main_tiny_imagenet.py

# Copyright (c) 2015-present, Facebook, Inc.
# All rights reserved.
import argparse
import datetime
import numpy as np
import time
import torch
import torch.backends.cudnn as cudnn
import json

from pathlib import Path

from timm.data import Mixup
from timm.models import create_model
from timm.loss import LabelSmoothingCrossEntropy, SoftTargetCrossEntropy
from timm.scheduler import create_scheduler
from timm.optim import create_optimizer
from timm.utils import NativeScaler, get_state_dict, ModelEma

from datasets import build_dataset
from trainer import train_one_epoch, evaluate
from losses import DistillationLoss
from samplers import RASampler
import models
import utils
from imagenet_model import ShiftingTransformer
import os.path
from os import path

def get_args_parser():
    parser = argparse.ArgumentParser('DeiT training and evaluation script', add_help=False)
    parser.add_argument('--batch-size', default=64, type=int)
    parser.add_argument('--epochs', default=300, type=int)

    # Model parameters
    parser.add_argument('--model', default='deit_base_patch16_224', type=str, metavar='MODEL',
                        help='Name of model to train')
    parser.add_argument('--input-size', default=32, type=int, help='images input size')

    parser.add_argument('--drop', type=float, default=0.0, metavar='PCT',
                        help='Dropout rate (default: 0.)')
    parser.add_argument('--drop-path', type=float, default=0.2, metavar='PCT',
                        help='Drop path rate (default: 0.1)')

    parser.add_argument('--model-ema', action='store_true')
    parser.add_argument('--no-model-ema', action='store_false', dest='model_ema')
    parser.set_defaults(model_ema=False)
    parser.add_argument('--model-ema-decay', type=float, default=0.99996, help='')
    parser.add_argument('--model-ema-force-cpu', action='store_true', default=False, help='')

    # Optimizer parameters
    parser.add_argument('--opt', default='adamw', type=str, metavar='OPTIMIZER',
                        help='Optimizer (default: "adamw"')
    parser.add_argument('--opt-eps', default=1e-8, type=float, metavar='EPSILON',
                        help='Optimizer Epsilon (default: 1e-8)')
    parser.add_argument('--opt-betas', default=None, type=float, nargs='+', metavar='BETA',
                        help='Optimizer Betas (default: None, use opt default)')
    parser.add_argument('--clip-grad', type=float, default=None, metavar='NORM',
                        help='Clip gradient norm (default: None, no clipping)')
    parser.add_argument('--momentum', type=float, default=0.9, metavar='M',
                        help='SGD momentum (default: 0.9)')
    parser.add_argument('--weight-decay', type=float, default=0.05,
                        help='weight decay (default: 0.05)')
    # Learning rate schedule parameters
    parser.add_argument('--sched', default='cosine', type=str, metavar='SCHEDULER',
                        help='LR scheduler (default: "cosine"')
    parser.add_argument('--lr', type=float, default=5e-4, metavar='LR',
                        help='learning rate (default: 2.5e-4)')
    parser.add_argument('--lr-noise', type=float, nargs='+', default=None, metavar='pct, pct',
                        help='learning rate noise on/off epoch percentages')
    parser.add_argument('--lr-noise-pct', type=float, default=0.67, metavar='PERCENT',
                        help='learning rate noise limit percent (default: 0.67)')
    parser.add_argument('--lr-noise-std', type=float, default=1.0, metavar='STDDEV',
                        help='learning rate noise std-dev (default: 1.0)')
    parser.add_argument('--warmup-lr', type=float, default=5e-7, metavar='LR',
                        help='warmup learning rate (default: 1e-6)')
    parser.add_argument('--min-lr', type=float, default=5e-6, metavar='LR',
                        help='lower lr bound for cyclic schedulers that hit 0 (1e-5)')

    parser.add_argument('--decay-epochs', type=float, default=30, metavar='N',
                        help='epoch interval to decay LR')
    parser.add_argument('--warmup-epochs', type=int, default=20, metavar='N',
                        help='epochs to warmup LR, if scheduler supports')
    parser.add_argument('--cooldown-epochs', type=int, default=10, metavar='N',
                        help='epochs to cooldown LR at min_lr, after cyclic schedule ends')
    parser.add_argument('--patience-epochs', type=int, default=10, metavar='N',
                        help='patience epochs for Plateau LR scheduler (default: 10')
    parser.add_argument('--decay-rate', '--dr', type=float, default=0.1, metavar='RATE',
                        help='LR decay rate (default: 0.1)')

    # Augmentation parameters
    parser.add_argument('--color-jitter', type=float, default=0.4, metavar='PCT',
                        help='Color jitter factor (default: 0.4)')
    parser.add_argument('--aa', type=str, default='rand-m9-mstd0.5-inc1', metavar='NAME',
                        help='Use AutoAugment policy. "v0" or "original". " + \
                             "(default: rand-m9-mstd0.5-inc1)'),
    parser.add_argument('--smoothing', type=float, default=0.1, help='Label smoothing (default: 0.1)')
    parser.add_argument('--train-interpolation', type=str, default='bicubic',
                        help='Training interpolation (random, bilinear, bicubic default: "bicubic")')

    parser.add_argument('--repeated-aug', action='store_true')
    parser.add_argument('--no-repeated-aug', action='store_false', dest='repeated_aug')
    parser.set_defaults(repeated_aug=False)

    # * Random Erase params
    parser.add_argument('--reprob', type=float, default=0.25, metavar='PCT',
                        help='Random erase prob (default: 0.25)')
    parser.add_argument('--remode', type=str, default='pixel',
                        help='Random erase mode (default: "pixel")')
    parser.add_argument('--recount', type=int, default=1,
                        help='Random erase count (default: 1)')
    parser.add_argument('--resplit', action='store_true', default=False,
                        help='Do not random erase first (clean) augmentation split')

    # * Mixup params
    parser.add_argument('--mixup', type=float, default=0.8,
                        help='mixup alpha, mixup enabled if > 0. (default: 0.8)')
    parser.add_argument('--cutmix', type=float, default=1.0,
                        help='cutmix alpha, cutmix enabled if > 0. (default: 1.0)')
    parser.add_argument('--cutmix-minmax', type=float, nargs='+', default=None,
                        help='cutmix min/max ratio, overrides alpha and enables cutmix if set (default: None)')
    parser.add_argument('--mixup-prob', type=float, default=1.0,
                        help='Probability of performing mixup or cutmix when either/both is enabled')
    parser.add_argument('--mixup-switch-prob', type=float, default=0.5,
                        help='Probability of switching to cutmix when both mixup and cutmix enabled')
    parser.add_argument('--mixup-mode', type=str, default='batch',
                        help='How to apply mixup/cutmix params. Per "batch", "pair", or "elem"')
    # Distillation parameters
    parser.add_argument('--teacher-model', default='regnety_160', type=str, metavar='MODEL',
                        help='Name of teacher model to train (default: "regnety_160"')
    parser.add_argument('--teacher-path', type=str, default='')
    parser.add_argument('--distillation-type', default='none', choices=['none', 'soft', 'hard'], type=str, help="")
    parser.add_argument('--distillation-alpha', default=0.5, type=float, help="")
    parser.add_argument('--distillation-tau', default=1.0, type=float, help="")

    # * Finetuning params
    parser.add_argument('--finetune', default='', help='finetune from checkpoint')

    # Dataset parameters
    parser.add_argument('--data-path', default='/datasets01/imagenet_full_size/061417/', type=str,
                        help='dataset path')
    parser.add_argument('--data-set', default='IMNET', choices=['cifar10', 'cifar100', 'IMNET', 'INAT', 'INAT19', 'IMNET_SUBSET'],
                        type=str, help='Image Net dataset path')
    parser.add_argument('--inat-category', default='name',
                        choices=['kingdom', 'phylum', 'class', 'order', 'supercategory', 'family', 'genus', 'name'],
                        type=str, help='semantic granularity')
    parser.add_argument('--data-seed', type=int, default=1020)

    parser.add_argument('--output_dir', default='',
                        help='path where to save, empty for no saving')
    parser.add_argument('--device', default='cuda',
                        help='device to use for training / testing')
    parser.add_argument('--seed', default=0, type=int)
    parser.add_argument('--resume', default='', help='resume from checkpoint')
    parser.add_argument('--start_epoch', default=0, type=int, metavar='N',
                        help='start epoch')
    parser.add_argument('--eval', action='store_true', help='Perform evaluation only')
    parser.add_argument('--dist-eval', action='store_true', default=False, help='Enabling distributed evaluation')
    parser.add_argument('--num_workers', default=10, type=int)
    parser.add_argument('--pin-mem', action='store_true',
                        help='Pin CPU memory in DataLoader for more efficient (sometimes) transfer to GPU.')
    parser.add_argument('--no-pin-mem', action='store_false', dest='pin_mem',
                        help='')
    parser.set_defaults(pin_mem=True)

    # distributed training parameters
    parser.add_argument('--world_size', default=1, type=int,
                        help='number of distributed processes')
    parser.add_argument('--dist_url', default='env://', help='url used to set up distributed training')

    #our model params
    parser.add_argument('--patch-size', default=7, type=int)
    parser.add_argument('--kernel-size', default=9, type=int)
    parser.add_argument('--stride-size', default=7, type=int)
    parser.add_argument('--padding-size', default=1, type=int)

    parser.add_argument('--scaling-factor', default=4, type=int)
    parser.add_argument('--num-variants', default=10, type=int)
    parser.add_argument('--embedding-dim', default=192, type=int)
    parser.add_argument('--num-heads', default=3, type=int)
    parser.add_argument('--num-hierarchies', default=4, type=int)
    return parser


def main(args):
    utils.init_distributed_mode(args)

    print(args)

    if args.distillation_type != 'none' and args.finetune and not args.eval:
        raise NotImplementedError("Finetuning with distillation not yet supported")

    device = torch.device(args.device)

    # fix the seed for reproducibility
    seed = args.seed + utils.get_rank()
    torch.manual_seed(seed)
    np.random.seed(seed)
    # random.seed(seed)

    cudnn.benchmark = True
    if args.data_set == 'cifar10' or args.data_set == 'cifar100':
        args.input_size = 32
    elif args.data_set == 'IMNET' or args.data_set == 'IMNET_SUBSET':
        args.input_size = 224
    dataset_train, args.nb_classes = build_dataset(is_train=True, args=args)
    dataset_val, _ = build_dataset(is_train=False, args=args)
    print("train dataset: ", len(dataset_train))

    if True:  # args.distributed:
        num_tasks = utils.get_world_size()
        global_rank = utils.get_rank()
        if args.repeated_aug:
            sampler_train = RASampler(
                dataset_train, num_replicas=num_tasks, rank=global_rank, shuffle=True
            )
        else:
            print("no repeated augmentations")
            sampler_train = torch.utils.data.DistributedSampler(
                dataset_train, num_replicas=num_tasks, rank=global_rank, shuffle=True
            )
        if args.dist_eval:
            if len(dataset_val) % num_tasks != 0:
                print('Warning: Enabling distributed evaluation with an eval dataset not divisible by process number. '
                      'This will slightly alter validation results as extra duplicate entries are added to achieve '
                      'equal num of samples per-process.')
            sampler_val = torch.utils.data.DistributedSampler(
                dataset_val, num_replicas=num_tasks, rank=global_rank, shuffle=False)
        else:
            sampler_val = torch.utils.data.SequentialSampler(dataset_val)
    else:
        sampler_train = torch.utils.data.RandomSampler(dataset_train)
        sampler_val = torch.utils.data.SequentialSampler(dataset_val)

    data_loader_train = torch.utils.data.DataLoader(
        dataset_train, sampler=sampler_train,
        batch_size=args.batch_size,
        num_workers=args.num_workers,
        pin_memory=False,
        drop_last=True,
    )

    data_loader_val = torch.utils.data.DataLoader(
        dataset_val, sampler=sampler_val,
        batch_size=int(1.5 * args.batch_size),
        num_workers=args.num_workers,
        pin_memory=False,
        drop_last=False
    )

    mixup_fn = None
    mixup_active = args.mixup > 0 or args.cutmix > 0. or args.cutmix_minmax is not None
    if mixup_active:
        mixup_fn = Mixup(
            mixup_alpha=args.mixup, cutmix_alpha=args.cutmix, cutmix_minmax=args.cutmix_minmax,
            prob=args.mixup_prob, switch_prob=args.mixup_switch_prob, mode=args.mixup_mode,
            label_smoothing=args.smoothing, num_classes=args.nb_classes)

    print(f"Creating model: {args.model}")


    image_size = args.input_size
    patch_size = args.patch_size
    embedding_dim = ((192, 192), (192, 192), (192, 192), (192, 192))
    heads = (4, 4, 4, 4)
    num_hierarchies = 4  # number of hierarchies
    num_layers_per_block = (2, 4, 4,4)  # the number of transformer blocks at each heirarchy, starting from the bottom
    assert len(num_layers_per_block)  == num_hierarchies, 'the number of levels does not correspond to the num layer per block'

    num_classes = args.nb_classes
    kernel_size = args.kernel_size
    stride_size = args.stride_size
    padding_size = args.padding_size
    init_patch_embed_size = 1
    scaling_factor = 4
    num_variants=args.num_variants
    model = ShiftingTransformer(
        scaling_factor=scaling_factor,
        output_dir=args.output_dir,
        num_variants=num_variants,
        image_size=image_size,
        patch_size=patch_size,
        embedding_dim=embedding_dim,
        heads=heads,
        num_hierarchies=num_hierarchies,  # number of hierarchies
        num_layers_per_block=num_layers_per_block,  # the number of transformer blocks at each heirarchy, starting from the bottom
        num_classes=num_classes,
        init_patch_embed_size =init_patch_embed_size,
        kernel_size = kernel_size,
        stride_size = stride_size,
        padding_size = padding_size
    )
    print( "image size: ", image_size, "patch size: ", patch_size, "kernel size: ", kernel_size, "stide size: ", stride_size,
          "padding size: ", padding_size, "embedding_dim: ", embedding_dim, "num heads: ", heads, "num hierarchies: ", num_hierarchies,
          "num layers per block: ", num_layers_per_block, "num classes: ", num_classes)
    with open(str(args.output_dir) + "/architecture_configuration.txt", "w+") as text_file:
        text_file.write(str(args))
        text_file.write("\n")
        text_file.write("\n")
        text_file.write(" image size: ")
        text_file.write(str(image_size))
        text_file.write(" patch size: ")
        text_file.write(str(patch_size))
        text_file.write(" kernel size: ")
        text_file.write(str(kernel_size))
        text_file.write(" stide size: ")
        text_file.write(str(stride_size))
        text_file.write("\n")
        text_file.write( "padding size: ")
        text_file.write(str(padding_size))
        text_file.write(" embedding_dim: ")
        text_file.write(str(embedding_dim))
        text_file.write(" num heads: ")
        text_file.write(str(heads))
        text_file.write("\n")
        text_file.write(" num hierarchies: ")
        text_file.write(str(num_hierarchies))
        text_file.write(" num layers per block: ")
        text_file.write(str(num_layers_per_block))
        text_file.write(" num classes: ")
        text_file.write(str(num_classes))

    if args.finetune:
        if args.finetune.startswith('https'):
            checkpoint = torch.hub.load_state_dict_from_url(
                args.finetune, map_location='cpu', check_hash=True)
        else:
            checkpoint = torch.load(args.finetune, map_location='cpu')

        checkpoint_model = checkpoint['model']
        state_dict = model.state_dict()
        for k in ['head.weight', 'head.bias', 'head_dist.weight', 'head_dist.bias']:
            if k in checkpoint_model and checkpoint_model[k].shape != state_dict[k].shape:
                print(f"Removing key {k} from pretrained checkpoint")
                del checkpoint_model[k]

        # interpolate position embedding
        pos_embed_checkpoint = checkpoint_model['pos_embed']
        embedding_size = pos_embed_checkpoint.shape[-1]
        num_patches = model.patch_embed.num_patches
        num_extra_tokens = model.pos_embed.shape[-2] - num_patches
        # height (== width) for the checkpoint position embedding
        orig_size = int((pos_embed_checkpoint.shape[-2] - num_extra_tokens) ** 0.5)
        # height (== width) for the new position embedding
        new_size = int(num_patches ** 0.5)
        # class_token and dist_token are kept unchanged
        extra_tokens = pos_embed_checkpoint[:, :num_extra_tokens]
        # only the position tokens are interpolated
        pos_tokens = pos_embed_checkpoint[:, num_extra_tokens:]
        pos_tokens = pos_tokens.reshape(-1, orig_size, orig_size, embedding_size).permute(0, 3, 1, 2)
        pos_tokens = torch.nn.functional.interpolate(
            pos_tokens, size=(new_size, new_size), mode='bicubic', align_corners=False)
        pos_tokens = pos_tokens.permute(0, 2, 3, 1).flatten(1, 2)
        new_pos_embed = torch.cat((extra_tokens, pos_tokens), dim=1)
        checkpoint_model['pos_embed'] = new_pos_embed

        model.load_state_dict(checkpoint_model, strict=False)

    model.to(device)
    # new=torch.zeros((3,32,32)).to(device)
    model_ema = None
    if args.model_ema:
        print("model ema true")
        # Important to create EMA model after cuda(), DP wrapper, and AMP but before SyncBN and DDP wrapper
        model_ema = ModelEma(
            model,
            decay=args.model_ema_decay,
            device='cpu' if args.model_ema_force_cpu else '',
            resume='')

    model_without_ddp = model
    if args.distributed:
        model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.gpu], find_unused_parameters=False)
        model_without_ddp = model.module
    n_parameters = sum(p.numel() for p in model.parameters() if p.requires_grad)

    linear_scaled_lr = args.lr * args.batch_size * utils.get_world_size() / 512
    linear_scaled_warmup_lr = args.warmup_lr * args.batch_size * utils.get_world_size() / 512.0
    linear_scaled_min_lr = args.min_lr * args.batch_size * utils.get_world_size() / 512.0
    args.lr = linear_scaled_lr
    args.warmup_lr = linear_scaled_warmup_lr
    args.min_lr = linear_scaled_min_lr

    print("lr: ", args.lr)
    print('number of params:', n_parameters)

    with open(str(args.output_dir) + "/architecture_configuration.txt", "a+") as text_file:
        text_file.write('\nnumber of params: ')
        text_file.write(str(n_parameters))
        text_file.write("\nlr: ")
        text_file.write(str(args.lr))
        text_file.close()

    optimizer = create_optimizer(args, model_without_ddp)
    loss_scaler = NativeScaler()

    lr_scheduler, _ = create_scheduler(args, optimizer)

    criterion = LabelSmoothingCrossEntropy()

    if args.mixup > 0.:
        # smoothing is handled with mixup label transform
        criterion = SoftTargetCrossEntropy()
    elif args.smoothing:
        criterion = LabelSmoothingCrossEntropy(smoothing=args.smoothing)
    else:
        criterion = torch.nn.CrossEntropyLoss()

    teacher_model = None
    if args.distillation_type != 'none':
        assert args.teacher_path, 'need to specify teacher-path when using distillation'
        print(f"Creating teacher model: {args.teacher_model}")
        teacher_model = create_model(
            args.teacher_model,
            pretrained=False,
            num_classes=args.nb_classes,
            global_pool='avg',
        )
        if args.teacher_path.startswith('https'):
            checkpoint = torch.hub.load_state_dict_from_url(
                args.teacher_path, map_location='cpu', check_hash=True)
        else:
            checkpoint = torch.load(args.teacher_path, map_location='cpu')
        teacher_model.load_state_dict(checkpoint['model'])
        teacher_model.to(device)
        teacher_model.eval()

    # wrap the criterion in our custom DistillationLoss, which
    # just dispatches to the original criterion if args.distillation_type is 'none'
    criterion = DistillationLoss(
        criterion, teacher_model, args.distillation_type, args.distillation_alpha, args.distillation_tau
    )
    output_dir = Path(args.output_dir)
    if args.resume or path.exists(str(args.output_dir)+'/checkpoint.pth'):
        if args.resume.startswith('https'):
            checkpoint = torch.hub.load_state_dict_from_url(
                args.resume, map_location='cpu', check_hash=True)
        elif "epoch" in args.resume:
            checkpoint = torch.load(args.resume, map_location='cpu')
        else:
            # if path.exists(str(args.output_dir)+'/checkpoint_last.pth'):
            #     args.resume = str(args.output_dir) +'/checkpoint_last.pth'
            # else:
            args.resume = str(args.output_dir) +'/checkpoint.pth'
            checkpoint = torch.load(args.resume, map_location='cpu')
        model_without_ddp.load_state_dict(checkpoint['model'])
        for name, param in model.named_parameters():
            if param.requires_grad:
                print (name)
        if not args.eval and 'optimizer' in checkpoint and 'lr_scheduler' in checkpoint and 'epoch' in checkpoint:
            optimizer.load_state_dict(checkpoint['optimizer'])
            lr_scheduler.load_state_dict(checkpoint['lr_scheduler'])
            args.start_epoch = checkpoint['epoch'] + 1
            if args.model_ema:
                utils._load_checkpoint_for_ema(model_ema, checkpoint['model_ema'])
            if 'scaler' in checkpoint:
                loss_scaler.load_state_dict(checkpoint['scaler'])

    if args.eval:
        test_stats = evaluate(data_loader_val, model, device)
        print(f"Accuracy of the network on the {len(dataset_val)} test images: {test_stats['acc1']:.1f}%")
        return

    print(f"Start training for {args.epochs} epochs")
    start_time = time.time()
    max_accuracy = 0.0
    # for name, param in model.named_parameters():
    #     if param.requires_grad:
    #         print(name)
    for epoch in range(args.start_epoch, args.epochs):
        if args.distributed:
            data_loader_train.sampler.set_epoch(epoch)

        train_stats = train_one_epoch(args.model,
            model, criterion, data_loader_train,
            optimizer, device, epoch, loss_scaler,
            args.clip_grad, model_ema, mixup_fn,
            set_training_mode=args.finetune == ''  # keep in eval mode during finetuning
        )

        lr_scheduler.step(epoch)
        # save_iter = epoch%100
        if args.output_dir:
            if epoch%20 == 0:
                checkpoint_paths = [str(output_dir) + '/checkpoint.pth',
                                    str(output_dir) + '/checkpoint_epoch_' + str(epoch) + '.pth']
                for checkpoint_path in checkpoint_paths:
                    utils.save_on_master({
                        'model': model_without_ddp.state_dict(),
                        'optimizer': optimizer.state_dict(),
                        'lr_scheduler': lr_scheduler.state_dict(),
                        'epoch': epoch,
                        'scaler': loss_scaler.state_dict(),
                        'args': args,
                    }, checkpoint_path)
            else:
                checkpoint_paths = [str(output_dir) + '/checkpoint.pth']
                for checkpoint_path in checkpoint_paths:
                    utils.save_on_master({
                        'model': model_without_ddp.state_dict(),
                        'optimizer': optimizer.state_dict(),
                        'lr_scheduler': lr_scheduler.state_dict(),
                        'epoch': epoch,
                        'scaler': loss_scaler.state_dict(),
                        'args': args,
                    }, checkpoint_path)

        test_stats = evaluate(data_loader_val, model, device)
        print(f"Accuracy of the network on the {len(dataset_val)} test images: {test_stats['acc1']:.1f}%")
        max_accuracy = max(max_accuracy, test_stats["acc1"])
        print(f'Max accuracy: {max_accuracy:.2f}%')

        log_stats = {**{f'train_{k}': v for k, v in train_stats.items()},
        **{f'test_{k}': v for k, v in test_stats.items()},
                     'epoch': epoch,
                     'n_parameters': n_parameters}

        if args.output_dir and utils.is_main_process():
            with (output_dir / "log.txt").open("a") as f:
                f.write(json.dumps(log_stats) + "\n")

    total_time = time.time() - start_time
    total_time_str = str(datetime.timedelta(seconds=int(total_time)))
    print('Training time {}'.format(total_time_str))


if __name__ == '__main__':
    parser = argparse.ArgumentParser('DeiT training and evaluation script', parents=[get_args_parser()])
    args = parser.parse_args()
    if args.output_dir:
        Path(args.output_dir).mkdir(parents=True, exist_ok=True)
    main(args)