + '\n') + + if save_img or save_crop or view_img: # Add bbox to image + c = int(cls) # integer class + label = None if hide_labels else (names[c] if hide_conf else f'{names[c]} {conf:.2f}') + annotator.box_label(xyxy, label, color=colors(c, True)) + if save_crop: + save_one_box(xyxy, imc, file=save_dir / 'crops' / names[c] / f'{p.stem}.jpg', BGR=True) + + # Stream results + im0 = annotator.result() + if view_img: + if platform.system() == 'Linux' and p not in windows: + windows.append(p) + cv2.namedWindow(str(p), cv2.WINDOW_NORMAL | cv2.WINDOW_KEEPRATIO) # allow window resize (Linux) + cv2.resizeWindow(str(p), im0.shape[1], im0.shape[0]) + cv2.imshow(str(p), im0) + if cv2.waitKey(1) == ord('q'): # 1 millisecond + exit() + + # Save results (image with detections) + if save_img: + if dataset.mode == 'image': + cv2.imwrite(save_path, im0) + else: # 'video' or 'stream' + if vid_path[i] != save_path: # new video + vid_path[i] = save_path + if isinstance(vid_writer[i], cv2.VideoWriter): + vid_writer[i].release() # release previous video writer + if vid_cap: # video + fps = vid_cap.get(cv2.CAP_PROP_FPS) + w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH)) + h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT)) + else: # stream + fps, w, h = 30, im0.shape[1], im0.shape[0] + save_path = str(Path(save_path).with_suffix('.mp4')) # force *.mp4 suffix on results videos + vid_writer[i] = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h)) + vid_writer[i].write(im0) + + # Print time (inference-only) + LOGGER.info(f"{s}{'' if len(det) else '(no detections), '}{dt[1].dt * 1E3:.1f}ms") + + # Print results + t = tuple(x.t / seen * 1E3 for x in dt) # speeds per image + LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {(1, 3, *imgsz)}' % t) + if save_txt or save_img: + s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else '' + LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}{s}") + if update: + strip_optimizer(weights[0]) # update model (to fix SourceChangeWarning) + + +def parse_opt(): + parser = argparse.ArgumentParser() + parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s-seg.pt', help='model path(s)') + parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob, 0 for webcam') + parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='(optional) dataset.yaml path') + parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640], help='inference size h,w') + parser.add_argument('--conf-thres', type=float, default=0.25, help='confidence threshold') + parser.add_argument('--iou-thres', type=float, default=0.45, help='NMS IoU threshold') + parser.add_argument('--max-det', type=int, default=1000, help='maximum detections per image') + parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--view-img', action='store_true', help='show results') + parser.add_argument('--save-txt', action='store_true', help='save results to *.txt') + parser.add_argument('--save-conf', action='store_true', help='save confidences in --save-txt labels') + parser.add_argument('--save-crop', action='store_true', help='save cropped prediction boxes') + parser.add_argument('--nosave', action='store_true', help='do not save images/videos') + parser.add_argument('--classes', nargs='+', type=int, help='filter by class: --classes 0, or --classes 0 2 3') + parser.add_argument('--agnostic-nms', action='store_true', help='class-agnostic NMS') + parser.add_argument('--augment', action='store_true', help='augmented inference') + parser.add_argument('--visualize', action='store_true', help='visualize features') + parser.add_argument('--update', action='store_true', help='update all models') + parser.add_argument('--project', default=ROOT / 'runs/predict-seg', help='save results to project/name') + parser.add_argument('--name', default='exp', help='save results to project/name') + parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') + parser.add_argument('--line-thickness', default=3, type=int, help='bounding box thickness (pixels)') + parser.add_argument('--hide-labels', default=False, action='store_true', help='hide labels') + parser.add_argument('--hide-conf', default=False, action='store_true', help='hide confidences') + parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') + parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') + parser.add_argument('--vid-stride', type=int, default=1, help='video frame-rate stride') + parser.add_argument('--retina-masks', action='store_true', help='whether to plot masks in native resolution') + opt = parser.parse_args() + opt.imgsz *= 2 if len(opt.imgsz) == 1 else 1 # expand + print_args(vars(opt)) + return opt + + +def main(opt): + check_requirements(exclude=('tensorboard', 'thop')) + run(**vars(opt)) + + +if __name__ == "__main__": + opt = parse_opt() + main(opt) diff --git a/segment/train.py b/segment/train.py new file mode 100644 index 000000000000..bda379176151 --- /dev/null +++ b/segment/train.py @@ -0,0 +1,676 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Train a YOLOv5 segment model on a segment dataset +Models and datasets download automatically from the latest YOLOv5 release. + +Usage - Single-GPU training: + $ python segment/train.py --data coco128-seg.yaml --weights yolov5s-seg.pt --img 640 # from pretrained (recommended) + $ python segment/train.py --data coco128-seg.yaml --weights '' --cfg yolov5s-seg.yaml --img 640 # from scratch + +Usage - Multi-GPU DDP training: + $ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 segment/train.py --data coco128-seg.yaml --weights yolov5s-seg.pt --img 640 --device 0,1,2,3 + +Models: https://github.com/ultralytics/yolov5/tree/master/models +Datasets: https://github.com/ultralytics/yolov5/tree/master/data +Tutorial: https://github.com/ultralytics/yolov5/wiki/Train-Custom-Data +""" + +import argparse +import math +import os +import random +import sys +import time +from copy import deepcopy +from datetime import datetime +from pathlib import Path + +import numpy as np +import torch +import torch.distributed as dist +import torch.nn as nn +import yaml +from torch.optim import lr_scheduler +from tqdm import tqdm + +FILE = Path(__file__).resolve() +ROOT = FILE.parents[1] # YOLOv5 root directory +if str(ROOT) not in sys.path: + sys.path.append(str(ROOT)) # add ROOT to PATH +ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative + +import torch.nn.functional as F + +import segment.val as validate # for end-of-epoch mAP +from models.experimental import attempt_load +from models.yolo import SegmentationModel +from utils.autoanchor import check_anchors +from utils.autobatch import check_train_batch_size +from utils.callbacks import Callbacks +from utils.downloads import attempt_download, is_url +from utils.general import (LOGGER, check_amp, check_dataset, check_file, check_git_status, check_img_size, + check_requirements, check_suffix, check_yaml, colorstr, get_latest_run, increment_path, + init_seeds, intersect_dicts, labels_to_class_weights, labels_to_image_weights, one_cycle, + print_args, print_mutation, strip_optimizer, yaml_save) +from utils.loggers import GenericLogger +from utils.plots import plot_evolve, plot_labels +from utils.segment.dataloaders import create_dataloader +from utils.segment.loss import ComputeLoss +from utils.segment.metrics import KEYS, fitness +from utils.segment.plots import plot_images_and_masks, plot_results_with_masks +from utils.torch_utils import (EarlyStopping, ModelEMA, de_parallel, select_device, smart_DDP, smart_optimizer, + smart_resume, torch_distributed_zero_first) + +LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html +RANK = int(os.getenv('RANK', -1)) +WORLD_SIZE = int(os.getenv('WORLD_SIZE', 1)) + + +def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictionary + save_dir, epochs, batch_size, weights, single_cls, evolve, data, cfg, resume, noval, nosave, workers, freeze, mask_ratio = \ + Path(opt.save_dir), opt.epochs, opt.batch_size, opt.weights, opt.single_cls, opt.evolve, opt.data, opt.cfg, \ + opt.resume, opt.noval, opt.nosave, opt.workers, opt.freeze, opt.mask_ratio + # callbacks.run('on_pretrain_routine_start') + + # Directories + w = save_dir / 'weights' # weights dir + (w.parent if evolve else w).mkdir(parents=True, exist_ok=True) # make dir + last, best = w / 'last.pt', w / 'best.pt' + + # Hyperparameters + if isinstance(hyp, str): + with open(hyp, errors='ignore') as f: + hyp = yaml.safe_load(f) # load hyps dict + LOGGER.info(colorstr('hyperparameters: ') + ', '.join(f'{k}={v}' for k, v in hyp.items())) + opt.hyp = hyp.copy() # for saving hyps to checkpoints + + # Save run settings + if not evolve: + yaml_save(save_dir / 'hyp.yaml', hyp) + yaml_save(save_dir / 'opt.yaml', vars(opt)) + + # Loggers + data_dict = None + if RANK in {-1, 0}: + logger = GenericLogger(opt=opt, console_logger=LOGGER) + # loggers = Loggers(save_dir, weights, opt, hyp, LOGGER) # loggers instance + # if loggers.clearml: + # data_dict = loggers.clearml.data_dict # None if no ClearML dataset or filled in by ClearML + # if loggers.wandb: + # data_dict = loggers.wandb.data_dict + # if resume: + # weights, epochs, hyp, batch_size = opt.weights, opt.epochs, opt.hyp, opt.batch_size + # + # # Register actions + # for k in methods(loggers): + # callbacks.register_action(k, callback=getattr(loggers, k)) + + # Config + plots = not evolve and not opt.noplots # create plots + overlap = not opt.no_overlap + cuda = device.type != 'cpu' + init_seeds(opt.seed + 1 + RANK, deterministic=True) + with torch_distributed_zero_first(LOCAL_RANK): + data_dict = data_dict or check_dataset(data) # check if None + train_path, val_path = data_dict['train'], data_dict['val'] + nc = 1 if single_cls else int(data_dict['nc']) # number of classes + names = {0: 'item'} if single_cls and len(data_dict['names']) != 1 else data_dict['names'] # class names + is_coco = isinstance(val_path, str) and val_path.endswith('coco/val2017.txt') # COCO dataset + + # Model + check_suffix(weights, '.pt') # check weights + pretrained = weights.endswith('.pt') + if pretrained: + with torch_distributed_zero_first(LOCAL_RANK): + weights = attempt_download(weights) # download if not found locally + ckpt = torch.load(weights, map_location='cpu') # load checkpoint to CPU to avoid CUDA memory leak + model = SegmentationModel(cfg or ckpt['model'].yaml, ch=3, nc=nc, anchors=hyp.get('anchors')).to(device) + exclude = ['anchor'] if (cfg or hyp.get('anchors')) and not resume else [] # exclude keys + csd = ckpt['model'].float().state_dict() # checkpoint state_dict as FP32 + csd = intersect_dicts(csd, model.state_dict(), exclude=exclude) # intersect + model.load_state_dict(csd, strict=False) # load + LOGGER.info(f'Transferred {len(csd)}/{len(model.state_dict())} items from {weights}') # report + else: + model = SegmentationModel(cfg, ch=3, nc=nc, anchors=hyp.get('anchors')).to(device) # create + amp = check_amp(model) # check AMP + + # Freeze + freeze = [f'model.{x}.' for x in (freeze if len(freeze) > 1 else range(freeze[0]))] # layers to freeze + for k, v in model.named_parameters(): + v.requires_grad = True # train all layers + # v.register_hook(lambda x: torch.nan_to_num(x)) # NaN to 0 (commented for erratic training results) + if any(x in k for x in freeze): + LOGGER.info(f'freezing {k}') + v.requires_grad = False + + # Image size + gs = max(int(model.stride.max()), 32) # grid size (max stride) + imgsz = check_img_size(opt.imgsz, gs, floor=gs * 2) # verify imgsz is gs-multiple + + # Batch size + if RANK == -1 and batch_size == -1: # single-GPU only, estimate best batch size + batch_size = check_train_batch_size(model, imgsz, amp) + logger.update_params({"batch_size": batch_size}) + # loggers.on_params_update({"batch_size": batch_size}) + + # Optimizer + nbs = 64 # nominal batch size + accumulate = max(round(nbs / batch_size), 1) # accumulate loss before optimizing + hyp['weight_decay'] *= batch_size * accumulate / nbs # scale weight_decay + optimizer = smart_optimizer(model, opt.optimizer, hyp['lr0'], hyp['momentum'], hyp['weight_decay']) + + # Scheduler + if opt.cos_lr: + lf = one_cycle(1, hyp['lrf'], epochs) # cosine 1->hyp['lrf'] + else: + lf = lambda x: (1 - x / epochs) * (1.0 - hyp['lrf']) + hyp['lrf'] # linear + scheduler = lr_scheduler.LambdaLR(optimizer, lr_lambda=lf) # plot_lr_scheduler(optimizer, scheduler, epochs) + + # EMA + ema = ModelEMA(model) if RANK in {-1, 0} else None + + # Resume + best_fitness, start_epoch = 0.0, 0 + if pretrained: + if resume: + best_fitness, start_epoch, epochs = smart_resume(ckpt, optimizer, ema, weights, epochs, resume) + del ckpt, csd + + # DP mode + if cuda and RANK == -1 and torch.cuda.device_count() > 1: + LOGGER.warning('WARNING: DP not recommended, use torch.distributed.run for best DDP Multi-GPU results.\n' + 'See Multi-GPU Tutorial at https://github.com/ultralytics/yolov5/issues/475 to get started.') + model = torch.nn.DataParallel(model) + + # SyncBatchNorm + if opt.sync_bn and cuda and RANK != -1: + model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(model).to(device) + LOGGER.info('Using SyncBatchNorm()') + + # Trainloader + train_loader, dataset = create_dataloader( + train_path, + imgsz, + batch_size // WORLD_SIZE, + gs, + single_cls, + hyp=hyp, + augment=True, + cache=None if opt.cache == 'val' else opt.cache, + rect=opt.rect, + rank=LOCAL_RANK, + workers=workers, + image_weights=opt.image_weights, + quad=opt.quad, + prefix=colorstr('train: '), + shuffle=True, + mask_downsample_ratio=mask_ratio, + overlap_mask=overlap, + ) + labels = np.concatenate(dataset.labels, 0) + mlc = int(labels[:, 0].max()) # max label class + assert mlc < nc, f'Label class {mlc} exceeds nc={nc} in {data}. Possible class labels are 0-{nc - 1}' + + # Process 0 + if RANK in {-1, 0}: + val_loader = create_dataloader(val_path, + imgsz, + batch_size // WORLD_SIZE * 2, + gs, + single_cls, + hyp=hyp, + cache=None if noval else opt.cache, + rect=True, + rank=-1, + workers=workers * 2, + pad=0.5, + mask_downsample_ratio=mask_ratio, + overlap_mask=overlap, + prefix=colorstr('val: '))[0] + + if not resume: + if not opt.noautoanchor: + check_anchors(dataset, model=model, thr=hyp['anchor_t'], imgsz=imgsz) # run AutoAnchor + model.half().float() # pre-reduce anchor precision + + if plots: + plot_labels(labels, names, save_dir) + # callbacks.run('on_pretrain_routine_end', labels, names) + + # DDP mode + if cuda and RANK != -1: + model = smart_DDP(model) + + # Model attributes + nl = de_parallel(model).model[-1].nl # number of detection layers (to scale hyps) + hyp['box'] *= 3 / nl # scale to layers + hyp['cls'] *= nc / 80 * 3 / nl # scale to classes and layers + hyp['obj'] *= (imgsz / 640) ** 2 * 3 / nl # scale to image size and layers + hyp['label_smoothing'] = opt.label_smoothing + model.nc = nc # attach number of classes to model + model.hyp = hyp # attach hyperparameters to model + model.class_weights = labels_to_class_weights(dataset.labels, nc).to(device) * nc # attach class weights + model.names = names + + # Start training + t0 = time.time() + nb = len(train_loader) # number of batches + nw = max(round(hyp['warmup_epochs'] * nb), 100) # number of warmup iterations, max(3 epochs, 100 iterations) + # nw = min(nw, (epochs - start_epoch) / 2 * nb) # limit warmup to < 1/2 of training + last_opt_step = -1 + maps = np.zeros(nc) # mAP per class + results = (0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) # P, R, mAP@.5, mAP@.5-.95, val_loss(box, obj, cls) + scheduler.last_epoch = start_epoch - 1 # do not move + scaler = torch.cuda.amp.GradScaler(enabled=amp) + stopper, stop = EarlyStopping(patience=opt.patience), False + compute_loss = ComputeLoss(model, overlap=overlap) # init loss class + # callbacks.run('on_train_start') + LOGGER.info(f'Image sizes {imgsz} train, {imgsz} val\n' + f'Using {train_loader.num_workers * WORLD_SIZE} dataloader workers\n' + f"Logging results to {colorstr('bold', save_dir)}\n" + f'Starting training for {epochs} epochs...') + for epoch in range(start_epoch, epochs): # epoch ------------------------------------------------------------------ + # callbacks.run('on_train_epoch_start') + model.train() + + # Update image weights (optional, single-GPU only) + if opt.image_weights: + cw = model.class_weights.cpu().numpy() * (1 - maps) ** 2 / nc # class weights + iw = labels_to_image_weights(dataset.labels, nc=nc, class_weights=cw) # image weights + dataset.indices = random.choices(range(dataset.n), weights=iw, k=dataset.n) # rand weighted idx + + # Update mosaic border (optional) + # b = int(random.uniform(0.25 * imgsz, 0.75 * imgsz + gs) // gs * gs) + # dataset.mosaic_border = [b - imgsz, -b] # height, width borders + + mloss = torch.zeros(4, device=device) # mean losses + if RANK != -1: + train_loader.sampler.set_epoch(epoch) + pbar = enumerate(train_loader) + LOGGER.info(('\n' + '%11s' * 8) % + ('Epoch', 'GPU_mem', 'box_loss', 'seg_loss', 'obj_loss', 'cls_loss', 'Instances', 'Size')) + if RANK in {-1, 0}: + pbar = tqdm(pbar, total=nb, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar + optimizer.zero_grad() + for i, (imgs, targets, paths, _, masks) in pbar: # batch ------------------------------------------------------ + # callbacks.run('on_train_batch_start') + ni = i + nb * epoch # number integrated batches (since train start) + imgs = imgs.to(device, non_blocking=True).float() / 255 # uint8 to float32, 0-255 to 0.0-1.0 + + # Warmup + if ni <= nw: + xi = [0, nw] # x interp + # compute_loss.gr = np.interp(ni, xi, [0.0, 1.0]) # iou loss ratio (obj_loss = 1.0 or iou) + accumulate = max(1, np.interp(ni, xi, [1, nbs / batch_size]).round()) + for j, x in enumerate(optimizer.param_groups): + # bias lr falls from 0.1 to lr0, all other lrs rise from 0.0 to lr0 + x['lr'] = np.interp(ni, xi, [hyp['warmup_bias_lr'] if j == 0 else 0.0, x['initial_lr'] * lf(epoch)]) + if 'momentum' in x: + x['momentum'] = np.interp(ni, xi, [hyp['warmup_momentum'], hyp['momentum']]) + + # Multi-scale + if opt.multi_scale: + sz = random.randrange(imgsz * 0.5, imgsz * 1.5 + gs) // gs * gs # size + sf = sz / max(imgs.shape[2:]) # scale factor + if sf != 1: + ns = [math.ceil(x * sf / gs) * gs for x in imgs.shape[2:]] # new shape (stretched to gs-multiple) + imgs = nn.functional.interpolate(imgs, size=ns, mode='bilinear', align_corners=False) + + # Forward + with torch.cuda.amp.autocast(amp): + pred = model(imgs) # forward + loss, loss_items = compute_loss(pred, targets.to(device), masks=masks.to(device).float()) + if RANK != -1: + loss *= WORLD_SIZE # gradient averaged between devices in DDP mode + if opt.quad: + loss *= 4. + + # Backward + scaler.scale(loss).backward() + + # Optimize - https://pytorch.org/docs/master/notes/amp_examples.html + if ni - last_opt_step >= accumulate: + scaler.unscale_(optimizer) # unscale gradients + torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm=10.0) # clip gradients + scaler.step(optimizer) # optimizer.step + scaler.update() + optimizer.zero_grad() + if ema: + ema.update(model) + last_opt_step = ni + + # Log + if RANK in {-1, 0}: + mloss = (mloss * i + loss_items) / (i + 1) # update mean losses + mem = f'{torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0:.3g}G' # (GB) + pbar.set_description(('%11s' * 2 + '%11.4g' * 6) % + (f'{epoch}/{epochs - 1}', mem, *mloss, targets.shape[0], imgs.shape[-1])) + # callbacks.run('on_train_batch_end', model, ni, imgs, targets, paths) + # if callbacks.stop_training: + # return + + # Mosaic plots + if plots: + if ni < 3: + plot_images_and_masks(imgs, targets, masks, paths, save_dir / f"train_batch{ni}.jpg") + if ni == 10: + files = sorted(save_dir.glob('train*.jpg')) + logger.log_images(files, "Mosaics", epoch) + # end batch ------------------------------------------------------------------------------------------------ + + # Scheduler + lr = [x['lr'] for x in optimizer.param_groups] # for loggers + scheduler.step() + + if RANK in {-1, 0}: + # mAP + # callbacks.run('on_train_epoch_end', epoch=epoch) + ema.update_attr(model, include=['yaml', 'nc', 'hyp', 'names', 'stride', 'class_weights']) + final_epoch = (epoch + 1 == epochs) or stopper.possible_stop + if not noval or final_epoch: # Calculate mAP + results, maps, _ = validate.run(data_dict, + batch_size=batch_size // WORLD_SIZE * 2, + imgsz=imgsz, + half=amp, + model=ema.ema, + single_cls=single_cls, + dataloader=val_loader, + save_dir=save_dir, + plots=False, + callbacks=callbacks, + compute_loss=compute_loss, + mask_downsample_ratio=mask_ratio, + overlap=overlap) + + # Update best mAP + fi = fitness(np.array(results).reshape(1, -1)) # weighted combination of [P, R, mAP@.5, mAP@.5-.95] + stop = stopper(epoch=epoch, fitness=fi) # early stop check + if fi > best_fitness: + best_fitness = fi + log_vals = list(mloss) + list(results) + lr + # callbacks.run('on_fit_epoch_end', log_vals, epoch, best_fitness, fi) + # Log val metrics and media + metrics_dict = dict(zip(KEYS, log_vals)) + logger.log_metrics(metrics_dict, epoch) + + # Save model + if (not nosave) or (final_epoch and not evolve): # if save + ckpt = { + 'epoch': epoch, + 'best_fitness': best_fitness, + 'model': deepcopy(de_parallel(model)).half(), + 'ema': deepcopy(ema.ema).half(), + 'updates': ema.updates, + 'optimizer': optimizer.state_dict(), + # 'wandb_id': loggers.wandb.wandb_run.id if loggers.wandb else None, + 'opt': vars(opt), + 'date': datetime.now().isoformat()} + + # Save last, best and delete + torch.save(ckpt, last) + if best_fitness == fi: + torch.save(ckpt, best) + if opt.save_period > 0 and epoch % opt.save_period == 0: + torch.save(ckpt, w / f'epoch{epoch}.pt') + logger.log_model(w / f'epoch{epoch}.pt') + del ckpt + # callbacks.run('on_model_save', last, epoch, final_epoch, best_fitness, fi) + + # EarlyStopping + if RANK != -1: # if DDP training + broadcast_list = [stop if RANK == 0 else None] + dist.broadcast_object_list(broadcast_list, 0) # broadcast 'stop' to all ranks + if RANK != 0: + stop = broadcast_list[0] + if stop: + break # must break all DDP ranks + + # end epoch ---------------------------------------------------------------------------------------------------- + # end training ----------------------------------------------------------------------------------------------------- + if RANK in {-1, 0}: + LOGGER.info(f'\n{epoch - start_epoch + 1} epochs completed in {(time.time() - t0) / 3600:.3f} hours.') + for f in last, best: + if f.exists(): + strip_optimizer(f) # strip optimizers + if f is best: + LOGGER.info(f'\nValidating {f}...') + results, _, _ = validate.run( + data_dict, + batch_size=batch_size // WORLD_SIZE * 2, + imgsz=imgsz, + model=attempt_load(f, device).half(), + iou_thres=0.65 if is_coco else 0.60, # best pycocotools at iou 0.65 + single_cls=single_cls, + dataloader=val_loader, + save_dir=save_dir, + save_json=is_coco, + verbose=True, + plots=plots, + callbacks=callbacks, + compute_loss=compute_loss, + mask_downsample_ratio=mask_ratio, + overlap=overlap) # val best model with plots + if is_coco: + # callbacks.run('on_fit_epoch_end', list(mloss) + list(results) + lr, epoch, best_fitness, fi) + metrics_dict = dict(zip(KEYS, list(mloss) + list(results) + lr)) + logger.log_metrics(metrics_dict, epoch) + + # callbacks.run('on_train_end', last, best, epoch, results) + # on train end callback using genericLogger + logger.log_metrics(dict(zip(KEYS[4:16], results)), epochs) + if not opt.evolve: + logger.log_model(best, epoch) + if plots: + plot_results_with_masks(file=save_dir / 'results.csv') # save results.png + files = ['results.png', 'confusion_matrix.png', *(f'{x}_curve.png' for x in ('F1', 'PR', 'P', 'R'))] + files = [(save_dir / f) for f in files if (save_dir / f).exists()] # filter + LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}") + logger.log_images(files, "Results", epoch + 1) + logger.log_images(sorted(save_dir.glob('val*.jpg')), "Validation", epoch + 1) + torch.cuda.empty_cache() + return results + + +def parse_opt(known=False): + parser = argparse.ArgumentParser() + parser.add_argument('--weights', type=str, default=ROOT / 'yolov5s-seg.pt', help='initial weights path') + parser.add_argument('--cfg', type=str, default='', help='model.yaml path') + parser.add_argument('--data', type=str, default=ROOT / 'data/coco128-seg.yaml', help='dataset.yaml path') + parser.add_argument('--hyp', type=str, default=ROOT / 'data/hyps/hyp.scratch-low.yaml', help='hyperparameters path') + parser.add_argument('--epochs', type=int, default=300, help='total training epochs') + parser.add_argument('--batch-size', type=int, default=16, help='total batch size for all GPUs, -1 for autobatch') + parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='train, val image size (pixels)') + parser.add_argument('--rect', action='store_true', help='rectangular training') + parser.add_argument('--resume', nargs='?', const=True, default=False, help='resume most recent training') + parser.add_argument('--nosave', action='store_true', help='only save final checkpoint') + parser.add_argument('--noval', action='store_true', help='only validate final epoch') + parser.add_argument('--noautoanchor', action='store_true', help='disable AutoAnchor') + parser.add_argument('--noplots', action='store_true', help='save no plot files') + parser.add_argument('--evolve', type=int, nargs='?', const=300, help='evolve hyperparameters for x generations') + parser.add_argument('--bucket', type=str, default='', help='gsutil bucket') + parser.add_argument('--cache', type=str, nargs='?', const='ram', help='--cache images in "ram" (default) or "disk"') + parser.add_argument('--image-weights', action='store_true', help='use weighted image selection for training') + parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--multi-scale', action='store_true', help='vary img-size +/- 50%%') + parser.add_argument('--single-cls', action='store_true', help='train multi-class data as single-class') + parser.add_argument('--optimizer', type=str, choices=['SGD', 'Adam', 'AdamW'], default='SGD', help='optimizer') + parser.add_argument('--sync-bn', action='store_true', help='use SyncBatchNorm, only available in DDP mode') + parser.add_argument('--workers', type=int, default=8, help='max dataloader workers (per RANK in DDP mode)') + parser.add_argument('--project', default=ROOT / 'runs/train-seg', help='save to project/name') + parser.add_argument('--name', default='exp', help='save to project/name') + parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') + parser.add_argument('--quad', action='store_true', help='quad dataloader') + parser.add_argument('--cos-lr', action='store_true', help='cosine LR scheduler') + parser.add_argument('--label-smoothing', type=float, default=0.0, help='Label smoothing epsilon') + parser.add_argument('--patience', type=int, default=100, help='EarlyStopping patience (epochs without improvement)') + parser.add_argument('--freeze', nargs='+', type=int, default=[0], help='Freeze layers: backbone=10, first3=0 1 2') + parser.add_argument('--save-period', type=int, default=-1, help='Save checkpoint every x epochs (disabled if < 1)') + parser.add_argument('--seed', type=int, default=0, help='Global training seed') + parser.add_argument('--local_rank', type=int, default=-1, help='Automatic DDP Multi-GPU argument, do not modify') + + # Instance Segmentation Args + parser.add_argument('--mask-ratio', type=int, default=4, help='Downsample the truth masks to saving memory') + parser.add_argument('--no-overlap', action='store_true', help='Overlap masks train faster at slightly less mAP') + + # Weights & Biases arguments + # parser.add_argument('--entity', default=None, help='W&B: Entity') + # parser.add_argument('--upload_dataset', nargs='?', const=True, default=False, help='W&B: Upload data, "val" option') + # parser.add_argument('--bbox_interval', type=int, default=-1, help='W&B: Set bounding-box image logging interval') + # parser.add_argument('--artifact_alias', type=str, default='latest', help='W&B: Version of dataset artifact to use') + + return parser.parse_known_args()[0] if known else parser.parse_args() + + +def main(opt, callbacks=Callbacks()): + # Checks + if RANK in {-1, 0}: + print_args(vars(opt)) + check_git_status() + check_requirements() + + # Resume + if opt.resume and not opt.evolve: # resume from specified or most recent last.pt + last = Path(check_file(opt.resume) if isinstance(opt.resume, str) else get_latest_run()) + opt_yaml = last.parent.parent / 'opt.yaml' # train options yaml + opt_data = opt.data # original dataset + if opt_yaml.is_file(): + with open(opt_yaml, errors='ignore') as f: + d = yaml.safe_load(f) + else: + d = torch.load(last, map_location='cpu')['opt'] + opt = argparse.Namespace(**d) # replace + opt.cfg, opt.weights, opt.resume = '', str(last), True # reinstate + if is_url(opt_data): + opt.data = check_file(opt_data) # avoid HUB resume auth timeout + else: + opt.data, opt.cfg, opt.hyp, opt.weights, opt.project = \ + check_file(opt.data), check_yaml(opt.cfg), check_yaml(opt.hyp), str(opt.weights), str(opt.project) # checks + assert len(opt.cfg) or len(opt.weights), 'either --cfg or --weights must be specified' + if opt.evolve: + if opt.project == str(ROOT / 'runs/train'): # if default project name, rename to runs/evolve + opt.project = str(ROOT / 'runs/evolve') + opt.exist_ok, opt.resume = opt.resume, False # pass resume to exist_ok and disable resume + if opt.name == 'cfg': + opt.name = Path(opt.cfg).stem # use model.yaml as name + opt.save_dir = str(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok)) + + # DDP mode + device = select_device(opt.device, batch_size=opt.batch_size) + if LOCAL_RANK != -1: + msg = 'is not compatible with YOLOv5 Multi-GPU DDP training' + assert not opt.image_weights, f'--image-weights {msg}' + assert not opt.evolve, f'--evolve {msg}' + assert opt.batch_size != -1, f'AutoBatch with --batch-size -1 {msg}, please pass a valid --batch-size' + assert opt.batch_size % WORLD_SIZE == 0, f'--batch-size {opt.batch_size} must be multiple of WORLD_SIZE' + assert torch.cuda.device_count() > LOCAL_RANK, 'insufficient CUDA devices for DDP command' + torch.cuda.set_device(LOCAL_RANK) + device = torch.device('cuda', LOCAL_RANK) + dist.init_process_group(backend="nccl" if dist.is_nccl_available() else "gloo") + + # Train + if not opt.evolve: + train(opt.hyp, opt, device, callbacks) + + # Evolve hyperparameters (optional) + else: + # Hyperparameter evolution metadata (mutation scale 0-1, lower_limit, upper_limit) + meta = { + 'lr0': (1, 1e-5, 1e-1), # initial learning rate (SGD=1E-2, Adam=1E-3) + 'lrf': (1, 0.01, 1.0), # final OneCycleLR learning rate (lr0 * lrf) + 'momentum': (0.3, 0.6, 0.98), # SGD momentum/Adam beta1 + 'weight_decay': (1, 0.0, 0.001), # optimizer weight decay + 'warmup_epochs': (1, 0.0, 5.0), # warmup epochs (fractions ok) + 'warmup_momentum': (1, 0.0, 0.95), # warmup initial momentum + 'warmup_bias_lr': (1, 0.0, 0.2), # warmup initial bias lr + 'box': (1, 0.02, 0.2), # box loss gain + 'cls': (1, 0.2, 4.0), # cls loss gain + 'cls_pw': (1, 0.5, 2.0), # cls BCELoss positive_weight + 'obj': (1, 0.2, 4.0), # obj loss gain (scale with pixels) + 'obj_pw': (1, 0.5, 2.0), # obj BCELoss positive_weight + 'iou_t': (0, 0.1, 0.7), # IoU training threshold + 'anchor_t': (1, 2.0, 8.0), # anchor-multiple threshold + 'anchors': (2, 2.0, 10.0), # anchors per output grid (0 to ignore) + 'fl_gamma': (0, 0.0, 2.0), # focal loss gamma (efficientDet default gamma=1.5) + 'hsv_h': (1, 0.0, 0.1), # image HSV-Hue augmentation (fraction) + 'hsv_s': (1, 0.0, 0.9), # image HSV-Saturation augmentation (fraction) + 'hsv_v': (1, 0.0, 0.9), # image HSV-Value augmentation (fraction) + 'degrees': (1, 0.0, 45.0), # image rotation (+/- deg) + 'translate': (1, 0.0, 0.9), # image translation (+/- fraction) + 'scale': (1, 0.0, 0.9), # image scale (+/- gain) + 'shear': (1, 0.0, 10.0), # image shear (+/- deg) + 'perspective': (0, 0.0, 0.001), # image perspective (+/- fraction), range 0-0.001 + 'flipud': (1, 0.0, 1.0), # image flip up-down (probability) + 'fliplr': (0, 0.0, 1.0), # image flip left-right (probability) + 'mosaic': (1, 0.0, 1.0), # image mixup (probability) + 'mixup': (1, 0.0, 1.0), # image mixup (probability) + 'copy_paste': (1, 0.0, 1.0)} # segment copy-paste (probability) + + with open(opt.hyp, errors='ignore') as f: + hyp = yaml.safe_load(f) # load hyps dict + if 'anchors' not in hyp: # anchors commented in hyp.yaml + hyp['anchors'] = 3 + if opt.noautoanchor: + del hyp['anchors'], meta['anchors'] + opt.noval, opt.nosave, save_dir = True, True, Path(opt.save_dir) # only val/save final epoch + # ei = [isinstance(x, (int, float)) for x in hyp.values()] # evolvable indices + evolve_yaml, evolve_csv = save_dir / 'hyp_evolve.yaml', save_dir / 'evolve.csv' + if opt.bucket: + os.system(f'gsutil cp gs://{opt.bucket}/evolve.csv {evolve_csv}') # download evolve.csv if exists + + for _ in range(opt.evolve): # generations to evolve + if evolve_csv.exists(): # if evolve.csv exists: select best hyps and mutate + # Select parent(s) + parent = 'single' # parent selection method: 'single' or 'weighted' + x = np.loadtxt(evolve_csv, ndmin=2, delimiter=',', skiprows=1) + n = min(5, len(x)) # number of previous results to consider + x = x[np.argsort(-fitness(x))][:n] # top n mutations + w = fitness(x) - fitness(x).min() + 1E-6 # weights (sum > 0) + if parent == 'single' or len(x) == 1: + # x = x[random.randint(0, n - 1)] # random selection + x = x[random.choices(range(n), weights=w)[0]] # weighted selection + elif parent == 'weighted': + x = (x * w.reshape(n, 1)).sum(0) / w.sum() # weighted combination + + # Mutate + mp, s = 0.8, 0.2 # mutation probability, sigma + npr = np.random + npr.seed(int(time.time())) + g = np.array([meta[k][0] for k in hyp.keys()]) # gains 0-1 + ng = len(meta) + v = np.ones(ng) + while all(v == 1): # mutate until a change occurs (prevent duplicates) + v = (g * (npr.random(ng) < mp) * npr.randn(ng) * npr.random() * s + 1).clip(0.3, 3.0) + for i, k in enumerate(hyp.keys()): # plt.hist(v.ravel(), 300) + hyp[k] = float(x[i + 7] * v[i]) # mutate + + # Constrain to limits + for k, v in meta.items(): + hyp[k] = max(hyp[k], v[1]) # lower limit + hyp[k] = min(hyp[k], v[2]) # upper limit + hyp[k] = round(hyp[k], 5) # significant digits + + # Train mutation + results = train(hyp.copy(), opt, device, callbacks) + callbacks = Callbacks() + # Write mutation results + print_mutation(results, hyp.copy(), save_dir, opt.bucket) + + # Plot results + plot_evolve(evolve_csv) + LOGGER.info(f'Hyperparameter evolution finished {opt.evolve} generations\n' + f"Results saved to {colorstr('bold', save_dir)}\n" + f'Usage example: $ python train.py --hyp {evolve_yaml}') + + +def run(**kwargs): + # Usage: import train; train.run(data='coco128.yaml', imgsz=320, weights='yolov5m.pt') + opt = parse_opt(True) + for k, v in kwargs.items(): + setattr(opt, k, v) + main(opt) + return opt + + +if __name__ == "__main__": + opt = parse_opt() + main(opt) diff --git a/segment/val.py b/segment/val.py new file mode 100644 index 000000000000..138aa00aaed3 --- /dev/null +++ b/segment/val.py @@ -0,0 +1,471 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Validate a trained YOLOv5 segment model on a segment dataset + +Usage: + $ bash data/scripts/get_coco.sh --val --segments # download COCO-segments val split (1G, 5000 images) + $ python segment/val.py --weights yolov5s-seg.pt --data coco.yaml --img 640- # validate COCO-segments + +Usage - formats: + $ python segment/val.py --weights yolov5s-seg.pt # PyTorch + yolov5s-seg.torchscript # TorchScript + yolov5s-seg.onnx # ONNX Runtime or OpenCV DNN with --dnn + yolov5s-seg.xml # OpenVINO + yolov5s-seg.engine # TensorRT + yolov5s-seg.mlmodel # CoreML (macOS-only) + yolov5s-seg_saved_model # TensorFlow SavedModel + yolov5s-seg.pb # TensorFlow GraphDef + yolov5s-seg.tflite # TensorFlow Lite + yolov5s-seg_edgetpu.tflite # TensorFlow Edge TPU + yolov5s-seg_paddle_model # PaddlePaddle +""" + +import argparse +import json +import os +import sys +from multiprocessing.pool import ThreadPool +from pathlib import Path + +import numpy as np +import torch +from tqdm import tqdm + +FILE = Path(__file__).resolve() +ROOT = FILE.parents[1] # YOLOv5 root directory +if str(ROOT) not in sys.path: + sys.path.append(str(ROOT)) # add ROOT to PATH +ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative + +import torch.nn.functional as F + +from models.common import DetectMultiBackend +from models.yolo import SegmentationModel +from utils.callbacks import Callbacks +from utils.general import (LOGGER, NUM_THREADS, Profile, check_dataset, check_img_size, check_requirements, check_yaml, + coco80_to_coco91_class, colorstr, increment_path, non_max_suppression, print_args, + scale_coords, xywh2xyxy, xyxy2xywh) +from utils.metrics import ConfusionMatrix, box_iou +from utils.plots import output_to_target, plot_val_study +from utils.segment.dataloaders import create_dataloader +from utils.segment.general import mask_iou, process_mask, process_mask_upsample, scale_image +from utils.segment.metrics import Metrics, ap_per_class_box_and_mask +from utils.segment.plots import plot_images_and_masks +from utils.torch_utils import de_parallel, select_device, smart_inference_mode + + +def save_one_txt(predn, save_conf, shape, file): + # Save one txt result + gn = torch.tensor(shape)[[1, 0, 1, 0]] # normalization gain whwh + for *xyxy, conf, cls in predn.tolist(): + xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh + line = (cls, *xywh, conf) if save_conf else (cls, *xywh) # label format + with open(file, 'a') as f: + f.write(('%g ' * len(line)).rstrip() % line + '\n') + + +def save_one_json(predn, jdict, path, class_map, pred_masks): + # Save one JSON result {"image_id": 42, "category_id": 18, "bbox": [258.15, 41.29, 348.26, 243.78], "score": 0.236} + from pycocotools.mask import encode + + def single_encode(x): + rle = encode(np.asarray(x[:, :, None], order="F", dtype="uint8"))[0] + rle["counts"] = rle["counts"].decode("utf-8") + return rle + + image_id = int(path.stem) if path.stem.isnumeric() else path.stem + box = xyxy2xywh(predn[:, :4]) # xywh + box[:, :2] -= box[:, 2:] / 2 # xy center to top-left corner + pred_masks = np.transpose(pred_masks, (2, 0, 1)) + with ThreadPool(NUM_THREADS) as pool: + rles = pool.map(single_encode, pred_masks) + for i, (p, b) in enumerate(zip(predn.tolist(), box.tolist())): + jdict.append({ + 'image_id': image_id, + 'category_id': class_map[int(p[5])], + 'bbox': [round(x, 3) for x in b], + 'score': round(p[4], 5), + 'segmentation': rles[i]}) + + +def process_batch(detections, labels, iouv, pred_masks=None, gt_masks=None, overlap=False, masks=False): + """ + Return correct prediction matrix + Arguments: + detections (array[N, 6]), x1, y1, x2, y2, conf, class + labels (array[M, 5]), class, x1, y1, x2, y2 + Returns: + correct (array[N, 10]), for 10 IoU levels + """ + if masks: + if overlap: + nl = len(labels) + index = torch.arange(nl, device=gt_masks.device).view(nl, 1, 1) + 1 + gt_masks = gt_masks.repeat(nl, 1, 1) # shape(1,640,640) -> (n,640,640) + gt_masks = torch.where(gt_masks == index, 1.0, 0.0) + if gt_masks.shape[1:] != pred_masks.shape[1:]: + gt_masks = F.interpolate(gt_masks[None], pred_masks.shape[1:], mode="bilinear", align_corners=False)[0] + gt_masks = gt_masks.gt_(0.5) + iou = mask_iou(gt_masks.view(gt_masks.shape[0], -1), pred_masks.view(pred_masks.shape[0], -1)) + else: # boxes + iou = box_iou(labels[:, 1:], detections[:, :4]) + + correct = np.zeros((detections.shape[0], iouv.shape[0])).astype(bool) + correct_class = labels[:, 0:1] == detections[:, 5] + for i in range(len(iouv)): + x = torch.where((iou >= iouv[i]) & correct_class) # IoU > threshold and classes match + if x[0].shape[0]: + matches = torch.cat((torch.stack(x, 1), iou[x[0], x[1]][:, None]), 1).cpu().numpy() # [label, detect, iou] + if x[0].shape[0] > 1: + matches = matches[matches[:, 2].argsort()[::-1]] + matches = matches[np.unique(matches[:, 1], return_index=True)[1]] + # matches = matches[matches[:, 2].argsort()[::-1]] + matches = matches[np.unique(matches[:, 0], return_index=True)[1]] + correct[matches[:, 1].astype(int), i] = True + return torch.tensor(correct, dtype=torch.bool, device=iouv.device) + + +@smart_inference_mode() +def run( + data, + weights=None, # model.pt path(s) + batch_size=32, # batch size + imgsz=640, # inference size (pixels) + conf_thres=0.001, # confidence threshold + iou_thres=0.6, # NMS IoU threshold + max_det=300, # maximum detections per image + task='val', # train, val, test, speed or study + device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu + workers=8, # max dataloader workers (per RANK in DDP mode) + single_cls=False, # treat as single-class dataset + augment=False, # augmented inference + verbose=False, # verbose output + save_txt=False, # save results to *.txt + save_hybrid=False, # save label+prediction hybrid results to *.txt + save_conf=False, # save confidences in --save-txt labels + save_json=False, # save a COCO-JSON results file + project=ROOT / 'runs/val-seg', # save to project/name + name='exp', # save to project/name + exist_ok=False, # existing project/name ok, do not increment + half=True, # use FP16 half-precision inference + dnn=False, # use OpenCV DNN for ONNX inference + model=None, + dataloader=None, + save_dir=Path(''), + plots=True, + overlap=False, + mask_downsample_ratio=1, + compute_loss=None, + callbacks=Callbacks(), +): + if save_json: + check_requirements(['pycocotools']) + process = process_mask_upsample # more accurate + else: + process = process_mask # faster + + # Initialize/load model and set device + training = model is not None + if training: # called by train.py + device, pt, jit, engine = next(model.parameters()).device, True, False, False # get model device, PyTorch model + half &= device.type != 'cpu' # half precision only supported on CUDA + model.half() if half else model.float() + nm = de_parallel(model).model[-1].nm # number of masks + else: # called directly + device = select_device(device, batch_size=batch_size) + + # Directories + save_dir = increment_path(Path(project) / name, exist_ok=exist_ok) # increment run + (save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir + + # Load model + model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data, fp16=half) + stride, pt, jit, engine = model.stride, model.pt, model.jit, model.engine + imgsz = check_img_size(imgsz, s=stride) # check image size + half = model.fp16 # FP16 supported on limited backends with CUDA + nm = de_parallel(model).model.model[-1].nm if isinstance(model, SegmentationModel) else 32 # number of masks + if engine: + batch_size = model.batch_size + else: + device = model.device + if not (pt or jit): + batch_size = 1 # export.py models default to batch-size 1 + LOGGER.info(f'Forcing --batch-size 1 square inference (1,3,{imgsz},{imgsz}) for non-PyTorch models') + + # Data + data = check_dataset(data) # check + + # Configure + model.eval() + cuda = device.type != 'cpu' + is_coco = isinstance(data.get('val'), str) and data['val'].endswith(f'coco{os.sep}val2017.txt') # COCO dataset + nc = 1 if single_cls else int(data['nc']) # number of classes + iouv = torch.linspace(0.5, 0.95, 10, device=device) # iou vector for mAP@0.5:0.95 + niou = iouv.numel() + + # Dataloader + if not training: + if pt and not single_cls: # check --weights are trained on --data + ncm = model.model.nc + assert ncm == nc, f'{weights} ({ncm} classes) trained on different --data than what you passed ({nc} ' \ + f'classes). Pass correct combination of --weights and --data that are trained together.' + model.warmup(imgsz=(1 if pt else batch_size, 3, imgsz, imgsz)) # warmup + pad = 0.0 if task in ('speed', 'benchmark') else 0.5 + rect = False if task == 'benchmark' else pt # square inference for benchmarks + task = task if task in ('train', 'val', 'test') else 'val' # path to train/val/test images + dataloader = create_dataloader(data[task], + imgsz, + batch_size, + stride, + single_cls, + pad=pad, + rect=rect, + workers=workers, + prefix=colorstr(f'{task}: '), + overlap_mask=overlap, + mask_downsample_ratio=mask_downsample_ratio)[0] + + seen = 0 + confusion_matrix = ConfusionMatrix(nc=nc) + names = model.names if hasattr(model, 'names') else model.module.names # get class names + if isinstance(names, (list, tuple)): # old format + names = dict(enumerate(names)) + class_map = coco80_to_coco91_class() if is_coco else list(range(1000)) + s = ('%22s' + '%11s' * 10) % ('Class', 'Images', 'Instances', 'Box(P', "R", "mAP50", "mAP50-95)", "Mask(P", "R", + "mAP50", "mAP50-95)") + dt = Profile(), Profile(), Profile() + metrics = Metrics() + loss = torch.zeros(4, device=device) + jdict, stats = [], [] + # callbacks.run('on_val_start') + pbar = tqdm(dataloader, desc=s, bar_format='{l_bar}{bar:10}{r_bar}{bar:-10b}') # progress bar + for batch_i, (im, targets, paths, shapes, masks) in enumerate(pbar): + # callbacks.run('on_val_batch_start') + with dt[0]: + if cuda: + im = im.to(device, non_blocking=True) + targets = targets.to(device) + masks = masks.to(device) + masks = masks.float() + im = im.half() if half else im.float() # uint8 to fp16/32 + im /= 255 # 0 - 255 to 0.0 - 1.0 + nb, _, height, width = im.shape # batch size, channels, height, width + + # Inference + with dt[1]: + preds, protos, train_out = model(im) if compute_loss else (*model(im, augment=augment)[:2], None) + + # Loss + if compute_loss: + loss += compute_loss((train_out, protos), targets, masks)[1] # box, obj, cls + + # NMS + targets[:, 2:] *= torch.tensor((width, height, width, height), device=device) # to pixels + lb = [targets[targets[:, 0] == i, 1:] for i in range(nb)] if save_hybrid else [] # for autolabelling + with dt[2]: + preds = non_max_suppression(preds, + conf_thres, + iou_thres, + labels=lb, + multi_label=True, + agnostic=single_cls, + max_det=max_det, + nm=nm) + + # Metrics + plot_masks = [] # masks for plotting + for si, (pred, proto) in enumerate(zip(preds, protos)): + labels = targets[targets[:, 0] == si, 1:] + nl, npr = labels.shape[0], pred.shape[0] # number of labels, predictions + path, shape = Path(paths[si]), shapes[si][0] + correct_masks = torch.zeros(npr, niou, dtype=torch.bool, device=device) # init + correct_bboxes = torch.zeros(npr, niou, dtype=torch.bool, device=device) # init + seen += 1 + + if npr == 0: + if nl: + stats.append((correct_masks, correct_bboxes, *torch.zeros((2, 0), device=device), labels[:, 0])) + if plots: + confusion_matrix.process_batch(detections=None, labels=labels[:, 0]) + continue + + # Masks + midx = [si] if overlap else targets[:, 0] == si + gt_masks = masks[midx] + pred_masks = process(proto, pred[:, 6:], pred[:, :4], shape=im[si].shape[1:]) + + # Predictions + if single_cls: + pred[:, 5] = 0 + predn = pred.clone() + scale_coords(im[si].shape[1:], predn[:, :4], shape, shapes[si][1]) # native-space pred + + # Evaluate + if nl: + tbox = xywh2xyxy(labels[:, 1:5]) # target boxes + scale_coords(im[si].shape[1:], tbox, shape, shapes[si][1]) # native-space labels + labelsn = torch.cat((labels[:, 0:1], tbox), 1) # native-space labels + correct_bboxes = process_batch(predn, labelsn, iouv) + correct_masks = process_batch(predn, labelsn, iouv, pred_masks, gt_masks, overlap=overlap, masks=True) + if plots: + confusion_matrix.process_batch(predn, labelsn) + stats.append((correct_masks, correct_bboxes, pred[:, 4], pred[:, 5], labels[:, 0])) # (conf, pcls, tcls) + + pred_masks = torch.as_tensor(pred_masks, dtype=torch.uint8) + if plots and batch_i < 3: + plot_masks.append(pred_masks[:15].cpu()) # filter top 15 to plot + + # Save/log + if save_txt: + save_one_txt(predn, save_conf, shape, file=save_dir / 'labels' / f'{path.stem}.txt') + if save_json: + pred_masks = scale_image(im[si].shape[1:], + pred_masks.permute(1, 2, 0).contiguous().cpu().numpy(), shape, shapes[si][1]) + save_one_json(predn, jdict, path, class_map, pred_masks) # append to COCO-JSON dictionary + # callbacks.run('on_val_image_end', pred, predn, path, names, im[si]) + + # Plot images + if plots and batch_i < 3: + if len(plot_masks): + plot_masks = torch.cat(plot_masks, dim=0) + plot_images_and_masks(im, targets, masks, paths, save_dir / f'val_batch{batch_i}_labels.jpg', names) + plot_images_and_masks(im, output_to_target(preds, max_det=15), plot_masks, paths, + save_dir / f'val_batch{batch_i}_pred.jpg', names) # pred + + # callbacks.run('on_val_batch_end') + + # Compute metrics + stats = [torch.cat(x, 0).cpu().numpy() for x in zip(*stats)] # to numpy + if len(stats) and stats[0].any(): + results = ap_per_class_box_and_mask(*stats, plot=plots, save_dir=save_dir, names=names) + metrics.update(results) + nt = np.bincount(stats[4].astype(int), minlength=nc) # number of targets per class + + # Print results + pf = '%22s' + '%11i' * 2 + '%11.3g' * 8 # print format + LOGGER.info(pf % ("all", seen, nt.sum(), *metrics.mean_results())) + if nt.sum() == 0: + LOGGER.warning(f'WARNING: no labels found in {task} set, can not compute metrics without labels ⚠️') + + # Print results per class + if (verbose or (nc < 50 and not training)) and nc > 1 and len(stats): + for i, c in enumerate(metrics.ap_class_index): + LOGGER.info(pf % (names[c], seen, nt[c], *metrics.class_result(i))) + + # Print speeds + t = tuple(x.t / seen * 1E3 for x in dt) # speeds per image + if not training: + shape = (batch_size, 3, imgsz, imgsz) + LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {shape}' % t) + + # Plots + if plots: + confusion_matrix.plot(save_dir=save_dir, names=list(names.values())) + # callbacks.run('on_val_end') + + mp_bbox, mr_bbox, map50_bbox, map_bbox, mp_mask, mr_mask, map50_mask, map_mask = metrics.mean_results() + + # Save JSON + if save_json and len(jdict): + w = Path(weights[0] if isinstance(weights, list) else weights).stem if weights is not None else '' # weights + anno_json = str(Path(data.get('path', '../coco')) / 'annotations/instances_val2017.json') # annotations json + pred_json = str(save_dir / f"{w}_predictions.json") # predictions json + LOGGER.info(f'\nEvaluating pycocotools mAP... saving {pred_json}...') + with open(pred_json, 'w') as f: + json.dump(jdict, f) + + try: # https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocoEvalDemo.ipynb + from pycocotools.coco import COCO + from pycocotools.cocoeval import COCOeval + + anno = COCO(anno_json) # init annotations api + pred = anno.loadRes(pred_json) # init predictions api + results = [] + for eval in COCOeval(anno, pred, 'bbox'), COCOeval(anno, pred, 'segm'): + if is_coco: + eval.params.imgIds = [int(Path(x).stem) for x in dataloader.dataset.im_files] # img ID to evaluate + eval.evaluate() + eval.accumulate() + eval.summarize() + results.extend(eval.stats[:2]) # update results (mAP@0.5:0.95, mAP@0.5) + map_bbox, map50_bbox, map_mask, map50_mask = results + except Exception as e: + LOGGER.info(f'pycocotools unable to run: {e}') + + # Return results + model.float() # for training + if not training: + s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else '' + LOGGER.info(f"Results saved to {colorstr('bold', save_dir)}{s}") + final_metric = mp_bbox, mr_bbox, map50_bbox, map_bbox, mp_mask, mr_mask, map50_mask, map_mask + return (*final_metric, *(loss.cpu() / len(dataloader)).tolist()), metrics.get_maps(nc), t + + +def parse_opt(): + parser = argparse.ArgumentParser() + parser.add_argument('--data', type=str, default=ROOT / 'data/coco128-seg.yaml', help='dataset.yaml path') + parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s-seg.pt', help='model path(s)') + parser.add_argument('--batch-size', type=int, default=32, help='batch size') + parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='inference size (pixels)') + parser.add_argument('--conf-thres', type=float, default=0.001, help='confidence threshold') + parser.add_argument('--iou-thres', type=float, default=0.6, help='NMS IoU threshold') + parser.add_argument('--max-det', type=int, default=300, help='maximum detections per image') + parser.add_argument('--task', default='val', help='train, val, test, speed or study') + parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') + parser.add_argument('--workers', type=int, default=8, help='max dataloader workers (per RANK in DDP mode)') + parser.add_argument('--single-cls', action='store_true', help='treat as single-class dataset') + parser.add_argument('--augment', action='store_true', help='augmented inference') + parser.add_argument('--verbose', action='store_true', help='report mAP by class') + parser.add_argument('--save-txt', action='store_true', help='save results to *.txt') + parser.add_argument('--save-hybrid', action='store_true', help='save label+prediction hybrid results to *.txt') + parser.add_argument('--save-conf', action='store_true', help='save confidences in --save-txt labels') + parser.add_argument('--save-json', action='store_true', help='save a COCO-JSON results file') + parser.add_argument('--project', default=ROOT / 'runs/val-seg', help='save results to project/name') + parser.add_argument('--name', default='exp', help='save to project/name') + parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment') + parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference') + parser.add_argument('--dnn', action='store_true', help='use OpenCV DNN for ONNX inference') + opt = parser.parse_args() + opt.data = check_yaml(opt.data) # check YAML + # opt.save_json |= opt.data.endswith('coco.yaml') + opt.save_txt |= opt.save_hybrid + print_args(vars(opt)) + return opt + + +def main(opt): + check_requirements(requirements=ROOT / 'requirements.txt', exclude=('tensorboard', 'thop')) + + if opt.task in ('train', 'val', 'test'): # run normally + if opt.conf_thres > 0.001: # https://github.com/ultralytics/yolov5/issues/1466 + LOGGER.info(f'WARNING: confidence threshold {opt.conf_thres} > 0.001 produces invalid results ⚠️') + if opt.save_hybrid: + LOGGER.info('WARNING: --save-hybrid will return high mAP from hybrid labels, not from predictions alone ⚠️') + run(**vars(opt)) + + else: + weights = opt.weights if isinstance(opt.weights, list) else [opt.weights] + opt.half = True # FP16 for fastest results + if opt.task == 'speed': # speed benchmarks + # python val.py --task speed --data coco.yaml --batch 1 --weights yolov5n.pt yolov5s.pt... + opt.conf_thres, opt.iou_thres, opt.save_json = 0.25, 0.45, False + for opt.weights in weights: + run(**vars(opt), plots=False) + + elif opt.task == 'study': # speed vs mAP benchmarks + # python val.py --task study --data coco.yaml --iou 0.7 --weights yolov5n.pt yolov5s.pt... + for opt.weights in weights: + f = f'study_{Path(opt.data).stem}_{Path(opt.weights).stem}.txt' # filename to save to + x, y = list(range(256, 1536 + 128, 128)), [] # x axis (image sizes), y axis + for opt.imgsz in x: # img-size + LOGGER.info(f'\nRunning {f} --imgsz {opt.imgsz}...') + r, _, t = run(**vars(opt), plots=False) + y.append(r + t) # results and times + np.savetxt(f, y, fmt='%10.4g') # save + os.system('zip -r study.zip study_*.txt') + plot_val_study(x=x) # plot + + +if __name__ == "__main__": + opt = parse_opt() + main(opt) diff --git a/utils/dataloaders.py b/utils/dataloaders.py old mode 100755 new mode 100644 index d8ef11fd94b4..c04be853c580 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -484,6 +484,7 @@ def __init__(self, self.im_files = [self.im_files[i] for i in irect] self.label_files = [self.label_files[i] for i in irect] self.labels = [self.labels[i] for i in irect] + self.segments = [self.segments[i] for i in irect] self.shapes = s[irect] # wh ar = ar[irect] diff --git a/utils/general.py b/utils/general.py old mode 100755 new mode 100644 index f5fb2c93a3d5..8633511f89f5 --- a/utils/general.py +++ b/utils/general.py @@ -798,15 +798,18 @@ def clip_coords(boxes, shape): boxes[:, [1, 3]] = boxes[:, [1, 3]].clip(0, shape[0]) # y1, y2 -def non_max_suppression(prediction, - conf_thres=0.25, - iou_thres=0.45, - classes=None, - agnostic=False, - multi_label=False, - labels=(), - max_det=300): - """Non-Maximum Suppression (NMS) on inference results to reject overlapping bounding boxes +def non_max_suppression( + prediction, + conf_thres=0.25, + iou_thres=0.45, + classes=None, + agnostic=False, + multi_label=False, + labels=(), + max_det=300, + nm=0, # number of masks +): + """Non-Maximum Suppression (NMS) on inference results to reject overlapping detections Returns: list of detections, on (n,6) tensor per image [xyxy, conf, cls] @@ -816,7 +819,7 @@ def non_max_suppression(prediction, prediction = prediction[0] # select only inference output bs = prediction.shape[0] # batch size - nc = prediction.shape[2] - 5 # number of classes + nc = prediction.shape[2] - nm - 5 # number of classes xc = prediction[..., 4] > conf_thres # candidates # Checks @@ -827,13 +830,14 @@ def non_max_suppression(prediction, # min_wh = 2 # (pixels) minimum box width and height max_wh = 7680 # (pixels) maximum box width and height max_nms = 30000 # maximum number of boxes into torchvision.ops.nms() - time_limit = 0.3 + 0.03 * bs # seconds to quit after + time_limit = 0.5 + 0.05 * bs # seconds to quit after redundant = True # require redundant detections multi_label &= nc > 1 # multiple labels per box (adds 0.5ms/img) merge = False # use merge-NMS t = time.time() - output = [torch.zeros((0, 6), device=prediction.device)] * bs + mi = 5 + nc # mask start index + output = [torch.zeros((0, 6 + nm), device=prediction.device)] * bs for xi, x in enumerate(prediction): # image index, image inference # Apply constraints # x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0 # width-height @@ -842,7 +846,7 @@ def non_max_suppression(prediction, # Cat apriori labels if autolabelling if labels and len(labels[xi]): lb = labels[xi] - v = torch.zeros((len(lb), nc + 5), device=x.device) + v = torch.zeros((len(lb), nc + nm + 5), device=x.device) v[:, :4] = lb[:, 1:5] # box v[:, 4] = 1.0 # conf v[range(len(lb)), lb[:, 0].long() + 5] = 1.0 # cls @@ -855,16 +859,17 @@ def non_max_suppression(prediction, # Compute conf x[:, 5:] *= x[:, 4:5] # conf = obj_conf * cls_conf - # Box (center x, center y, width, height) to (x1, y1, x2, y2) - box = xywh2xyxy(x[:, :4]) + # Box/Mask + box = xywh2xyxy(x[:, :4]) # center_x, center_y, width, height) to (x1, y1, x2, y2) + mask = x[:, mi:] # zero columns if no masks # Detections matrix nx6 (xyxy, conf, cls) if multi_label: - i, j = (x[:, 5:] > conf_thres).nonzero(as_tuple=False).T - x = torch.cat((box[i], x[i, j + 5, None], j[:, None].float()), 1) + i, j = (x[:, 5:mi] > conf_thres).nonzero(as_tuple=False).T + x = torch.cat((box[i], x[i, 5 + j, None], j[:, None].float(), mask[i]), 1) else: # best class only - conf, j = x[:, 5:].max(1, keepdim=True) - x = torch.cat((box, conf, j.float()), 1)[conf.view(-1) > conf_thres] + conf, j = x[:, 5:mi].max(1, keepdim=True) + x = torch.cat((box, conf, j.float(), mask), 1)[conf.view(-1) > conf_thres] # Filter by class if classes is not None: @@ -880,6 +885,8 @@ def non_max_suppression(prediction, continue elif n > max_nms: # excess boxes x = x[x[:, 4].argsort(descending=True)[:max_nms]] # sort by confidence + else: + x = x[x[:, 4].argsort(descending=True)] # sort by confidence # Batched NMS c = x[:, 5:6] * (0 if agnostic else max_wh) # classes diff --git a/utils/metrics.py b/utils/metrics.py index ee7d33982cfc..001813cbcd65 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -28,7 +28,7 @@ def smooth(y, f=0.05): return np.convolve(yp, np.ones(nf) / nf, mode='valid') # y-smoothed -def ap_per_class(tp, conf, pred_cls, target_cls, plot=False, save_dir='.', names=(), eps=1e-16): +def ap_per_class(tp, conf, pred_cls, target_cls, plot=False, save_dir='.', names=(), eps=1e-16, prefix=""): """ Compute the average precision, given the recall and precision curves. Source: https://github.com/rafaelpadilla/Object-Detection-Metrics. # Arguments @@ -83,10 +83,10 @@ def ap_per_class(tp, conf, pred_cls, target_cls, plot=False, save_dir='.', names names = [v for k, v in names.items() if k in unique_classes] # list: only classes that have data names = dict(enumerate(names)) # to dict if plot: - plot_pr_curve(px, py, ap, Path(save_dir) / 'PR_curve.png', names) - plot_mc_curve(px, f1, Path(save_dir) / 'F1_curve.png', names, ylabel='F1') - plot_mc_curve(px, p, Path(save_dir) / 'P_curve.png', names, ylabel='Precision') - plot_mc_curve(px, r, Path(save_dir) / 'R_curve.png', names, ylabel='Recall') + plot_pr_curve(px, py, ap, Path(save_dir) / f'{prefix}PR_curve.png', names) + plot_mc_curve(px, f1, Path(save_dir) / f'{prefix}F1_curve.png', names, ylabel='F1') + plot_mc_curve(px, p, Path(save_dir) / f'{prefix}P_curve.png', names, ylabel='Precision') + plot_mc_curve(px, r, Path(save_dir) / f'{prefix}R_curve.png', names, ylabel='Recall') i = smooth(f1.mean(0), 0.1).argmax() # max F1 index p, r, f1 = p[:, i], r[:, i], f1[:, i] diff --git a/utils/plots.py b/utils/plots.py index 0530d0abdf48..d8d5b225a774 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -23,6 +23,7 @@ from utils.general import (CONFIG_DIR, FONT, LOGGER, check_font, check_requirements, clip_coords, increment_path, is_ascii, xywh2xyxy, xyxy2xywh) from utils.metrics import fitness +from utils.segment.general import scale_image # Settings RANK = int(os.getenv('RANK', -1)) @@ -113,6 +114,52 @@ def box_label(self, box, label='', color=(128, 128, 128), txt_color=(255, 255, 2 thickness=tf, lineType=cv2.LINE_AA) + def masks(self, masks, colors, im_gpu=None, alpha=0.5): + """Plot masks at once. + Args: + masks (tensor): predicted masks on cuda, shape: [n, h, w] + colors (List[List[Int]]): colors for predicted masks, [[r, g, b] * n] + im_gpu (tensor): img is in cuda, shape: [3, h, w], range: [0, 1] + alpha (float): mask transparency: 0.0 fully transparent, 1.0 opaque + """ + if self.pil: + # convert to numpy first + self.im = np.asarray(self.im).copy() + if im_gpu is None: + # Add multiple masks of shape(h,w,n) with colors list([r,g,b], [r,g,b], ...) + if len(masks) == 0: + return + if isinstance(masks, torch.Tensor): + masks = torch.as_tensor(masks, dtype=torch.uint8) + masks = masks.permute(1, 2, 0).contiguous() + masks = masks.cpu().numpy() + # masks = np.ascontiguousarray(masks.transpose(1, 2, 0)) + masks = scale_image(masks.shape[:2], masks, self.im.shape) + masks = np.asarray(masks, dtype=np.float32) + colors = np.asarray(colors, dtype=np.float32) # shape(n,3) + s = masks.sum(2, keepdims=True).clip(0, 1) # add all masks together + masks = (masks @ colors).clip(0, 255) # (h,w,n) @ (n,3) = (h,w,3) + self.im[:] = masks * alpha + self.im * (1 - s * alpha) + else: + if len(masks) == 0: + self.im[:] = im_gpu.permute(1, 2, 0).contiguous().cpu().numpy() * 255 + colors = torch.tensor(colors, device=im_gpu.device, dtype=torch.float32) / 255.0 + colors = colors[:, None, None] # shape(n,1,1,3) + masks = masks.unsqueeze(3) # shape(n,h,w,1) + masks_color = masks * (colors * alpha) # shape(n,h,w,3) + + inv_alph_masks = (1 - masks * alpha).cumprod(0) # shape(n,h,w,1) + mcs = (masks_color * inv_alph_masks).sum(0) * 2 # mask color summand shape(n,h,w,3) + + im_gpu = im_gpu.flip(dims=[0]) # flip channel + im_gpu = im_gpu.permute(1, 2, 0).contiguous() # shape(h,w,3) + im_gpu = im_gpu * inv_alph_masks[-1] + mcs + im_mask = (im_gpu * 255).byte().cpu().numpy() + self.im[:] = scale_image(im_gpu.shape, im_mask, self.im.shape) + if self.pil: + # convert im back to PIL and update draw + self.fromarray(self.im) + def rectangle(self, xy, fill=None, outline=None, width=1): # Add rectangle to image (PIL-only) self.draw.rectangle(xy, fill, outline, width) @@ -124,6 +171,11 @@ def text(self, xy, text, txt_color=(255, 255, 255), anchor='top'): xy[1] += 1 - h self.draw.text(xy, text, fill=txt_color, font=self.font) + def fromarray(self, im): + # Update self.im from a numpy array + self.im = im if isinstance(im, Image.Image) else Image.fromarray(im) + self.draw = ImageDraw.Draw(self.im) + def result(self): # Return annotated image as array return np.asarray(self.im) @@ -180,26 +232,31 @@ def butter_lowpass(cutoff, fs, order): return filtfilt(b, a, data) # forward-backward filter -def output_to_target(output): - # Convert model output to target format [batch_id, class_id, x, y, w, h, conf] +def output_to_target(output, max_det=300): + # Convert model output to target format [batch_id, class_id, x, y, w, h, conf] for plotting targets = [] for i, o in enumerate(output): - targets.extend([i, cls, *list(*xyxy2xywh(np.array(box)[None])), conf] for *box, conf, cls in o.cpu().numpy()) - return np.array(targets) + box, conf, cls = o[:max_det, :6].cpu().split((4, 1, 1), 1) + j = torch.full((conf.shape[0], 1), i) + targets.append(torch.cat((j, cls, xyxy2xywh(box), conf), 1)) + return torch.cat(targets, 0).numpy() @threaded -def plot_images(images, targets, paths=None, fname='images.jpg', names=None, max_size=1920, max_subplots=16): +def plot_images(images, targets, paths=None, fname='images.jpg', names=None): # Plot image grid with labels if isinstance(images, torch.Tensor): images = images.cpu().float().numpy() if isinstance(targets, torch.Tensor): targets = targets.cpu().numpy() - if np.max(images[0]) <= 1: - images *= 255 # de-normalise (optional) + + max_size = 1920 # max image size + max_subplots = 16 # max image subplots, i.e. 4x4 bs, _, h, w = images.shape # batch size, _, height, width bs = min(bs, max_subplots) # limit plot images ns = np.ceil(bs ** 0.5) # number of subplots (square) + if np.max(images[0]) <= 1: + images *= 255 # de-normalise (optional) # Build Image mosaic = np.full((int(ns * h), int(ns * w), 3), 255, dtype=np.uint8) # init diff --git a/utils/segment/__init__.py b/utils/segment/__init__.py new file mode 100644 index 000000000000..e69de29bb2d1 diff --git a/utils/segment/augmentations.py b/utils/segment/augmentations.py new file mode 100644 index 000000000000..169addedf0f5 --- /dev/null +++ b/utils/segment/augmentations.py @@ -0,0 +1,104 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Image augmentation functions +""" + +import math +import random + +import cv2 +import numpy as np + +from ..augmentations import box_candidates +from ..general import resample_segments, segment2box + + +def mixup(im, labels, segments, im2, labels2, segments2): + # Applies MixUp augmentation https://arxiv.org/pdf/1710.09412.pdf + r = np.random.beta(32.0, 32.0) # mixup ratio, alpha=beta=32.0 + im = (im * r + im2 * (1 - r)).astype(np.uint8) + labels = np.concatenate((labels, labels2), 0) + segments = np.concatenate((segments, segments2), 0) + return im, labels, segments + + +def random_perspective(im, + targets=(), + segments=(), + degrees=10, + translate=.1, + scale=.1, + shear=10, + perspective=0.0, + border=(0, 0)): + # torchvision.transforms.RandomAffine(degrees=(-10, 10), translate=(.1, .1), scale=(.9, 1.1), shear=(-10, 10)) + # targets = [cls, xyxy] + + height = im.shape[0] + border[0] * 2 # shape(h,w,c) + width = im.shape[1] + border[1] * 2 + + # Center + C = np.eye(3) + C[0, 2] = -im.shape[1] / 2 # x translation (pixels) + C[1, 2] = -im.shape[0] / 2 # y translation (pixels) + + # Perspective + P = np.eye(3) + P[2, 0] = random.uniform(-perspective, perspective) # x perspective (about y) + P[2, 1] = random.uniform(-perspective, perspective) # y perspective (about x) + + # Rotation and Scale + R = np.eye(3) + a = random.uniform(-degrees, degrees) + # a += random.choice([-180, -90, 0, 90]) # add 90deg rotations to small rotations + s = random.uniform(1 - scale, 1 + scale) + # s = 2 ** random.uniform(-scale, scale) + R[:2] = cv2.getRotationMatrix2D(angle=a, center=(0, 0), scale=s) + + # Shear + S = np.eye(3) + S[0, 1] = math.tan(random.uniform(-shear, shear) * math.pi / 180) # x shear (deg) + S[1, 0] = math.tan(random.uniform(-shear, shear) * math.pi / 180) # y shear (deg) + + # Translation + T = np.eye(3) + T[0, 2] = (random.uniform(0.5 - translate, 0.5 + translate) * width) # x translation (pixels) + T[1, 2] = (random.uniform(0.5 - translate, 0.5 + translate) * height) # y translation (pixels) + + # Combined rotation matrix + M = T @ S @ R @ P @ C # order of operations (right to left) is IMPORTANT + if (border[0] != 0) or (border[1] != 0) or (M != np.eye(3)).any(): # image changed + if perspective: + im = cv2.warpPerspective(im, M, dsize=(width, height), borderValue=(114, 114, 114)) + else: # affine + im = cv2.warpAffine(im, M[:2], dsize=(width, height), borderValue=(114, 114, 114)) + + # Visualize + # import matplotlib.pyplot as plt + # ax = plt.subplots(1, 2, figsize=(12, 6))[1].ravel() + # ax[0].imshow(im[:, :, ::-1]) # base + # ax[1].imshow(im2[:, :, ::-1]) # warped + + # Transform label coordinates + n = len(targets) + new_segments = [] + if n: + new = np.zeros((n, 4)) + segments = resample_segments(segments) # upsample + for i, segment in enumerate(segments): + xy = np.ones((len(segment), 3)) + xy[:, :2] = segment + xy = xy @ M.T # transform + xy = (xy[:, :2] / xy[:, 2:3] if perspective else xy[:, :2]) # perspective rescale or affine + + # clip + new[i] = segment2box(xy, width, height) + new_segments.append(xy) + + # filter candidates + i = box_candidates(box1=targets[:, 1:5].T * s, box2=new.T, area_thr=0.01) + targets = targets[i] + targets[:, 1:5] = new[i] + new_segments = np.array(new_segments)[i] + + return im, targets, new_segments diff --git a/utils/segment/dataloaders.py b/utils/segment/dataloaders.py new file mode 100644 index 000000000000..f6fe642d077f --- /dev/null +++ b/utils/segment/dataloaders.py @@ -0,0 +1,330 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Dataloaders +""" + +import os +import random + +import cv2 +import numpy as np +import torch +from torch.utils.data import DataLoader, distributed + +from ..augmentations import augment_hsv, copy_paste, letterbox +from ..dataloaders import InfiniteDataLoader, LoadImagesAndLabels, seed_worker +from ..general import LOGGER, xyn2xy, xywhn2xyxy, xyxy2xywhn +from ..torch_utils import torch_distributed_zero_first +from .augmentations import mixup, random_perspective + + +def create_dataloader(path, + imgsz, + batch_size, + stride, + single_cls=False, + hyp=None, + augment=False, + cache=False, + pad=0.0, + rect=False, + rank=-1, + workers=8, + image_weights=False, + quad=False, + prefix='', + shuffle=False, + mask_downsample_ratio=1, + overlap_mask=False): + if rect and shuffle: + LOGGER.warning('WARNING: --rect is incompatible with DataLoader shuffle, setting shuffle=False') + shuffle = False + with torch_distributed_zero_first(rank): # init dataset *.cache only once if DDP + dataset = LoadImagesAndLabelsAndMasks( + path, + imgsz, + batch_size, + augment=augment, # augmentation + hyp=hyp, # hyperparameters + rect=rect, # rectangular batches + cache_images=cache, + single_cls=single_cls, + stride=int(stride), + pad=pad, + image_weights=image_weights, + prefix=prefix, + downsample_ratio=mask_downsample_ratio, + overlap=overlap_mask) + + batch_size = min(batch_size, len(dataset)) + nd = torch.cuda.device_count() # number of CUDA devices + nw = min([os.cpu_count() // max(nd, 1), batch_size if batch_size > 1 else 0, workers]) # number of workers + sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle) + loader = DataLoader if image_weights else InfiniteDataLoader # only DataLoader allows for attribute updates + # generator = torch.Generator() + # generator.manual_seed(0) + return loader( + dataset, + batch_size=batch_size, + shuffle=shuffle and sampler is None, + num_workers=nw, + sampler=sampler, + pin_memory=True, + collate_fn=LoadImagesAndLabelsAndMasks.collate_fn4 if quad else LoadImagesAndLabelsAndMasks.collate_fn, + worker_init_fn=seed_worker, + # generator=generator, + ), dataset + + +class LoadImagesAndLabelsAndMasks(LoadImagesAndLabels): # for training/testing + + def __init__( + self, + path, + img_size=640, + batch_size=16, + augment=False, + hyp=None, + rect=False, + image_weights=False, + cache_images=False, + single_cls=False, + stride=32, + pad=0, + prefix="", + downsample_ratio=1, + overlap=False, + ): + super().__init__(path, img_size, batch_size, augment, hyp, rect, image_weights, cache_images, single_cls, + stride, pad, prefix) + self.downsample_ratio = downsample_ratio + self.overlap = overlap + + def __getitem__(self, index): + index = self.indices[index] # linear, shuffled, or image_weights + + hyp = self.hyp + mosaic = self.mosaic and random.random() < hyp['mosaic'] + masks = [] + if mosaic: + # Load mosaic + img, labels, segments = self.load_mosaic(index) + shapes = None + + # MixUp augmentation + if random.random() < hyp["mixup"]: + img, labels, segments = mixup(img, labels, segments, *self.load_mosaic(random.randint(0, self.n - 1))) + + else: + # Load image + img, (h0, w0), (h, w) = self.load_image(index) + + # Letterbox + shape = self.batch_shapes[self.batch[index]] if self.rect else self.img_size # final letterboxed shape + img, ratio, pad = letterbox(img, shape, auto=False, scaleup=self.augment) + shapes = (h0, w0), ((h / h0, w / w0), pad) # for COCO mAP rescaling + + labels = self.labels[index].copy() + # [array, array, ....], array.shape=(num_points, 2), xyxyxyxy + segments = self.segments[index].copy() + if len(segments): + for i_s in range(len(segments)): + segments[i_s] = xyn2xy( + segments[i_s], + ratio[0] * w, + ratio[1] * h, + padw=pad[0], + padh=pad[1], + ) + if labels.size: # normalized xywh to pixel xyxy format + labels[:, 1:] = xywhn2xyxy(labels[:, 1:], ratio[0] * w, ratio[1] * h, padw=pad[0], padh=pad[1]) + + if self.augment: + img, labels, segments = random_perspective( + img, + labels, + segments=segments, + degrees=hyp["degrees"], + translate=hyp["translate"], + scale=hyp["scale"], + shear=hyp["shear"], + perspective=hyp["perspective"], + return_seg=True, + ) + + nl = len(labels) # number of labels + if nl: + labels[:, 1:5] = xyxy2xywhn(labels[:, 1:5], w=img.shape[1], h=img.shape[0], clip=True, eps=1e-3) + if self.overlap: + masks, sorted_idx = polygons2masks_overlap(img.shape[:2], + segments, + downsample_ratio=self.downsample_ratio) + masks = masks[None] # (640, 640) -> (1, 640, 640) + labels = labels[sorted_idx] + else: + masks = polygons2masks(img.shape[:2], segments, color=1, downsample_ratio=self.downsample_ratio) + + masks = (torch.from_numpy(masks) if len(masks) else torch.zeros(1 if self.overlap else nl, img.shape[0] // + self.downsample_ratio, img.shape[1] // + self.downsample_ratio)) + # TODO: albumentations support + if self.augment: + # Albumentations + # there are some augmentation that won't change boxes and masks, + # so just be it for now. + img, labels = self.albumentations(img, labels) + nl = len(labels) # update after albumentations + + # HSV color-space + augment_hsv(img, hgain=hyp["hsv_h"], sgain=hyp["hsv_s"], vgain=hyp["hsv_v"]) + + # Flip up-down + if random.random() < hyp["flipud"]: + img = np.flipud(img) + if nl: + labels[:, 2] = 1 - labels[:, 2] + masks = torch.flip(masks, dims=[1]) + + # Flip left-right + if random.random() < hyp["fliplr"]: + img = np.fliplr(img) + if nl: + labels[:, 1] = 1 - labels[:, 1] + masks = torch.flip(masks, dims=[2]) + + # Cutouts # labels = cutout(img, labels, p=0.5) + + labels_out = torch.zeros((nl, 6)) + if nl: + labels_out[:, 1:] = torch.from_numpy(labels) + + # Convert + img = img.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB + img = np.ascontiguousarray(img) + + return (torch.from_numpy(img), labels_out, self.im_files[index], shapes, masks) + + def load_mosaic(self, index): + # YOLOv5 4-mosaic loader. Loads 1 image + 3 random images into a 4-image mosaic + labels4, segments4 = [], [] + s = self.img_size + yc, xc = (int(random.uniform(-x, 2 * s + x)) for x in self.mosaic_border) # mosaic center x, y + + # 3 additional image indices + indices = [index] + random.choices(self.indices, k=3) # 3 additional image indices + for i, index in enumerate(indices): + # Load image + img, _, (h, w) = self.load_image(index) + + # place img in img4 + if i == 0: # top left + img4 = np.full((s * 2, s * 2, img.shape[2]), 114, dtype=np.uint8) # base image with 4 tiles + x1a, y1a, x2a, y2a = max(xc - w, 0), max(yc - h, 0), xc, yc # xmin, ymin, xmax, ymax (large image) + x1b, y1b, x2b, y2b = w - (x2a - x1a), h - (y2a - y1a), w, h # xmin, ymin, xmax, ymax (small image) + elif i == 1: # top right + x1a, y1a, x2a, y2a = xc, max(yc - h, 0), min(xc + w, s * 2), yc + x1b, y1b, x2b, y2b = 0, h - (y2a - y1a), min(w, x2a - x1a), h + elif i == 2: # bottom left + x1a, y1a, x2a, y2a = max(xc - w, 0), yc, xc, min(s * 2, yc + h) + x1b, y1b, x2b, y2b = w - (x2a - x1a), 0, w, min(y2a - y1a, h) + elif i == 3: # bottom right + x1a, y1a, x2a, y2a = xc, yc, min(xc + w, s * 2), min(s * 2, yc + h) + x1b, y1b, x2b, y2b = 0, 0, min(w, x2a - x1a), min(y2a - y1a, h) + + img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b] # img4[ymin:ymax, xmin:xmax] + padw = x1a - x1b + padh = y1a - y1b + + labels, segments = self.labels[index].copy(), self.segments[index].copy() + + if labels.size: + labels[:, 1:] = xywhn2xyxy(labels[:, 1:], w, h, padw, padh) # normalized xywh to pixel xyxy format + segments = [xyn2xy(x, w, h, padw, padh) for x in segments] + labels4.append(labels) + segments4.extend(segments) + + # Concat/clip labels + labels4 = np.concatenate(labels4, 0) + for x in (labels4[:, 1:], *segments4): + np.clip(x, 0, 2 * s, out=x) # clip when using random_perspective() + # img4, labels4 = replicate(img4, labels4) # replicate + + # Augment + img4, labels4, segments4 = copy_paste(img4, labels4, segments4, p=self.hyp["copy_paste"]) + img4, labels4, segments4 = random_perspective(img4, + labels4, + segments4, + degrees=self.hyp["degrees"], + translate=self.hyp["translate"], + scale=self.hyp["scale"], + shear=self.hyp["shear"], + perspective=self.hyp["perspective"], + border=self.mosaic_border) # border to remove + return img4, labels4, segments4 + + @staticmethod + def collate_fn(batch): + img, label, path, shapes, masks = zip(*batch) # transposed + batched_masks = torch.cat(masks, 0) + for i, l in enumerate(label): + l[:, 0] = i # add target image index for build_targets() + return torch.stack(img, 0), torch.cat(label, 0), path, shapes, batched_masks + + +def polygon2mask(img_size, polygons, color=1, downsample_ratio=1): + """ + Args: + img_size (tuple): The image size. + polygons (np.ndarray): [N, M], N is the number of polygons, + M is the number of points(Be divided by 2). + """ + mask = np.zeros(img_size, dtype=np.uint8) + polygons = np.asarray(polygons) + polygons = polygons.astype(np.int32) + shape = polygons.shape + polygons = polygons.reshape(shape[0], -1, 2) + cv2.fillPoly(mask, polygons, color=color) + nh, nw = (img_size[0] // downsample_ratio, img_size[1] // downsample_ratio) + # NOTE: fillPoly firstly then resize is trying the keep the same way + # of loss calculation when mask-ratio=1. + mask = cv2.resize(mask, (nw, nh)) + return mask + + +def polygons2masks(img_size, polygons, color, downsample_ratio=1): + """ + Args: + img_size (tuple): The image size. + polygons (list[np.ndarray]): each polygon is [N, M], + N is the number of polygons, + M is the number of points(Be divided by 2). + """ + masks = [] + for si in range(len(polygons)): + mask = polygon2mask(img_size, [polygons[si].reshape(-1)], color, downsample_ratio) + masks.append(mask) + return np.array(masks) + + +def polygons2masks_overlap(img_size, segments, downsample_ratio=1): + """Return a (640, 640) overlap mask.""" + masks = np.zeros((img_size[0] // downsample_ratio, img_size[1] // downsample_ratio), dtype=np.uint8) + areas = [] + ms = [] + for si in range(len(segments)): + mask = polygon2mask( + img_size, + [segments[si].reshape(-1)], + downsample_ratio=downsample_ratio, + color=1, + ) + ms.append(mask) + areas.append(mask.sum()) + areas = np.asarray(areas) + index = np.argsort(-areas) + ms = np.array(ms)[index] + for i in range(len(segments)): + mask = ms[i] * (i + 1) + masks = masks + mask + masks = np.clip(masks, a_min=0, a_max=i + 1) + return masks, index diff --git a/utils/segment/general.py b/utils/segment/general.py new file mode 100644 index 000000000000..36547ed0889c --- /dev/null +++ b/utils/segment/general.py @@ -0,0 +1,120 @@ +import cv2 +import torch +import torch.nn.functional as F + + +def crop_mask(masks, boxes): + """ + "Crop" predicted masks by zeroing out everything not in the predicted bbox. + Vectorized by Chong (thanks Chong). + + Args: + - masks should be a size [h, w, n] tensor of masks + - boxes should be a size [n, 4] tensor of bbox coords in relative point form + """ + + n, h, w = masks.shape + x1, y1, x2, y2 = torch.chunk(boxes[:, :, None], 4, 1) # x1 shape(1,1,n) + r = torch.arange(w, device=masks.device, dtype=x1.dtype)[None, None, :] # rows shape(1,w,1) + c = torch.arange(h, device=masks.device, dtype=x1.dtype)[None, :, None] # cols shape(h,1,1) + + return masks * ((r >= x1) * (r < x2) * (c >= y1) * (c < y2)) + + +def process_mask_upsample(protos, masks_in, bboxes, shape): + """ + Crop after upsample. + proto_out: [mask_dim, mask_h, mask_w] + out_masks: [n, mask_dim], n is number of masks after nms + bboxes: [n, 4], n is number of masks after nms + shape:input_image_size, (h, w) + + return: h, w, n + """ + + c, mh, mw = protos.shape # CHW + masks = (masks_in @ protos.float().view(c, -1)).sigmoid().view(-1, mh, mw) + masks = F.interpolate(masks[None], shape, mode='bilinear', align_corners=False)[0] # CHW + masks = crop_mask(masks, bboxes) # CHW + return masks.gt_(0.5) + + +def process_mask(protos, masks_in, bboxes, shape, upsample=False): + """ + Crop before upsample. + proto_out: [mask_dim, mask_h, mask_w] + out_masks: [n, mask_dim], n is number of masks after nms + bboxes: [n, 4], n is number of masks after nms + shape:input_image_size, (h, w) + + return: h, w, n + """ + + c, mh, mw = protos.shape # CHW + ih, iw = shape + masks = (masks_in @ protos.float().view(c, -1)).sigmoid().view(-1, mh, mw) # CHW + + downsampled_bboxes = bboxes.clone() + downsampled_bboxes[:, 0] *= mw / iw + downsampled_bboxes[:, 2] *= mw / iw + downsampled_bboxes[:, 3] *= mh / ih + downsampled_bboxes[:, 1] *= mh / ih + + masks = crop_mask(masks, downsampled_bboxes) # CHW + if upsample: + masks = F.interpolate(masks[None], shape, mode='bilinear', align_corners=False)[0] # CHW + return masks.gt_(0.5) + + +def scale_image(im1_shape, masks, im0_shape, ratio_pad=None): + """ + img1_shape: model input shape, [h, w] + img0_shape: origin pic shape, [h, w, 3] + masks: [h, w, num] + """ + # Rescale coordinates (xyxy) from im1_shape to im0_shape + if ratio_pad is None: # calculate from im0_shape + gain = min(im1_shape[0] / im0_shape[0], im1_shape[1] / im0_shape[1]) # gain = old / new + pad = (im1_shape[1] - im0_shape[1] * gain) / 2, (im1_shape[0] - im0_shape[0] * gain) / 2 # wh padding + else: + pad = ratio_pad[1] + top, left = int(pad[1]), int(pad[0]) # y, x + bottom, right = int(im1_shape[0] - pad[1]), int(im1_shape[1] - pad[0]) + + if len(masks.shape) < 2: + raise ValueError(f'"len of masks shape" should be 2 or 3, but got {len(masks.shape)}') + masks = masks[top:bottom, left:right] + # masks = masks.permute(2, 0, 1).contiguous() + # masks = F.interpolate(masks[None], im0_shape[:2], mode='bilinear', align_corners=False)[0] + # masks = masks.permute(1, 2, 0).contiguous() + masks = cv2.resize(masks, (im0_shape[1], im0_shape[0])) + + if len(masks.shape) == 2: + masks = masks[:, :, None] + return masks + + +def mask_iou(mask1, mask2, eps=1e-7): + """ + mask1: [N, n] m1 means number of predicted objects + mask2: [M, n] m2 means number of gt objects + Note: n means image_w x image_h + + return: masks iou, [N, M] + """ + intersection = torch.matmul(mask1, mask2.t()).clamp(0) + union = (mask1.sum(1)[:, None] + mask2.sum(1)[None]) - intersection # (area1 + area2) - intersection + return intersection / (union + eps) + + +def masks_iou(mask1, mask2, eps=1e-7): + """ + mask1: [N, n] m1 means number of predicted objects + mask2: [N, n] m2 means number of gt objects + Note: n means image_w x image_h + + return: masks iou, (N, ) + """ + intersection = (mask1 * mask2).sum(1).clamp(0) # (N, ) + union = (mask1.sum(1) + mask2.sum(1))[None] - intersection # (area1 + area2) - intersection + return intersection / (union + eps) diff --git a/utils/segment/loss.py b/utils/segment/loss.py new file mode 100644 index 000000000000..b45b2c27e0a0 --- /dev/null +++ b/utils/segment/loss.py @@ -0,0 +1,186 @@ +import torch +import torch.nn as nn +import torch.nn.functional as F + +from ..general import xywh2xyxy +from ..loss import FocalLoss, smooth_BCE +from ..metrics import bbox_iou +from ..torch_utils import de_parallel +from .general import crop_mask + + +class ComputeLoss: + # Compute losses + def __init__(self, model, autobalance=False, overlap=False): + self.sort_obj_iou = False + self.overlap = overlap + device = next(model.parameters()).device # get model device + h = model.hyp # hyperparameters + self.device = device + + # Define criteria + BCEcls = nn.BCEWithLogitsLoss(pos_weight=torch.tensor([h['cls_pw']], device=device)) + BCEobj = nn.BCEWithLogitsLoss(pos_weight=torch.tensor([h['obj_pw']], device=device)) + + # Class label smoothing https://arxiv.org/pdf/1902.04103.pdf eqn 3 + self.cp, self.cn = smooth_BCE(eps=h.get('label_smoothing', 0.0)) # positive, negative BCE targets + + # Focal loss + g = h['fl_gamma'] # focal loss gamma + if g > 0: + BCEcls, BCEobj = FocalLoss(BCEcls, g), FocalLoss(BCEobj, g) + + m = de_parallel(model).model[-1] # Detect() module + self.balance = {3: [4.0, 1.0, 0.4]}.get(m.nl, [4.0, 1.0, 0.25, 0.06, 0.02]) # P3-P7 + self.ssi = list(m.stride).index(16) if autobalance else 0 # stride 16 index + self.BCEcls, self.BCEobj, self.gr, self.hyp, self.autobalance = BCEcls, BCEobj, 1.0, h, autobalance + self.na = m.na # number of anchors + self.nc = m.nc # number of classes + self.nl = m.nl # number of layers + self.nm = m.nm # number of masks + self.anchors = m.anchors + self.device = device + + def __call__(self, preds, targets, masks): # predictions, targets, model + p, proto = preds + bs, nm, mask_h, mask_w = proto.shape # batch size, number of masks, mask height, mask width + lcls = torch.zeros(1, device=self.device) + lbox = torch.zeros(1, device=self.device) + lobj = torch.zeros(1, device=self.device) + lseg = torch.zeros(1, device=self.device) + tcls, tbox, indices, anchors, tidxs, xywhn = self.build_targets(p, targets) # targets + + # Losses + for i, pi in enumerate(p): # layer index, layer predictions + b, a, gj, gi = indices[i] # image, anchor, gridy, gridx + tobj = torch.zeros(pi.shape[:4], dtype=pi.dtype, device=self.device) # target obj + + n = b.shape[0] # number of targets + if n: + pxy, pwh, _, pcls, pmask = pi[b, a, gj, gi].split((2, 2, 1, self.nc, nm), 1) # subset of predictions + + # Box regression + pxy = pxy.sigmoid() * 2 - 0.5 + pwh = (pwh.sigmoid() * 2) ** 2 * anchors[i] + pbox = torch.cat((pxy, pwh), 1) # predicted box + iou = bbox_iou(pbox, tbox[i], CIoU=True).squeeze() # iou(prediction, target) + lbox += (1.0 - iou).mean() # iou loss + + # Objectness + iou = iou.detach().clamp(0).type(tobj.dtype) + if self.sort_obj_iou: + j = iou.argsort() + b, a, gj, gi, iou = b[j], a[j], gj[j], gi[j], iou[j] + if self.gr < 1: + iou = (1.0 - self.gr) + self.gr * iou + tobj[b, a, gj, gi] = iou # iou ratio + + # Classification + if self.nc > 1: # cls loss (only if multiple classes) + t = torch.full_like(pcls, self.cn, device=self.device) # targets + t[range(n), tcls[i]] = self.cp + lcls += self.BCEcls(pcls, t) # BCE + + # Mask regression + if tuple(masks.shape[-2:]) != (mask_h, mask_w): # downsample + masks = F.interpolate(masks[None], (mask_h, mask_w), mode="nearest")[0] + marea = xywhn[i][:, 2:].prod(1) # mask width, height normalized + mxyxy = xywh2xyxy(xywhn[i] * torch.tensor([mask_w, mask_h, mask_w, mask_h], device=self.device)) + for bi in b.unique(): + j = b == bi # matching index + if self.overlap: + mask_gti = torch.where(masks[bi][None] == tidxs[i][j].view(-1, 1, 1), 1.0, 0.0) + else: + mask_gti = masks[tidxs[i]][j] + lseg += self.single_mask_loss(mask_gti, pmask[j], proto[bi], mxyxy[j], marea[j]) + + obji = self.BCEobj(pi[..., 4], tobj) + lobj += obji * self.balance[i] # obj loss + if self.autobalance: + self.balance[i] = self.balance[i] * 0.9999 + 0.0001 / obji.detach().item() + + if self.autobalance: + self.balance = [x / self.balance[self.ssi] for x in self.balance] + lbox *= self.hyp["box"] + lobj *= self.hyp["obj"] + lcls *= self.hyp["cls"] + lseg *= self.hyp["box"] / bs + + loss = lbox + lobj + lcls + lseg + return loss * bs, torch.cat((lbox, lseg, lobj, lcls)).detach() + + def single_mask_loss(self, gt_mask, pred, proto, xyxy, area): + # Mask loss for one image + pred_mask = (pred @ proto.view(self.nm, -1)).view(-1, *proto.shape[1:]) # (n,32) @ (32,80,80) -> (n,80,80) + loss = F.binary_cross_entropy_with_logits(pred_mask, gt_mask, reduction="none") + return (crop_mask(loss, xyxy).mean(dim=(1, 2)) / area).mean() + + def build_targets(self, p, targets): + # Build targets for compute_loss(), input targets(image,class,x,y,w,h) + na, nt = self.na, targets.shape[0] # number of anchors, targets + tcls, tbox, indices, anch, tidxs, xywhn = [], [], [], [], [], [] + gain = torch.ones(8, device=self.device) # normalized to gridspace gain + ai = torch.arange(na, device=self.device).float().view(na, 1).repeat(1, nt) # same as .repeat_interleave(nt) + if self.overlap: + batch = p[0].shape[0] + ti = [] + for i in range(batch): + num = (targets[:, 0] == i).sum() # find number of targets of each image + ti.append(torch.arange(num, device=self.device).float().view(1, num).repeat(na, 1) + 1) # (na, num) + ti = torch.cat(ti, 1) # (na, nt) + else: + ti = torch.arange(nt, device=self.device).float().view(1, nt).repeat(na, 1) + targets = torch.cat((targets.repeat(na, 1, 1), ai[..., None], ti[..., None]), 2) # append anchor indices + + g = 0.5 # bias + off = torch.tensor( + [ + [0, 0], + [1, 0], + [0, 1], + [-1, 0], + [0, -1], # j,k,l,m + # [1, 1], [1, -1], [-1, 1], [-1, -1], # jk,jm,lk,lm + ], + device=self.device).float() * g # offsets + + for i in range(self.nl): + anchors, shape = self.anchors[i], p[i].shape + gain[2:6] = torch.tensor(shape)[[3, 2, 3, 2]] # xyxy gain + + # Match targets to anchors + t = targets * gain # shape(3,n,7) + if nt: + # Matches + r = t[..., 4:6] / anchors[:, None] # wh ratio + j = torch.max(r, 1 / r).max(2)[0] < self.hyp['anchor_t'] # compare + # j = wh_iou(anchors, t[:, 4:6]) > model.hyp['iou_t'] # iou(3,n)=wh_iou(anchors(3,2), gwh(n,2)) + t = t[j] # filter + + # Offsets + gxy = t[:, 2:4] # grid xy + gxi = gain[[2, 3]] - gxy # inverse + j, k = ((gxy % 1 < g) & (gxy > 1)).T + l, m = ((gxi % 1 < g) & (gxi > 1)).T + j = torch.stack((torch.ones_like(j), j, k, l, m)) + t = t.repeat((5, 1, 1))[j] + offsets = (torch.zeros_like(gxy)[None] + off[:, None])[j] + else: + t = targets[0] + offsets = 0 + + # Define + bc, gxy, gwh, at = t.chunk(4, 1) # (image, class), grid xy, grid wh, anchors + (a, tidx), (b, c) = at.long().T, bc.long().T # anchors, image, class + gij = (gxy - offsets).long() + gi, gj = gij.T # grid indices + + # Append + indices.append((b, a, gj.clamp_(0, shape[2] - 1), gi.clamp_(0, shape[3] - 1))) # image, anchor, grid + tbox.append(torch.cat((gxy - gij, gwh), 1)) # box + anch.append(anchors[a]) # anchors + tcls.append(c) # class + tidxs.append(tidx) + xywhn.append(torch.cat((gxy, gwh), 1) / gain[2:6]) # xywh normalized + + return tcls, tbox, indices, anch, tidxs, xywhn diff --git a/utils/segment/metrics.py b/utils/segment/metrics.py new file mode 100644 index 000000000000..b09ce23fb9e3 --- /dev/null +++ b/utils/segment/metrics.py @@ -0,0 +1,210 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" +Model validation metrics +""" + +import numpy as np + +from ..metrics import ap_per_class + + +def fitness(x): + # Model fitness as a weighted combination of metrics + w = [0.0, 0.0, 0.1, 0.9, 0.0, 0.0, 0.1, 0.9] + return (x[:, :8] * w).sum(1) + + +def ap_per_class_box_and_mask( + tp_m, + tp_b, + conf, + pred_cls, + target_cls, + plot=False, + save_dir=".", + names=(), +): + """ + Args: + tp_b: tp of boxes. + tp_m: tp of masks. + other arguments see `func: ap_per_class`. + """ + results_boxes = ap_per_class(tp_b, + conf, + pred_cls, + target_cls, + plot=plot, + save_dir=save_dir, + names=names, + prefix="Box")[2:] + results_masks = ap_per_class(tp_m, + conf, + pred_cls, + target_cls, + plot=plot, + save_dir=save_dir, + names=names, + prefix="Mask")[2:] + + results = { + "boxes": { + "p": results_boxes[0], + "r": results_boxes[1], + "ap": results_boxes[3], + "f1": results_boxes[2], + "ap_class": results_boxes[4]}, + "masks": { + "p": results_masks[0], + "r": results_masks[1], + "ap": results_masks[3], + "f1": results_masks[2], + "ap_class": results_masks[4]}} + return results + + +class Metric: + + def __init__(self) -> None: + self.p = [] # (nc, ) + self.r = [] # (nc, ) + self.f1 = [] # (nc, ) + self.all_ap = [] # (nc, 10) + self.ap_class_index = [] # (nc, ) + + @property + def ap50(self): + """AP@0.5 of all classes. + Return: + (nc, ) or []. + """ + return self.all_ap[:, 0] if len(self.all_ap) else [] + + @property + def ap(self): + """AP@0.5:0.95 + Return: + (nc, ) or []. + """ + return self.all_ap.mean(1) if len(self.all_ap) else [] + + @property + def mp(self): + """mean precision of all classes. + Return: + float. + """ + return self.p.mean() if len(self.p) else 0.0 + + @property + def mr(self): + """mean recall of all classes. + Return: + float. + """ + return self.r.mean() if len(self.r) else 0.0 + + @property + def map50(self): + """Mean AP@0.5 of all classes. + Return: + float. + """ + return self.all_ap[:, 0].mean() if len(self.all_ap) else 0.0 + + @property + def map(self): + """Mean AP@0.5:0.95 of all classes. + Return: + float. + """ + return self.all_ap.mean() if len(self.all_ap) else 0.0 + + def mean_results(self): + """Mean of results, return mp, mr, map50, map""" + return (self.mp, self.mr, self.map50, self.map) + + def class_result(self, i): + """class-aware result, return p[i], r[i], ap50[i], ap[i]""" + return (self.p[i], self.r[i], self.ap50[i], self.ap[i]) + + def get_maps(self, nc): + maps = np.zeros(nc) + self.map + for i, c in enumerate(self.ap_class_index): + maps[c] = self.ap[i] + return maps + + def update(self, results): + """ + Args: + results: tuple(p, r, ap, f1, ap_class) + """ + p, r, all_ap, f1, ap_class_index = results + self.p = p + self.r = r + self.all_ap = all_ap + self.f1 = f1 + self.ap_class_index = ap_class_index + + +class Metrics: + """Metric for boxes and masks.""" + + def __init__(self) -> None: + self.metric_box = Metric() + self.metric_mask = Metric() + + def update(self, results): + """ + Args: + results: Dict{'boxes': Dict{}, 'masks': Dict{}} + """ + self.metric_box.update(list(results["boxes"].values())) + self.metric_mask.update(list(results["masks"].values())) + + def mean_results(self): + return self.metric_box.mean_results() + self.metric_mask.mean_results() + + def class_result(self, i): + return self.metric_box.class_result(i) + self.metric_mask.class_result(i) + + def get_maps(self, nc): + return self.metric_box.get_maps(nc) + self.metric_mask.get_maps(nc) + + @property + def ap_class_index(self): + # boxes and masks have the same ap_class_index + return self.metric_box.ap_class_index + + +KEYS = [ + "train/box_loss", + "train/seg_loss", # train loss + "train/obj_loss", + "train/cls_loss", + "metrics/precision(B)", + "metrics/recall(B)", + "metrics/mAP_0.5(B)", + "metrics/mAP_0.5:0.95(B)", # metrics + "metrics/precision(M)", + "metrics/recall(M)", + "metrics/mAP_0.5(M)", + "metrics/mAP_0.5:0.95(M)", # metrics + "val/box_loss", + "val/seg_loss", # val loss + "val/obj_loss", + "val/cls_loss", + "x/lr0", + "x/lr1", + "x/lr2",] + +BEST_KEYS = [ + "best/epoch", + "best/precision(B)", + "best/recall(B)", + "best/mAP_0.5(B)", + "best/mAP_0.5:0.95(B)", + "best/precision(M)", + "best/recall(M)", + "best/mAP_0.5(M)", + "best/mAP_0.5:0.95(M)",] diff --git a/utils/segment/plots.py b/utils/segment/plots.py new file mode 100644 index 000000000000..e882c14390f0 --- /dev/null +++ b/utils/segment/plots.py @@ -0,0 +1,143 @@ +import contextlib +import math +from pathlib import Path + +import cv2 +import matplotlib.pyplot as plt +import numpy as np +import pandas as pd +import torch + +from .. import threaded +from ..general import xywh2xyxy +from ..plots import Annotator, colors + + +@threaded +def plot_images_and_masks(images, targets, masks, paths=None, fname='images.jpg', names=None): + # Plot image grid with labels + if isinstance(images, torch.Tensor): + images = images.cpu().float().numpy() + if isinstance(targets, torch.Tensor): + targets = targets.cpu().numpy() + if isinstance(masks, torch.Tensor): + masks = masks.cpu().numpy().astype(int) + + max_size = 1920 # max image size + max_subplots = 16 # max image subplots, i.e. 4x4 + bs, _, h, w = images.shape # batch size, _, height, width + bs = min(bs, max_subplots) # limit plot images + ns = np.ceil(bs ** 0.5) # number of subplots (square) + if np.max(images[0]) <= 1: + images *= 255 # de-normalise (optional) + + # Build Image + mosaic = np.full((int(ns * h), int(ns * w), 3), 255, dtype=np.uint8) # init + for i, im in enumerate(images): + if i == max_subplots: # if last batch has fewer images than we expect + break + x, y = int(w * (i // ns)), int(h * (i % ns)) # block origin + im = im.transpose(1, 2, 0) + mosaic[y:y + h, x:x + w, :] = im + + # Resize (optional) + scale = max_size / ns / max(h, w) + if scale < 1: + h = math.ceil(scale * h) + w = math.ceil(scale * w) + mosaic = cv2.resize(mosaic, tuple(int(x * ns) for x in (w, h))) + + # Annotate + fs = int((h + w) * ns * 0.01) # font size + annotator = Annotator(mosaic, line_width=round(fs / 10), font_size=fs, pil=True, example=names) + for i in range(i + 1): + x, y = int(w * (i // ns)), int(h * (i % ns)) # block origin + annotator.rectangle([x, y, x + w, y + h], None, (255, 255, 255), width=2) # borders + if paths: + annotator.text((x + 5, y + 5 + h), text=Path(paths[i]).name[:40], txt_color=(220, 220, 220)) # filenames + if len(targets) > 0: + idx = targets[:, 0] == i + ti = targets[idx] # image targets + + boxes = xywh2xyxy(ti[:, 2:6]).T + classes = ti[:, 1].astype('int') + labels = ti.shape[1] == 6 # labels if no conf column + conf = None if labels else ti[:, 6] # check for confidence presence (label vs pred) + + if boxes.shape[1]: + if boxes.max() <= 1.01: # if normalized with tolerance 0.01 + boxes[[0, 2]] *= w # scale to pixels + boxes[[1, 3]] *= h + elif scale < 1: # absolute coords need scale if image scales + boxes *= scale + boxes[[0, 2]] += x + boxes[[1, 3]] += y + for j, box in enumerate(boxes.T.tolist()): + cls = classes[j] + color = colors(cls) + cls = names[cls] if names else cls + if labels or conf[j] > 0.25: # 0.25 conf thresh + label = f'{cls}' if labels else f'{cls} {conf[j]:.1f}' + annotator.box_label(box, label, color=color) + + # Plot masks + if len(masks): + if masks.max() > 1.0: # mean that masks are overlap + image_masks = masks[[i]] # (1, 640, 640) + nl = len(ti) + index = np.arange(nl).reshape(nl, 1, 1) + 1 + image_masks = np.repeat(image_masks, nl, axis=0) + image_masks = np.where(image_masks == index, 1.0, 0.0) + else: + image_masks = masks[idx] + + im = np.asarray(annotator.im).copy() + for j, box in enumerate(boxes.T.tolist()): + if labels or conf[j] > 0.25: # 0.25 conf thresh + color = colors(classes[j]) + mh, mw = image_masks[j].shape + if mh != h or mw != w: + mask = image_masks[j].astype(np.uint8) + mask = cv2.resize(mask, (w, h)) + mask = mask.astype(np.bool) + else: + mask = image_masks[j].astype(np.bool) + with contextlib.suppress(Exception): + im[y:y + h, x:x + w, :][mask] = im[y:y + h, x:x + w, :][mask] * 0.4 + np.array(color) * 0.6 + annotator.fromarray(im) + annotator.im.save(fname) # save + + +def plot_results_with_masks(file="path/to/results.csv", dir="", best=True): + # Plot training results.csv. Usage: from utils.plots import *; plot_results('path/to/results.csv') + save_dir = Path(file).parent if file else Path(dir) + fig, ax = plt.subplots(2, 8, figsize=(18, 6), tight_layout=True) + ax = ax.ravel() + files = list(save_dir.glob("results*.csv")) + assert len(files), f"No results.csv files found in {save_dir.resolve()}, nothing to plot." + for f in files: + try: + data = pd.read_csv(f) + index = np.argmax(0.9 * data.values[:, 8] + 0.1 * data.values[:, 7] + 0.9 * data.values[:, 12] + + 0.1 * data.values[:, 11]) + s = [x.strip() for x in data.columns] + x = data.values[:, 0] + for i, j in enumerate([1, 2, 3, 4, 5, 6, 9, 10, 13, 14, 15, 16, 7, 8, 11, 12]): + y = data.values[:, j] + # y[y == 0] = np.nan # don't show zero values + ax[i].plot(x, y, marker=".", label=f.stem, linewidth=2, markersize=2) + if best: + # best + ax[i].scatter(index, y[index], color="r", label=f"best:{index}", marker="*", linewidth=3) + ax[i].set_title(s[j] + f"\n{round(y[index], 5)}") + else: + # last + ax[i].scatter(x[-1], y[-1], color="r", label="last", marker="*", linewidth=3) + ax[i].set_title(s[j] + f"\n{round(y[-1], 5)}") + # if j in [8, 9, 10]: # share train and val loss y axes + # ax[i].get_shared_y_axes().join(ax[i], ax[i - 5]) + except Exception as e: + print(f"Warning: Plotting error for {f}: {e}") + ax[1].legend() + fig.savefig(save_dir / "results.png", dpi=200) + plt.close() diff --git a/val.py b/val.py index 4b0bdddae3b1..6a0f18e28392 100644 --- a/val.py +++ b/val.py @@ -71,12 +71,12 @@ def save_one_json(predn, jdict, path, class_map): def process_batch(detections, labels, iouv): """ - Return correct predictions matrix. Both sets of boxes are in (x1, y1, x2, y2) format. + Return correct prediction matrix Arguments: - detections (Array[N, 6]), x1, y1, x2, y2, conf, class - labels (Array[M, 5]), class, x1, y1, x2, y2 + detections (array[N, 6]), x1, y1, x2, y2, conf, class + labels (array[M, 5]), class, x1, y1, x2, y2 Returns: - correct (Array[N, 10]), for 10 IoU levels + correct (array[N, 10]), for 10 IoU levels """ correct = np.zeros((detections.shape[0], iouv.shape[0])).astype(bool) iou = box_iou(labels[:, 1:], detections[:, :4]) @@ -102,6 +102,7 @@ def run( imgsz=640, # inference size (pixels) conf_thres=0.001, # confidence threshold iou_thres=0.6, # NMS IoU threshold + max_det=300, # maximum detections per image task='val', # train, val, test, speed or study device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu workers=8, # max dataloader workers (per RANK in DDP mode) @@ -187,7 +188,7 @@ def run( if isinstance(names, (list, tuple)): # old format names = dict(enumerate(names)) class_map = coco80_to_coco91_class() if is_coco else list(range(1000)) - s = ('%22s' + '%11s' * 6) % ('Class', 'Images', 'Instances', 'P', 'R', 'mAP@.5', 'mAP@.5:.95') + s = ('%22s' + '%11s' * 6) % ('Class', 'Images', 'Instances', 'P', 'R', 'mAP50', 'mAP50-95') dt, p, r, f1, mp, mr, map50, map = (Profile(), Profile(), Profile()), 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 loss = torch.zeros(3, device=device) jdict, stats, ap, ap_class = [], [], [], [] @@ -205,7 +206,7 @@ def run( # Inference with dt[1]: - out, train_out = model(im) if compute_loss else (model(im, augment=augment), None) + preds, train_out = model(im) if compute_loss else (model(im, augment=augment), None) # Loss if compute_loss: @@ -215,10 +216,16 @@ def run( targets[:, 2:] *= torch.tensor((width, height, width, height), device=device) # to pixels lb = [targets[targets[:, 0] == i, 1:] for i in range(nb)] if save_hybrid else [] # for autolabelling with dt[2]: - out = non_max_suppression(out, conf_thres, iou_thres, labels=lb, multi_label=True, agnostic=single_cls) + preds = non_max_suppression(preds, + conf_thres, + iou_thres, + labels=lb, + multi_label=True, + agnostic=single_cls, + max_det=max_det) # Metrics - for si, pred in enumerate(out): + for si, pred in enumerate(preds): labels = targets[targets[:, 0] == si, 1:] nl, npr = labels.shape[0], pred.shape[0] # number of labels, predictions path, shape = Path(paths[si]), shapes[si][0] @@ -258,9 +265,9 @@ def run( # Plot images if plots and batch_i < 3: plot_images(im, targets, paths, save_dir / f'val_batch{batch_i}_labels.jpg', names) # labels - plot_images(im, output_to_target(out), paths, save_dir / f'val_batch{batch_i}_pred.jpg', names) # pred + plot_images(im, output_to_target(preds), paths, save_dir / f'val_batch{batch_i}_pred.jpg', names) # pred - callbacks.run('on_val_batch_end', batch_i, im, targets, paths, shapes, out) + callbacks.run('on_val_batch_end', batch_i, im, targets, paths, shapes, preds) # Compute metrics stats = [torch.cat(x, 0).cpu().numpy() for x in zip(*stats)] # to numpy @@ -332,11 +339,12 @@ def run( def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path') - parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s.pt', help='model.pt path(s)') + parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s.pt', help='model path(s)') parser.add_argument('--batch-size', type=int, default=32, help='batch size') parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='inference size (pixels)') parser.add_argument('--conf-thres', type=float, default=0.001, help='confidence threshold') parser.add_argument('--iou-thres', type=float, default=0.6, help='NMS IoU threshold') + parser.add_argument('--max-det', type=int, default=300, help='maximum detections per image') parser.add_argument('--task', default='val', help='train, val, test, speed or study') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--workers', type=int, default=8, help='max dataloader workers (per RANK in DDP mode)') From 58ad5ca5ce6b4fb3da6420bcc7b11a09e20674fd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 16 Sep 2022 00:21:13 +0200 Subject: [PATCH 580/661] Fix val.py zero-TP bug (#9431) Resolves https://github.com/ultralytics/yolov5/issues/9400 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- val.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/val.py b/val.py index 6a0f18e28392..e003d2144b7f 100644 --- a/val.py +++ b/val.py @@ -189,7 +189,8 @@ def run( names = dict(enumerate(names)) class_map = coco80_to_coco91_class() if is_coco else list(range(1000)) s = ('%22s' + '%11s' * 6) % ('Class', 'Images', 'Instances', 'P', 'R', 'mAP50', 'mAP50-95') - dt, p, r, f1, mp, mr, map50, map = (Profile(), Profile(), Profile()), 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 + tp, fp, p, r, f1, mp, mr, map50, ap50, map = 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 + dt = Profile(), Profile(), Profile() # profiling times loss = torch.zeros(3, device=device) jdict, stats, ap, ap_class = [], [], [], [] callbacks.run('on_val_start') From a1e5f9a97de2a3ace012315208c686744ced2782 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 16 Sep 2022 00:55:21 +0200 Subject: [PATCH 581/661] New model.yaml `activation:` field (#9371) * New model.yaml `activation:` field Add optional model yaml activation field to define model-wide activations, i.e.: ```yaml activation: nn.LeakyReLU(0.1) # activation with arguments activation: nn.SiLU() # activation with no arguments ``` Signed-off-by: Glenn Jocher * Update yolo.py Signed-off-by: Glenn Jocher * Add example models * l to m models * update * Add yolov5s-LeakyReLU.yaml * Update yolov5s-LeakyReLU.yaml Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- models/common.py | 8 +++-- models/hub/yolov5s-LeakyReLU.yaml | 49 +++++++++++++++++++++++++++++++ models/yolo.py | 6 +++- 3 files changed, 59 insertions(+), 4 deletions(-) create mode 100644 models/hub/yolov5s-LeakyReLU.yaml diff --git a/models/common.py b/models/common.py index 0d90ff4f8827..debbc2d03f60 100644 --- a/models/common.py +++ b/models/common.py @@ -39,11 +39,13 @@ def autopad(k, p=None, d=1): # kernel, padding, dilation class Conv(nn.Module): # Standard convolution with args(ch_in, ch_out, kernel, stride, padding, groups, dilation, activation) + act = nn.SiLU() # default activation + def __init__(self, c1, c2, k=1, s=1, p=None, g=1, d=1, act=True): super().__init__() self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p, d), groups=g, dilation=d, bias=False) self.bn = nn.BatchNorm2d(c2) - self.act = nn.SiLU() if act is True else (act if isinstance(act, nn.Module) else nn.Identity()) + self.act = self.act if act is True else act if isinstance(act, nn.Module) else nn.Identity() def forward(self, x): return self.act(self.bn(self.conv(x))) @@ -54,8 +56,8 @@ def forward_fuse(self, x): class DWConv(Conv): # Depth-wise convolution - def __init__(self, c1, c2, k=1, s=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups - super().__init__(c1, c2, k, s, g=math.gcd(c1, c2), act=act) + def __init__(self, c1, c2, k=1, s=1, d=1, act=True): # ch_in, ch_out, kernel, stride, dilation, activation + super().__init__(c1, c2, k, s, g=math.gcd(c1, c2), d=d, act=act) class DWConvTranspose2d(nn.ConvTranspose2d): diff --git a/models/hub/yolov5s-LeakyReLU.yaml b/models/hub/yolov5s-LeakyReLU.yaml new file mode 100644 index 000000000000..3a179bf3311c --- /dev/null +++ b/models/hub/yolov5s-LeakyReLU.yaml @@ -0,0 +1,49 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license + +# Parameters +nc: 80 # number of classes +activation: nn.LeakyReLU(0.1) # <----- Conv() activation used throughout entire YOLOv5 model +depth_multiple: 0.33 # model depth multiple +width_multiple: 0.50 # layer channel multiple +anchors: + - [10,13, 16,30, 33,23] # P3/8 + - [30,61, 62,45, 59,119] # P4/16 + - [116,90, 156,198, 373,326] # P5/32 + +# YOLOv5 v6.0 backbone +backbone: + # [from, number, module, args] + [[-1, 1, Conv, [64, 6, 2, 2]], # 0-P1/2 + [-1, 1, Conv, [128, 3, 2]], # 1-P2/4 + [-1, 3, C3, [128]], + [-1, 1, Conv, [256, 3, 2]], # 3-P3/8 + [-1, 6, C3, [256]], + [-1, 1, Conv, [512, 3, 2]], # 5-P4/16 + [-1, 9, C3, [512]], + [-1, 1, Conv, [1024, 3, 2]], # 7-P5/32 + [-1, 3, C3, [1024]], + [-1, 1, SPPF, [1024, 5]], # 9 + ] + +# YOLOv5 v6.0 head +head: + [[-1, 1, Conv, [512, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 6], 1, Concat, [1]], # cat backbone P4 + [-1, 3, C3, [512, False]], # 13 + + [-1, 1, Conv, [256, 1, 1]], + [-1, 1, nn.Upsample, [None, 2, 'nearest']], + [[-1, 4], 1, Concat, [1]], # cat backbone P3 + [-1, 3, C3, [256, False]], # 17 (P3/8-small) + + [-1, 1, Conv, [256, 3, 2]], + [[-1, 14], 1, Concat, [1]], # cat head P4 + [-1, 3, C3, [512, False]], # 20 (P4/16-medium) + + [-1, 1, Conv, [512, 3, 2]], + [[-1, 10], 1, Concat, [1]], # cat head P5 + [-1, 3, C3, [1024, False]], # 23 (P5/32-large) + + [[17, 20, 23], 1, Detect, [nc, anchors]], # Detect(P3, P4, P5) + ] diff --git a/models/yolo.py b/models/yolo.py index a0702a7c0257..46039c36d7e1 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -297,8 +297,12 @@ def _from_yaml(self, cfg): def parse_model(d, ch): # model_dict, input_channels(3) + # Parse a YOLOv5 model.yaml dictionary LOGGER.info(f"\n{'':>3}{'from':>18}{'n':>3}{'params':>10} {'module':<40}{'arguments':<30}") - anchors, nc, gd, gw = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple'] + anchors, nc, gd, gw, act = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple'], d.get('activation') + if act: + Conv.act = eval(act) # redefine default activation, i.e. Conv.act = nn.SiLU() + LOGGER.info(f"{colorstr('activation:')} {act}") # print na = (len(anchors[0]) // 2) if isinstance(anchors, list) else anchors # number of anchors no = na * (nc + 5) # number of outputs = anchors * (classes + 5) From c7a2d6bcf4f7e88db53f3d09a8484391dac7bc89 Mon Sep 17 00:00:00 2001 From: Hoyeong-GenGenAI <5404902+hotohoto@users.noreply.github.com> Date: Fri, 16 Sep 2022 18:53:18 +0900 Subject: [PATCH 582/661] Fix tick labels for background FN/FP (#9414) * Fix tick labels for background FN/FP In the confusion matrix. * Remove FP/FN from the background labels of the confusion matrix * Update metrics.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: Glenn Jocher --- utils/metrics.py | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/utils/metrics.py b/utils/metrics.py index 001813cbcd65..021a46ce5d37 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -170,12 +170,12 @@ def process_batch(self, detections, labels): if n and sum(j) == 1: self.matrix[detection_classes[m1[j]], gc] += 1 # correct else: - self.matrix[self.nc, gc] += 1 # background FP + self.matrix[self.nc, gc] += 1 # true background if n: for i, dc in enumerate(detection_classes): if not any(m1 == i): - self.matrix[dc, self.nc] += 1 # background FN + self.matrix[dc, self.nc] += 1 # predicted background def matrix(self): return self.matrix @@ -197,6 +197,7 @@ def plot(self, normalize=True, save_dir='', names=()): nc, nn = self.nc, len(names) # number of classes, names sn.set(font_scale=1.0 if nc < 50 else 0.8) # for label size labels = (0 < nn < 99) and (nn == nc) # apply names to ticklabels + ticklabels = (names + ['background']) if labels else "auto" with warnings.catch_warnings(): warnings.simplefilter('ignore') # suppress empty matrix RuntimeWarning: All-NaN slice encountered sn.heatmap(array, @@ -208,8 +209,8 @@ def plot(self, normalize=True, save_dir='', names=()): fmt='.2f', square=True, vmin=0.0, - xticklabels=names + ['background FP'] if labels else "auto", - yticklabels=names + ['background FN'] if labels else "auto").set_facecolor((1, 1, 1)) + xticklabels=ticklabels, + yticklabels=ticklabels).set_facecolor((1, 1, 1)) ax.set_ylabel('True') ax.set_ylabel('Predicted') ax.set_title('Confusion Matrix') From 03f2ca8eff8918b98169256d055353a1f15b8e32 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 16 Sep 2022 12:31:43 +0200 Subject: [PATCH 583/661] Fix TensorRT exports to ONNX opset 12 (#9441) * Fix TensorRT exports to ONNX opset 12 Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/export.py b/export.py index 1b25f3f8221b..cc4386ae4916 100644 --- a/export.py +++ b/export.py @@ -251,7 +251,7 @@ def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose model.model[-1].anchor_grid = grid else: # TensorRT >= 8 check_version(trt.__version__, '8.0.0', hard=True) # require tensorrt>=8.0.0 - export_onnx(model, im, file, 13, False, dynamic, simplify) # opset 13 + export_onnx(model, im, file, 12, False, dynamic, simplify) # opset 12 onnx = file.with_suffix('.onnx') LOGGER.info(f'\n{prefix} starting export with TensorRT {trt.__version__}...') @@ -274,11 +274,10 @@ def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose inputs = [network.get_input(i) for i in range(network.num_inputs)] outputs = [network.get_output(i) for i in range(network.num_outputs)] - LOGGER.info(f'{prefix} Network Description:') for inp in inputs: - LOGGER.info(f'{prefix}\tinput "{inp.name}" with shape {inp.shape} and dtype {inp.dtype}') + LOGGER.info(f'{prefix} input "{inp.name}" with shape{inp.shape} {inp.dtype}') for out in outputs: - LOGGER.info(f'{prefix}\toutput "{out.name}" with shape {out.shape} and dtype {out.dtype}') + LOGGER.info(f'{prefix} output "{out.name}" with shape{out.shape} {out.dtype}') if dynamic: if im.shape[0] <= 1: @@ -288,7 +287,7 @@ def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose profile.set_shape(inp.name, (1, *im.shape[1:]), (max(1, im.shape[0] // 2), *im.shape[1:]), im.shape) config.add_optimization_profile(profile) - LOGGER.info(f'{prefix} building FP{16 if builder.platform_has_fast_fp16 and half else 32} engine in {f}') + LOGGER.info(f'{prefix} building FP{16 if builder.platform_has_fast_fp16 and half else 32} engine as {f}') if builder.platform_has_fast_fp16 and half: config.set_flag(trt.BuilderFlag.FP16) with builder.build_engine(network, config) as engine, open(f, 'wb') as t: From 2ac4b634c745cc46c4728e682c6da66f79f6416a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 16 Sep 2022 17:25:44 +0200 Subject: [PATCH 584/661] AutoShape explicit arguments fix (#9443) * AutoShape explicit arguments fix Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- models/common.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/models/common.py b/models/common.py index debbc2d03f60..85b82e10a4e1 100644 --- a/models/common.py +++ b/models/common.py @@ -633,7 +633,7 @@ def forward(self, ims, size=640, augment=False, profile=False): autocast = self.amp and (p.device.type != 'cpu') # Automatic Mixed Precision (AMP) inference if isinstance(ims, torch.Tensor): # torch with amp.autocast(autocast): - return self.model(ims.to(p.device).type_as(p), augment, profile) # inference + return self.model(ims.to(p.device).type_as(p), augment=augment) # inference # Pre-process n, ims = (len(ims), list(ims)) if isinstance(ims, (list, tuple)) else (1, [ims]) # number, list of images @@ -662,7 +662,7 @@ def forward(self, ims, size=640, augment=False, profile=False): with amp.autocast(autocast): # Inference with dt[1]: - y = self.model(x, augment, profile) # forward + y = self.model(x, augment=augment) # forward # Post-process with dt[2]: @@ -696,7 +696,7 @@ def __init__(self, ims, pred, files, times=(0, 0, 0), names=None, shape=None): self.xywhn = [x / g for x, g in zip(self.xywh, gn)] # xywh normalized self.n = len(self.pred) # number of images (batch size) self.t = tuple(x.t / self.n * 1E3 for x in times) # timestamps (ms) - self.s = shape # inference BCHW shape + self.s = tuple(shape) # inference BCHW shape def display(self, pprint=False, show=False, save=False, crop=False, render=False, labels=True, save_dir=Path('')): crops = [] @@ -726,7 +726,7 @@ def display(self, pprint=False, show=False, save=False, crop=False, render=False im = Image.fromarray(im.astype(np.uint8)) if isinstance(im, np.ndarray) else im # from np if pprint: - print(s.rstrip(', ')) + LOGGER.info(s.rstrip(', ')) if show: im.show(self.files[i]) # show if save: @@ -743,7 +743,7 @@ def display(self, pprint=False, show=False, save=False, crop=False, render=False def print(self): self.display(pprint=True) # print results - print(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {tuple(self.s)}' % self.t) + LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {self.s}' % self.t) def show(self, labels=True): self.display(show=True, labels=labels) # show results From fe10b4abc054cba1b5fab1d3598b3caf77b53859 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 16 Sep 2022 18:36:55 +0200 Subject: [PATCH 585/661] Update Detections() instance printing (#9445) * Update Detections() instance printing Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- models/common.py | 39 +++++++++++++++++++++------------------ 1 file changed, 21 insertions(+), 18 deletions(-) diff --git a/models/common.py b/models/common.py index 85b82e10a4e1..9c08120fe7f6 100644 --- a/models/common.py +++ b/models/common.py @@ -698,14 +698,15 @@ def __init__(self, ims, pred, files, times=(0, 0, 0), names=None, shape=None): self.t = tuple(x.t / self.n * 1E3 for x in times) # timestamps (ms) self.s = tuple(shape) # inference BCHW shape - def display(self, pprint=False, show=False, save=False, crop=False, render=False, labels=True, save_dir=Path('')): - crops = [] + def _run(self, pprint=False, show=False, save=False, crop=False, render=False, labels=True, save_dir=Path('')): + s, crops = '', [] for i, (im, pred) in enumerate(zip(self.ims, self.pred)): - s = f'image {i + 1}/{len(self.pred)}: {im.shape[0]}x{im.shape[1]} ' # string + s += f'\nimage {i + 1}/{len(self.pred)}: {im.shape[0]}x{im.shape[1]} ' # string if pred.shape[0]: for c in pred[:, -1].unique(): n = (pred[:, -1] == c).sum() # detections per class s += f"{n} {self.names[int(c)]}{'s' * (n > 1)}, " # add to string + s = s.rstrip(', ') if show or save or render or crop: annotator = Annotator(im, example=str(self.names)) for *box, conf, cls in reversed(pred): # xyxy, confidence, class @@ -725,8 +726,6 @@ def display(self, pprint=False, show=False, save=False, crop=False, render=False s += '(no detections)' im = Image.fromarray(im.astype(np.uint8)) if isinstance(im, np.ndarray) else im # from np - if pprint: - LOGGER.info(s.rstrip(', ')) if show: im.show(self.files[i]) # show if save: @@ -736,28 +735,27 @@ def display(self, pprint=False, show=False, save=False, crop=False, render=False LOGGER.info(f"Saved {self.n} image{'s' * (self.n > 1)} to {colorstr('bold', save_dir)}") if render: self.ims[i] = np.asarray(im) + if pprint: + s = s.lstrip('\n') + return f'{s}\nSpeed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {self.s}' % self.t if crop: if save: LOGGER.info(f'Saved results to {save_dir}\n') return crops - def print(self): - self.display(pprint=True) # print results - LOGGER.info(f'Speed: %.1fms pre-process, %.1fms inference, %.1fms NMS per image at shape {self.s}' % self.t) - def show(self, labels=True): - self.display(show=True, labels=labels) # show results + self._run(show=True, labels=labels) # show results def save(self, labels=True, save_dir='runs/detect/exp'): save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) # increment save_dir - self.display(save=True, labels=labels, save_dir=save_dir) # save results + self._run(save=True, labels=labels, save_dir=save_dir) # save results def crop(self, save=True, save_dir='runs/detect/exp'): save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) if save else None - return self.display(crop=True, save=save, save_dir=save_dir) # crop results + return self._run(crop=True, save=save, save_dir=save_dir) # crop results def render(self, labels=True): - self.display(render=True, labels=labels) # render results + self._run(render=True, labels=labels) # render results return self.ims def pandas(self): @@ -779,12 +777,17 @@ def tolist(self): # setattr(d, k, getattr(d, k)[0]) # pop out of list return x - def __len__(self): - return self.n # override len(results) + def print(self): + LOGGER.info(self.__str__()) + + def __len__(self): # override len(results) + return self.n + + def __str__(self): # override print(results) + return self._run(pprint=True) # print results - def __str__(self): - self.print() # override print(results) - return '' + def __repr__(self): + return f'YOLOv5 {self.__class__} instance\n' + self.__str__() class Proto(nn.Module): From db06f495db02501ef94efe46171d952642dec880 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 16 Sep 2022 20:44:56 +0200 Subject: [PATCH 586/661] AutoUpdate TensorFlow in export.py (#9447) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- export.py | 1 + 1 file changed, 1 insertion(+) diff --git a/export.py b/export.py index cc4386ae4916..a575c73e375f 100644 --- a/export.py +++ b/export.py @@ -309,6 +309,7 @@ def export_saved_model(model, keras=False, prefix=colorstr('TensorFlow SavedModel:')): # YOLOv5 TensorFlow SavedModel export + check_requirements('tensorflow' if torch.cuda.is_available() else 'tensorflow-cpu') import tensorflow as tf from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2 From 5e1a9553fbed73995c9b81e63ba41cc70fdf89de Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 16 Sep 2022 21:46:07 +0200 Subject: [PATCH 587/661] AutoBatch `cudnn.benchmark=True` fix (#9448) * AutoBatch `cudnn.benchmark=True` fix May resolve https://github.com/ultralytics/yolov5/issues/9287 Signed-off-by: Glenn Jocher * Update autobatch.py Signed-off-by: Glenn Jocher * Update autobatch.py Signed-off-by: Glenn Jocher * Update general.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/autobatch.py | 3 +++ utils/general.py | 2 +- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/utils/autobatch.py b/utils/autobatch.py index 641b055b9fe3..3204fd26fc41 100644 --- a/utils/autobatch.py +++ b/utils/autobatch.py @@ -33,6 +33,9 @@ def autobatch(model, imgsz=640, fraction=0.8, batch_size=16): if device.type == 'cpu': LOGGER.info(f'{prefix}CUDA not detected, using default CPU batch-size {batch_size}') return batch_size + if torch.backends.cudnn.benchmark: + LOGGER.info(f'{prefix} ⚠️ Requires torch.backends.cudnn.benchmark=False, using default batch-size {batch_size}') + return batch_size # Inspect CUDA memory gb = 1 << 30 # bytes to GiB (1024 ** 3) diff --git a/utils/general.py b/utils/general.py index 8633511f89f5..af95b3dc2b8b 100644 --- a/utils/general.py +++ b/utils/general.py @@ -223,7 +223,7 @@ def init_seeds(seed=0, deterministic=False): torch.manual_seed(seed) torch.cuda.manual_seed(seed) torch.cuda.manual_seed_all(seed) # for Multi-GPU, exception safe - torch.backends.cudnn.benchmark = True # for faster training + # torch.backends.cudnn.benchmark = True # AutoBatch problem https://github.com/ultralytics/yolov5/issues/9287 if deterministic and check_version(torch.__version__, '1.12.0'): # https://github.com/ultralytics/yolov5/pull/8213 torch.use_deterministic_algorithms(True) torch.backends.cudnn.deterministic = True From 4a4308001ce1699fca2d9566b652e2388a088973 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 17 Sep 2022 15:19:43 +0200 Subject: [PATCH 588/661] Do not move downloaded zips (#9455) * Do not move downloaded zips Prevent multiple downloads on HUB of same dataset @kalenmike Signed-off-by: Glenn Jocher * Update general.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/utils/general.py b/utils/general.py index af95b3dc2b8b..4d080f282ed0 100644 --- a/utils/general.py +++ b/utils/general.py @@ -568,10 +568,10 @@ def download(url, dir='.', unzip=True, delete=True, curl=False, threads=1, retry def download_one(url, dir): # Download 1 file success = True - f = dir / Path(url).name # filename - if Path(url).is_file(): # exists in current path - Path(url).rename(f) # move to dir - elif not f.exists(): + if Path(url).is_file(): + f = Path(url) # filename + else: # does not exist + f = dir / Path(url).name LOGGER.info(f'Downloading {url} to {f}...') for i in range(retry + 1): if curl: From 6a9fffd19a96799c683c94d2d4da8c453e819116 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 17 Sep 2022 15:42:24 +0200 Subject: [PATCH 589/661] Update general.py (#9454) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) diff --git a/utils/general.py b/utils/general.py index 4d080f282ed0..38856b6bfa1c 100644 --- a/utils/general.py +++ b/utils/general.py @@ -469,8 +469,7 @@ def check_dataset(data, autodownload=True): # Read yaml (optional) if isinstance(data, (str, Path)): - with open(data, errors='ignore') as f: - data = yaml.safe_load(f) # dictionary + data = yaml_load(data) # dictionary # Checks for k in 'train', 'val', 'names': @@ -485,7 +484,13 @@ def check_dataset(data, autodownload=True): path = (ROOT / path).resolve() for k in 'train', 'val', 'test': if data.get(k): # prepend path - data[k] = str(path / data[k]) if isinstance(data[k], str) else [str(path / x) for x in data[k]] + if isinstance(data[k], str): + x = (path / data[k]).resolve() + if not x.exists() and data[k].startswith('../'): + x = (path / data[k][3:]).resolve() + data[k] = str(x) + else: + data[k] = [str((path / x).resolve()) for x in data[k]] # Parse yaml train, val, test, s = (data.get(x) for x in ('train', 'val', 'test', 'download')) @@ -496,13 +501,12 @@ def check_dataset(data, autodownload=True): if not s or not autodownload: raise Exception('Dataset not found ❌') t = time.time() - root = path.parent if 'path' in data else '..' # unzip directory i.e. '../' if s.startswith('http') and s.endswith('.zip'): # URL f = Path(s).name # filename LOGGER.info(f'Downloading {s} to {f}...') torch.hub.download_url_to_file(s, f) - Path(root).mkdir(parents=True, exist_ok=True) # create root - ZipFile(f).extractall(path=root) # unzip + Path(DATASETS_DIR).mkdir(parents=True, exist_ok=True) # create root + ZipFile(f).extractall(path=DATASETS_DIR) # unzip Path(f).unlink() # remove zip r = None # success elif s.startswith('bash '): # bash script @@ -511,7 +515,7 @@ def check_dataset(data, autodownload=True): else: # python script r = exec(s, {'yaml': data}) # return None dt = f'({round(time.time() - t, 1)}s)' - s = f"success ✅ {dt}, saved to {colorstr('bold', root)}" if r in (0, None) else f"failure {dt} ❌" + s = f"success ✅ {dt}, saved to {colorstr('bold', DATASETS_DIR)}" if r in (0, None) else f"failure {dt} ❌" LOGGER.info(f"Dataset download {s}") check_font('Arial.ttf' if is_ascii(data['names']) else 'Arial.Unicode.ttf', progress=True) # download fonts return data # dictionary From 060837406542c5c65301b8fde641f4d92a1f395e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 17 Sep 2022 23:17:59 +0200 Subject: [PATCH 590/661] `Detect()` and `Segment()` fixes for CoreML and Paddle (#9458) * Detect() and Segment() fixes for CoreML and Paddle Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- models/yolo.py | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index 46039c36d7e1..0dca6353a356 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -64,17 +64,17 @@ def forward(self, x): if self.dynamic or self.grid[i].shape[2:4] != x[i].shape[2:4]: self.grid[i], self.anchor_grid[i] = self._make_grid(nx, ny, i) - y = x[i].clone() - y[..., :5 + self.nc].sigmoid_() - if self.inplace: - y[..., 0:2] = (y[..., 0:2] * 2 + self.grid[i]) * self.stride[i] # xy - y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh - else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953 - xy, wh, etc = y.split((2, 2, self.no - 4), 4) # tensor_split((2, 4, 5), 4) if torch 1.8.0 + if isinstance(self, Segment): # (boxes + masks) + xy, wh, conf, mask = x[i].split((2, 2, self.nc + 1, self.no - self.nc - 5), 4) + xy = (xy.sigmoid() * 2 + self.grid[i]) * self.stride[i] # xy + wh = (wh.sigmoid() * 2) ** 2 * self.anchor_grid[i] # wh + y = torch.cat((xy, wh, conf.sigmoid(), mask), 4) + else: # Detect (boxes only) + xy, wh, conf = x[i].sigmoid().split((2, 2, self.nc + 1), 4) xy = (xy * 2 + self.grid[i]) * self.stride[i] # xy wh = (wh * 2) ** 2 * self.anchor_grid[i] # wh - y = torch.cat((xy, wh, etc), 4) - z.append(y.view(bs, -1, self.no)) + y = torch.cat((xy, wh, conf), 4) + z.append(y.view(bs, self.na * nx * ny, self.no)) return x if self.training else (torch.cat(z, 1),) if self.export else (torch.cat(z, 1), x) From afb9860522e5023d64f4fd36fb78b6f26011f760 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 00:17:31 +0200 Subject: [PATCH 591/661] Add Paddle exports to benchmarks (#9459) * Add Paddle exports to benchmarks Signed-off-by: Glenn Jocher * Update plots.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- benchmarks.py | 2 +- models/common.py | 10 ++++------ utils/segment/plots.py | 4 ++-- 3 files changed, 7 insertions(+), 9 deletions(-) diff --git a/benchmarks.py b/benchmarks.py index 58e083c95d55..161af73c1eda 100644 --- a/benchmarks.py +++ b/benchmarks.py @@ -65,7 +65,7 @@ def run( model_type = type(attempt_load(weights, fuse=False)) # DetectionModel, SegmentationModel, etc. for i, (name, f, suffix, cpu, gpu) in export.export_formats().iterrows(): # index, (name, file, suffix, CPU, GPU) try: - assert i not in (9, 10, 11), 'inference not supported' # Edge TPU, TF.js and Paddle are unsupported + assert i not in (9, 10), 'inference not supported' # Edge TPU and TF.js are unsupported assert i != 5 or platform.system() == 'Darwin', 'inference only supported on macOS>=10.13' # CoreML if 'cpu' in device.type: assert cpu, 'inference not supported on CPU' diff --git a/models/common.py b/models/common.py index 9c08120fe7f6..2b61307ad46b 100644 --- a/models/common.py +++ b/models/common.py @@ -460,8 +460,8 @@ def wrap_frozen_graph(gd, inputs, outputs): if cuda: config.enable_use_gpu(memory_pool_init_size_mb=2048, device_id=0) predictor = pdi.create_predictor(config) - input_names = predictor.get_input_names() - input_handle = predictor.get_input_handle(input_names[0]) + input_handle = predictor.get_input_handle(predictor.get_input_names()[0]) + output_names = predictor.get_output_names() else: raise NotImplementedError(f'ERROR: {w} is not a supported format') @@ -517,12 +517,10 @@ def forward(self, im, augment=False, visualize=False): k = 'var_' + str(sorted(int(k.replace('var_', '')) for k in y)[-1]) # output key y = y[k] # output elif self.paddle: # PaddlePaddle - im = im.cpu().numpy().astype("float32") + im = im.cpu().numpy().astype(np.float32) self.input_handle.copy_from_cpu(im) self.predictor.run() - output_names = self.predictor.get_output_names() - output_handle = self.predictor.get_output_handle(output_names[0]) - y = output_handle.copy_to_cpu() + y = [self.predictor.get_output_handle(x).copy_to_cpu() for x in self.output_names] else: # TensorFlow (SavedModel, GraphDef, Lite, Edge TPU) im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) if self.saved_model: # SavedModel diff --git a/utils/segment/plots.py b/utils/segment/plots.py index e882c14390f0..9b90900b3772 100644 --- a/utils/segment/plots.py +++ b/utils/segment/plots.py @@ -99,9 +99,9 @@ def plot_images_and_masks(images, targets, masks, paths=None, fname='images.jpg' if mh != h or mw != w: mask = image_masks[j].astype(np.uint8) mask = cv2.resize(mask, (w, h)) - mask = mask.astype(np.bool) + mask = mask.astype(bool) else: - mask = image_masks[j].astype(np.bool) + mask = image_masks[j].astype(bool) with contextlib.suppress(Exception): im[y:y + h, x:x + w, :][mask] = im[y:y + h, x:x + w, :][mask] * 0.4 + np.array(color) * 0.6 annotator.fromarray(im) From e8a9c5ae41b53f756e46de1190831b14b53c3b24 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 00:57:48 +0200 Subject: [PATCH 592/661] Add `macos-latest` runner for CoreML benchmarks (#9453) * Add `macos-latest` runner for CoreML benchmarks Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- models/common.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/models/common.py b/models/common.py index 2b61307ad46b..825a4c4e2633 100644 --- a/models/common.py +++ b/models/common.py @@ -514,8 +514,7 @@ def forward(self, im, augment=False, visualize=False): conf, cls = y['confidence'].max(1), y['confidence'].argmax(1).astype(np.float) y = np.concatenate((box, conf.reshape(-1, 1), cls.reshape(-1, 1)), 1) else: - k = 'var_' + str(sorted(int(k.replace('var_', '')) for k in y)[-1]) # output key - y = y[k] # output + y = list(reversed(y.values())) # reversed for segmentation models (pred, proto) elif self.paddle: # PaddlePaddle im = im.cpu().numpy().astype(np.float32) self.input_handle.copy_from_cpu(im) From 8ae81a6c87ebbf6a25c4dc2c77ef443b1d84098a Mon Sep 17 00:00:00 2001 From: Junjie Zhang <46258221+Oswells@users.noreply.github.com> Date: Sun, 18 Sep 2022 18:27:43 +0800 Subject: [PATCH 593/661] Fix cutout bug (#9452) * fix cutout bug Signed-off-by: Junjie Zhang <46258221+Oswells@users.noreply.github.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Junjie Zhang <46258221+Oswells@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- utils/augmentations.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/augmentations.py b/utils/augmentations.py index a5587351f75b..f49110f43c6a 100644 --- a/utils/augmentations.py +++ b/utils/augmentations.py @@ -12,7 +12,7 @@ import torchvision.transforms as T import torchvision.transforms.functional as TF -from utils.general import LOGGER, check_version, colorstr, resample_segments, segment2box +from utils.general import LOGGER, check_version, colorstr, resample_segments, segment2box, xywhn2xyxy from utils.metrics import bbox_ioa IMAGENET_MEAN = 0.485, 0.456, 0.406 # RGB mean @@ -281,7 +281,7 @@ def cutout(im, labels, p=0.5): # return unobscured labels if len(labels) and s > 0.03: box = np.array([xmin, ymin, xmax, ymax], dtype=np.float32) - ioa = bbox_ioa(box, labels[:, 1:5]) # intersection over area + ioa = bbox_ioa(box, xywhn2xyxy(labels[:, 1:5], w, h)) # intersection over area labels = labels[ioa < 0.60] # remove >60% obscured labels return labels From 95cef1ae6b3bdf4ced616a2b6f3c9655803e9ea7 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 12:42:23 +0200 Subject: [PATCH 594/661] Optimize imports (#9464) * Optimize imports * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Reformat * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- segment/train.py | 2 -- utils/loggers/clearml/clearml_utils.py | 1 + utils/loggers/comet/hpo.py | 2 +- 3 files changed, 2 insertions(+), 3 deletions(-) diff --git a/segment/train.py b/segment/train.py index bda379176151..8abd0944551d 100644 --- a/segment/train.py +++ b/segment/train.py @@ -39,8 +39,6 @@ sys.path.append(str(ROOT)) # add ROOT to PATH ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative -import torch.nn.functional as F - import segment.val as validate # for end-of-epoch mAP from models.experimental import attempt_load from models.yolo import SegmentationModel diff --git a/utils/loggers/clearml/clearml_utils.py b/utils/loggers/clearml/clearml_utils.py index 1e136907367d..eb1c12ce6cac 100644 --- a/utils/loggers/clearml/clearml_utils.py +++ b/utils/loggers/clearml/clearml_utils.py @@ -11,6 +11,7 @@ try: import clearml from clearml import Dataset, Task + assert hasattr(clearml, '__version__') # verify package import not local dir except (ImportError, AssertionError): clearml = None diff --git a/utils/loggers/comet/hpo.py b/utils/loggers/comet/hpo.py index eab4df9978cf..7dd5c92e8de1 100644 --- a/utils/loggers/comet/hpo.py +++ b/utils/loggers/comet/hpo.py @@ -14,7 +14,7 @@ if str(ROOT) not in sys.path: sys.path.append(str(ROOT)) # add ROOT to PATH -from train import parse_opt, train +from train import train from utils.callbacks import Callbacks from utils.general import increment_path from utils.torch_utils import select_device From dc42e6ef2232979e6f0f606da670f42c6d59108c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 14:45:08 +0200 Subject: [PATCH 595/661] TensorRT SegmentationModel fix (#9465) * TensorRT SegmentationModel fix * TensorRT SegmentationModel fix * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * TensorRT SegmentationModel fix * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * TensorRT SegmentationModel fix * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * TensorRT SegmentationModel fix * TensorRT SegmentationModel fix * fix * sort output names * Update ci-testing.yml Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 23 ++++++++++++----------- models/common.py | 27 ++++++++++++++++----------- 2 files changed, 28 insertions(+), 22 deletions(-) diff --git a/export.py b/export.py index a575c73e375f..9955870e9e43 100644 --- a/export.py +++ b/export.py @@ -66,7 +66,7 @@ ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.experimental import attempt_load -from models.yolo import ClassificationModel, Detect +from models.yolo import ClassificationModel, Detect, DetectionModel, SegmentationModel from utils.dataloaders import LoadImages from utils.general import (LOGGER, Profile, check_dataset, check_img_size, check_requirements, check_version, check_yaml, colorstr, file_size, get_default_args, print_args, url2file, yaml_save) @@ -134,6 +134,15 @@ def export_onnx(model, im, file, opset, dynamic, simplify, prefix=colorstr('ONNX LOGGER.info(f'\n{prefix} starting export with onnx {onnx.__version__}...') f = file.with_suffix('.onnx') + output_names = ['output0', 'output1'] if isinstance(model, SegmentationModel) else ['output0'] + if dynamic: + dynamic = {'images': {0: 'batch', 2: 'height', 3: 'width'}} # shape(1,3,640,640) + if isinstance(model, SegmentationModel): + dynamic['output0'] = {0: 'batch', 1: 'anchors'} # shape(1,25200,85) + dynamic['output1'] = {0: 'batch', 2: 'mask_height', 3: 'mask_width'} # shape(1,32,160,160) + elif isinstance(model, DetectionModel): + dynamic['output0'] = {0: 'batch', 1: 'anchors'} # shape(1,25200,85) + torch.onnx.export( model.cpu() if dynamic else model, # --dynamic only compatible with cpu im.cpu() if dynamic else im, @@ -142,16 +151,8 @@ def export_onnx(model, im, file, opset, dynamic, simplify, prefix=colorstr('ONNX opset_version=opset, do_constant_folding=True, input_names=['images'], - output_names=['output'], - dynamic_axes={ - 'images': { - 0: 'batch', - 2: 'height', - 3: 'width'}, # shape(1,3,640,640) - 'output': { - 0: 'batch', - 1: 'anchors'} # shape(1,25200,85) - } if dynamic else None) + output_names=output_names, + dynamic_axes=dynamic or None) # Checks model_onnx = onnx.load(f) # load onnx model diff --git a/models/common.py b/models/common.py index 825a4c4e2633..d0bc65e02f91 100644 --- a/models/common.py +++ b/models/common.py @@ -390,18 +390,21 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, model = runtime.deserialize_cuda_engine(f.read()) context = model.create_execution_context() bindings = OrderedDict() + output_names = [] fp16 = False # default updated below dynamic = False - for index in range(model.num_bindings): - name = model.get_binding_name(index) - dtype = trt.nptype(model.get_binding_dtype(index)) - if model.binding_is_input(index): - if -1 in tuple(model.get_binding_shape(index)): # dynamic + for i in range(model.num_bindings): + name = model.get_binding_name(i) + dtype = trt.nptype(model.get_binding_dtype(i)) + if model.binding_is_input(i): + if -1 in tuple(model.get_binding_shape(i)): # dynamic dynamic = True - context.set_binding_shape(index, tuple(model.get_profile_shape(0, index)[2])) + context.set_binding_shape(i, tuple(model.get_profile_shape(0, i)[2])) if dtype == np.float16: fp16 = True - shape = tuple(context.get_binding_shape(index)) + else: # output + output_names.append(name) + shape = tuple(context.get_binding_shape(i)) im = torch.from_numpy(np.empty(shape, dtype=dtype)).to(device) bindings[name] = Binding(name, dtype, shape, im, int(im.data_ptr())) binding_addrs = OrderedDict((n, d.ptr) for n, d in bindings.items()) @@ -495,15 +498,17 @@ def forward(self, im, augment=False, visualize=False): y = list(self.executable_network([im]).values()) elif self.engine: # TensorRT if self.dynamic and im.shape != self.bindings['images'].shape: - i_in, i_out = (self.model.get_binding_index(x) for x in ('images', 'output')) - self.context.set_binding_shape(i_in, im.shape) # reshape if dynamic + i = self.model.get_binding_index('images') + self.context.set_binding_shape(i, im.shape) # reshape if dynamic self.bindings['images'] = self.bindings['images']._replace(shape=im.shape) - self.bindings['output'].data.resize_(tuple(self.context.get_binding_shape(i_out))) + for name in self.output_names: + i = self.model.get_binding_index(name) + self.bindings[name].data.resize_(tuple(self.context.get_binding_shape(i))) s = self.bindings['images'].shape assert im.shape == s, f"input size {im.shape} {'>' if self.dynamic else 'not equal to'} max model size {s}" self.binding_addrs['images'] = int(im.data_ptr()) self.context.execute_v2(list(self.binding_addrs.values())) - y = self.bindings['output'].data + y = [self.bindings[x].data for x in sorted(self.output_names)] elif self.coreml: # CoreML im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) im = Image.fromarray((im[0] * 255).astype('uint8')) From 4d50cd3469d75b18e99ce1e831ca024e3d25a2d7 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 15:02:04 +0200 Subject: [PATCH 596/661] `Conv()` dilation argument fix (#9466) Resolves https://github.com/ultralytics/yolov5/issues/9384 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- models/common.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/models/common.py b/models/common.py index d0bc65e02f91..33db74dcd9ae 100644 --- a/models/common.py +++ b/models/common.py @@ -232,7 +232,7 @@ class Focus(nn.Module): # Focus wh information into c-space def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True): # ch_in, ch_out, kernel, stride, padding, groups super().__init__() - self.conv = Conv(c1 * 4, c2, k, s, p, g, act) + self.conv = Conv(c1 * 4, c2, k, s, p, g, act=act) # self.contract = Contract(gain=2) def forward(self, x): # x(b,c,w,h) -> y(b,4c,w/2,h/2) @@ -245,8 +245,8 @@ class GhostConv(nn.Module): def __init__(self, c1, c2, k=1, s=1, g=1, act=True): # ch_in, ch_out, kernel, stride, groups super().__init__() c_ = c2 // 2 # hidden channels - self.cv1 = Conv(c1, c_, k, s, None, g, act) - self.cv2 = Conv(c_, c_, 5, 1, None, c_, act) + self.cv1 = Conv(c1, c_, k, s, None, g, act=act) + self.cv2 = Conv(c_, c_, 5, 1, None, c_, act=act) def forward(self, x): y = self.cv1(x) From 295c5e9d3ce70f5dbdb897c2da6a58e58f7c1125 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 16:13:22 +0200 Subject: [PATCH 597/661] Update ClassificationModel default training `imgsz=224` (#9469) Update ClassificationModel default training imgsz=224 To match classify/val.py and classify/predict.py Helps https://github.com/ultralytics/yolov5/issues/9462 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- classify/train.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/classify/train.py b/classify/train.py index 223367260bad..23c90e0a5274 100644 --- a/classify/train.py +++ b/classify/train.py @@ -3,7 +3,7 @@ Train a YOLOv5 classifier model on a classification dataset Usage - Single-GPU training: - $ python classify/train.py --model yolov5s-cls.pt --data imagenette160 --epochs 5 --img 128 + $ python classify/train.py --model yolov5s-cls.pt --data imagenette160 --epochs 5 --img 224 Usage - Multi-GPU DDP training: $ python -m torch.distributed.run --nproc_per_node 4 --master_port 1 classify/train.py --model yolov5s-cls.pt --data imagenet --epochs 5 --img 224 --device 0,1,2,3 @@ -272,7 +272,7 @@ def parse_opt(known=False): parser.add_argument('--data', type=str, default='imagenette160', help='cifar10, cifar100, mnist, imagenet, ...') parser.add_argument('--epochs', type=int, default=10, help='total training epochs') parser.add_argument('--batch-size', type=int, default=64, help='total batch size for all GPUs') - parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=128, help='train, val image size (pixels)') + parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=224, help='train, val image size (pixels)') parser.add_argument('--nosave', action='store_true', help='only save final checkpoint') parser.add_argument('--cache', type=str, nargs='?', const='ram', help='--cache images in "ram" (default) or "disk"') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') From ca9c993d6c3c9f59c44d28b22d8968709cd11693 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 16:15:25 +0200 Subject: [PATCH 598/661] =?UTF-8?q?Standardize=20warnings=20with=20`WARNIN?= =?UTF-8?q?G=20=20=E2=9A=A0=EF=B8=8F=20...`=20(#9467)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * Standardize warnings * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- benchmarks.py | 2 +- classify/train.py | 2 +- export.py | 2 +- hubconf.py | 2 +- segment/train.py | 2 +- segment/val.py | 6 +++--- train.py | 2 +- utils/__init__.py | 10 ++++++++-- utils/autoanchor.py | 4 ++-- utils/autobatch.py | 2 +- utils/dataloaders.py | 18 +++++++++--------- utils/general.py | 21 ++++++++------------- utils/loggers/__init__.py | 4 ++-- utils/metrics.py | 2 +- utils/segment/dataloaders.py | 2 +- utils/torch_utils.py | 2 +- val.py | 6 +++--- 17 files changed, 45 insertions(+), 44 deletions(-) diff --git a/benchmarks.py b/benchmarks.py index 161af73c1eda..b3b58eb3257c 100644 --- a/benchmarks.py +++ b/benchmarks.py @@ -91,7 +91,7 @@ def run( except Exception as e: if hard_fail: assert type(e) is AssertionError, f'Benchmark --hard-fail for {name}: {e}' - LOGGER.warning(f'WARNING: Benchmark failure for {name}: {e}') + LOGGER.warning(f'WARNING ⚠️ Benchmark failure for {name}: {e}') y.append([name, None, None, None]) # mAP, t_inference if pt_only and i == 0: break # break after PyTorch diff --git a/classify/train.py b/classify/train.py index 23c90e0a5274..178ebcdfff53 100644 --- a/classify/train.py +++ b/classify/train.py @@ -114,7 +114,7 @@ def train(opt, device): m = hub.list('ultralytics/yolov5') # + hub.list('pytorch/vision') # models raise ModuleNotFoundError(f'--model {opt.model} not found. Available models are: \n' + '\n'.join(m)) if isinstance(model, DetectionModel): - LOGGER.warning("WARNING: pass YOLOv5 classifier model with '-cls' suffix, i.e. '--model yolov5s-cls.pt'") + LOGGER.warning("WARNING ⚠️ pass YOLOv5 classifier model with '-cls' suffix, i.e. '--model yolov5s-cls.pt'") model = ClassificationModel(model=model, nc=nc, cutoff=opt.cutoff or 10) # convert to classification model reshape_classifier_output(model, nc) # update class count for m in model.modules(): diff --git a/export.py b/export.py index 9955870e9e43..ac9b13db8ec0 100644 --- a/export.py +++ b/export.py @@ -282,7 +282,7 @@ def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose if dynamic: if im.shape[0] <= 1: - LOGGER.warning(f"{prefix}WARNING: --dynamic model requires maximum --batch-size argument") + LOGGER.warning(f"{prefix}WARNING ⚠️ --dynamic model requires maximum --batch-size argument") profile = builder.create_optimization_profile() for inp in inputs: profile.set_shape(inp.name, (1, *im.shape[1:]), (max(1, im.shape[0] // 2), *im.shape[1:]), im.shape) diff --git a/hubconf.py b/hubconf.py index 2f05565629a5..4224760a4732 100644 --- a/hubconf.py +++ b/hubconf.py @@ -47,7 +47,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo model = DetectMultiBackend(path, device=device, fuse=autoshape) # detection model if autoshape: if model.pt and isinstance(model.model, ClassificationModel): - LOGGER.warning('WARNING: ⚠️ YOLOv5 v6.2 ClassificationModel is not yet AutoShape compatible. ' + LOGGER.warning('WARNING ⚠️ YOLOv5 v6.2 ClassificationModel is not yet AutoShape compatible. ' 'You must pass torch tensors in BCHW to this model, i.e. shape(1,3,224,224).') else: model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS diff --git a/segment/train.py b/segment/train.py index 8abd0944551d..5121c5fa784a 100644 --- a/segment/train.py +++ b/segment/train.py @@ -176,7 +176,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # DP mode if cuda and RANK == -1 and torch.cuda.device_count() > 1: - LOGGER.warning('WARNING: DP not recommended, use torch.distributed.run for best DDP Multi-GPU results.\n' + LOGGER.warning('WARNING ⚠️ DP not recommended, use torch.distributed.run for best DDP Multi-GPU results.\n' 'See Multi-GPU Tutorial at https://github.com/ultralytics/yolov5/issues/475 to get started.') model = torch.nn.DataParallel(model) diff --git a/segment/val.py b/segment/val.py index 138aa00aaed3..59ab76672a30 100644 --- a/segment/val.py +++ b/segment/val.py @@ -345,7 +345,7 @@ def run( pf = '%22s' + '%11i' * 2 + '%11.3g' * 8 # print format LOGGER.info(pf % ("all", seen, nt.sum(), *metrics.mean_results())) if nt.sum() == 0: - LOGGER.warning(f'WARNING: no labels found in {task} set, can not compute metrics without labels ⚠️') + LOGGER.warning(f'WARNING ⚠️ no labels found in {task} set, can not compute metrics without labels') # Print results per class if (verbose or (nc < 50 and not training)) and nc > 1 and len(stats): @@ -438,9 +438,9 @@ def main(opt): if opt.task in ('train', 'val', 'test'): # run normally if opt.conf_thres > 0.001: # https://github.com/ultralytics/yolov5/issues/1466 - LOGGER.info(f'WARNING: confidence threshold {opt.conf_thres} > 0.001 produces invalid results ⚠️') + LOGGER.warning(f'WARNING ⚠️ confidence threshold {opt.conf_thres} > 0.001 produces invalid results') if opt.save_hybrid: - LOGGER.info('WARNING: --save-hybrid will return high mAP from hybrid labels, not from predictions alone ⚠️') + LOGGER.warning('WARNING ⚠️ --save-hybrid returns high mAP from hybrid labels, not from predictions alone') run(**vars(opt)) else: diff --git a/train.py b/train.py index 4eff6e5d645a..9efece250581 100644 --- a/train.py +++ b/train.py @@ -173,7 +173,7 @@ def train(hyp, opt, device, callbacks): # hyp is path/to/hyp.yaml or hyp dictio # DP mode if cuda and RANK == -1 and torch.cuda.device_count() > 1: - LOGGER.warning('WARNING: DP not recommended, use torch.distributed.run for best DDP Multi-GPU results.\n' + LOGGER.warning('WARNING ⚠️ DP not recommended, use torch.distributed.run for best DDP Multi-GPU results.\n' 'See Multi-GPU Tutorial at https://github.com/ultralytics/yolov5/issues/475 to get started.') model = torch.nn.DataParallel(model) diff --git a/utils/__init__.py b/utils/__init__.py index 46225c2208ce..8403a6149827 100644 --- a/utils/__init__.py +++ b/utils/__init__.py @@ -4,9 +4,15 @@ """ import contextlib +import platform import threading +def emojis(str=''): + # Return platform-dependent emoji-safe version of string + return str.encode().decode('ascii', 'ignore') if platform.system() == 'Windows' else str + + class TryExcept(contextlib.ContextDecorator): # YOLOv5 TryExcept class. Usage: @TryExcept() decorator or 'with TryExcept():' context manager def __init__(self, msg=''): @@ -17,7 +23,7 @@ def __enter__(self): def __exit__(self, exc_type, value, traceback): if value: - print(f'{self.msg}{value}') + print(emojis(f'{self.msg}{value}')) return True @@ -38,7 +44,7 @@ def notebook_init(verbose=True): import os import shutil - from utils.general import check_font, check_requirements, emojis, is_colab + from utils.general import check_font, check_requirements, is_colab from utils.torch_utils import select_device # imports check_requirements(('psutil', 'IPython')) diff --git a/utils/autoanchor.py b/utils/autoanchor.py index 0b49ab3319c0..7e7e9985d68a 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -122,7 +122,7 @@ def print_results(k, verbose=True): # Filter i = (wh0 < 3.0).any(1).sum() if i: - LOGGER.info(f'{PREFIX}WARNING: Extremely small objects found: {i} of {len(wh0)} labels are < 3 pixels in size') + LOGGER.info(f'{PREFIX}WARNING ⚠️ Extremely small objects found: {i} of {len(wh0)} labels are <3 pixels in size') wh = wh0[(wh0 >= 2.0).any(1)].astype(np.float32) # filter > 2 pixels # wh = wh * (npr.rand(wh.shape[0], 1) * 0.9 + 0.1) # multiply by random scale 0-1 @@ -134,7 +134,7 @@ def print_results(k, verbose=True): k = kmeans(wh / s, n, iter=30)[0] * s # points assert n == len(k) # kmeans may return fewer points than requested if wh is insufficient or too similar except Exception: - LOGGER.warning(f'{PREFIX}WARNING: switching strategies from kmeans to random init') + LOGGER.warning(f'{PREFIX}WARNING ⚠️ switching strategies from kmeans to random init') k = np.sort(npr.rand(n * 2)).reshape(n, 2) * img_size # random init wh, wh0 = (torch.tensor(x, dtype=torch.float32) for x in (wh, wh0)) k = print_results(k, verbose=False) diff --git a/utils/autobatch.py b/utils/autobatch.py index 3204fd26fc41..49435f51a244 100644 --- a/utils/autobatch.py +++ b/utils/autobatch.py @@ -65,7 +65,7 @@ def autobatch(model, imgsz=640, fraction=0.8, batch_size=16): b = batch_sizes[max(i - 1, 0)] # select prior safe point if b < 1 or b > 1024: # b outside of safe range b = batch_size - LOGGER.warning(f'{prefix}WARNING: ⚠️ CUDA anomaly detected, recommend restart environment and retry command.') + LOGGER.warning(f'{prefix}WARNING ⚠️ CUDA anomaly detected, recommend restart environment and retry command.') fraction = np.polyval(p, b) / t # actual fraction predicted LOGGER.info(f'{prefix}Using batch-size {b} for {d} {t * fraction:.2f}G/{t:.2f}G ({fraction * 100:.0f}%) ✅') diff --git a/utils/dataloaders.py b/utils/dataloaders.py index c04be853c580..5c3460eb0d6e 100644 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -116,7 +116,7 @@ def create_dataloader(path, prefix='', shuffle=False): if rect and shuffle: - LOGGER.warning('WARNING: --rect is incompatible with DataLoader shuffle, setting shuffle=False') + LOGGER.warning('WARNING ⚠️ --rect is incompatible with DataLoader shuffle, setting shuffle=False') shuffle = False with torch_distributed_zero_first(rank): # init dataset *.cache only once if DDP dataset = LoadImagesAndLabels( @@ -328,7 +328,7 @@ def __init__(self, sources='streams.txt', img_size=640, stride=32, auto=True, tr self.auto = auto and self.rect self.transforms = transforms # optional if not self.rect: - LOGGER.warning('WARNING: Stream shapes differ. For optimal performance supply similarly-shaped streams.') + LOGGER.warning('WARNING ⚠️ Stream shapes differ. For optimal performance supply similarly-shaped streams.') def update(self, i, cap, stream): # Read stream `i` frames in daemon thread @@ -341,7 +341,7 @@ def update(self, i, cap, stream): if success: self.imgs[i] = im else: - LOGGER.warning('WARNING: Video stream unresponsive, please check your IP camera connection.') + LOGGER.warning('WARNING ⚠️ Video stream unresponsive, please check your IP camera connection.') self.imgs[i] = np.zeros_like(self.imgs[i]) cap.open(stream) # re-open stream if signal was lost time.sleep(0.0) # wait time @@ -543,7 +543,7 @@ def cache_labels(self, path=Path('./labels.cache'), prefix=''): if msgs: LOGGER.info('\n'.join(msgs)) if nf == 0: - LOGGER.warning(f'{prefix}WARNING: No labels found in {path}. {HELP_URL}') + LOGGER.warning(f'{prefix}WARNING ⚠️ No labels found in {path}. {HELP_URL}') x['hash'] = get_hash(self.label_files + self.im_files) x['results'] = nf, nm, ne, nc, len(self.im_files) x['msgs'] = msgs # warnings @@ -553,7 +553,7 @@ def cache_labels(self, path=Path('./labels.cache'), prefix=''): path.with_suffix('.cache.npy').rename(path) # remove .npy suffix LOGGER.info(f'{prefix}New cache created: {path}') except Exception as e: - LOGGER.warning(f'{prefix}WARNING: Cache directory {path.parent} is not writeable: {e}') # not writeable + LOGGER.warning(f'{prefix}WARNING ⚠️ Cache directory {path.parent} is not writeable: {e}') # not writeable return x def __len__(self): @@ -917,7 +917,7 @@ def verify_image_label(args): f.seek(-2, 2) if f.read() != b'\xff\xd9': # corrupt JPEG ImageOps.exif_transpose(Image.open(im_file)).save(im_file, 'JPEG', subsampling=0, quality=100) - msg = f'{prefix}WARNING: {im_file}: corrupt JPEG restored and saved' + msg = f'{prefix}WARNING ⚠️ {im_file}: corrupt JPEG restored and saved' # verify labels if os.path.isfile(lb_file): @@ -939,7 +939,7 @@ def verify_image_label(args): lb = lb[i] # remove duplicates if segments: segments = [segments[x] for x in i] - msg = f'{prefix}WARNING: {im_file}: {nl - len(i)} duplicate labels removed' + msg = f'{prefix}WARNING ⚠️ {im_file}: {nl - len(i)} duplicate labels removed' else: ne = 1 # label empty lb = np.zeros((0, 5), dtype=np.float32) @@ -949,7 +949,7 @@ def verify_image_label(args): return im_file, lb, shape, segments, nm, nf, ne, nc, msg except Exception as e: nc = 1 - msg = f'{prefix}WARNING: {im_file}: ignoring corrupt image/label: {e}' + msg = f'{prefix}WARNING ⚠️ {im_file}: ignoring corrupt image/label: {e}' return [None, None, None, None, nm, nf, ne, nc, msg] @@ -1012,7 +1012,7 @@ def _hub_ops(self, f, max_dim=1920): im = im.resize((int(im.width * r), int(im.height * r))) im.save(f_new, 'JPEG', quality=50, optimize=True) # save except Exception as e: # use OpenCV - print(f'WARNING: HUB ops PIL failure {f}: {e}') + LOGGER.info(f'WARNING ⚠️ HUB ops PIL failure {f}: {e}') im = cv2.imread(f) im_height, im_width = im.shape[:2] r = max_dim / max(im_height, im_width) # ratio diff --git a/utils/general.py b/utils/general.py index 38856b6bfa1c..fd0b4090a0fa 100644 --- a/utils/general.py +++ b/utils/general.py @@ -34,7 +34,7 @@ import torchvision import yaml -from utils import TryExcept +from utils import TryExcept, emojis from utils.downloads import gsutil_getsize from utils.metrics import box_iou, fitness @@ -248,11 +248,6 @@ def get_latest_run(search_dir='.'): return max(last_list, key=os.path.getctime) if last_list else '' -def emojis(str=''): - # Return platform-dependent emoji-safe version of string - return str.encode().decode('ascii', 'ignore') if platform.system() == 'Windows' else str - - def file_age(path=__file__): # Return days since last file update dt = (datetime.now() - datetime.fromtimestamp(Path(path).stat().st_mtime)) # delta @@ -333,7 +328,7 @@ def check_version(current='0.0.0', minimum='0.0.0', name='version ', pinned=Fals # Check version vs. required version current, minimum = (pkg.parse_version(x) for x in (current, minimum)) result = (current == minimum) if pinned else (current >= minimum) # bool - s = f'WARNING: ⚠️ {name}{minimum} is required by YOLOv5, but {name}{current} is currently installed' # string + s = f'WARNING ⚠️ {name}{minimum} is required by YOLOv5, but {name}{current} is currently installed' # string if hard: assert result, emojis(s) # assert min requirements met if verbose and not result: @@ -373,7 +368,7 @@ def check_requirements(requirements=ROOT / 'requirements.txt', exclude=(), insta f"{prefix} ⚠️ {colorstr('bold', 'Restart runtime or rerun command for updates to take effect')}\n" LOGGER.info(s) except Exception as e: - LOGGER.warning(f'{prefix} {e}') + LOGGER.warning(f'{prefix} ❌ {e}') def check_img_size(imgsz, s=32, floor=0): @@ -384,7 +379,7 @@ def check_img_size(imgsz, s=32, floor=0): imgsz = list(imgsz) # convert to list if tuple new_size = [max(make_divisible(x, int(s)), floor) for x in imgsz] if new_size != imgsz: - LOGGER.warning(f'WARNING: --img-size {imgsz} must be multiple of max stride {s}, updating to {new_size}') + LOGGER.warning(f'WARNING ⚠️ --img-size {imgsz} must be multiple of max stride {s}, updating to {new_size}') return new_size @@ -399,7 +394,7 @@ def check_imshow(): cv2.waitKey(1) return True except Exception as e: - LOGGER.warning(f'WARNING: Environment does not support cv2.imshow() or PIL Image.show() image displays\n{e}') + LOGGER.warning(f'WARNING ⚠️ Environment does not support cv2.imshow() or PIL Image.show() image displays\n{e}') return False @@ -589,9 +584,9 @@ def download_one(url, dir): if success: break elif i < retry: - LOGGER.warning(f'Download failure, retrying {i + 1}/{retry} {url}...') + LOGGER.warning(f'⚠️ Download failure, retrying {i + 1}/{retry} {url}...') else: - LOGGER.warning(f'Failed to download {url}...') + LOGGER.warning(f'❌ Failed to download {url}...') if unzip and success and f.suffix in ('.zip', '.tar', '.gz'): LOGGER.info(f'Unzipping {f}...') @@ -908,7 +903,7 @@ def non_max_suppression( output[xi] = x[i] if (time.time() - t) > time_limit: - LOGGER.warning(f'WARNING: NMS time limit {time_limit:.3f}s exceeded') + LOGGER.warning(f'WARNING ⚠️ NMS time limit {time_limit:.3f}s exceeded') break # time limit exceeded return output diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index f29debb76907..941d09e19e2d 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -11,7 +11,7 @@ import torch from torch.utils.tensorboard import SummaryWriter -from utils.general import colorstr, cv2 +from utils.general import LOGGER, colorstr, cv2 from utils.loggers.clearml.clearml_utils import ClearmlLogger from utils.loggers.wandb.wandb_utils import WandbLogger from utils.plots import plot_images, plot_labels, plot_results @@ -393,7 +393,7 @@ def log_tensorboard_graph(tb, model, imgsz=(640, 640)): warnings.simplefilter('ignore') # suppress jit trace warning tb.add_graph(torch.jit.trace(de_parallel(model), im, strict=False), []) except Exception as e: - print(f'WARNING: TensorBoard graph visualization failure {e}') + LOGGER.warning(f'WARNING ⚠️ TensorBoard graph visualization failure {e}') def web_project_name(project): diff --git a/utils/metrics.py b/utils/metrics.py index 021a46ce5d37..ed611d7d38fa 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -186,7 +186,7 @@ def tp_fp(self): # fn = self.matrix.sum(0) - tp # false negatives (missed detections) return tp[:-1], fp[:-1] # remove background class - @TryExcept('WARNING: ConfusionMatrix plot failure: ') + @TryExcept('WARNING ⚠️ ConfusionMatrix plot failure: ') def plot(self, normalize=True, save_dir='', names=()): import seaborn as sn diff --git a/utils/segment/dataloaders.py b/utils/segment/dataloaders.py index f6fe642d077f..d137caa5ab27 100644 --- a/utils/segment/dataloaders.py +++ b/utils/segment/dataloaders.py @@ -37,7 +37,7 @@ def create_dataloader(path, mask_downsample_ratio=1, overlap_mask=False): if rect and shuffle: - LOGGER.warning('WARNING: --rect is incompatible with DataLoader shuffle, setting shuffle=False') + LOGGER.warning('WARNING ⚠️ --rect is incompatible with DataLoader shuffle, setting shuffle=False') shuffle = False with torch_distributed_zero_first(rank): # init dataset *.cache only once if DDP dataset = LoadImagesAndLabelsAndMasks( diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 8a3366ca3e27..9f257d06ac60 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -47,7 +47,7 @@ def smartCrossEntropyLoss(label_smoothing=0.0): if check_version(torch.__version__, '1.10.0'): return nn.CrossEntropyLoss(label_smoothing=label_smoothing) if label_smoothing > 0: - LOGGER.warning(f'WARNING: label smoothing {label_smoothing} requires torch>=1.10.0') + LOGGER.warning(f'WARNING ⚠️ label smoothing {label_smoothing} requires torch>=1.10.0') return nn.CrossEntropyLoss() diff --git a/val.py b/val.py index e003d2144b7f..3ab4bc3fdb58 100644 --- a/val.py +++ b/val.py @@ -282,7 +282,7 @@ def run( pf = '%22s' + '%11i' * 2 + '%11.3g' * 4 # print format LOGGER.info(pf % ('all', seen, nt.sum(), mp, mr, map50, map)) if nt.sum() == 0: - LOGGER.warning(f'WARNING: no labels found in {task} set, can not compute metrics without labels ⚠️') + LOGGER.warning(f'WARNING ⚠️ no labels found in {task} set, can not compute metrics without labels') # Print results per class if (verbose or (nc < 50 and not training)) and nc > 1 and len(stats): @@ -374,9 +374,9 @@ def main(opt): if opt.task in ('train', 'val', 'test'): # run normally if opt.conf_thres > 0.001: # https://github.com/ultralytics/yolov5/issues/1466 - LOGGER.info(f'WARNING: confidence threshold {opt.conf_thres} > 0.001 produces invalid results ⚠️') + LOGGER.info(f'WARNING ⚠️ confidence threshold {opt.conf_thres} > 0.001 produces invalid results') if opt.save_hybrid: - LOGGER.info('WARNING: --save-hybrid will return high mAP from hybrid labels, not from predictions alone ⚠️') + LOGGER.info('WARNING ⚠️ --save-hybrid will return high mAP from hybrid labels, not from predictions alone') run(**vars(opt)) else: From 92b52424d468feb48c51c3dde173d5d2c606a44b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 17:34:34 +0200 Subject: [PATCH 599/661] TensorFlow macOS AutoUpdate (#9471) * TensorFlow macOS AutoUpdate * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 11 ++++++++--- requirements.txt | 2 +- 2 files changed, 9 insertions(+), 4 deletions(-) diff --git a/export.py b/export.py index ac9b13db8ec0..ae292afe06f6 100644 --- a/export.py +++ b/export.py @@ -72,6 +72,8 @@ check_yaml, colorstr, file_size, get_default_args, print_args, url2file, yaml_save) from utils.torch_utils import select_device, smart_inference_mode +MACOS = platform.system() == 'Darwin' # macOS environment + def export_formats(): # YOLOv5 export formats @@ -224,7 +226,7 @@ def export_coreml(model, im, file, int8, half, prefix=colorstr('CoreML:')): ct_model = ct.convert(ts, inputs=[ct.ImageType('image', shape=im.shape, scale=1 / 255, bias=[0, 0, 0])]) bits, mode = (8, 'kmeans_lut') if int8 else (16, 'linear') if half else (32, None) if bits < 32: - if platform.system() == 'Darwin': # quantization only supported on macOS + if MACOS: # quantization only supported on macOS with warnings.catch_warnings(): warnings.filterwarnings("ignore", category=DeprecationWarning) # suppress numpy==1.20 float warning ct_model = ct.models.neural_network.quantization_utils.quantize_weights(ct_model, bits, mode) @@ -310,8 +312,11 @@ def export_saved_model(model, keras=False, prefix=colorstr('TensorFlow SavedModel:')): # YOLOv5 TensorFlow SavedModel export - check_requirements('tensorflow' if torch.cuda.is_available() else 'tensorflow-cpu') - import tensorflow as tf + try: + import tensorflow as tf + except Exception: + check_requirements(f"tensorflow{'' if torch.cuda.is_available() else '-macos' if MACOS else '-cpu'}") + import tensorflow as tf from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2 from models.tf import TFModel diff --git a/requirements.txt b/requirements.txt index 44fe1ce697b7..835346f218a4 100644 --- a/requirements.txt +++ b/requirements.txt @@ -30,7 +30,7 @@ seaborn>=0.11.0 # nvidia-pyindex # TensorRT export # nvidia-tensorrt # TensorRT export # scikit-learn==0.19.2 # CoreML quantization -# tensorflow>=2.4.1 # TFLite export (or tensorflow-cpu, tensorflow-aarch64) +# tensorflow>=2.4.1 # TF exports (-cpu, -aarch64, -macos) # tensorflowjs>=3.9.0 # TF.js export # openvino-dev # OpenVINO export From 120e27e38efd4351b5e5bb5d735635f4cbf1bc86 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 19:34:10 +0200 Subject: [PATCH 600/661] `classify/predict --save-txt` fix (#9478) Classify --save-txt Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- classify/predict.py | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 878cf48b6fef..4857c69766e7 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -119,13 +119,15 @@ def run( for i, prob in enumerate(pred): # per image seen += 1 if webcam: # batch_size >= 1 - p, im0 = path[i], im0s[i].copy() + p, im0, frame = path[i], im0s[i].copy(), dataset.count s += f'{i}: ' else: - p, im0 = path, im0s.copy() + p, im0, frame = path, im0s.copy(), getattr(dataset, 'frame', 0) p = Path(p) # to Path save_path = str(save_dir / p.name) # im.jpg + txt_path = str(save_dir / 'labels' / p.stem) + ('' if dataset.mode == 'image' else f'_{frame}') # im.txt + s += '%gx%g ' % im.shape[2:] # print string annotator = Annotator(im0, example=str(names), pil=True) @@ -134,9 +136,12 @@ def run( s += f"{', '.join(f'{names[j]} {prob[j]:.2f}' for j in top5i)}, " # Write results + text = '\n'.join(f'{prob[j]:.2f} {names[j]}' for j in top5i) if save_img or view_img: # Add bbox to image - text = '\n'.join(f'{prob[j]:.2f} {names[j]}' for j in top5i) annotator.text((32, 32), text, txt_color=(255, 255, 255)) + if save_txt: # Write to file + with open(f'{txt_path}.txt', 'a') as f: + f.write(text + '\n') # Stream results im0 = annotator.result() @@ -188,7 +193,7 @@ def parse_opt(): parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[224], help='inference size h,w') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') parser.add_argument('--view-img', action='store_true', help='show results') - parser.add_argument('--save-txt', action='store_true', help='save results to *.txt') + parser.add_argument('--save-txt', action='store_false', help='save results to *.txt') parser.add_argument('--nosave', action='store_true', help='do not save images/videos') parser.add_argument('--augment', action='store_true', help='augmented inference') parser.add_argument('--visualize', action='store_true', help='visualize features') From fda8aa551d0b732153c2e0848dd6abd887a41cd1 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 18 Sep 2022 19:52:46 +0200 Subject: [PATCH 601/661] TensorFlow SegmentationModel support (#9472) * TensorFlow SegmentationModel support * TensorFlow SegmentationModel support * TensorFlow SegmentationModel support * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * TFLite fixes * GraphDef fixes * Update ci-testing.yml Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- .github/workflows/ci-testing.yml | 2 +- export.py | 2 +- models/common.py | 29 ++++++++++++++++++++--------- models/tf.py | 15 ++++++++------- 4 files changed, 30 insertions(+), 18 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index 537ba96e7225..fffc92d1b72f 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -43,7 +43,7 @@ jobs: python benchmarks.py --data coco128.yaml --weights ${{ matrix.model }}.pt --img 320 --hard-fail 0.29 - name: Benchmark SegmentationModel run: | - python benchmarks.py --data coco128-seg.yaml --weights ${{ matrix.model }}-seg.pt --img 320 + python benchmarks.py --data coco128-seg.yaml --weights ${{ matrix.model }}-seg.pt --img 320 --hard-fail 0.22 Tests: timeout-minutes: 60 diff --git a/export.py b/export.py index ae292afe06f6..fe4e53d06cc3 100644 --- a/export.py +++ b/export.py @@ -341,7 +341,7 @@ def export_saved_model(model, m = m.get_concrete_function(spec) frozen_func = convert_variables_to_constants_v2(m) tfm = tf.Module() - tfm.__call__ = tf.function(lambda x: frozen_func(x)[:4] if tf_nms else frozen_func(x)[0], [spec]) + tfm.__call__ = tf.function(lambda x: frozen_func(x)[:4] if tf_nms else frozen_func(x), [spec]) tfm.__call__(im) tf.saved_model.save(tfm, f, diff --git a/models/common.py b/models/common.py index 33db74dcd9ae..fac95a82fdb9 100644 --- a/models/common.py +++ b/models/common.py @@ -427,10 +427,17 @@ def wrap_frozen_graph(gd, inputs, outputs): ge = x.graph.as_graph_element return x.prune(tf.nest.map_structure(ge, inputs), tf.nest.map_structure(ge, outputs)) + def gd_outputs(gd): + name_list, input_list = [], [] + for node in gd.node: # tensorflow.core.framework.node_def_pb2.NodeDef + name_list.append(node.name) + input_list.extend(node.input) + return sorted(f'{x}:0' for x in list(set(name_list) - set(input_list)) if not x.startswith('NoOp')) + gd = tf.Graph().as_graph_def() # TF GraphDef with open(w, 'rb') as f: gd.ParseFromString(f.read()) - frozen_func = wrap_frozen_graph(gd, inputs="x:0", outputs="Identity:0") + frozen_func = wrap_frozen_graph(gd, inputs="x:0", outputs=gd_outputs(gd)) elif tflite or edgetpu: # https://www.tensorflow.org/lite/guide/python#install_tensorflow_lite_for_python try: # https://coral.ai/docs/edgetpu/tflite-python/#update-existing-tf-lite-code-for-the-edge-tpu from tflite_runtime.interpreter import Interpreter, load_delegate @@ -528,22 +535,26 @@ def forward(self, im, augment=False, visualize=False): else: # TensorFlow (SavedModel, GraphDef, Lite, Edge TPU) im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) if self.saved_model: # SavedModel - y = (self.model(im, training=False) if self.keras else self.model(im)).numpy() + y = self.model(im, training=False) if self.keras else self.model(im) elif self.pb: # GraphDef - y = self.frozen_func(x=self.tf.constant(im)).numpy() + y = self.frozen_func(x=self.tf.constant(im)) else: # Lite or Edge TPU - input, output = self.input_details[0], self.output_details[0] + input = self.input_details[0] int8 = input['dtype'] == np.uint8 # is TFLite quantized uint8 model if int8: scale, zero_point = input['quantization'] im = (im / scale + zero_point).astype(np.uint8) # de-scale self.interpreter.set_tensor(input['index'], im) self.interpreter.invoke() - y = self.interpreter.get_tensor(output['index']) - if int8: - scale, zero_point = output['quantization'] - y = (y.astype(np.float32) - zero_point) * scale # re-scale - y[..., :4] *= [w, h, w, h] # xywh normalized to pixels + y = [] + for output in self.output_details: + x = self.interpreter.get_tensor(output['index']) + if int8: + scale, zero_point = output['quantization'] + x = (x.astype(np.float32) - zero_point) * scale # re-scale + y.append(x) + y = [x if isinstance(x, np.ndarray) else x.numpy() for x in y] + y[0][..., :4] *= [w, h, w, h] # xywh normalized to pixels if isinstance(y, (list, tuple)): return self.from_numpy(y[0]) if len(y) == 1 else [self.from_numpy(x) for x in y] diff --git a/models/tf.py b/models/tf.py index 8cce147059d3..ae58ca738e2e 100644 --- a/models/tf.py +++ b/models/tf.py @@ -299,15 +299,15 @@ def call(self, inputs): x[i] = tf.reshape(x[i], [-1, ny * nx, self.na, self.no]) if not self.training: # inference - y = tf.sigmoid(x[i]) + y = x[i] grid = tf.transpose(self.grid[i], [0, 2, 1, 3]) - 0.5 anchor_grid = tf.transpose(self.anchor_grid[i], [0, 2, 1, 3]) * 4 - xy = (y[..., 0:2] * 2 + grid) * self.stride[i] # xy - wh = y[..., 2:4] ** 2 * anchor_grid + xy = (tf.sigmoid(y[..., 0:2]) * 2 + grid) * self.stride[i] # xy + wh = tf.sigmoid(y[..., 2:4]) ** 2 * anchor_grid # Normalize xywh to 0-1 to reduce calibration error xy /= tf.constant([[self.imgsz[1], self.imgsz[0]]], dtype=tf.float32) wh /= tf.constant([[self.imgsz[1], self.imgsz[0]]], dtype=tf.float32) - y = tf.concat([xy, wh, y[..., 4:]], -1) + y = tf.concat([xy, wh, tf.sigmoid(y[..., 4:5 + self.nc]), y[..., 5 + self.nc:]], -1) z.append(tf.reshape(y, [-1, self.na * ny * nx, self.no])) return tf.transpose(x, [0, 2, 1, 3]) if self.training else (tf.concat(z, 1), x) @@ -333,8 +333,9 @@ def __init__(self, nc=80, anchors=(), nm=32, npr=256, ch=(), imgsz=(640, 640), w def call(self, x): p = self.proto(x[0]) + p = tf.transpose(p, [0, 3, 1, 2]) # from shape(1,160,160,32) to shape(1,32,160,160) x = self.detect(self, x) - return (x, p) if self.training else ((x[0], p),) + return (x, p) if self.training else (x[0], p) class TFProto(keras.layers.Layer): @@ -485,8 +486,8 @@ def predict(self, conf_thres, clip_boxes=False) return nms, x[1] - return x[0] # output only first tensor [1,6300,85] = [xywh, conf, class0, class1, ...] - # x = x[0][0] # [x(1,6300,85), ...] to x(6300,85) + return x # output [1,6300,85] = [xywh, conf, class0, class1, ...] + # x = x[0] # [x(1,6300,85), ...] to x(6300,85) # xywh = x[..., :4] # x(6300,4) boxes # conf = x[..., 4:5] # x(6300,1) confidences # cls = tf.reshape(tf.cast(tf.argmax(x[..., 5:], axis=1), tf.float32), (-1, 1)) # x(6300,1) classes From f038ad71729960facad54407e1b353b0e81242e2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 19 Sep 2022 12:18:55 +0200 Subject: [PATCH 602/661] AutoBatch report include reserved+allocated (#9491) May resolve https://github.com/ultralytics/yolov5/issues/9287#issuecomment-1250767031 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/autobatch.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/utils/autobatch.py b/utils/autobatch.py index 49435f51a244..bdeb91c3d2bd 100644 --- a/utils/autobatch.py +++ b/utils/autobatch.py @@ -19,7 +19,7 @@ def check_train_batch_size(model, imgsz=640, amp=True): def autobatch(model, imgsz=640, fraction=0.8, batch_size=16): - # Automatically estimate best batch size to use `fraction` of available CUDA memory + # Automatically estimate best YOLOv5 batch size to use `fraction` of available CUDA memory # Usage: # import torch # from utils.autobatch import autobatch @@ -67,6 +67,6 @@ def autobatch(model, imgsz=640, fraction=0.8, batch_size=16): b = batch_size LOGGER.warning(f'{prefix}WARNING ⚠️ CUDA anomaly detected, recommend restart environment and retry command.') - fraction = np.polyval(p, b) / t # actual fraction predicted + fraction = (np.polyval(p, b) + r + a) / t # actual fraction predicted LOGGER.info(f'{prefix}Using batch-size {b} for {d} {t * fraction:.2f}G/{t:.2f}G ({fraction * 100:.0f}%) ✅') return b From 868c0e9bbb45b031e7bfd73c6d3983bcce07b9c1 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 19 Sep 2022 13:31:24 +0200 Subject: [PATCH 603/661] Update Detect() grid init `for` loop (#9494) May resolve threaded inference issue in https://github.com/ultralytics/yolov5/pull/9425#issuecomment-1250802928 by avoiding memory sharing on init. Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- models/yolo.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/models/yolo.py b/models/yolo.py index 0dca6353a356..1d0da2a6e010 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -47,8 +47,8 @@ def __init__(self, nc=80, anchors=(), ch=(), inplace=True): # detection layer self.no = nc + 5 # number of outputs per anchor self.nl = len(anchors) # number of detection layers self.na = len(anchors[0]) // 2 # number of anchors - self.grid = [torch.empty(1)] * self.nl # init grid - self.anchor_grid = [torch.empty(1)] * self.nl # init anchor grid + self.grid = [torch.empty(0) for _ in range(self.nl)] # init grid + self.anchor_grid = [torch.empty(0) for _ in range(self.nl)] # init anchor grid self.register_buffer('anchors', torch.tensor(anchors).float().view(self.nl, -1, 2)) # shape(nl,na,2) self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv self.inplace = inplace # use inplace ops (e.g. slice assignment) From 11640698977724daf7982c9da398c2ee2f2b6e91 Mon Sep 17 00:00:00 2001 From: mucunwuxian Date: Mon, 19 Sep 2022 21:01:46 +0900 Subject: [PATCH 604/661] Accelerate video inference (#9487) * The following code is slow, "self.cap.set(cv2.CAP_PROP_POS_FRAMES, self.vid_stride * (self.frame + 1)) # read at vid_stride". * adjust... * Update dataloaders.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: Glenn Jocher --- utils/dataloaders.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 5c3460eb0d6e..5b03b4eb9759 100644 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -232,8 +232,9 @@ def __next__(self): if self.video_flag[self.count]: # Read video self.mode = 'video' - ret_val, im0 = self.cap.read() - self.cap.set(cv2.CAP_PROP_POS_FRAMES, self.vid_stride * (self.frame + 1)) # read at vid_stride + for _ in range(self.vid_stride): + self.cap.grab() + ret_val, im0 = self.cap.retrieve() while not ret_val: self.count += 1 self.cap.release() From 0b724c5b851b32bb3a8fbfab3cc2d68f93b4661e Mon Sep 17 00:00:00 2001 From: Dhruv Nair Date: Mon, 19 Sep 2022 11:26:19 -0400 Subject: [PATCH 605/661] Comet Image Logging Fix (#9498) fix issues with image logging --- utils/loggers/comet/__init__.py | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/utils/loggers/comet/__init__.py b/utils/loggers/comet/__init__.py index 4ee86dd70d6e..3b3142b002c5 100644 --- a/utils/loggers/comet/__init__.py +++ b/utils/loggers/comet/__init__.py @@ -22,6 +22,7 @@ comet_ml = None COMET_PROJECT_NAME = None +import PIL import torch import torchvision.transforms as T import yaml @@ -131,6 +132,8 @@ def __init__(self, opt, hyp, run_id=None, job_type="Training", **experiment_kwar else: self.iou_thres = IOU_THRES + self.log_parameters({"val_iou_threshold": self.iou_thres, "val_conf_threshold": self.conf_thres}) + self.comet_log_predictions = COMET_LOG_PREDICTIONS if self.opt.bbox_interval == -1: self.comet_log_prediction_interval = 1 if self.opt.epochs < 10 else self.opt.epochs // 10 @@ -139,6 +142,7 @@ def __init__(self, opt, hyp, run_id=None, job_type="Training", **experiment_kwar if self.comet_log_predictions: self.metadata_dict = {} + self.logged_image_names = [] self.comet_log_per_class_metrics = COMET_LOG_PER_CLASS_METRICS @@ -249,11 +253,12 @@ def log_predictions(self, image, labelsn, path, shape, predn): filtered_detections = detections[mask] filtered_labels = labelsn[mask] - processed_image = (image * 255).to(torch.uint8) - image_id = path.split("/")[-1].split(".")[0] image_name = f"{image_id}_curr_epoch_{self.experiment.curr_epoch}" - self.log_image(to_pil(processed_image), name=image_name) + if image_name not in self.logged_image_names: + native_scale_image = PIL.Image.open(path) + self.log_image(native_scale_image, name=image_name) + self.logged_image_names.append(image_name) metadata = [] for cls, *xyxy in filtered_labels.tolist(): From 0171198f38f36c55090c91c49a7b5abacd571324 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 19 Sep 2022 20:38:11 +0200 Subject: [PATCH 606/661] Fix visualization title bug (#9500) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/plots.py | 1 - 1 file changed, 1 deletion(-) diff --git a/utils/plots.py b/utils/plots.py index d8d5b225a774..51bb7d6c20af 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -204,7 +204,6 @@ def feature_visualization(x, module_type, stage, n=32, save_dir=Path('runs/detec ax[i].axis('off') LOGGER.info(f'Saving {f}... ({n}/{channels})') - plt.title('Features') plt.savefig(f, dpi=300, bbox_inches='tight') plt.close() np.save(str(f.with_suffix('.npy')), x[0].cpu().numpy()) # npy save From 63368e71d23e453ded1d94094a2b43b75c1a54fa Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E6=9B=BE=E9=80=B8=E5=A4=AB=EF=BC=88Zeng=20Yifu=EF=BC=89?= <41098760+Zengyf-CVer@users.noreply.github.com> Date: Tue, 20 Sep 2022 07:11:29 +0800 Subject: [PATCH 607/661] Add paddle tips (#9502) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * Update export.py Signed-off-by: 曾逸夫（Zeng Yifu） <41098760+Zengyf-CVer@users.noreply.github.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: 曾逸夫（Zeng Yifu） <41098760+Zengyf-CVer@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/export.py b/export.py index fe4e53d06cc3..04c2ed9c802d 100644 --- a/export.py +++ b/export.py @@ -596,10 +596,11 @@ def parse_opt(): parser.add_argument('--topk-all', type=int, default=100, help='TF.js NMS: topk for all classes to keep') parser.add_argument('--iou-thres', type=float, default=0.45, help='TF.js NMS: IoU threshold') parser.add_argument('--conf-thres', type=float, default=0.25, help='TF.js NMS: confidence threshold') - parser.add_argument('--include', - nargs='+', - default=['torchscript'], - help='torchscript, onnx, openvino, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs') + parser.add_argument( + '--include', + nargs='+', + default=['torchscript'], + help='torchscript, onnx, openvino, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle') opt = parser.parse_args() print_args(vars(opt)) return opt From 095f601d9d32ea0f0afd47554c068659939ecf4e Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 20 Sep 2022 12:22:02 +0200 Subject: [PATCH 608/661] Segmentation `polygons2masks_overlap()` in `np.int32` (#9493) * Segmentation `polygons2masks_overlap()` in `np.int32` May resolve https://github.com/ultralytics/yolov5/issues/9461 WARNING: Masks should be uint8 for fastest speed, change needs profiling results to determine impact. @AyushExel @Laughing-q Signed-off-by: Glenn Jocher * Update dataloaders.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/segment/dataloaders.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/utils/segment/dataloaders.py b/utils/segment/dataloaders.py index d137caa5ab27..49575f065752 100644 --- a/utils/segment/dataloaders.py +++ b/utils/segment/dataloaders.py @@ -308,7 +308,8 @@ def polygons2masks(img_size, polygons, color, downsample_ratio=1): def polygons2masks_overlap(img_size, segments, downsample_ratio=1): """Return a (640, 640) overlap mask.""" - masks = np.zeros((img_size[0] // downsample_ratio, img_size[1] // downsample_ratio), dtype=np.uint8) + masks = np.zeros((img_size[0] // downsample_ratio, img_size[1] // downsample_ratio), + dtype=np.int32 if len(segments) > 255 else np.uint8) areas = [] ms = [] for si in range(len(segments)): From f8b74631e50bcac1bef8a52283102a5feb7217a6 Mon Sep 17 00:00:00 2001 From: FeiGeChuanShu <774074168@qq.com> Date: Tue, 20 Sep 2022 19:04:45 +0800 Subject: [PATCH 609/661] Fix `random_perspective` param bug in segment (#9512) * fix random_perspective param bug when mosaic=False Signed-off-by: FeiGeChuanShu <774074168@qq.com> * Update dataloaders.py * Update dataloaders.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: FeiGeChuanShu <774074168@qq.com> Co-authored-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/segment/dataloaders.py | 19 ++++++++----------- 1 file changed, 8 insertions(+), 11 deletions(-) diff --git a/utils/segment/dataloaders.py b/utils/segment/dataloaders.py index 49575f065752..97ef8556068e 100644 --- a/utils/segment/dataloaders.py +++ b/utils/segment/dataloaders.py @@ -140,17 +140,14 @@ def __getitem__(self, index): labels[:, 1:] = xywhn2xyxy(labels[:, 1:], ratio[0] * w, ratio[1] * h, padw=pad[0], padh=pad[1]) if self.augment: - img, labels, segments = random_perspective( - img, - labels, - segments=segments, - degrees=hyp["degrees"], - translate=hyp["translate"], - scale=hyp["scale"], - shear=hyp["shear"], - perspective=hyp["perspective"], - return_seg=True, - ) + img, labels, segments = random_perspective(img, + labels, + segments=segments, + degrees=hyp["degrees"], + translate=hyp["translate"], + scale=hyp["scale"], + shear=hyp["shear"], + perspective=hyp["perspective"]) nl = len(labels) # number of labels if nl: From e233c038ed63780843446dd7bf00d5cc6a2711fe Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 20 Sep 2022 16:38:04 +0200 Subject: [PATCH 610/661] Remove `check_requirements('flatbuffers==1.12')` (#9514) * Remove `check_requirements('flatbuffers==1.12')` Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- export.py | 2 -- 1 file changed, 2 deletions(-) diff --git a/export.py b/export.py index 04c2ed9c802d..a2aa5e830c33 100644 --- a/export.py +++ b/export.py @@ -534,8 +534,6 @@ def run( if coreml: # CoreML f[4], _ = export_coreml(model, im, file, int8, half) if any((saved_model, pb, tflite, edgetpu, tfjs)): # TensorFlow formats - if int8 or edgetpu: # TFLite --int8 bug https://github.com/ultralytics/yolov5/issues/5707 - check_requirements('flatbuffers==1.12') # required before `import tensorflow` assert not tflite or not tfjs, 'TFLite and TF.js models must be exported separately, please pass only one type.' assert not isinstance(model, ClassificationModel), 'ClassificationModel export to TF formats not yet supported.' f[5], s_model = export_saved_model(model.cpu(), From bd35191033d52a9e48e6c8faaeaaa009243b988f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 20 Sep 2022 18:47:14 +0200 Subject: [PATCH 611/661] Fix TF Lite exports (#9517) * Update tf.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> From c0d97138456f2257f608c4120c8fd65abcf69326 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 20 Sep 2022 19:01:03 +0200 Subject: [PATCH 612/661] TFLite fix 2 (#9518) * TFLite fix 2 Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- models/tf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/tf.py b/models/tf.py index ae58ca738e2e..0520c30a96df 100644 --- a/models/tf.py +++ b/models/tf.py @@ -310,7 +310,7 @@ def call(self, inputs): y = tf.concat([xy, wh, tf.sigmoid(y[..., 4:5 + self.nc]), y[..., 5 + self.nc:]], -1) z.append(tf.reshape(y, [-1, self.na * ny * nx, self.no])) - return tf.transpose(x, [0, 2, 1, 3]) if self.training else (tf.concat(z, 1), x) + return tf.transpose(x, [0, 2, 1, 3]) if self.training else (tf.concat(z, 1),) @staticmethod def _make_grid(nx=20, ny=20): From 77dcf55168d59131f75b8187c6be27172eec00ec Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 20 Sep 2022 22:57:42 +0200 Subject: [PATCH 613/661] FROM nvcr.io/nvidia/pytorch:22.08-py3 (#9520) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/docker/Dockerfile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index 4b9367cc27db..764ee278c22e 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -3,7 +3,7 @@ # Image is CUDA-optimized for YOLOv5 single/multi-GPU training and inference # Start FROM NVIDIA PyTorch image https://ngc.nvidia.com/catalog/containers/nvidia:pytorch -FROM nvcr.io/nvidia/pytorch:22.07-py3 +FROM nvcr.io/nvidia/pytorch:22.08-py3 RUN rm -rf /opt/pytorch # remove 1.2GB dir # Downloads to user config dir From 6ebef288944ea3a8152f8e0c98a2aee0bd922144 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 21 Sep 2022 15:12:12 +0200 Subject: [PATCH 614/661] Remove scikit-learn constraint on coremltools 6.0 (#9530) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 835346f218a4..75e7cc9e94d3 100644 --- a/requirements.txt +++ b/requirements.txt @@ -29,7 +29,7 @@ seaborn>=0.11.0 # onnx-simplifier>=0.4.1 # ONNX simplifier # nvidia-pyindex # TensorRT export # nvidia-tensorrt # TensorRT export -# scikit-learn==0.19.2 # CoreML quantization +# scikit-learn # CoreML quantization # tensorflow>=2.4.1 # TF exports (-cpu, -aarch64, -macos) # tensorflowjs>=3.9.0 # TF.js export # openvino-dev # OpenVINO export From 499a6bf5736a1b78341dfd142bd7c82f71ebf459 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 21 Sep 2022 15:14:54 +0200 Subject: [PATCH 615/661] Update scikit-learn constraint per coremltools 6.0 (#9531) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 75e7cc9e94d3..17db73678fc1 100644 --- a/requirements.txt +++ b/requirements.txt @@ -29,7 +29,7 @@ seaborn>=0.11.0 # onnx-simplifier>=0.4.1 # ONNX simplifier # nvidia-pyindex # TensorRT export # nvidia-tensorrt # TensorRT export -# scikit-learn # CoreML quantization +# scikit-learn<=1.1.2 # CoreML quantization # tensorflow>=2.4.1 # TF exports (-cpu, -aarch64, -macos) # tensorflowjs>=3.9.0 # TF.js export # openvino-dev # OpenVINO export From db6847431b489a6b8d36c14f05e08970025d01a2 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 21 Sep 2022 17:55:25 +0200 Subject: [PATCH 616/661] Update `coremltools>=6.0` (#9532) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 17db73678fc1..55c1f2428e3f 100644 --- a/requirements.txt +++ b/requirements.txt @@ -24,7 +24,7 @@ pandas>=1.1.4 seaborn>=0.11.0 # Export -------------------------------------- -# coremltools>=5.2 # CoreML export +# coremltools>=6.0 # CoreML export # onnx>=1.9.0 # ONNX export # onnx-simplifier>=0.4.1 # ONNX simplifier # nvidia-pyindex # TensorRT export From 6f0284763b0f66467dc04e5a5d87e5a68d1d49cd Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 21 Sep 2022 19:53:26 +0200 Subject: [PATCH 617/661] Update albumentations (#9503) * Add `RandomResizedCrop(ratio)` * Update ratio * Update ratio * Update ratio * Update ratio * Update ratio * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Create augmentations.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update augmentations.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/augmentations.py | 27 +++++++++++++++------------ utils/dataloaders.py | 2 +- 2 files changed, 16 insertions(+), 13 deletions(-) diff --git a/utils/augmentations.py b/utils/augmentations.py index f49110f43c6a..7c8e0bcdede6 100644 --- a/utils/augmentations.py +++ b/utils/augmentations.py @@ -21,7 +21,7 @@ class Albumentations: # YOLOv5 Albumentations class (optional, only used if package is installed) - def __init__(self): + def __init__(self, size=640): self.transform = None prefix = colorstr('albumentations: ') try: @@ -29,6 +29,7 @@ def __init__(self): check_version(A.__version__, '1.0.3', hard=True) # version requirement T = [ + A.RandomResizedCrop(height=size, width=size, scale=(0.8, 1.0), ratio=(0.9, 1.11), p=0.0), A.Blur(p=0.01), A.MedianBlur(p=0.01), A.ToGray(p=0.01), @@ -303,15 +304,17 @@ def box_candidates(box1, box2, wh_thr=2, ar_thr=100, area_thr=0.1, eps=1e-16): return (w2 > wh_thr) & (h2 > wh_thr) & (w2 * h2 / (w1 * h1 + eps) > area_thr) & (ar < ar_thr) # candidates -def classify_albumentations(augment=True, - size=224, - scale=(0.08, 1.0), - hflip=0.5, - vflip=0.0, - jitter=0.4, - mean=IMAGENET_MEAN, - std=IMAGENET_STD, - auto_aug=False): +def classify_albumentations( + augment=True, + size=224, + scale=(0.08, 1.0), + ratio=(0.75, 1.0 / 0.75), # 0.75, 1.33 + hflip=0.5, + vflip=0.0, + jitter=0.4, + mean=IMAGENET_MEAN, + std=IMAGENET_STD, + auto_aug=False): # YOLOv5 classification Albumentations (optional, only used if package is installed) prefix = colorstr('albumentations: ') try: @@ -319,7 +322,7 @@ def classify_albumentations(augment=True, from albumentations.pytorch import ToTensorV2 check_version(A.__version__, '1.0.3', hard=True) # version requirement if augment: # Resize and crop - T = [A.RandomResizedCrop(height=size, width=size, scale=scale)] + T = [A.RandomResizedCrop(height=size, width=size, scale=scale, ratio=ratio)] if auto_aug: # TODO: implement AugMix, AutoAug & RandAug in albumentation LOGGER.info(f'{prefix}auto augmentations are currently not supported') @@ -338,7 +341,7 @@ def classify_albumentations(augment=True, return A.Compose(T) except ImportError: # package not installed, skip - pass + LOGGER.warning(f'{prefix}⚠️ not found, install with `pip install albumentations` (recommended)') except Exception as e: LOGGER.info(f'{prefix}{e}') diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 5b03b4eb9759..ee79bd0bc5a5 100644 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -404,7 +404,7 @@ def __init__(self, self.mosaic_border = [-img_size // 2, -img_size // 2] self.stride = stride self.path = path - self.albumentations = Albumentations() if augment else None + self.albumentations = Albumentations(size=img_size) if augment else None try: f = [] # image files From 999482b45163c1b808a187b02183f324a9c782cb Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 21 Sep 2022 23:08:52 +0200 Subject: [PATCH 618/661] import re (#9535) * import re Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- export.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/export.py b/export.py index a2aa5e830c33..e3cf392b0101 100644 --- a/export.py +++ b/export.py @@ -48,6 +48,7 @@ import json import os import platform +import re import subprocess import sys import time @@ -427,8 +428,6 @@ def export_edgetpu(file, prefix=colorstr('Edge TPU:')): def export_tfjs(file, prefix=colorstr('TensorFlow.js:')): # YOLOv5 TensorFlow.js export check_requirements('tensorflowjs') - import re - import tensorflowjs as tfjs LOGGER.info(f'\n{prefix} starting export with tensorflowjs {tfjs.__version__}...') From 489920ab30b217fed14d3ddd31c23e9afc5be238 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 22 Sep 2022 00:34:35 +0200 Subject: [PATCH 619/661] TF.js fix (#9536) * TF.js fix May resolve https://github.com/ultralytics/yolov5/issues/9534 Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update ci-testing.yml Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- models/tf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/models/tf.py b/models/tf.py index 0520c30a96df..1446d8841646 100644 --- a/models/tf.py +++ b/models/tf.py @@ -485,7 +485,7 @@ def predict(self, iou_thres, conf_thres, clip_boxes=False) - return nms, x[1] + return (nms,) return x # output [1,6300,85] = [xywh, conf, class0, class1, ...] # x = x[0] # [x(1,6300,85), ...] to x(6300,85) # xywh = x[..., :4] # x(6300,4) boxes From b25d5a75f2c89aace5cae342f3fe29dfdd46e401 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 22 Sep 2022 23:23:40 +0200 Subject: [PATCH 620/661] Refactor dataset batch-size (#9551) --- classify/predict.py | 3 +-- detect.py | 3 +-- segment/predict.py | 3 +-- 3 files changed, 3 insertions(+), 6 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 4857c69766e7..ef59ff6f550a 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -91,10 +91,9 @@ def run( if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) - bs = len(dataset) # batch_size else: dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) - bs = 1 # batch_size + bs = len(dataset) # batch_size vid_path, vid_writer = [None] * bs, [None] * bs # Run inference diff --git a/detect.py b/detect.py index 310d169281bf..4015b9ae0d7f 100644 --- a/detect.py +++ b/detect.py @@ -99,10 +99,9 @@ def run( if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) - bs = len(dataset) # batch_size else: dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) - bs = 1 # batch_size + bs = len(dataset) # batch_size vid_path, vid_writer = [None] * bs, [None] * bs # Run inference diff --git a/segment/predict.py b/segment/predict.py index ba4cf2905255..2ea6bd9327e0 100644 --- a/segment/predict.py +++ b/segment/predict.py @@ -101,10 +101,9 @@ def run( if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) - bs = len(dataset) # batch_size else: dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) - bs = 1 # batch_size + bs = len(dataset) # batch_size vid_path, vid_writer = [None] * bs, [None] * bs # Run inference From 30fa9b610a3a6d9dc6a9e5961388710e5af0b704 Mon Sep 17 00:00:00 2001 From: zombob <2613669+zombob@users.noreply.github.com> Date: Fri, 23 Sep 2022 05:58:14 +0800 Subject: [PATCH 621/661] Add `--source screen` for screenshot inference (#9542) * add screenshot as source * fix: screen number support * Fix: mutiple screen specific area * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * parse screen args in LoadScreenshots * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * sequence+ '_' as file name for save-txt save-crop * screenshot as stream * Update requirements.txt Signed-off-by: Glenn Jocher * Update dataloaders.py Signed-off-by: Glenn Jocher * Update dataloaders.py Signed-off-by: Glenn Jocher * Update detect.py Signed-off-by: Glenn Jocher * Update detect.py Signed-off-by: Glenn Jocher * Update detect.py Signed-off-by: Glenn Jocher * Update dataloaders.py Signed-off-by: Glenn Jocher * Update detect.py Signed-off-by: Glenn Jocher * Update detect.py Signed-off-by: Glenn Jocher * Update predict.py Signed-off-by: Glenn Jocher * Update detect.py Signed-off-by: Glenn Jocher * Update predict.py Signed-off-by: Glenn Jocher * Update README.md Signed-off-by: Glenn Jocher * Update tutorial.ipynb Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: xin Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- README.md | 1 + classify/predict.py | 9 +++++--- detect.py | 9 +++++--- requirements.txt | 1 + segment/predict.py | 9 +++++--- tutorial.ipynb | 1 + utils/dataloaders.py | 49 ++++++++++++++++++++++++++++++++++++++++++++ 7 files changed, 70 insertions(+), 9 deletions(-) diff --git a/README.md b/README.md index da8bf1dad862..1d43111d56e7 100644 --- a/README.md +++ b/README.md @@ -107,6 +107,7 @@ the latest YOLOv5 [release](https://github.com/ultralytics/yolov5/releases) and python detect.py --source 0 # webcam img.jpg # image vid.mp4 # video + screen # screenshot path/ # directory 'path/*.jpg' # glob 'https://youtu.be/Zgi9g1ksQHc' # YouTube diff --git a/classify/predict.py b/classify/predict.py index ef59ff6f550a..011e7b83f09b 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -42,7 +42,7 @@ from models.common import DetectMultiBackend from utils.augmentations import classify_transforms -from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadStreams +from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadScreenshots, LoadStreams from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, increment_path, print_args, strip_optimizer) from utils.plots import Annotator @@ -52,7 +52,7 @@ @smart_inference_mode() def run( weights=ROOT / 'yolov5s-cls.pt', # model.pt path(s) - source=ROOT / 'data/images', # file/dir/URL/glob, 0 for webcam + source=ROOT / 'data/images', # file/dir/URL/glob/screen/0(webcam) data=ROOT / 'data/coco128.yaml', # dataset.yaml path imgsz=(224, 224), # inference size (height, width) device='', # cuda device, i.e. 0 or 0,1,2,3 or cpu @@ -74,6 +74,7 @@ def run( is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS) is_url = source.lower().startswith(('rtsp://', 'rtmp://', 'http://', 'https://')) webcam = source.isnumeric() or source.endswith('.txt') or (is_url and not is_file) + screenshot = source.lower().startswith('screen') if is_url and is_file: source = check_file(source) # download @@ -91,6 +92,8 @@ def run( if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) + elif screenshot: + dataset = LoadScreenshots(source, img_size=imgsz, stride=stride, auto=pt) else: dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) bs = len(dataset) # batch_size @@ -187,7 +190,7 @@ def run( def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s-cls.pt', help='model path(s)') - parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob, 0 for webcam') + parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob/screen/0(webcam)') parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='(optional) dataset.yaml path') parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[224], help='inference size h,w') parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu') diff --git a/detect.py b/detect.py index 4015b9ae0d7f..9036b26263e5 100644 --- a/detect.py +++ b/detect.py @@ -40,7 +40,7 @@ ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.common import DetectMultiBackend -from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadStreams +from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadScreenshots, LoadStreams from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, increment_path, non_max_suppression, print_args, scale_coords, strip_optimizer, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box @@ -50,7 +50,7 @@ @smart_inference_mode() def run( weights=ROOT / 'yolov5s.pt', # model.pt path(s) - source=ROOT / 'data/images', # file/dir/URL/glob, 0 for webcam + source=ROOT / 'data/images', # file/dir/URL/glob/screen/0(webcam) data=ROOT / 'data/coco128.yaml', # dataset.yaml path imgsz=(640, 640), # inference size (height, width) conf_thres=0.25, # confidence threshold @@ -82,6 +82,7 @@ def run( is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS) is_url = source.lower().startswith(('rtsp://', 'rtmp://', 'http://', 'https://')) webcam = source.isnumeric() or source.endswith('.txt') or (is_url and not is_file) + screenshot = source.lower().startswith('screen') if is_url and is_file: source = check_file(source) # download @@ -99,6 +100,8 @@ def run( if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) + elif screenshot: + dataset = LoadScreenshots(source, img_size=imgsz, stride=stride, auto=pt) else: dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) bs = len(dataset) # batch_size @@ -212,7 +215,7 @@ def run( def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s.pt', help='model path(s)') - parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob, 0 for webcam') + parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob/screen/0(webcam)') parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='(optional) dataset.yaml path') parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640], help='inference size h,w') parser.add_argument('--conf-thres', type=float, default=0.25, help='confidence threshold') diff --git a/requirements.txt b/requirements.txt index 55c1f2428e3f..914da54e73fc 100644 --- a/requirements.txt +++ b/requirements.txt @@ -38,6 +38,7 @@ seaborn>=0.11.0 ipython # interactive notebook psutil # system utilization thop>=0.1.1 # FLOPs computation +# mss # screenshots # albumentations>=1.0.3 # pycocotools>=2.0 # COCO mAP # roboflow diff --git a/segment/predict.py b/segment/predict.py index 2ea6bd9327e0..43cebc706371 100644 --- a/segment/predict.py +++ b/segment/predict.py @@ -40,7 +40,7 @@ ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from models.common import DetectMultiBackend -from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadStreams +from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadScreenshots, LoadStreams from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, increment_path, non_max_suppression, print_args, scale_coords, strip_optimizer, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box @@ -51,7 +51,7 @@ @smart_inference_mode() def run( weights=ROOT / 'yolov5s-seg.pt', # model.pt path(s) - source=ROOT / 'data/images', # file/dir/URL/glob, 0 for webcam + source=ROOT / 'data/images', # file/dir/URL/glob/screen/0(webcam) data=ROOT / 'data/coco128.yaml', # dataset.yaml path imgsz=(640, 640), # inference size (height, width) conf_thres=0.25, # confidence threshold @@ -84,6 +84,7 @@ def run( is_file = Path(source).suffix[1:] in (IMG_FORMATS + VID_FORMATS) is_url = source.lower().startswith(('rtsp://', 'rtmp://', 'http://', 'https://')) webcam = source.isnumeric() or source.endswith('.txt') or (is_url and not is_file) + screenshot = source.lower().startswith('screen') if is_url and is_file: source = check_file(source) # download @@ -101,6 +102,8 @@ def run( if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) + elif screenshot: + dataset = LoadScreenshots(source, img_size=imgsz, stride=stride, auto=pt) else: dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) bs = len(dataset) # batch_size @@ -222,7 +225,7 @@ def run( def parse_opt(): parser = argparse.ArgumentParser() parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s-seg.pt', help='model path(s)') - parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob, 0 for webcam') + parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob/screen/0(webcam)') parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='(optional) dataset.yaml path') parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640], help='inference size h,w') parser.add_argument('--conf-thres', type=float, default=0.25, help='confidence threshold') diff --git a/tutorial.ipynb b/tutorial.ipynb index 957437b2be6d..f87cccd99df8 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -445,6 +445,7 @@ "python detect.py --source 0 # webcam\n", " img.jpg # image \n", " vid.mp4 # video\n", + " screen # screenshot\n", " path/ # directory\n", " 'path/*.jpg' # glob\n", " 'https://youtu.be/Zgi9g1ksQHc' # YouTube\n", diff --git a/utils/dataloaders.py b/utils/dataloaders.py index ee79bd0bc5a5..7aee0b891161 100644 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -185,6 +185,55 @@ def __iter__(self): yield from iter(self.sampler) +class LoadScreenshots: + # YOLOv5 screenshot dataloader, i.e. `python detect.py --source "screen 0 100 100 512 256"` + def __init__(self, source, img_size=640, stride=32, auto=True, transforms=None): + # source = [screen_number left top width height] (pixels) + check_requirements('mss') + import mss + + source, *params = source.split() + self.screen, left, top, width, height = 0, None, None, None, None # default to full screen 0 + if len(params) == 1: + self.screen = int(params[0]) + elif len(params) == 4: + left, top, width, height = (int(x) for x in params) + elif len(params) == 5: + self.screen, left, top, width, height = (int(x) for x in params) + self.img_size = img_size + self.stride = stride + self.transforms = transforms + self.auto = auto + self.mode = 'stream' + self.frame = 0 + self.sct = mss.mss() + + # Parse monitor shape + monitor = self.sct.monitors[self.screen] + self.top = monitor["top"] if top is None else (monitor["top"] + top) + self.left = monitor["left"] if left is None else (monitor["left"] + left) + self.width = width or monitor["width"] + self.height = height or monitor["height"] + self.monitor = {"left": self.left, "top": self.top, "width": self.width, "height": self.height} + + def __iter__(self): + return self + + def __next__(self): + # mss screen capture: get raw pixels from the screen as np array + im0 = np.array(self.sct.grab(self.monitor))[:, :, :3] # [:, :, :3] BGRA to BGR + s = f"screen {self.screen} (LTWH): {self.left},{self.top},{self.width},{self.height}: " + + if self.transforms: + im = self.transforms(im0) # transforms + else: + im = letterbox(im0, self.img_size, stride=self.stride, auto=self.auto)[0] # padded resize + im = im.transpose((2, 0, 1))[::-1] # HWC to CHW, BGR to RGB + im = np.ascontiguousarray(im) # contiguous + self.frame += 1 + return str(self.screen), im, im0, None, s # screen, img, original img, im0s, s + + class LoadImages: # YOLOv5 image/video dataloader, i.e. `python detect.py --source image.jpg/vid.mp4` def __init__(self, path, img_size=640, stride=32, auto=True, transforms=None, vid_stride=1): From 1320ce183e3997c4e3a7bf23c22b9edb222519a4 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 23 Sep 2022 23:20:19 +0200 Subject: [PATCH 622/661] Update `is_url()` (#9566) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/downloads.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/utils/downloads.py b/utils/downloads.py index dd2698f995a4..bd495068522d 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -16,13 +16,13 @@ import torch -def is_url(url, check_online=True): - # Check if online file exists +def is_url(url, check_exists=True): + # Check if string is URL and check if URL exists try: url = str(url) result = urllib.parse.urlparse(url) assert all([result.scheme, result.netloc, result.path]) # check if is url - return (urllib.request.urlopen(url).getcode() == 200) if check_online else True # check if exists online + return (urllib.request.urlopen(url).getcode() == 200) if check_exists else True # check if exists online except (AssertionError, urllib.request.HTTPError): return False From d669a74623f273f74213a88b5233964d1ab3ea08 Mon Sep 17 00:00:00 2001 From: Gaz Iqbal Date: Fri, 23 Sep 2022 15:56:42 -0700 Subject: [PATCH 623/661] Detect.py supports running against a Triton container (#9228) * update coco128-seg comments * Enables detect.py to use Triton for inference Triton Inference Server is an open source inference serving software that streamlines AI inferencing. https://github.com/triton-inference-server/server The user can now provide a "--triton-url" argument to detect.py to use a local or remote Triton server for inference. For e.g., http://localhost:8000 will use http over port 8000 and grpc://localhost:8001 will use grpc over port 8001. Note, it is not necessary to specify a weights file to use Triton. A Triton container can be created by first exporting the Yolov5 model to a Triton supported runtime. Onnx, Torchscript, TensorRT are supported by both Triton and the export.py script. The exported model can then be containerized via the OctoML CLI. See https://github.com/octoml/octo-cli#getting-started for a guide. * added triton client to requirements * fixed support for TFSavedModels in Triton * reverted change * Test CoreML update Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher * Use pathlib Signed-off-by: Glenn Jocher * Refacto DetectMultiBackend to directly accept triton url as --weights http://... Signed-off-by: Glenn Jocher * Deploy category Signed-off-by: Glenn Jocher * Update detect.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher * Update common.py Signed-off-by: Glenn Jocher * Update predict.py Signed-off-by: Glenn Jocher * Update predict.py Signed-off-by: Glenn Jocher * Update predict.py Signed-off-by: Glenn Jocher * Update triton.py Signed-off-by: Glenn Jocher * Update triton.py Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Add printout and requirements check * Cleanup Signed-off-by: Glenn Jocher * triton fixes * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fixed triton model query over grpc * Update check_requirements('tritonclient[all]') * group imports * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix likely remote URL bug * update comment * Update is_url() * Fix 2x download attempt on http://path/to/model.pt Signed-off-by: Glenn Jocher Co-authored-by: glennjocher Co-authored-by: Gaz Iqbal Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- classify/predict.py | 2 +- detect.py | 8 ++--- models/common.py | 44 +++++++++++++++-------- requirements.txt | 3 ++ segment/predict.py | 2 +- utils/downloads.py | 4 +-- utils/triton.py | 85 +++++++++++++++++++++++++++++++++++++++++++++ 7 files changed, 126 insertions(+), 22 deletions(-) create mode 100644 utils/triton.py diff --git a/classify/predict.py b/classify/predict.py index 011e7b83f09b..d3bec8eea7ba 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -104,7 +104,7 @@ def run( seen, windows, dt = 0, [], (Profile(), Profile(), Profile()) for path, im, im0s, vid_cap, s in dataset: with dt[0]: - im = torch.Tensor(im).to(device) + im = torch.Tensor(im).to(model.device) im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 if len(im.shape) == 3: im = im[None] # expand for batch dim diff --git a/detect.py b/detect.py index 9036b26263e5..e442ed75f4c7 100644 --- a/detect.py +++ b/detect.py @@ -49,7 +49,7 @@ @smart_inference_mode() def run( - weights=ROOT / 'yolov5s.pt', # model.pt path(s) + weights=ROOT / 'yolov5s.pt', # model path or triton URL source=ROOT / 'data/images', # file/dir/URL/glob/screen/0(webcam) data=ROOT / 'data/coco128.yaml', # dataset.yaml path imgsz=(640, 640), # inference size (height, width) @@ -108,11 +108,11 @@ def run( vid_path, vid_writer = [None] * bs, [None] * bs # Run inference - model.warmup(imgsz=(1 if pt else bs, 3, *imgsz)) # warmup + model.warmup(imgsz=(1 if pt or model.triton else bs, 3, *imgsz)) # warmup seen, windows, dt = 0, [], (Profile(), Profile(), Profile()) for path, im, im0s, vid_cap, s in dataset: with dt[0]: - im = torch.from_numpy(im).to(device) + im = torch.from_numpy(im).to(model.device) im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 im /= 255 # 0 - 255 to 0.0 - 1.0 if len(im.shape) == 3: @@ -214,7 +214,7 @@ def run( def parse_opt(): parser = argparse.ArgumentParser() - parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s.pt', help='model path(s)') + parser.add_argument('--weights', nargs='+', type=str, default=ROOT / 'yolov5s.pt', help='model path or triton URL') parser.add_argument('--source', type=str, default=ROOT / 'data/images', help='file/dir/URL/glob/screen/0(webcam)') parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='(optional) dataset.yaml path') parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640], help='inference size h,w') diff --git a/models/common.py b/models/common.py index fac95a82fdb9..177704849d3d 100644 --- a/models/common.py +++ b/models/common.py @@ -10,6 +10,7 @@ from collections import OrderedDict, namedtuple from copy import copy from pathlib import Path +from urllib.parse import urlparse import cv2 import numpy as np @@ -327,11 +328,13 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, super().__init__() w = str(weights[0] if isinstance(weights, list) else weights) - pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle = self._model_type(w) # type - w = attempt_download(w) # download if not local + pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle, triton = self._model_type(w) fp16 &= pt or jit or onnx or engine # FP16 + nhwc = coreml or saved_model or pb or tflite or edgetpu # BHWC formats (vs torch BCWH) stride = 32 # default stride cuda = torch.cuda.is_available() and device.type != 'cpu' # use CUDA + if not (pt or triton): + w = attempt_download(w) # download if not local if pt: # PyTorch model = attempt_load(weights if isinstance(weights, list) else w, device=device, inplace=True, fuse=fuse) @@ -342,7 +345,7 @@ def __init__(self, weights='yolov5s.pt', device=torch.device('cpu'), dnn=False, elif jit: # TorchScript LOGGER.info(f'Loading {w} for TorchScript inference...') extra_files = {'config.txt': ''} # model metadata - model = torch.jit.load(w, _extra_files=extra_files) + model = torch.jit.load(w, _extra_files=extra_files, map_location=device) model.half() if fp16 else model.float() if extra_files['config.txt']: # load metadata dict d = json.loads(extra_files['config.txt'], @@ -472,6 +475,12 @@ def gd_outputs(gd): predictor = pdi.create_predictor(config) input_handle = predictor.get_input_handle(predictor.get_input_names()[0]) output_names = predictor.get_output_names() + elif triton: # NVIDIA Triton Inference Server + LOGGER.info(f'Using {w} as Triton Inference Server...') + check_requirements('tritonclient[all]') + from utils.triton import TritonRemoteModel + model = TritonRemoteModel(url=w) + nhwc = model.runtime.startswith("tensorflow") else: raise NotImplementedError(f'ERROR: {w} is not a supported format') @@ -488,6 +497,8 @@ def forward(self, im, augment=False, visualize=False): b, ch, h, w = im.shape # batch, channel, height, width if self.fp16 and im.dtype != torch.float16: im = im.half() # to FP16 + if self.nhwc: + im = im.permute(0, 2, 3, 1) # torch BCHW to numpy BHWC shape(1,320,192,3) if self.pt: # PyTorch y = self.model(im, augment=augment, visualize=visualize) if augment or visualize else self.model(im) @@ -517,7 +528,7 @@ def forward(self, im, augment=False, visualize=False): self.context.execute_v2(list(self.binding_addrs.values())) y = [self.bindings[x].data for x in sorted(self.output_names)] elif self.coreml: # CoreML - im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) + im = im.cpu().numpy() im = Image.fromarray((im[0] * 255).astype('uint8')) # im = im.resize((192, 320), Image.ANTIALIAS) y = self.model.predict({'image': im}) # coordinates are xywh normalized @@ -532,8 +543,10 @@ def forward(self, im, augment=False, visualize=False): self.input_handle.copy_from_cpu(im) self.predictor.run() y = [self.predictor.get_output_handle(x).copy_to_cpu() for x in self.output_names] + elif self.triton: # NVIDIA Triton Inference Server + y = self.model(im) else: # TensorFlow (SavedModel, GraphDef, Lite, Edge TPU) - im = im.permute(0, 2, 3, 1).cpu().numpy() # torch BCHW to numpy BHWC shape(1,320,192,3) + im = im.cpu().numpy() if self.saved_model: # SavedModel y = self.model(im, training=False) if self.keras else self.model(im) elif self.pb: # GraphDef @@ -566,8 +579,8 @@ def from_numpy(self, x): def warmup(self, imgsz=(1, 3, 640, 640)): # Warmup model by running inference once - warmup_types = self.pt, self.jit, self.onnx, self.engine, self.saved_model, self.pb - if any(warmup_types) and self.device.type != 'cpu': + warmup_types = self.pt, self.jit, self.onnx, self.engine, self.saved_model, self.pb, self.triton + if any(warmup_types) and (self.device.type != 'cpu' or self.triton): im = torch.empty(*imgsz, dtype=torch.half if self.fp16 else torch.float, device=self.device) # input for _ in range(2 if self.jit else 1): # self.forward(im) # warmup @@ -575,14 +588,17 @@ def warmup(self, imgsz=(1, 3, 640, 640)): @staticmethod def _model_type(p='path/to/model.pt'): # Return model type from model path, i.e. path='path/to/model.onnx' -> type=onnx + # types = [pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle] from export import export_formats - sf = list(export_formats().Suffix) + ['.xml'] # export suffixes - check_suffix(p, sf) # checks - p = Path(p).name # eliminate trailing separators - pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle, xml2 = (s in p for s in sf) - xml |= xml2 # *_openvino_model or *.xml - tflite &= not edgetpu # *.tflite - return pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs, paddle + from utils.downloads import is_url + sf = list(export_formats().Suffix) # export suffixes + if not is_url(p, check=False): + check_suffix(p, sf) # checks + url = urlparse(p) # if url may be Triton inference server + types = [s in Path(p).name for s in sf] + types[8] &= not types[9] # tflite &= not edgetpu + triton = not any(types) and all([any(s in url.scheme for s in ["http", "grpc"]), url.netloc]) + return types + [triton] @staticmethod def _load_metadata(f=Path('path/to/meta.yaml')): diff --git a/requirements.txt b/requirements.txt index 914da54e73fc..4d6ec3509efa 100644 --- a/requirements.txt +++ b/requirements.txt @@ -34,6 +34,9 @@ seaborn>=0.11.0 # tensorflowjs>=3.9.0 # TF.js export # openvino-dev # OpenVINO export +# Deploy -------------------------------------- +# tritonclient[all]~=2.24.0 + # Extras -------------------------------------- ipython # interactive notebook psutil # system utilization diff --git a/segment/predict.py b/segment/predict.py index 43cebc706371..2e794c342de1 100644 --- a/segment/predict.py +++ b/segment/predict.py @@ -114,7 +114,7 @@ def run( seen, windows, dt = 0, [], (Profile(), Profile(), Profile()) for path, im, im0s, vid_cap, s in dataset: with dt[0]: - im = torch.from_numpy(im).to(device) + im = torch.from_numpy(im).to(model.device) im = im.half() if model.fp16 else im.float() # uint8 to fp16/32 im /= 255 # 0 - 255 to 0.0 - 1.0 if len(im.shape) == 3: diff --git a/utils/downloads.py b/utils/downloads.py index bd495068522d..433de84b51ca 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -16,13 +16,13 @@ import torch -def is_url(url, check_exists=True): +def is_url(url, check=True): # Check if string is URL and check if URL exists try: url = str(url) result = urllib.parse.urlparse(url) assert all([result.scheme, result.netloc, result.path]) # check if is url - return (urllib.request.urlopen(url).getcode() == 200) if check_exists else True # check if exists online + return (urllib.request.urlopen(url).getcode() == 200) if check else True # check if exists online except (AssertionError, urllib.request.HTTPError): return False diff --git a/utils/triton.py b/utils/triton.py new file mode 100644 index 000000000000..a94ef0ad197d --- /dev/null +++ b/utils/triton.py @@ -0,0 +1,85 @@ +# YOLOv5 🚀 by Ultralytics, GPL-3.0 license +""" Utils to interact with the Triton Inference Server +""" + +import typing +from urllib.parse import urlparse + +import torch + + +class TritonRemoteModel: + """ A wrapper over a model served by the Triton Inference Server. It can + be configured to communicate over GRPC or HTTP. It accepts Torch Tensors + as input and returns them as outputs. + """ + + def __init__(self, url: str): + """ + Keyword arguments: + url: Fully qualified address of the Triton server - for e.g. grpc://localhost:8000 + """ + + parsed_url = urlparse(url) + if parsed_url.scheme == "grpc": + from tritonclient.grpc import InferenceServerClient, InferInput + + self.client = InferenceServerClient(parsed_url.netloc) # Triton GRPC client + model_repository = self.client.get_model_repository_index() + self.model_name = model_repository.models[0].name + self.metadata = self.client.get_model_metadata(self.model_name, as_json=True) + + def create_input_placeholders() -> typing.List[InferInput]: + return [ + InferInput(i['name'], [int(s) for s in i["shape"]], i['datatype']) for i in self.metadata['inputs']] + + else: + from tritonclient.http import InferenceServerClient, InferInput + + self.client = InferenceServerClient(parsed_url.netloc) # Triton HTTP client + model_repository = self.client.get_model_repository_index() + self.model_name = model_repository[0]['name'] + self.metadata = self.client.get_model_metadata(self.model_name) + + def create_input_placeholders() -> typing.List[InferInput]: + return [ + InferInput(i['name'], [int(s) for s in i["shape"]], i['datatype']) for i in self.metadata['inputs']] + + self._create_input_placeholders_fn = create_input_placeholders + + @property + def runtime(self): + """Returns the model runtime""" + return self.metadata.get("backend", self.metadata.get("platform")) + + def __call__(self, *args, **kwargs) -> typing.Union[torch.Tensor, typing.Tuple[torch.Tensor, ...]]: + """ Invokes the model. Parameters can be provided via args or kwargs. + args, if provided, are assumed to match the order of inputs of the model. + kwargs are matched with the model input names. + """ + inputs = self._create_inputs(*args, **kwargs) + response = self.client.infer(model_name=self.model_name, inputs=inputs) + result = [] + for output in self.metadata['outputs']: + tensor = torch.as_tensor(response.as_numpy(output['name'])) + result.append(tensor) + return result[0] if len(result) == 1 else result + + def _create_inputs(self, *args, **kwargs): + args_len, kwargs_len = len(args), len(kwargs) + if not args_len and not kwargs_len: + raise RuntimeError("No inputs provided.") + if args_len and kwargs_len: + raise RuntimeError("Cannot specify args and kwargs at the same time") + + placeholders = self._create_input_placeholders_fn() + if args_len: + if args_len != len(placeholders): + raise RuntimeError(f"Expected {len(placeholders)} inputs, got {args_len}.") + for input, value in zip(placeholders, args): + input.set_data_from_numpy(value.cpu().numpy()) + else: + for input in placeholders: + value = kwargs[input.name] + input.set_data_from_numpy(value.cpu().numpy()) + return placeholders From c8e52304cf5c34653570c5c3953ba061bc33c1af Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sat, 24 Sep 2022 16:02:41 +0200 Subject: [PATCH 624/661] New `scale_segments()` function (#9570) * Rename scale_coords to scale_boxes * add scale_segments --- detect.py | 4 +-- models/common.py | 4 +-- segment/predict.py | 4 +-- segment/val.py | 6 ++--- utils/general.py | 46 ++++++++++++++++++++++++++------- utils/loggers/comet/__init__.py | 8 +++--- utils/plots.py | 4 +-- val.py | 6 ++--- 8 files changed, 54 insertions(+), 28 deletions(-) diff --git a/detect.py b/detect.py index e442ed75f4c7..4971033b35fb 100644 --- a/detect.py +++ b/detect.py @@ -42,7 +42,7 @@ from models.common import DetectMultiBackend from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadScreenshots, LoadStreams from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, - increment_path, non_max_suppression, print_args, scale_coords, strip_optimizer, xyxy2xywh) + increment_path, non_max_suppression, print_args, scale_boxes, strip_optimizer, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box from utils.torch_utils import select_device, smart_inference_mode @@ -148,7 +148,7 @@ def run( annotator = Annotator(im0, line_width=line_thickness, example=str(names)) if len(det): # Rescale boxes from img_size to im0 size - det[:, :4] = scale_coords(im.shape[2:], det[:, :4], im0.shape).round() + det[:, :4] = scale_boxes(im.shape[2:], det[:, :4], im0.shape).round() # Print results for c in det[:, 5].unique(): diff --git a/models/common.py b/models/common.py index 177704849d3d..273e73d9e729 100644 --- a/models/common.py +++ b/models/common.py @@ -23,7 +23,7 @@ from utils.dataloaders import exif_transpose, letterbox from utils.general import (LOGGER, ROOT, Profile, check_requirements, check_suffix, check_version, colorstr, - increment_path, make_divisible, non_max_suppression, scale_coords, xywh2xyxy, xyxy2xywh, + increment_path, make_divisible, non_max_suppression, scale_boxes, xywh2xyxy, xyxy2xywh, yaml_load) from utils.plots import Annotator, colors, save_one_box from utils.torch_utils import copy_attr, smart_inference_mode @@ -703,7 +703,7 @@ def forward(self, ims, size=640, augment=False, profile=False): self.multi_label, max_det=self.max_det) # NMS for i in range(n): - scale_coords(shape1, y[i][:, :4], shape0[i]) + scale_boxes(shape1, y[i][:, :4], shape0[i]) return Detections(ims, y, files, dt, self.names, x.shape) diff --git a/segment/predict.py b/segment/predict.py index 2e794c342de1..2241204715b5 100644 --- a/segment/predict.py +++ b/segment/predict.py @@ -42,7 +42,7 @@ from models.common import DetectMultiBackend from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadScreenshots, LoadStreams from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, - increment_path, non_max_suppression, print_args, scale_coords, strip_optimizer, xyxy2xywh) + increment_path, non_max_suppression, print_args, scale_boxes, strip_optimizer, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box from utils.segment.general import process_mask from utils.torch_utils import select_device, smart_inference_mode @@ -152,7 +152,7 @@ def run( masks = process_mask(proto[i], det[:, 6:], det[:, :4], im.shape[2:], upsample=True) # HWC # Rescale boxes from img_size to im0 size - det[:, :4] = scale_coords(im.shape[2:], det[:, :4], im0.shape).round() + det[:, :4] = scale_boxes(im.shape[2:], det[:, :4], im0.shape).round() # Print results for c in det[:, 5].unique(): diff --git a/segment/val.py b/segment/val.py index 59ab76672a30..0a37998c1771 100644 --- a/segment/val.py +++ b/segment/val.py @@ -44,7 +44,7 @@ from utils.callbacks import Callbacks from utils.general import (LOGGER, NUM_THREADS, Profile, check_dataset, check_img_size, check_requirements, check_yaml, coco80_to_coco91_class, colorstr, increment_path, non_max_suppression, print_args, - scale_coords, xywh2xyxy, xyxy2xywh) + scale_boxes, xywh2xyxy, xyxy2xywh) from utils.metrics import ConfusionMatrix, box_iou from utils.plots import output_to_target, plot_val_study from utils.segment.dataloaders import create_dataloader @@ -298,12 +298,12 @@ def run( if single_cls: pred[:, 5] = 0 predn = pred.clone() - scale_coords(im[si].shape[1:], predn[:, :4], shape, shapes[si][1]) # native-space pred + scale_boxes(im[si].shape[1:], predn[:, :4], shape, shapes[si][1]) # native-space pred # Evaluate if nl: tbox = xywh2xyxy(labels[:, 1:5]) # target boxes - scale_coords(im[si].shape[1:], tbox, shape, shapes[si][1]) # native-space labels + scale_boxes(im[si].shape[1:], tbox, shape, shapes[si][1]) # native-space labels labelsn = torch.cat((labels[:, 0:1], tbox), 1) # native-space labels correct_bboxes = process_batch(predn, labelsn, iouv) correct_masks = process_batch(predn, labelsn, iouv, pred_masks, gt_masks, overlap=overlap, masks=True) diff --git a/utils/general.py b/utils/general.py index fd0b4090a0fa..87e7e20df1ab 100644 --- a/utils/general.py +++ b/utils/general.py @@ -725,7 +725,7 @@ def xywhn2xyxy(x, w=640, h=640, padw=0, padh=0): def xyxy2xywhn(x, w=640, h=640, clip=False, eps=0.0): # Convert nx4 boxes from [x1, y1, x2, y2] to [x, y, w, h] normalized where xy1=top-left, xy2=bottom-right if clip: - clip_coords(x, (h - eps, w - eps)) # warning: inplace clip + clip_boxes(x, (h - eps, w - eps)) # warning: inplace clip y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x) y[:, 0] = ((x[:, 0] + x[:, 2]) / 2) / w # x center y[:, 1] = ((x[:, 1] + x[:, 3]) / 2) / h # y center @@ -769,7 +769,23 @@ def resample_segments(segments, n=1000): return segments -def scale_coords(img1_shape, coords, img0_shape, ratio_pad=None): +def scale_boxes(img1_shape, boxes, img0_shape, ratio_pad=None): + # Rescale boxes (xyxy) from img1_shape to img0_shape + if ratio_pad is None: # calculate from img0_shape + gain = min(img1_shape[0] / img0_shape[0], img1_shape[1] / img0_shape[1]) # gain = old / new + pad = (img1_shape[1] - img0_shape[1] * gain) / 2, (img1_shape[0] - img0_shape[0] * gain) / 2 # wh padding + else: + gain = ratio_pad[0][0] + pad = ratio_pad[1] + + boxes[:, [0, 2]] -= pad[0] # x padding + boxes[:, [1, 3]] -= pad[1] # y padding + boxes[:, :4] /= gain + clip_boxes(boxes, img0_shape) + return boxes + + +def scale_segments(img1_shape, segments, img0_shape, ratio_pad=None): # Rescale coords (xyxy) from img1_shape to img0_shape if ratio_pad is None: # calculate from img0_shape gain = min(img1_shape[0] / img0_shape[0], img1_shape[1] / img0_shape[1]) # gain = old / new @@ -778,15 +794,15 @@ def scale_coords(img1_shape, coords, img0_shape, ratio_pad=None): gain = ratio_pad[0][0] pad = ratio_pad[1] - coords[:, [0, 2]] -= pad[0] # x padding - coords[:, [1, 3]] -= pad[1] # y padding - coords[:, :4] /= gain - clip_coords(coords, img0_shape) - return coords + segments[:, 0] -= pad[0] # x padding + segments[:, 1] -= pad[1] # y padding + segments /= gain + clip_segments(segments, img0_shape) + return segments -def clip_coords(boxes, shape): - # Clip bounding xyxy bounding boxes to image shape (height, width) +def clip_boxes(boxes, shape): + # Clip boxes (xyxy) to image shape (height, width) if isinstance(boxes, torch.Tensor): # faster individually boxes[:, 0].clamp_(0, shape[1]) # x1 boxes[:, 1].clamp_(0, shape[0]) # y1 @@ -797,6 +813,16 @@ def clip_coords(boxes, shape): boxes[:, [1, 3]] = boxes[:, [1, 3]].clip(0, shape[0]) # y1, y2 +def clip_segments(boxes, shape): + # Clip segments (xy1,xy2,...) to image shape (height, width) + if isinstance(boxes, torch.Tensor): # faster individually + boxes[:, 0].clamp_(0, shape[1]) # x + boxes[:, 1].clamp_(0, shape[0]) # y + else: # np.array (faster grouped) + boxes[:, 0] = boxes[:, 0].clip(0, shape[1]) # x + boxes[:, 1] = boxes[:, 1].clip(0, shape[0]) # y + + def non_max_suppression( prediction, conf_thres=0.25, @@ -980,7 +1006,7 @@ def apply_classifier(x, model, img, im0): d[:, :4] = xywh2xyxy(b).long() # Rescale boxes from img_size to im0 size - scale_coords(img.shape[2:], d[:, :4], im0[i].shape) + scale_boxes(img.shape[2:], d[:, :4], im0[i].shape) # Classes pred_cls1 = d[:, 5].long() diff --git a/utils/loggers/comet/__init__.py b/utils/loggers/comet/__init__.py index 3b3142b002c5..ba5cecc8e096 100644 --- a/utils/loggers/comet/__init__.py +++ b/utils/loggers/comet/__init__.py @@ -28,7 +28,7 @@ import yaml from utils.dataloaders import img2label_paths -from utils.general import check_dataset, scale_coords, xywh2xyxy +from utils.general import check_dataset, scale_boxes, xywh2xyxy from utils.metrics import box_iou COMET_PREFIX = "comet://" @@ -293,14 +293,14 @@ def preprocess_prediction(self, image, labels, shape, pred): pred[:, 5] = 0 predn = pred.clone() - scale_coords(image.shape[1:], predn[:, :4], shape[0], shape[1]) + scale_boxes(image.shape[1:], predn[:, :4], shape[0], shape[1]) labelsn = None if nl: tbox = xywh2xyxy(labels[:, 1:5]) # target boxes - scale_coords(image.shape[1:], tbox, shape[0], shape[1]) # native-space labels + scale_boxes(image.shape[1:], tbox, shape[0], shape[1]) # native-space labels labelsn = torch.cat((labels[:, 0:1], tbox), 1) # native-space labels - scale_coords(image.shape[1:], predn[:, :4], shape[0], shape[1]) # native-space pred + scale_boxes(image.shape[1:], predn[:, :4], shape[0], shape[1]) # native-space pred return predn, labelsn diff --git a/utils/plots.py b/utils/plots.py index 51bb7d6c20af..36df271c60e1 100644 --- a/utils/plots.py +++ b/utils/plots.py @@ -20,7 +20,7 @@ from PIL import Image, ImageDraw, ImageFont from utils import TryExcept, threaded -from utils.general import (CONFIG_DIR, FONT, LOGGER, check_font, check_requirements, clip_coords, increment_path, +from utils.general import (CONFIG_DIR, FONT, LOGGER, check_font, check_requirements, clip_boxes, increment_path, is_ascii, xywh2xyxy, xyxy2xywh) from utils.metrics import fitness from utils.segment.general import scale_image @@ -565,7 +565,7 @@ def save_one_box(xyxy, im, file=Path('im.jpg'), gain=1.02, pad=10, square=False, b[:, 2:] = b[:, 2:].max(1)[0].unsqueeze(1) # attempt rectangle to square b[:, 2:] = b[:, 2:] * gain + pad # box wh * gain + pad xyxy = xywh2xyxy(b).long() - clip_coords(xyxy, im.shape) + clip_boxes(xyxy, im.shape) crop = im[int(xyxy[0, 1]):int(xyxy[0, 3]), int(xyxy[0, 0]):int(xyxy[0, 2]), ::(1 if BGR else -1)] if save: file.parent.mkdir(parents=True, exist_ok=True) # make directory diff --git a/val.py b/val.py index 3ab4bc3fdb58..c0954498d2fb 100644 --- a/val.py +++ b/val.py @@ -40,7 +40,7 @@ from utils.dataloaders import create_dataloader from utils.general import (LOGGER, Profile, check_dataset, check_img_size, check_requirements, check_yaml, coco80_to_coco91_class, colorstr, increment_path, non_max_suppression, print_args, - scale_coords, xywh2xyxy, xyxy2xywh) + scale_boxes, xywh2xyxy, xyxy2xywh) from utils.metrics import ConfusionMatrix, ap_per_class, box_iou from utils.plots import output_to_target, plot_images, plot_val_study from utils.torch_utils import select_device, smart_inference_mode @@ -244,12 +244,12 @@ def run( if single_cls: pred[:, 5] = 0 predn = pred.clone() - scale_coords(im[si].shape[1:], predn[:, :4], shape, shapes[si][1]) # native-space pred + scale_boxes(im[si].shape[1:], predn[:, :4], shape, shapes[si][1]) # native-space pred # Evaluate if nl: tbox = xywh2xyxy(labels[:, 1:5]) # target boxes - scale_coords(im[si].shape[1:], tbox, shape, shapes[si][1]) # native-space labels + scale_boxes(im[si].shape[1:], tbox, shape, shapes[si][1]) # native-space labels labelsn = torch.cat((labels[:, 0:1], tbox), 1) # native-space labels correct = process_batch(predn, labelsn, iouv) if plots: From f11a8a62d27c2740af5df940973d231fd5fcb038 Mon Sep 17 00:00:00 2001 From: Forever518 <1423429527@qq.com> Date: Sun, 25 Sep 2022 01:35:07 +0800 Subject: [PATCH 625/661] generator seed fix for DDP mAP drop (#9545) * Try to fix DDP mAP drop by setting generator's seed to RANK * Fix default activation bug * Update dataloaders.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update dataloaders.py Co-authored-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- models/common.py | 4 ++-- models/yolo.py | 2 +- utils/dataloaders.py | 5 +++-- utils/segment/dataloaders.py | 8 +++++--- 4 files changed, 11 insertions(+), 8 deletions(-) diff --git a/models/common.py b/models/common.py index 273e73d9e729..2fe99be8972b 100644 --- a/models/common.py +++ b/models/common.py @@ -40,13 +40,13 @@ def autopad(k, p=None, d=1): # kernel, padding, dilation class Conv(nn.Module): # Standard convolution with args(ch_in, ch_out, kernel, stride, padding, groups, dilation, activation) - act = nn.SiLU() # default activation + default_act = nn.SiLU() # default activation def __init__(self, c1, c2, k=1, s=1, p=None, g=1, d=1, act=True): super().__init__() self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p, d), groups=g, dilation=d, bias=False) self.bn = nn.BatchNorm2d(c2) - self.act = self.act if act is True else act if isinstance(act, nn.Module) else nn.Identity() + self.act = self.default_act if act is True else act if isinstance(act, nn.Module) else nn.Identity() def forward(self, x): return self.act(self.bn(self.conv(x))) diff --git a/models/yolo.py b/models/yolo.py index 1d0da2a6e010..ed21c067ee93 100644 --- a/models/yolo.py +++ b/models/yolo.py @@ -301,7 +301,7 @@ def parse_model(d, ch): # model_dict, input_channels(3) LOGGER.info(f"\n{'':>3}{'from':>18}{'n':>3}{'params':>10} {'module':<40}{'arguments':<30}") anchors, nc, gd, gw, act = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple'], d.get('activation') if act: - Conv.act = eval(act) # redefine default activation, i.e. Conv.act = nn.SiLU() + Conv.default_act = eval(act) # redefine default activation, i.e. Conv.default_act = nn.SiLU() LOGGER.info(f"{colorstr('activation:')} {act}") # print na = (len(anchors[0]) // 2) if isinstance(anchors, list) else anchors # number of anchors no = na * (nc + 5) # number of outputs = anchors * (classes + 5) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 7aee0b891161..6cd1da6b9cf9 100644 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -40,6 +40,7 @@ VID_FORMATS = 'asf', 'avi', 'gif', 'm4v', 'mkv', 'mov', 'mp4', 'mpeg', 'mpg', 'ts', 'wmv' # include video suffixes BAR_FORMAT = '{l_bar}{bar:10}{r_bar}{bar:-10b}' # tqdm bar format LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1)) # https://pytorch.org/docs/stable/elastic/run.html +RANK = int(os.getenv('RANK', -1)) PIN_MEMORY = str(os.getenv('PIN_MEMORY', True)).lower() == 'true' # global pin_memory for dataloaders # Get orientation exif tag @@ -139,7 +140,7 @@ def create_dataloader(path, sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle) loader = DataLoader if image_weights else InfiniteDataLoader # only DataLoader allows for attribute updates generator = torch.Generator() - generator.manual_seed(0) + generator.manual_seed(6148914691236517205 + RANK) return loader(dataset, batch_size=batch_size, shuffle=shuffle and sampler is None, @@ -1169,7 +1170,7 @@ def create_classification_dataloader(path, nw = min([os.cpu_count() // max(nd, 1), batch_size if batch_size > 1 else 0, workers]) sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle) generator = torch.Generator() - generator.manual_seed(0) + generator.manual_seed(6148914691236517205 + RANK) return InfiniteDataLoader(dataset, batch_size=batch_size, shuffle=shuffle and sampler is None, diff --git a/utils/segment/dataloaders.py b/utils/segment/dataloaders.py index 97ef8556068e..a63d6ec013fd 100644 --- a/utils/segment/dataloaders.py +++ b/utils/segment/dataloaders.py @@ -17,6 +17,8 @@ from ..torch_utils import torch_distributed_zero_first from .augmentations import mixup, random_perspective +RANK = int(os.getenv('RANK', -1)) + def create_dataloader(path, imgsz, @@ -61,8 +63,8 @@ def create_dataloader(path, nw = min([os.cpu_count() // max(nd, 1), batch_size if batch_size > 1 else 0, workers]) # number of workers sampler = None if rank == -1 else distributed.DistributedSampler(dataset, shuffle=shuffle) loader = DataLoader if image_weights else InfiniteDataLoader # only DataLoader allows for attribute updates - # generator = torch.Generator() - # generator.manual_seed(0) + generator = torch.Generator() + generator.manual_seed(6148914691236517205 + RANK) return loader( dataset, batch_size=batch_size, @@ -72,7 +74,7 @@ def create_dataloader(path, pin_memory=True, collate_fn=LoadImagesAndLabelsAndMasks.collate_fn4 if quad else LoadImagesAndLabelsAndMasks.collate_fn, worker_init_fn=seed_worker, - # generator=generator, + generator=generator, ), dataset From 55fbac933bc25b3151082021fa3f10790b3b936a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 25 Sep 2022 02:59:25 +0200 Subject: [PATCH 626/661] Update default GitHub assets (#9573) * Update default GitHub assets Signed-off-by: Glenn Jocher * Update downloads.py Signed-off-by: Glenn Jocher * Update downloads.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/downloads.py | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/utils/downloads.py b/utils/downloads.py index 433de84b51ca..73b8334cb94a 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -87,9 +87,7 @@ def github_assets(repository, version='latest'): return file # GitHub assets - assets = [ - 'yolov5n.pt', 'yolov5s.pt', 'yolov5m.pt', 'yolov5l.pt', 'yolov5x.pt', 'yolov5n6.pt', 'yolov5s6.pt', - 'yolov5m6.pt', 'yolov5l6.pt', 'yolov5x6.pt'] + assets = [f'yolov5{size}{suffix}.pt' for size in 'nsmlx' for suffix in ('', '6', '-cls', '-seg')] # default try: tag, assets = github_assets(repo, release) except Exception: @@ -107,7 +105,6 @@ def github_assets(repository, version='latest'): safe_download( file, url=f'https://github.com/{repo}/releases/download/{tag}/{name}', - url2=f'https://storage.googleapis.com/{repo}/{tag}/{name}', # backup url (optional) min_bytes=1E5, error_msg=f'{file} missing, try downloading from https://github.com/{repo}/releases/{tag} or {url3}') From ee91dc9bb32d2dddc46c633b711a778a6c603143 Mon Sep 17 00:00:00 2001 From: "David A. Macey" Date: Sun, 25 Sep 2022 08:47:16 -0400 Subject: [PATCH 627/661] Update requirements.txt comment https://pytorch.org/get-started/locally/ (#9576) * Update Requirements with PyTorch CUDA Added --extra-index-url https://download.pytorch.org/whl/cu116 URL to requirements file for ease of creating venv with CUDA enabled PyTorch. Otherwise CPU PyTorch is installed an unable to use local GPUs. Signed-off-by: David A. Macey * Update requirements.txt Signed-off-by: Glenn Jocher * Update requirements.txt Signed-off-by: Glenn Jocher Signed-off-by: David A. Macey Signed-off-by: Glenn Jocher Co-authored-by: Glenn Jocher --- requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements.txt b/requirements.txt index 4d6ec3509efa..0436f415c642 100644 --- a/requirements.txt +++ b/requirements.txt @@ -9,7 +9,7 @@ Pillow>=7.1.2 PyYAML>=5.3.1 requests>=2.23.0 scipy>=1.4.1 -torch>=1.7.0 +torch>=1.7.0 # see https://pytorch.org/get-started/locally/ (recommended) torchvision>=0.8.1 tqdm>=4.64.0 # protobuf<=3.20.1 # https://github.com/ultralytics/yolov5/issues/8012 From 2787ad701fbb308cfb494ae8fb68b0fcea0e4077 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 25 Sep 2022 14:52:49 +0200 Subject: [PATCH 628/661] Add segment line predictions (#9571) * Add segment line predictions Signed-off-by: Glenn Jocher * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- segment/predict.py | 20 ++++++++++++-------- utils/segment/general.py | 14 ++++++++++++++ 2 files changed, 26 insertions(+), 8 deletions(-) diff --git a/segment/predict.py b/segment/predict.py index 2241204715b5..607a8697d731 100644 --- a/segment/predict.py +++ b/segment/predict.py @@ -42,9 +42,10 @@ from models.common import DetectMultiBackend from utils.dataloaders import IMG_FORMATS, VID_FORMATS, LoadImages, LoadScreenshots, LoadStreams from utils.general import (LOGGER, Profile, check_file, check_img_size, check_imshow, check_requirements, colorstr, cv2, - increment_path, non_max_suppression, print_args, scale_boxes, strip_optimizer, xyxy2xywh) + increment_path, non_max_suppression, print_args, scale_boxes, scale_segments, + strip_optimizer, xyxy2xywh) from utils.plots import Annotator, colors, save_one_box -from utils.segment.general import process_mask +from utils.segment.general import masks2segments, process_mask from utils.torch_utils import select_device, smart_inference_mode @@ -145,14 +146,16 @@ def run( save_path = str(save_dir / p.name) # im.jpg txt_path = str(save_dir / 'labels' / p.stem) + ('' if dataset.mode == 'image' else f'_{frame}') # im.txt s += '%gx%g ' % im.shape[2:] # print string - gn = torch.tensor(im0.shape)[[1, 0, 1, 0]] # normalization gain whwh imc = im0.copy() if save_crop else im0 # for save_crop annotator = Annotator(im0, line_width=line_thickness, example=str(names)) if len(det): masks = process_mask(proto[i], det[:, 6:], det[:, :4], im.shape[2:], upsample=True) # HWC + det[:, :4] = scale_boxes(im.shape[2:], det[:, :4], im0.shape).round() # rescale boxes to im0 size - # Rescale boxes from img_size to im0 size - det[:, :4] = scale_boxes(im.shape[2:], det[:, :4], im0.shape).round() + # Segments + if save_txt: + segments = reversed(masks2segments(masks)) + segments = [scale_segments(im.shape[2:], x, im0.shape).round() for x in segments] # Print results for c in det[:, 5].unique(): @@ -165,10 +168,10 @@ def run( im_gpu=None if retina_masks else im[i]) # Write results - for *xyxy, conf, cls in reversed(det[:, :6]): + for j, (*xyxy, conf, cls) in enumerate(reversed(det[:, :6])): if save_txt: # Write to file - xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh - line = (cls, *xywh, conf) if save_conf else (cls, *xywh) # label format + segj = segments[j].reshape(-1) # (n,2) to (n*2) + line = (cls, *segj, conf) if save_conf else (cls, *segj) # label format with open(f'{txt_path}.txt', 'a') as f: f.write(('%g ' * len(line)).rstrip() % line + '\n') @@ -176,6 +179,7 @@ def run( c = int(cls) # integer class label = None if hide_labels else (names[c] if hide_conf else f'{names[c]} {conf:.2f}') annotator.box_label(xyxy, label, color=colors(c, True)) + annotator.draw.polygon(segments[j], outline=colors(c, True), width=3) if save_crop: save_one_box(xyxy, imc, file=save_dir / 'crops' / names[c] / f'{p.stem}.jpg', BGR=True) diff --git a/utils/segment/general.py b/utils/segment/general.py index 36547ed0889c..655123bdcfeb 100644 --- a/utils/segment/general.py +++ b/utils/segment/general.py @@ -1,4 +1,5 @@ import cv2 +import numpy as np import torch import torch.nn.functional as F @@ -118,3 +119,16 @@ def masks_iou(mask1, mask2, eps=1e-7): intersection = (mask1 * mask2).sum(1).clamp(0) # (N, ) union = (mask1.sum(1) + mask2.sum(1))[None] - intersection # (area1 + area2) - intersection return intersection / (union + eps) + + +def masks2segments(masks, strategy='largest'): + # Convert masks(n,160,160) into segments(n,xy) + segments = [] + for x in masks.int().numpy().astype('uint8'): + c = cv2.findContours(x, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[0] + if strategy == 'concat': # concatenate all segments + c = np.concatenate([x.reshape(-1, 2) for x in c]) + elif strategy == 'largest': # select largest segment + c = np.array(c[np.array([len(x) for x in c]).argmax()]).reshape(-1, 2) + segments.append(c.astype('float32')) + return segments From 966b0e09f0a5261e555c2a137af2ef9d58cc9779 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 25 Sep 2022 16:21:26 +0200 Subject: [PATCH 629/661] TensorRT detect.py inference fix (#9581) * Update * Update ci-testing.yml Signed-off-by: Glenn Jocher * Update ci-testing.yml Signed-off-by: Glenn Jocher * Segment fix * Segment fix Signed-off-by: Glenn Jocher --- .github/workflows/ci-testing.yml | 6 ++++++ classify/predict.py | 3 ++- detect.py | 3 ++- segment/predict.py | 5 +++-- 4 files changed, 13 insertions(+), 4 deletions(-) diff --git a/.github/workflows/ci-testing.yml b/.github/workflows/ci-testing.yml index fffc92d1b72f..1ec68e8412f9 100644 --- a/.github/workflows/ci-testing.yml +++ b/.github/workflows/ci-testing.yml @@ -44,6 +44,12 @@ jobs: - name: Benchmark SegmentationModel run: | python benchmarks.py --data coco128-seg.yaml --weights ${{ matrix.model }}-seg.pt --img 320 --hard-fail 0.22 + - name: Test predictions + run: | + python export.py --weights ${{ matrix.model }}-cls.pt --include onnx --img 224 + python detect.py --weights ${{ matrix.model }}.onnx --img 320 + python segment/predict.py --weights ${{ matrix.model }}-seg.onnx --img 320 + python classify/predict.py --weights ${{ matrix.model }}-cls.onnx --img 224 Tests: timeout-minutes: 60 diff --git a/classify/predict.py b/classify/predict.py index d3bec8eea7ba..9114aab1d703 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -89,14 +89,15 @@ def run( imgsz = check_img_size(imgsz, s=stride) # check image size # Dataloader + bs = 1 # batch_size if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) + bs = len(dataset) elif screenshot: dataset = LoadScreenshots(source, img_size=imgsz, stride=stride, auto=pt) else: dataset = LoadImages(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) - bs = len(dataset) # batch_size vid_path, vid_writer = [None] * bs, [None] * bs # Run inference diff --git a/detect.py b/detect.py index 4971033b35fb..8f48d8d28000 100644 --- a/detect.py +++ b/detect.py @@ -97,14 +97,15 @@ def run( imgsz = check_img_size(imgsz, s=stride) # check image size # Dataloader + bs = 1 # batch_size if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) + bs = len(dataset) elif screenshot: dataset = LoadScreenshots(source, img_size=imgsz, stride=stride, auto=pt) else: dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) - bs = len(dataset) # batch_size vid_path, vid_writer = [None] * bs, [None] * bs # Run inference diff --git a/segment/predict.py b/segment/predict.py index 607a8697d731..94117cd78633 100644 --- a/segment/predict.py +++ b/segment/predict.py @@ -100,14 +100,15 @@ def run( imgsz = check_img_size(imgsz, s=stride) # check image size # Dataloader + bs = 1 # batch_size if webcam: view_img = check_imshow() dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) + bs = len(dataset) elif screenshot: dataset = LoadScreenshots(source, img_size=imgsz, stride=stride, auto=pt) else: dataset = LoadImages(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) - bs = len(dataset) # batch_size vid_path, vid_writer = [None] * bs, [None] * bs # Run inference @@ -179,7 +180,7 @@ def run( c = int(cls) # integer class label = None if hide_labels else (names[c] if hide_conf else f'{names[c]} {conf:.2f}') annotator.box_label(xyxy, label, color=colors(c, True)) - annotator.draw.polygon(segments[j], outline=colors(c, True), width=3) + # annotator.draw.polygon(segments[j], outline=colors(c, True), width=3) if save_crop: save_one_box(xyxy, imc, file=save_dir / 'crops' / names[c] / f'{p.stem}.jpg', BGR=True) From 639d82fbabed66f347a17fd39cd058bcd26a4142 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 25 Sep 2022 20:12:57 +0200 Subject: [PATCH 630/661] Update Comet links (#9587) * Update Comet links Signed-off-by: Glenn Jocher * Update tutorial.ipynb Signed-off-by: Glenn Jocher * Update README.md Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- README.md | 4 ++-- tutorial.ipynb | 4 ++-- utils/loggers/comet/README.md | 6 +++--- 3 files changed, 7 insertions(+), 7 deletions(-) diff --git a/README.md b/README.md index 1d43111d56e7..1c5e123d61e7 100644 --- a/README.md +++ b/README.md @@ -168,7 +168,7 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12
- + @@ -186,7 +186,7 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 |Comet ⭐ NEW|Deci ⭐ NEW|ClearML ⭐ NEW|Roboflow|Weights & Biases |:-:|:-:|:-:|:-:|:-:| -|Visualize model metrics and predictions and upload models and datasets in realtime with [Comet](https://www.comet.com/site/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration)|Automatically compile and quantize YOLOv5 for better inference performance in one click at [Deci](https://bit.ly/yolov5-deci-platform)|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) +|Visualize model metrics and predictions and upload models and datasets in realtime with [Comet](https://bit.ly/yolov5-readme-comet)|Automatically compile and quantize YOLOv5 for better inference performance in one click at [Deci](https://bit.ly/yolov5-deci-platform)|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) ##
Why YOLOv5
diff --git a/tutorial.ipynb b/tutorial.ipynb index f87cccd99df8..8c78af2b84cd 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -865,7 +865,7 @@ "cell_type": "markdown", "source": [ "## Comet Logging and Visualization 🌟 NEW\n", - "[Comet](https://www.comet.com/site/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration) is now fully integrated with YOLOv5. Track and visualize model metrics in real time, save your hyperparameters, datasets, and model checkpoints, and visualize your model predictions with [Comet Custom Panels](https://www.comet.com/docs/v2/guides/comet-dashboard/code-panels/about-panels/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration)! Comet makes sure you never lose track of your work and makes it easy to share results and collaborate across teams of all sizes! \n", + "[Comet](https://bit.ly/yolov5-readme-comet) is now fully integrated with YOLOv5. Track and visualize model metrics in real time, save your hyperparameters, datasets, and model checkpoints, and visualize your model predictions with [Comet Custom Panels](https://bit.ly/yolov5-colab-comet-panels)! Comet makes sure you never lose track of your work and makes it easy to share results and collaborate across teams of all sizes! \n", "\n", "Getting started is easy:\n", "```shell\n", @@ -874,7 +874,7 @@ "python train.py --img 640 --epochs 3 --data coco128.yaml --weights yolov5s.pt # 3. train\n", "```\n", "\n", - "To learn more about all of the supported Comet features for this integration, check out the [Comet Tutorial](https://github.com/ultralytics/yolov5/tree/master/utils/loggers/comet). If you'd like to learn more about Comet, head over to our [documentation](https://www.comet.com/docs/v2/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration). Get started by trying out the Comet Colab Notebook:\n", + "To learn more about all of the supported Comet features for this integration, check out the [Comet Tutorial](https://github.com/ultralytics/yolov5/tree/master/utils/loggers/comet). If you'd like to learn more about Comet, head over to our [documentation](https://bit.ly/yolov5-colab-comet-docs). Get started by trying out the Comet Colab Notebook:\n", "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1RG0WOQyxlDlo5Km8GogJpIEJlg_5lyYO?usp=sharing)\n", "\n", "\"yolo-ui\"" diff --git a/utils/loggers/comet/README.md b/utils/loggers/comet/README.md index 7b0b8e0e2f09..3a51cb9b5a25 100644 --- a/utils/loggers/comet/README.md +++ b/utils/loggers/comet/README.md @@ -2,13 +2,13 @@ # YOLOv5 with Comet -This guide will cover how to use YOLOv5 with [Comet](https://www.comet.com/site/?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration) +This guide will cover how to use YOLOv5 with [Comet](https://bit.ly/yolov5-readme-comet) # About Comet Comet builds tools that help data scientists, engineers, and team leaders accelerate and optimize machine learning and deep learning models. -Track and visualize model metrics in real time, save your hyperparameters, datasets, and model checkpoints, and visualize your model predictions with [Comet Custom Panels](https://www.comet.com/examples/comet-example-yolov5?shareable=YcwMiJaZSXfcEXpGOHDD12vA1&ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration)! +Track and visualize model metrics in real time, save your hyperparameters, datasets, and model checkpoints, and visualize your model predictions with [Comet Custom Panels](https://bit.ly/yolov5-colab-comet-panels)! Comet makes sure you never lose track of your work and makes it easy to share results and collaborate across teams of all sizes! # Getting Started @@ -253,4 +253,4 @@ comet optimizer -j utils/loggers/comet/hpo.py \ Comet provides a number of ways to visualize the results of your sweep. Take a look at a [project with a completed sweep here](https://www.comet.com/examples/comet-example-yolov5/view/PrlArHGuuhDTKC1UuBmTtOSXD/panels?ref=yolov5&utm_source=yolov5&utm_medium=affilliate&utm_campaign=yolov5_comet_integration) -hyperparameter-yolo \ No newline at end of file +hyperparameter-yolo From 9006b41498a3bc512e293061e017a518f11e9902 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 25 Sep 2022 21:40:28 +0200 Subject: [PATCH 631/661] Add global YOLOv5_DATASETS_DIR (#9586) Update general.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index 87e7e20df1ab..de7871cb23f9 100644 --- a/utils/general.py +++ b/utils/general.py @@ -43,8 +43,8 @@ RANK = int(os.getenv('RANK', -1)) # Settings -DATASETS_DIR = ROOT.parent / 'datasets' # YOLOv5 datasets directory NUM_THREADS = min(8, max(1, os.cpu_count() - 1)) # number of YOLOv5 multiprocessing threads +DATASETS_DIR = Path(os.getenv('YOLOv5_DATASETS_DIR', ROOT.parent / 'datasets')) # global datasets directory AUTOINSTALL = str(os.getenv('YOLOv5_AUTOINSTALL', True)).lower() == 'true' # global auto-install mode VERBOSE = str(os.getenv('YOLOv5_VERBOSE', True)).lower() == 'true' # global verbose mode FONT = 'Arial.ttf' # https://ultralytics.com/assets/Arial.ttf From 9f1cf8dd1ca79b8128d73ac144e8899f51bc5816 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 25 Sep 2022 23:03:14 +0200 Subject: [PATCH 632/661] Add Paperspace Gradient badges (#9588) * Add Paperspace Gradient badges Signed-off-by: Glenn Jocher * Update README.md Signed-off-by: Glenn Jocher * Update tutorial.ipynb Signed-off-by: Glenn Jocher * Update tutorial.ipynb Signed-off-by: Glenn Jocher * Update tutorial.ipynb Signed-off-by: Glenn Jocher * Update README.md Signed-off-by: Glenn Jocher * Update tutorial.ipynb Signed-off-by: Glenn Jocher * Update greetings.yml Signed-off-by: Glenn Jocher * Update greetings.yml Signed-off-by: Glenn Jocher * Update README_cn.md Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- .github/README_cn.md | 4 ++-- .github/workflows/greetings.yml | 8 ++++---- README.md | 7 +++++-- tutorial.ipynb | 5 +++-- 4 files changed, 14 insertions(+), 10 deletions(-) diff --git a/.github/README_cn.md b/.github/README_cn.md index bb62714f003f..7e8aa6f7f087 100644 --- a/.github/README_cn.md +++ b/.github/README_cn.md @@ -12,13 +12,13 @@ [English](../README.md) | 简体中文
- CI CPU testing + YOLOv5 CI YOLOv5 Citation Docker Pulls
+ Run on Gradient Open In Colab Open In Kaggle - Join Forum

diff --git a/.github/workflows/greetings.yml b/.github/workflows/greetings.yml index 91bf190eb727..5e1589c340ed 100644 --- a/.github/workflows/greetings.yml +++ b/.github/workflows/greetings.yml @@ -44,14 +44,14 @@ jobs: YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including [CUDA](https://developer.nvidia.com/cuda)/[CUDNN](https://developer.nvidia.com/cudnn), [Python](https://www.python.org/) and [PyTorch](https://pytorch.org/) preinstalled): - - **Google Colab and Kaggle** notebooks with free GPU: Open In Colab Open In Kaggle + - **Notebooks** with free GPU: Run on Gradient Open In Colab Open In Kaggle - **Google Cloud** Deep Learning VM. See [GCP Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/GCP-Quickstart) - **Amazon** Deep Learning AMI. See [AWS Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/AWS-Quickstart) - **Docker Image**. See [Docker Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/Docker-Quickstart) Docker Pulls - ## Status - CI CPU testing + YOLOv5 CI + + If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 [training](https://github.com/ultralytics/yolov5/blob/master/train.py), [validation](https://github.com/ultralytics/yolov5/blob/master/val.py), [inference](https://github.com/ultralytics/yolov5/blob/master/detect.py), [export](https://github.com/ultralytics/yolov5/blob/master/export.py) and [benchmarks](https://github.com/ultralytics/yolov5/blob/master/benchmarks.py) on MacOS, Windows, and Ubuntu every 24 hours and on every commit. - If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), validation ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on macOS, Windows, and Ubuntu every 24 hours and on every commit. diff --git a/README.md b/README.md index 1c5e123d61e7..227735b52fac 100644 --- a/README.md +++ b/README.md @@ -12,13 +12,13 @@ English | [简体中文](.github/README_cn.md)
- CI CPU testing + YOLOv5 CI YOLOv5 Citation Docker Pulls
+ Run on Gradient Open In Colab Open In Kaggle - Join Forum

@@ -315,6 +315,9 @@ python export.py --weights yolov5s-cls.pt resnet50.pt efficientnet_b0.pt --inclu Get started in seconds with our verified environments. Click each icon below for details.
+ + + diff --git a/tutorial.ipynb b/tutorial.ipynb index 8c78af2b84cd..5d867fb36c93 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -375,6 +375,7 @@ "\n", "\n", "
\n", + " \"Run\n", " \"Open\n", " \"Open\n", "
\n", @@ -945,7 +946,7 @@ "\n", "YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including [CUDA](https://developer.nvidia.com/cuda)/[CUDNN](https://developer.nvidia.com/cudnn), [Python](https://www.python.org/) and [PyTorch](https://pytorch.org/) preinstalled):\n", "\n", - "- **Google Colab and Kaggle** notebooks with free GPU: \"Open \"Open\n", + "- **Notebooks** with free GPU: \"Run \"Open \"Open\n", "- **Google Cloud** Deep Learning VM. See [GCP Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/GCP-Quickstart)\n", "- **Amazon** Deep Learning AMI. See [AWS Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/AWS-Quickstart)\n", "- **Docker Image**. See [Docker Quickstart Guide](https://github.com/ultralytics/yolov5/wiki/Docker-Quickstart) \"Docker\n" @@ -959,7 +960,7 @@ "source": [ "# Status\n", "\n", - "![CI CPU testing](https://github.com/ultralytics/yolov5/workflows/CI%20CPU%20testing/badge.svg)\n", + "![YOLOv5 CI](https://github.com/ultralytics/yolov5/actions/workflows/ci-testing.yml/badge.svg)\n", "\n", "If this badge is green, all [YOLOv5 GitHub Actions](https://github.com/ultralytics/yolov5/actions) Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training ([train.py](https://github.com/ultralytics/yolov5/blob/master/train.py)), testing ([val.py](https://github.com/ultralytics/yolov5/blob/master/val.py)), inference ([detect.py](https://github.com/ultralytics/yolov5/blob/master/detect.py)) and export ([export.py](https://github.com/ultralytics/yolov5/blob/master/export.py)) on macOS, Windows, and Ubuntu every 24 hours and on every commit.\n" ] From 959a4665f820362c95f7435dc05175deeff19671 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 25 Sep 2022 23:26:15 +0200 Subject: [PATCH 633/661] #YOLOVISION22 announcement (#9590) * #YOLOVISION22 announcement Signed-off-by: Glenn Jocher * Update README.md Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- README.md | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/README.md b/README.md index 227735b52fac..56349867e4b6 100644 --- a/README.md +++ b/README.md @@ -1,4 +1,19 @@
+ + Hi, I'm [Glenn Jocher](https://www.linkedin.com/in/glenn-jocher/), author of [YOLOv5](https://github.com/ultralytics/yolov5) 🚀. + + I'd like to invite you to attend the world's first-ever YOLO conference: [#YOLOVISION22](https://ultralytics.com/yolo-vision)! + + This virtual event takes place on **September 27th, 2022** with talks from the world's leading vision AI experts from Google, SenseTime's MMLabs, Baidu's PaddlePaddle, Meituan's YOLOv6, Weight & Biases, Roboflow, Neural Magic, OctoML and of course Ultralytics YOLOv5 and many others. + + Save your spot at https://ultralytics.com/yolo-vision! + + + + +##
+
+

@@ -191,6 +206,8 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 ##

Why YOLOv5
+YOLOv5 has been designed to be super easy to get started and simple to learn. We prioritize real-world results. +

YOLOv5-P5 640 Figure (click to expand) From bfe052b8e1ab398e834a62b607e7d544e1a9876f Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Mon, 26 Sep 2022 12:39:08 +0200 Subject: [PATCH 634/661] Bump actions/stale from 5 to 6 (#9595) Bumps [actions/stale](https://github.com/actions/stale) from 5 to 6. - [Release notes](https://github.com/actions/stale/releases) - [Changelog](https://github.com/actions/stale/blob/main/CHANGELOG.md) - [Commits](https://github.com/actions/stale/compare/v5...v6) --- updated-dependencies: - dependency-name: actions/stale dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> --- .github/workflows/stale.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/stale.yml b/.github/workflows/stale.yml index 03d99790a4a7..9067c343608b 100644 --- a/.github/workflows/stale.yml +++ b/.github/workflows/stale.yml @@ -9,7 +9,7 @@ jobs: stale: runs-on: ubuntu-latest steps: - - uses: actions/stale@v5 + - uses: actions/stale@v6 with: repo-token: ${{ secrets.GITHUB_TOKEN }} stale-issue-message: | From bd9c0c42aee090b373db51c7393c972c26ed9913 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 26 Sep 2022 13:27:34 +0200 Subject: [PATCH 635/661] #YOLOVISION22 update (#9598) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 56349867e4b6..514270973137 100644 --- a/README.md +++ b/README.md @@ -4,7 +4,7 @@ I'd like to invite you to attend the world's first-ever YOLO conference: [#YOLOVISION22](https://ultralytics.com/yolo-vision)! - This virtual event takes place on **September 27th, 2022** with talks from the world's leading vision AI experts from Google, SenseTime's MMLabs, Baidu's PaddlePaddle, Meituan's YOLOv6, Weight & Biases, Roboflow, Neural Magic, OctoML and of course Ultralytics YOLOv5 and many others. + This virtual event takes place on **September 27th, 2022** with talks from the world's leading vision AI experts from Google, OpenMMLab's MMDetection, Baidu's PaddlePaddle, Meituan's YOLOv6, Weight & Biases, Roboflow, Neural Magic, OctoML and of course Ultralytics YOLOv5 and many others. Save your spot at https://ultralytics.com/yolo-vision! From c4c0ee8fc35937cfa940fdaaaf6b9660f5b355f5 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 26 Sep 2022 14:13:03 +0200 Subject: [PATCH 636/661] Apple MPS -> CPU NMS fallback strategy (#9600) Until more ops are fully supported this update will allow for seamless MPS inference (but slower MPS to CPU transfer before NMS, so slower NMS times). Partially resolves https://github.com/ultralytics/yolov5/issues/9596 Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/utils/general.py b/utils/general.py index de7871cb23f9..a855691d3a1f 100644 --- a/utils/general.py +++ b/utils/general.py @@ -843,6 +843,8 @@ def non_max_suppression( if isinstance(prediction, (list, tuple)): # YOLOv5 model in validation model, output = (inference_out, loss_out) prediction = prediction[0] # select only inference output + if 'mps' in prediction.device.type: # MPS not fully supported yet, convert tensors to CPU before NMS + prediction = prediction.cpu() bs = prediction.shape[0] # batch size nc = prediction.shape[2] - nm - 5 # number of classes xc = prediction[..., 4] > conf_thres # candidates From a5748e4b93ae6944ea813b26de6540e80141070b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 26 Sep 2022 20:10:24 +0200 Subject: [PATCH 637/661] Updated Segmentation and Classification usage (#9607) * Updated Segmentation and Classification usage Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- export.py | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/export.py b/export.py index e3cf392b0101..20c1fbc5c7b8 100644 --- a/export.py +++ b/export.py @@ -560,13 +560,20 @@ def run( # Finish f = [str(x) for x in f if x] # filter out '' and None if any(f): + tp = type(model) + dir = Path('segment' if tp is SegmentationModel else 'classify' if tp is ClassificationModel else '') + predict = 'detect.py' if tp is DetectionModel else 'predict.py' h = '--half' if half else '' # --half FP16 inference arg LOGGER.info(f'\nExport complete ({time.time() - t:.1f}s)' f"\nResults saved to {colorstr('bold', file.parent.resolve())}" - f"\nDetect: python detect.py --weights {f[-1]} {h}" - f"\nValidate: python val.py --weights {f[-1]} {h}" + f"\nDetect: python {dir / predict} --weights {f[-1]} {h}" + f"\nValidate: python {dir / 'val.py'} --weights {f[-1]} {h}" f"\nPyTorch Hub: model = torch.hub.load('ultralytics/yolov5', 'custom', '{f[-1]}')" f"\nVisualize: https://netron.app") + if tp is ClassificationModel: + LOGGER.warning("WARNING ⚠️ ClassificationModel not yet supported for PyTorch Hub AutoShape inference") + if tp is SegmentationModel: + LOGGER.warning("WARNING ⚠️ SegmentationModel not yet supported for PyTorch Hub AutoShape inference") return f # return list of exported files/dirs From 6b2c9d1d0f5f9acad86ff9e7043f094a071aa6fe Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 26 Sep 2022 20:46:50 +0200 Subject: [PATCH 638/661] Update export.py Usage examples (#9609) * Update export.py Usage examples Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- export.py | 15 ++++++--------- 1 file changed, 6 insertions(+), 9 deletions(-) diff --git a/export.py b/export.py index 20c1fbc5c7b8..cf37965cea6b 100644 --- a/export.py +++ b/export.py @@ -560,20 +560,17 @@ def run( # Finish f = [str(x) for x in f if x] # filter out '' and None if any(f): - tp = type(model) - dir = Path('segment' if tp is SegmentationModel else 'classify' if tp is ClassificationModel else '') - predict = 'detect.py' if tp is DetectionModel else 'predict.py' + cls, det, seg = (isinstance(model, x) for x in (ClassificationModel, DetectionModel, SegmentationModel)) # type + dir = Path('segment' if seg else 'classify' if cls else '') h = '--half' if half else '' # --half FP16 inference arg + s = "# WARNING ⚠️ ClassificationModel not yet supported for PyTorch Hub AutoShape inference" if cls else \ + "# WARNING ⚠️ SegmentationModel not yet supported for PyTorch Hub AutoShape inference" if seg else '' LOGGER.info(f'\nExport complete ({time.time() - t:.1f}s)' f"\nResults saved to {colorstr('bold', file.parent.resolve())}" - f"\nDetect: python {dir / predict} --weights {f[-1]} {h}" + f"\nDetect: python {dir / ('detect.py' if det else 'predict.py')} --weights {f[-1]} {h}" f"\nValidate: python {dir / 'val.py'} --weights {f[-1]} {h}" - f"\nPyTorch Hub: model = torch.hub.load('ultralytics/yolov5', 'custom', '{f[-1]}')" + f"\nPyTorch Hub: model = torch.hub.load('ultralytics/yolov5', 'custom', '{f[-1]}') {s}" f"\nVisualize: https://netron.app") - if tp is ClassificationModel: - LOGGER.warning("WARNING ⚠️ ClassificationModel not yet supported for PyTorch Hub AutoShape inference") - if tp is SegmentationModel: - LOGGER.warning("WARNING ⚠️ SegmentationModel not yet supported for PyTorch Hub AutoShape inference") return f # return list of exported files/dirs From 1460e5715700cdb130472e1314074ff648f811d8 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 27 Sep 2022 00:29:23 +0200 Subject: [PATCH 639/661] Fix `is_url('https://ultralytics.com')` (#9610) Failing on missing path, i.e. no 'www.' Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/downloads.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/downloads.py b/utils/downloads.py index 73b8334cb94a..60417c1f8835 100644 --- a/utils/downloads.py +++ b/utils/downloads.py @@ -21,7 +21,7 @@ def is_url(url, check=True): try: url = str(url) result = urllib.parse.urlparse(url) - assert all([result.scheme, result.netloc, result.path]) # check if is url + assert all([result.scheme, result.netloc]) # check if is url return (urllib.request.urlopen(url).getcode() == 200) if check else True # check if exists online except (AssertionError, urllib.request.HTTPError): return False From 7314363f26e23fc831a9a739b4031f9f0217084a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 27 Sep 2022 16:58:14 +0200 Subject: [PATCH 640/661] Add `results.save(save_dir='path', exist_ok=False)` (#9617) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- models/common.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/models/common.py b/models/common.py index 2fe99be8972b..d889d0292c61 100644 --- a/models/common.py +++ b/models/common.py @@ -775,12 +775,12 @@ def _run(self, pprint=False, show=False, save=False, crop=False, render=False, l def show(self, labels=True): self._run(show=True, labels=labels) # show results - def save(self, labels=True, save_dir='runs/detect/exp'): - save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) # increment save_dir + def save(self, labels=True, save_dir='runs/detect/exp', exist_ok=False): + save_dir = increment_path(save_dir, exist_ok, mkdir=True) # increment save_dir self._run(save=True, labels=labels, save_dir=save_dir) # save results - def crop(self, save=True, save_dir='runs/detect/exp'): - save_dir = increment_path(save_dir, exist_ok=save_dir != 'runs/detect/exp', mkdir=True) if save else None + def crop(self, save=True, save_dir='runs/detect/exp', exist_ok=False): + save_dir = increment_path(save_dir, exist_ok, mkdir=True) if save else None return self._run(crop=True, save=save, save_dir=save_dir) # crop results def render(self, labels=True): From 2373d5470e386a0c63c6ab77fbee6d699665e27b Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 27 Sep 2022 18:02:48 +0200 Subject: [PATCH 641/661] NMS MPS device wrapper (#9620) * NMS MPS device wrapper May resolve https://github.com/ultralytics/yolov5/issues/9613 Signed-off-by: Glenn Jocher * Update general.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/utils/general.py b/utils/general.py index a855691d3a1f..d31b043a113e 100644 --- a/utils/general.py +++ b/utils/general.py @@ -843,7 +843,9 @@ def non_max_suppression( if isinstance(prediction, (list, tuple)): # YOLOv5 model in validation model, output = (inference_out, loss_out) prediction = prediction[0] # select only inference output - if 'mps' in prediction.device.type: # MPS not fully supported yet, convert tensors to CPU before NMS + device = prediction.device + mps = 'mps' in device.type # Apple MPS + if mps: # MPS not fully supported yet, convert tensors to CPU before NMS prediction = prediction.cpu() bs = prediction.shape[0] # batch size nc = prediction.shape[2] - nm - 5 # number of classes @@ -930,6 +932,8 @@ def non_max_suppression( i = i[iou.sum(1) > 1] # require redundancy output[xi] = x[i] + if mps: + output[xi] = output[xi].to(device) if (time.time() - t) > time_limit: LOGGER.warning(f'WARNING ⚠️ NMS time limit {time_limit:.3f}s exceeded') break # time limit exceeded From 799e3d0cc92a9f431d97931641e7d0b46720699a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 28 Sep 2022 16:43:11 +0200 Subject: [PATCH 642/661] Add SegmentationModel unsupported warning (#9632) * Add SegmentationModel unsupported warning Signed-off-by: Glenn Jocher * Update hubconf.py Signed-off-by: Glenn Jocher * Update hubconf.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- hubconf.py | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/hubconf.py b/hubconf.py index 4224760a4732..95b95a5c30cc 100644 --- a/hubconf.py +++ b/hubconf.py @@ -30,7 +30,7 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo from models.common import AutoShape, DetectMultiBackend from models.experimental import attempt_load - from models.yolo import ClassificationModel, DetectionModel + from models.yolo import ClassificationModel, DetectionModel, SegmentationModel from utils.downloads import attempt_download from utils.general import LOGGER, check_requirements, intersect_dicts, logging from utils.torch_utils import select_device @@ -47,8 +47,11 @@ def _create(name, pretrained=True, channels=3, classes=80, autoshape=True, verbo model = DetectMultiBackend(path, device=device, fuse=autoshape) # detection model if autoshape: if model.pt and isinstance(model.model, ClassificationModel): - LOGGER.warning('WARNING ⚠️ YOLOv5 v6.2 ClassificationModel is not yet AutoShape compatible. ' + LOGGER.warning('WARNING ⚠️ YOLOv5 ClassificationModel is not yet AutoShape compatible. ' 'You must pass torch tensors in BCHW to this model, i.e. shape(1,3,224,224).') + elif model.pt and isinstance(model.model, SegmentationModel): + LOGGER.warning('WARNING ⚠️ YOLOv5 SegmentationModel is not yet AutoShape compatible. ' + 'You will not be able to run inference with this model.') else: model = AutoShape(model) # for file/URI/PIL/cv2/np inputs and NMS except Exception: From 0860e58557f26a0136dd8afbc82f408f31d15ecd Mon Sep 17 00:00:00 2001 From: Soumik Rakshit <19soumik.rakshit96@gmail.com> Date: Fri, 30 Sep 2022 02:31:45 +0530 Subject: [PATCH 643/661] Disabled upload_dataset flag temporarily due to an artifact related bug (#9652) * disabled upload_dataset flag temporarily due to an artifact related bug * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/loggers/wandb/wandb_utils.py | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/utils/loggers/wandb/wandb_utils.py b/utils/loggers/wandb/wandb_utils.py index e850d2ac8a7c..d2dd0fa7c6cd 100644 --- a/utils/loggers/wandb/wandb_utils.py +++ b/utils/loggers/wandb/wandb_utils.py @@ -132,6 +132,11 @@ def __init__(self, opt, run_id=None, job_type='Training'): job_type (str) -- To set the job_type for this run """ + # Temporary-fix + if opt.upload_dataset: + opt.upload_dataset = False + LOGGER.info("Uploading Dataset functionality is not being supported temporarily due to a bug.") + # Pre-training routine -- self.job_type = job_type self.wandb, self.wandb_run = wandb, None if not wandb else wandb.run From 82bec4c8785e123bbea01f6f2d4215c2077ac81f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 29 Sep 2022 23:35:39 +0200 Subject: [PATCH 644/661] Add NVIDIA Jetson Nano Deployment tutorial (#9656) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 514270973137..8b1c98b34e8f 100644 --- a/README.md +++ b/README.md @@ -163,6 +163,7 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 - [Multi-GPU Training](https://github.com/ultralytics/yolov5/issues/475) - [PyTorch Hub](https://github.com/ultralytics/yolov5/issues/36) 🌟 NEW - [TFLite, ONNX, CoreML, TensorRT Export](https://github.com/ultralytics/yolov5/issues/251) 🚀 +- [NVIDIA Jetson Nano Deployment](https://github.com/ultralytics/yolov5/issues/9627) 🌟 NEW - [Test-Time Augmentation (TTA)](https://github.com/ultralytics/yolov5/issues/303) - [Model Ensembling](https://github.com/ultralytics/yolov5/issues/318) - [Model Pruning/Sparsity](https://github.com/ultralytics/yolov5/issues/304) From 8a19437690548a158b78ab27b7f5b463a268fa19 Mon Sep 17 00:00:00 2001 From: Anant Sakhare <70131870+senhorinfinito@users.noreply.github.com> Date: Sat, 1 Oct 2022 20:12:31 +0530 Subject: [PATCH 645/661] =?UTF-8?q?Added=20cutout=20import=20from=20utils/?= =?UTF-8?q?augmentations.py=20to=20use=20Cutout=20Aug=20in=20=E2=80=A6=20(?= =?UTF-8?q?#9668)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * Added cutout import from utils/augmentations.py to use Cutout Aug in data loader by un-commenting line 679, 680, 681 * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- utils/dataloaders.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/dataloaders.py b/utils/dataloaders.py index 6cd1da6b9cf9..d849d5150f4b 100644 --- a/utils/dataloaders.py +++ b/utils/dataloaders.py @@ -29,7 +29,7 @@ from tqdm import tqdm from utils.augmentations import (Albumentations, augment_hsv, classify_albumentations, classify_transforms, copy_paste, - letterbox, mixup, random_perspective) + cutout, letterbox, mixup, random_perspective) from utils.general import (DATASETS_DIR, LOGGER, NUM_THREADS, check_dataset, check_requirements, check_yaml, clean_str, cv2, is_colab, is_kaggle, segments2boxes, xyn2xy, xywh2xyxy, xywhn2xyxy, xyxy2xywhn) from utils.torch_utils import torch_distributed_zero_first From 1158a50abd78808049327fdf60724b2b32726d88 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 2 Oct 2022 13:37:54 +0200 Subject: [PATCH 646/661] Simplify val.py benchmark mode with speed mode (#9674) Update --- benchmarks.py | 4 ++-- segment/val.py | 3 +-- val.py | 3 +-- 3 files changed, 4 insertions(+), 6 deletions(-) diff --git a/benchmarks.py b/benchmarks.py index b3b58eb3257c..ef5c882973f0 100644 --- a/benchmarks.py +++ b/benchmarks.py @@ -81,10 +81,10 @@ def run( # Validate if model_type == SegmentationModel: - result = val_seg(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) + result = val_seg(data, w, batch_size, imgsz, plots=False, device=device, task='speed', half=half) metric = result[0][7] # (box(p, r, map50, map), mask(p, r, map50, map), *loss(box, obj, cls)) else: # DetectionModel: - result = val_det(data, w, batch_size, imgsz, plots=False, device=device, task='benchmark', half=half) + result = val_det(data, w, batch_size, imgsz, plots=False, device=device, task='speed', half=half) metric = result[0][3] # (p, r, map50, map, *loss(box, obj, cls)) speed = result[2][1] # times (preprocess, inference, postprocess) y.append([name, round(file_size(w), 1), round(metric, 4), round(speed, 2)]) # MB, mAP, t_inference diff --git a/segment/val.py b/segment/val.py index 0a37998c1771..f1ec54638d61 100644 --- a/segment/val.py +++ b/segment/val.py @@ -210,8 +210,7 @@ def run( assert ncm == nc, f'{weights} ({ncm} classes) trained on different --data than what you passed ({nc} ' \ f'classes). Pass correct combination of --weights and --data that are trained together.' model.warmup(imgsz=(1 if pt else batch_size, 3, imgsz, imgsz)) # warmup - pad = 0.0 if task in ('speed', 'benchmark') else 0.5 - rect = False if task == 'benchmark' else pt # square inference for benchmarks + pad, rect = (0.0, False) if task == 'speed' else (0.5, pt) # square inference for benchmarks task = task if task in ('train', 'val', 'test') else 'val' # path to train/val/test images dataloader = create_dataloader(data[task], imgsz, diff --git a/val.py b/val.py index c0954498d2fb..ca838c0beb2f 100644 --- a/val.py +++ b/val.py @@ -169,8 +169,7 @@ def run( assert ncm == nc, f'{weights} ({ncm} classes) trained on different --data than what you passed ({nc} ' \ f'classes). Pass correct combination of --weights and --data that are trained together.' model.warmup(imgsz=(1 if pt else batch_size, 3, imgsz, imgsz)) # warmup - pad = 0.0 if task in ('speed', 'benchmark') else 0.5 - rect = False if task == 'benchmark' else pt # square inference for benchmarks + pad, rect = (0.0, False) if task == 'speed' else (0.5, pt) # square inference for benchmarks task = task if task in ('train', 'val', 'test') else 'val' # path to train/val/test images dataloader = create_dataloader(data[task], imgsz, From c98128fe71a8676037a0605ab389c7473c743d07 Mon Sep 17 00:00:00 2001 From: KristenKehrer <34010022+KristenKehrer@users.noreply.github.com> Date: Sun, 2 Oct 2022 18:25:10 -0400 Subject: [PATCH 647/661] Allow list for Comet artifact class 'names' field (#9654) * Update __init__.py In the Comet logger, when I run train.py, it wants to download the data artifact. It was requiring me to format the 'names' field in the data artifact metadata as a dictionary, so I've changed this so that it also accepts a list. Signed-off-by: KristenKehrer <34010022+KristenKehrer@users.noreply.github.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update utils/loggers/comet/__init__.py Co-authored-by: Dhruv Nair Signed-off-by: KristenKehrer <34010022+KristenKehrer@users.noreply.github.com> Signed-off-by: KristenKehrer <34010022+KristenKehrer@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Dhruv Nair Co-authored-by: Glenn Jocher --- utils/loggers/comet/__init__.py | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/utils/loggers/comet/__init__.py b/utils/loggers/comet/__init__.py index ba5cecc8e096..b0318f88d6a6 100644 --- a/utils/loggers/comet/__init__.py +++ b/utils/loggers/comet/__init__.py @@ -353,7 +353,14 @@ def download_dataset_artifact(self, artifact_path): metadata = logged_artifact.metadata data_dict = metadata.copy() data_dict["path"] = artifact_save_dir - data_dict["names"] = {int(k): v for k, v in metadata.get("names").items()} + + metadata_names = metadata.get("names") + if type(metadata_names) == dict: + data_dict["names"] = {int(k): v for k, v in metadata.get("names").items()} + elif type(metadata_names) == list: + data_dict["names"] = {int(k): v for k, v in zip(range(len(metadata_names)), metadata_names)} + else: + raise "Invalid 'names' field in dataset yaml file. Please use a list or dictionary" data_dict = self.update_data_paths(data_dict) return data_dict From 68d654d8c4d473aa81be91ac42f320009736992b Mon Sep 17 00:00:00 2001 From: "pre-commit-ci[bot]" <66853113+pre-commit-ci[bot]@users.noreply.github.com> Date: Tue, 4 Oct 2022 16:31:51 +0200 Subject: [PATCH 648/661] [pre-commit.ci] pre-commit suggestions (#9685) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit updates: - [github.com/asottile/pyupgrade: v2.37.3 → v2.38.2](https://github.com/asottile/pyupgrade/compare/v2.37.3...v2.38.2) Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- .pre-commit-config.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index ba8005535397..1cd102c26b41 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -24,7 +24,7 @@ repos: - id: check-docstring-first - repo: https://github.com/asottile/pyupgrade - rev: v2.37.3 + rev: v2.38.2 hooks: - id: pyupgrade name: Upgrade code From e4398cf179601d47207e9f526cf0760b82058930 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 4 Oct 2022 16:32:19 +0200 Subject: [PATCH 649/661] TensorRT `--dynamic` fix (#9691) * Update export.py Signed-off-by: Glenn Jocher * Update export.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- export.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/export.py b/export.py index cf37965cea6b..66d4d636133a 100644 --- a/export.py +++ b/export.py @@ -251,11 +251,11 @@ def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose if trt.__version__[0] == '7': # TensorRT 7 handling https://github.com/ultralytics/yolov5/issues/6012 grid = model.model[-1].anchor_grid model.model[-1].anchor_grid = [a[..., :1, :1, :] for a in grid] - export_onnx(model, im, file, 12, False, dynamic, simplify) # opset 12 + export_onnx(model, im, file, 12, dynamic, simplify) # opset 12 model.model[-1].anchor_grid = grid else: # TensorRT >= 8 check_version(trt.__version__, '8.0.0', hard=True) # require tensorrt>=8.0.0 - export_onnx(model, im, file, 12, False, dynamic, simplify) # opset 12 + export_onnx(model, im, file, 12, dynamic, simplify) # opset 12 onnx = file.with_suffix('.onnx') LOGGER.info(f'\n{prefix} starting export with TensorRT {trt.__version__}...') @@ -285,7 +285,7 @@ def export_engine(model, im, file, half, dynamic, simplify, workspace=4, verbose if dynamic: if im.shape[0] <= 1: - LOGGER.warning(f"{prefix}WARNING ⚠️ --dynamic model requires maximum --batch-size argument") + LOGGER.warning(f"{prefix} WARNING ⚠️ --dynamic model requires maximum --batch-size argument") profile = builder.create_optimization_profile() for inp in inputs: profile.set_shape(inp.name, (1, *im.shape[1:]), (max(1, im.shape[0] // 2), *im.shape[1:]), im.shape) From 7f097ddb6c9921d64fa504a8db79cf24fa0a913c Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 5 Oct 2022 22:29:46 +0200 Subject: [PATCH 650/661] FROM nvcr.io/nvidia/pytorch:22.09-py3 (#9711) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/docker/Dockerfile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index 764ee278c22e..9b93fad7b203 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -3,7 +3,7 @@ # Image is CUDA-optimized for YOLOv5 single/multi-GPU training and inference # Start FROM NVIDIA PyTorch image https://ngc.nvidia.com/catalog/containers/nvidia:pytorch -FROM nvcr.io/nvidia/pytorch:22.08-py3 +FROM nvcr.io/nvidia/pytorch:22.09-py3 RUN rm -rf /opt/pytorch # remove 1.2GB dir # Downloads to user config dir From 5ef69ef3e6180709bc292370ed314b6029ecabfc Mon Sep 17 00:00:00 2001 From: Paul Guerrie <97041392+paulguerrie@users.noreply.github.com> Date: Thu, 6 Oct 2022 14:55:15 -0600 Subject: [PATCH 651/661] Error in utils/segment/general `masks2segments()` (#9724) When running segmentation predict on gpu, the conversion from tensor to numpy fails. Calling `.cpu()` solves this problem. Signed-off-by: Paul Guerrie <97041392+paulguerrie@users.noreply.github.com> Signed-off-by: Paul Guerrie <97041392+paulguerrie@users.noreply.github.com> --- utils/segment/general.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utils/segment/general.py b/utils/segment/general.py index 655123bdcfeb..43bdc460f928 100644 --- a/utils/segment/general.py +++ b/utils/segment/general.py @@ -124,7 +124,7 @@ def masks_iou(mask1, mask2, eps=1e-7): def masks2segments(masks, strategy='largest'): # Convert masks(n,160,160) into segments(n,xy) segments = [] - for x in masks.int().numpy().astype('uint8'): + for x in masks.int().cpu().numpy().astype('uint8'): c = cv2.findContours(x, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[0] if strategy == 'concat': # concatenate all segments c = np.concatenate([x.reshape(-1, 2) for x in c]) From 209be932dec9e89b902f0ac2975fa599e9bc676f Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 9 Oct 2022 23:51:29 +0200 Subject: [PATCH 652/661] Fix segment evolution keys (#9742) * Update * Cleanup --- segment/train.py | 2 +- train.py | 4 +++- utils/general.py | 5 ++--- 3 files changed, 6 insertions(+), 5 deletions(-) diff --git a/segment/train.py b/segment/train.py index 5121c5fa784a..26f0d0c13c78 100644 --- a/segment/train.py +++ b/segment/train.py @@ -651,7 +651,7 @@ def main(opt, callbacks=Callbacks()): results = train(hyp.copy(), opt, device, callbacks) callbacks = Callbacks() # Write mutation results - print_mutation(results, hyp.copy(), save_dir, opt.bucket) + print_mutation(KEYS, results, hyp.copy(), save_dir, opt.bucket) # Plot results plot_evolve(evolve_csv) diff --git a/train.py b/train.py index 9efece250581..177e081c8c37 100644 --- a/train.py +++ b/train.py @@ -607,7 +607,9 @@ def main(opt, callbacks=Callbacks()): results = train(hyp.copy(), opt, device, callbacks) callbacks = Callbacks() # Write mutation results - print_mutation(results, hyp.copy(), save_dir, opt.bucket) + keys = ('metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95', 'val/box_loss', + 'val/obj_loss', 'val/cls_loss') + print_mutation(keys, results, hyp.copy(), save_dir, opt.bucket) # Plot results plot_evolve(evolve_csv) diff --git a/utils/general.py b/utils/general.py index d31b043a113e..e2faca9dbf2a 100644 --- a/utils/general.py +++ b/utils/general.py @@ -957,11 +957,10 @@ def strip_optimizer(f='best.pt', s=''): # from utils.general import *; strip_op LOGGER.info(f"Optimizer stripped from {f},{f' saved as {s},' if s else ''} {mb:.1f}MB") -def print_mutation(results, hyp, save_dir, bucket, prefix=colorstr('evolve: ')): +def print_mutation(keys, results, hyp, save_dir, bucket, prefix=colorstr('evolve: ')): evolve_csv = save_dir / 'evolve.csv' evolve_yaml = save_dir / 'hyp_evolve.yaml' - keys = ('metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95', 'val/box_loss', - 'val/obj_loss', 'val/cls_loss') + tuple(hyp.keys()) # [results + hyps] + keys = tuple(keys) + tuple(hyp.keys()) # [results + hyps] keys = tuple(x.strip() for x in keys) vals = results + tuple(hyp.values()) n = len(keys) From 2f1eb21ad6c0f715f38200c31e6e01a92c5acb25 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 10 Oct 2022 14:54:21 +0200 Subject: [PATCH 653/661] Remove #YOLOVISION22 notice (#9751) Update README.md Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- README.md | 15 --------------- 1 file changed, 15 deletions(-) diff --git a/README.md b/README.md index 8b1c98b34e8f..8c19e52c45d7 100644 --- a/README.md +++ b/README.md @@ -1,19 +1,4 @@
- - Hi, I'm [Glenn Jocher](https://www.linkedin.com/in/glenn-jocher/), author of [YOLOv5](https://github.com/ultralytics/yolov5) 🚀. - - I'd like to invite you to attend the world's first-ever YOLO conference: [#YOLOVISION22](https://ultralytics.com/yolo-vision)! - - This virtual event takes place on **September 27th, 2022** with talks from the world's leading vision AI experts from Google, OpenMMLab's MMDetection, Baidu's PaddlePaddle, Meituan's YOLOv6, Weight & Biases, Roboflow, Neural Magic, OctoML and of course Ultralytics YOLOv5 and many others. - - Save your spot at https://ultralytics.com/yolo-vision! - - - - -##
-
-

From 7a69035eb8a15f44a1dc8f1e07ee71b674e98271 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Wed, 12 Oct 2022 12:53:12 +0200 Subject: [PATCH 654/661] Update Loggers (#9760) * Update * Update tutorial.ipynb Signed-off-by: Glenn Jocher * Update tutorial.ipynb Signed-off-by: Glenn Jocher * Update tutorial.ipynb Signed-off-by: Glenn Jocher * Update requirements.txt Signed-off-by: Glenn Jocher * Update * Update README.md Signed-off-by: Glenn Jocher * Update Signed-off-by: Glenn Jocher --- README.md | 16 ++++++---------- requirements.txt | 2 +- tutorial.ipynb | 25 +++---------------------- utils/docker/Dockerfile | 2 +- utils/loggers/__init__.py | 14 +++++++------- utils/loggers/wandb/wandb_utils.py | 2 +- 6 files changed, 19 insertions(+), 42 deletions(-) diff --git a/README.md b/README.md index 8c19e52c45d7..8f45ccd229b5 100644 --- a/README.md +++ b/README.md @@ -155,7 +155,6 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12 - [Hyperparameter Evolution](https://github.com/ultralytics/yolov5/issues/607) - [Transfer Learning with Frozen Layers](https://github.com/ultralytics/yolov5/issues/1314) - [Architecture Summary](https://github.com/ultralytics/yolov5/issues/6998) 🌟 NEW -- [Weights & Biases Logging](https://github.com/ultralytics/yolov5/issues/1289) - [Roboflow for Datasets, Labeling, and Active Learning](https://github.com/ultralytics/yolov5/issues/4975)  🌟 NEW - [ClearML Logging](https://github.com/ultralytics/yolov5/tree/master/utils/loggers/clearml) 🌟 NEW - [Deci Platform](https://github.com/ultralytics/yolov5/wiki/Deci-Platform) 🌟 NEW @@ -171,23 +170,20 @@ python train.py --data coco.yaml --cfg yolov5n.yaml --weights '' --batch-size 12

- + - + - + - - -
-|Comet ⭐ NEW|Deci ⭐ NEW|ClearML ⭐ NEW|Roboflow|Weights & Biases -|:-:|:-:|:-:|:-:|:-:| -|Visualize model metrics and predictions and upload models and datasets in realtime with [Comet](https://bit.ly/yolov5-readme-comet)|Automatically compile and quantize YOLOv5 for better inference performance in one click at [Deci](https://bit.ly/yolov5-deci-platform)|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics) |Automatically track and visualize all your YOLOv5 training runs in the cloud with [Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_readme) +|Comet ⭐ NEW|Deci ⭐ NEW|ClearML ⭐ NEW|Roboflow| +|:-:|:-:|:-:|:-:| +|Visualize model metrics and predictions and upload models and datasets in realtime with [Comet](https://bit.ly/yolov5-readme-comet)|Automatically compile and quantize YOLOv5 for better inference performance in one click at [Deci](https://bit.ly/yolov5-deci-platform)|Automatically track, visualize and even remotely train YOLOv5 using [ClearML](https://cutt.ly/yolov5-readme-clearml) (open-source!)|Label and export your custom datasets directly to YOLOv5 for training with [Roboflow](https://roboflow.com/?ref=ultralytics)| ##
Why YOLOv5
diff --git a/requirements.txt b/requirements.txt index 0436f415c642..52f7b9ea57d2 100644 --- a/requirements.txt +++ b/requirements.txt @@ -16,8 +16,8 @@ tqdm>=4.64.0 # Logging ------------------------------------- tensorboard>=2.4.1 -# wandb # clearml +# comet # Plotting ------------------------------------ pandas>=1.1.4 diff --git a/tutorial.ipynb b/tutorial.ipynb index 5d867fb36c93..63abebc5b37f 100644 --- a/tutorial.ipynb +++ b/tutorial.ipynb @@ -655,7 +655,7 @@ "cell_type": "code", "source": [ "#@title Select YOLOv5 🚀 logger {run: 'auto'}\n", - "logger = 'TensorBoard' #@param ['TensorBoard', 'Comet', 'ClearML', 'W&B']\n", + "logger = 'TensorBoard' #@param ['TensorBoard', 'Comet', 'ClearML']\n", "\n", "if logger == 'TensorBoard':\n", " %load_ext tensorboard\n", @@ -664,10 +664,7 @@ " %pip install -q comet_ml\n", " import comet_ml; comet_ml.init()\n", "elif logger == 'ClearML':\n", - " %pip install -q clearml && clearml-init\n", - "elif logger == 'W&B':\n", - " %pip install -q wandb\n", - " import wandb; wandb.login()" + " %pip install -q clearml && clearml-init" ], "metadata": { "id": "i3oKtE4g-aNn" @@ -699,7 +696,7 @@ "YOLOv5 🚀 v6.2-56-g30e674b Python-3.7.13 torch-1.12.1+cu113 CUDA:0 (Tesla V100-SXM2-16GB, 16160MiB)\n", "\n", "\u001b[34m\u001b[1mhyperparameters: \u001b[0mlr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0\n", - "\u001b[34m\u001b[1mWeights & Biases: \u001b[0mrun 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs in Weights & Biases\n", + "\u001b[34m\u001b[1mComet: \u001b[0mrun 'pip install comet' to automatically track and visualize YOLOv5 🚀 runs with Comet\n", "\u001b[34m\u001b[1mClearML: \u001b[0mrun 'pip install clearml' to automatically track, visualize and remotely train YOLOv5 🚀 in ClearML\n", "\u001b[34m\u001b[1mTensorBoard: \u001b[0mStart with 'tensorboard --logdir runs/train', view at http://localhost:6006/\n", "\n", @@ -905,22 +902,6 @@ "id": "Lay2WsTjNJzP" } }, - { - "cell_type": "markdown", - "metadata": { - "id": "DLI1JmHU7B0l" - }, - "source": [ - "## Weights & Biases Logging\n", - "\n", - "[Weights & Biases](https://wandb.ai/site?utm_campaign=repo_yolo_notebook) (W&B) is integrated with YOLOv5 for real-time visualization and cloud logging of training runs. This allows for better run comparison and introspection, as well improved visibility and collaboration for teams. To enable W&B `pip install wandb`, and then train normally (you will be guided through setup on first use). \n", - "\n", - "During training you will see live updates at [https://wandb.ai/home](https://wandb.ai/home?utm_campaign=repo_yolo_notebook), and you can create and share detailed [Reports](https://wandb.ai/glenn-jocher/yolov5_tutorial/reports/YOLOv5-COCO128-Tutorial-Results--VmlldzozMDI5OTY) of your results. For more information see the [YOLOv5 Weights & Biases Tutorial](https://github.com/ultralytics/yolov5/issues/1289). \n", - "\n", - "\n", - "\"Weights" - ] - }, { "cell_type": "markdown", "metadata": { diff --git a/utils/docker/Dockerfile b/utils/docker/Dockerfile index 9b93fad7b203..be5c2fb71517 100644 --- a/utils/docker/Dockerfile +++ b/utils/docker/Dockerfile @@ -16,7 +16,7 @@ RUN apt update && apt install --no-install-recommends -y zip htop screen libgl1- COPY requirements.txt . RUN python -m pip install --upgrade pip wheel RUN pip uninstall -y Pillow torchtext torch torchvision -RUN pip install --no-cache -r requirements.txt albumentations wandb gsutil notebook Pillow>=9.1.0 \ +RUN pip install --no-cache -r requirements.txt albumentations comet clearml gsutil notebook Pillow>=9.1.0 \ 'opencv-python<4.6.0.66' \ --extra-index-url https://download.pytorch.org/whl/cu113 diff --git a/utils/loggers/__init__.py b/utils/loggers/__init__.py index 941d09e19e2d..bc8dd7621579 100644 --- a/utils/loggers/__init__.py +++ b/utils/loggers/__init__.py @@ -84,10 +84,10 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, self.csv = True # always log to csv # Messages - if not wandb: - prefix = colorstr('Weights & Biases: ') - s = f"{prefix}run 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs in Weights & Biases" - self.logger.info(s) + # if not wandb: + # prefix = colorstr('Weights & Biases: ') + # s = f"{prefix}run 'pip install wandb' to automatically track and visualize YOLOv5 🚀 runs in Weights & Biases" + # self.logger.info(s) if not clearml: prefix = colorstr('ClearML: ') s = f"{prefix}run 'pip install clearml' to automatically track, visualize and remotely train YOLOv5 🚀 in ClearML" @@ -110,9 +110,9 @@ def __init__(self, save_dir=None, weights=None, opt=None, hyp=None, logger=None, self.opt.hyp = self.hyp # add hyperparameters self.wandb = WandbLogger(self.opt, run_id) # temp warn. because nested artifacts not supported after 0.12.10 - if pkg.parse_version(wandb.__version__) >= pkg.parse_version('0.12.11'): - s = "YOLOv5 temporarily requires wandb version 0.12.10 or below. Some features may not work as expected." - self.logger.warning(s) + # if pkg.parse_version(wandb.__version__) >= pkg.parse_version('0.12.11'): + # s = "YOLOv5 temporarily requires wandb version 0.12.10 or below. Some features may not work as expected." + # self.logger.warning(s) else: self.wandb = None diff --git a/utils/loggers/wandb/wandb_utils.py b/utils/loggers/wandb/wandb_utils.py index d2dd0fa7c6cd..238f4edbf2a0 100644 --- a/utils/loggers/wandb/wandb_utils.py +++ b/utils/loggers/wandb/wandb_utils.py @@ -135,7 +135,7 @@ def __init__(self, opt, run_id=None, job_type='Training'): # Temporary-fix if opt.upload_dataset: opt.upload_dataset = False - LOGGER.info("Uploading Dataset functionality is not being supported temporarily due to a bug.") + # LOGGER.info("Uploading Dataset functionality is not being supported temporarily due to a bug.") # Pre-training routine -- self.job_type = job_type From 85ae985b6a232f3a3e2f7400243cec2ca0b5f8d1 Mon Sep 17 00:00:00 2001 From: Vladislav Veklenko <71467601+vladoossss@users.noreply.github.com> Date: Thu, 13 Oct 2022 01:44:01 +0200 Subject: [PATCH 655/661] update mask2segments and saving results (#9785) * update mask2segments and saving results * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update general.py Signed-off-by: Glenn Jocher * Update predict.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Glenn Jocher --- utils/segment/general.py | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/utils/segment/general.py b/utils/segment/general.py index 43bdc460f928..b526333dc5a1 100644 --- a/utils/segment/general.py +++ b/utils/segment/general.py @@ -126,9 +126,12 @@ def masks2segments(masks, strategy='largest'): segments = [] for x in masks.int().cpu().numpy().astype('uint8'): c = cv2.findContours(x, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[0] - if strategy == 'concat': # concatenate all segments - c = np.concatenate([x.reshape(-1, 2) for x in c]) - elif strategy == 'largest': # select largest segment - c = np.array(c[np.array([len(x) for x in c]).argmax()]).reshape(-1, 2) + if c: + if strategy == 'concat': # concatenate all segments + c = np.concatenate([x.reshape(-1, 2) for x in c]) + elif strategy == 'largest': # select largest segment + c = np.array(c[np.array([len(x) for x in c]).argmax()]).reshape(-1, 2) + else: + c = np.zeros((0, 2)) # no segments found segments.append(c.astype('float32')) return segments From 16f87bb38e76a5aa14ee93252042063b678ece86 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 13 Oct 2022 02:32:06 +0200 Subject: [PATCH 656/661] HUB VOC fix (#9792) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/general.py | 1 + 1 file changed, 1 insertion(+) diff --git a/utils/general.py b/utils/general.py index e2faca9dbf2a..d9d54d9e4f71 100644 --- a/utils/general.py +++ b/utils/general.py @@ -477,6 +477,7 @@ def check_dataset(data, autodownload=True): path = Path(extract_dir or data.get('path') or '') # optional 'path' default to '.' if not path.is_absolute(): path = (ROOT / path).resolve() + data['path'] = path # download scripts for k in 'train', 'val', 'test': if data.get(k): # prepend path if isinstance(data[k], str): From 15b75659ddc2552bd9239db8a3c940322da49b80 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Thu, 13 Oct 2022 15:27:16 +0200 Subject: [PATCH 657/661] Update hubconf.py local repo Usage example (#9803) * Update hubconf.py Signed-off-by: Glenn Jocher * Update hubconf.py Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- hubconf.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/hubconf.py b/hubconf.py index 95b95a5c30cc..2c6ec13f815c 100644 --- a/hubconf.py +++ b/hubconf.py @@ -4,8 +4,10 @@ Usage: import torch - model = torch.hub.load('ultralytics/yolov5', 'yolov5s') - model = torch.hub.load('ultralytics/yolov5:master', 'custom', 'path/to/yolov5s.onnx') # custom model from branch + model = torch.hub.load('ultralytics/yolov5', 'yolov5s') # official model + model = torch.hub.load('ultralytics/yolov5:master', 'yolov5s') # from branch + model = torch.hub.load('ultralytics/yolov5', 'custom', 'yolov5s.pt') # custom/local model + model = torch.hub.load('.', 'custom', 'yolov5s.pt', source='local') # local repo """ import torch From 2a19d070d8a92bbf44dca8a40c503ec7406228d9 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 14 Oct 2022 12:28:52 +0200 Subject: [PATCH 658/661] Fix xView dataloaders import (#9807) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- data/xView.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/data/xView.yaml b/data/xView.yaml index b134ceac8164..770ab7870449 100644 --- a/data/xView.yaml +++ b/data/xView.yaml @@ -87,7 +87,7 @@ download: | from PIL import Image from tqdm import tqdm - from utils.datasets import autosplit + from utils.dataloaders import autosplit from utils.general import download, xyxy2xywhn From df80e7c723b5722fe5b8d935ace73b8b28572ed4 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Fri, 14 Oct 2022 18:18:58 +0200 Subject: [PATCH 659/661] Argoverse HUB fix (#9809) Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- data/Argoverse.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/data/Argoverse.yaml b/data/Argoverse.yaml index e3e9ba161ed0..558151dc849e 100644 --- a/data/Argoverse.yaml +++ b/data/Argoverse.yaml @@ -63,7 +63,7 @@ download: | # Download - dir = Path('../datasets/Argoverse') # dataset root dir + dir = Path(yaml['path']) # dataset root dir urls = ['https://argoverse-hd.s3.us-east-2.amazonaws.com/Argoverse-HD-Full.zip'] download(urls, dir=dir, delete=False) From e42c89d4efc99bfbd8c5c208ffe67c11632da84a Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Sun, 16 Oct 2022 20:51:32 +0200 Subject: [PATCH 660/661] `smart_optimizer()` revert to weight with decay (#9817) If a parameter does not fall into any other category Signed-off-by: Glenn Jocher Signed-off-by: Glenn Jocher --- utils/torch_utils.py | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/utils/torch_utils.py b/utils/torch_utils.py index 9f257d06ac60..04a3873854ee 100644 --- a/utils/torch_utils.py +++ b/utils/torch_utils.py @@ -319,12 +319,13 @@ def smart_optimizer(model, name='Adam', lr=0.001, momentum=0.9, decay=1e-5): g = [], [], [] # optimizer parameter groups bn = tuple(v for k, v in nn.__dict__.items() if 'Norm' in k) # normalization layers, i.e. BatchNorm2d() for v in model.modules(): - if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter): # bias (no decay) - g[2].append(v.bias) - if isinstance(v, bn): # weight (no decay) - g[1].append(v.weight) - elif hasattr(v, 'weight') and isinstance(v.weight, nn.Parameter): # weight (with decay) - g[0].append(v.weight) + for p_name, p in v.named_parameters(recurse=0): + if p_name == 'bias': # bias (no decay) + g[2].append(p) + elif p_name == 'weight' and isinstance(v, bn): # weight (no decay) + g[1].append(p) + else: + g[0].append(p) # weight (with decay) if name == 'Adam': optimizer = torch.optim.Adam(g[2], lr=lr, betas=(momentum, 0.999)) # adjust beta1 to momentum From e3ff7806769444de864060494d1be8e18ce046a1 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Mon, 17 Oct 2022 14:34:33 +0200 Subject: [PATCH 661/661] Allow PyTorch Hub results to display in notebooks (#9825) * Allow PyTorch Hub results to display in notebooks * fix CI * fix CI * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix CI * fix CI * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix CI * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix CI * fix CI * fix CI * fix CI * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix CI * fix CI * fix CI Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> --- classify/predict.py | 2 +- detect.py | 2 +- models/common.py | 13 +++++++++---- segment/predict.py | 2 +- utils/__init__.py | 2 +- utils/autoanchor.py | 2 +- utils/general.py | 17 +++++++++++++---- utils/metrics.py | 2 +- 8 files changed, 28 insertions(+), 14 deletions(-) diff --git a/classify/predict.py b/classify/predict.py index 9114aab1d703..9373649bf27d 100644 --- a/classify/predict.py +++ b/classify/predict.py @@ -91,7 +91,7 @@ def run( # Dataloader bs = 1 # batch_size if webcam: - view_img = check_imshow() + view_img = check_imshow(warn=True) dataset = LoadStreams(source, img_size=imgsz, transforms=classify_transforms(imgsz[0]), vid_stride=vid_stride) bs = len(dataset) elif screenshot: diff --git a/detect.py b/detect.py index 8f48d8d28000..98af7235ea69 100644 --- a/detect.py +++ b/detect.py @@ -99,7 +99,7 @@ def run( # Dataloader bs = 1 # batch_size if webcam: - view_img = check_imshow() + view_img = check_imshow(warn=True) dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) bs = len(dataset) elif screenshot: diff --git a/models/common.py b/models/common.py index d889d0292c61..e6da429de3e5 100644 --- a/models/common.py +++ b/models/common.py @@ -18,16 +18,20 @@ import requests import torch import torch.nn as nn +from IPython.display import display from PIL import Image from torch.cuda import amp +from utils import TryExcept from utils.dataloaders import exif_transpose, letterbox -from utils.general import (LOGGER, ROOT, Profile, check_requirements, check_suffix, check_version, colorstr, - increment_path, make_divisible, non_max_suppression, scale_boxes, xywh2xyxy, xyxy2xywh, - yaml_load) +from utils.general import (LOGGER, ROOT, Profile, check_imshow, check_requirements, check_suffix, check_version, + colorstr, increment_path, make_divisible, non_max_suppression, scale_boxes, xywh2xyxy, + xyxy2xywh, yaml_load) from utils.plots import Annotator, colors, save_one_box from utils.torch_utils import copy_attr, smart_inference_mode +CHECK_IMSHOW = check_imshow() + def autopad(k, p=None, d=1): # kernel, padding, dilation # Pad to 'same' shape outputs @@ -756,7 +760,7 @@ def _run(self, pprint=False, show=False, save=False, crop=False, render=False, l im = Image.fromarray(im.astype(np.uint8)) if isinstance(im, np.ndarray) else im # from np if show: - im.show(self.files[i]) # show + im.show(self.files[i]) if CHECK_IMSHOW else display(im) if save: f = self.files[i] im.save(save_dir / f) # save @@ -772,6 +776,7 @@ def _run(self, pprint=False, show=False, save=False, crop=False, render=False, l LOGGER.info(f'Saved results to {save_dir}\n') return crops + @TryExcept('Showing images is not supported in this environment') def show(self, labels=True): self._run(show=True, labels=labels) # show results diff --git a/segment/predict.py b/segment/predict.py index 94117cd78633..44d6d3904c19 100644 --- a/segment/predict.py +++ b/segment/predict.py @@ -102,7 +102,7 @@ def run( # Dataloader bs = 1 # batch_size if webcam: - view_img = check_imshow() + view_img = check_imshow(warn=True) dataset = LoadStreams(source, img_size=imgsz, stride=stride, auto=pt, vid_stride=vid_stride) bs = len(dataset) elif screenshot: diff --git a/utils/__init__.py b/utils/__init__.py index 8403a6149827..0afe6f475625 100644 --- a/utils/__init__.py +++ b/utils/__init__.py @@ -23,7 +23,7 @@ def __enter__(self): def __exit__(self, exc_type, value, traceback): if value: - print(emojis(f'{self.msg}{value}')) + print(emojis(f"{self.msg}{': ' if self.msg else ''}{value}")) return True diff --git a/utils/autoanchor.py b/utils/autoanchor.py index 7e7e9985d68a..cfc4c276e3aa 100644 --- a/utils/autoanchor.py +++ b/utils/autoanchor.py @@ -26,7 +26,7 @@ def check_anchor_order(m): m.anchors[:] = m.anchors.flip(0) -@TryExcept(f'{PREFIX}ERROR: ') +@TryExcept(f'{PREFIX}ERROR') def check_anchors(dataset, model, thr=4.0, imgsz=640): # Check anchor fit to data, recompute if necessary m = model.module.model[-1] if hasattr(model, 'module') else model.model[-1] # Detect() diff --git a/utils/general.py b/utils/general.py index d9d54d9e4f71..76bc0b1d7a79 100644 --- a/utils/general.py +++ b/utils/general.py @@ -27,6 +27,7 @@ from zipfile import ZipFile import cv2 +import IPython import numpy as np import pandas as pd import pkg_resources as pkg @@ -73,6 +74,12 @@ def is_colab(): return 'COLAB_GPU' in os.environ +def is_notebook(): + # Is environment a Jupyter notebook? Verified on Colab, Jupyterlab, Kaggle, Paperspace + ipython_type = str(type(IPython.get_ipython())) + return 'colab' in ipython_type or 'zmqshell' in ipython_type + + def is_kaggle(): # Is environment a Kaggle Notebook? return os.environ.get('PWD') == '/kaggle/working' and os.environ.get('KAGGLE_URL_BASE') == 'https://www.kaggle.com' @@ -383,18 +390,20 @@ def check_img_size(imgsz, s=32, floor=0): return new_size -def check_imshow(): +def check_imshow(warn=False): # Check if environment supports image displays try: - assert not is_docker(), 'cv2.imshow() is disabled in Docker environments' - assert not is_colab(), 'cv2.imshow() is disabled in Google Colab environments' + assert not is_notebook() + assert not is_docker() + assert 'NoneType' not in str(type(IPython.get_ipython())) # SSH terminals, GitHub CI cv2.imshow('test', np.zeros((1, 1, 3))) cv2.waitKey(1) cv2.destroyAllWindows() cv2.waitKey(1) return True except Exception as e: - LOGGER.warning(f'WARNING ⚠️ Environment does not support cv2.imshow() or PIL Image.show() image displays\n{e}') + if warn: + LOGGER.warning(f'WARNING ⚠️ Environment does not support cv2.imshow() or PIL Image.show()\n{e}') return False diff --git a/utils/metrics.py b/utils/metrics.py index ed611d7d38fa..f0bc787e1518 100644 --- a/utils/metrics.py +++ b/utils/metrics.py @@ -186,7 +186,7 @@ def tp_fp(self): # fn = self.matrix.sum(0) - tp # false negatives (missed detections) return tp[:-1], fp[:-1] # remove background class - @TryExcept('WARNING ⚠️ ConfusionMatrix plot failure: ') + @TryExcept('WARNING ⚠️ ConfusionMatrix plot failure') def plot(self, normalize=True, save_dir='', names=()): import seaborn as sn