train_ctvis.py

try:
    # ignore ShapelyDeprecationWarning from fvcore
    from shapely.errors import ShapelyDeprecationWarning
    import warnings
    warnings.filterwarnings('ignore')
except ImportError:
    pass

import copy
import itertools
import logging
import os
from collections import OrderedDict
from typing import Any, Dict, List, Set

import torch
import weakref
import detectron2.utils.comm as comm
from detectron2.checkpoint import DetectionCheckpointer
from detectron2.config import get_cfg
from detectron2.data import MetadataCatalog
from detectron2.engine import (
    DefaultTrainer,
    default_argument_parser,
    default_setup,
    launch,
    create_ddp_model,
    AMPTrainer,
    SimpleTrainer
)
from detectron2.evaluation import (
    DatasetEvaluator,
    inference_on_dataset,
    print_csv_format,
    verify_results,
)
from detectron2.projects.deeplab import add_deeplab_config, build_lr_scheduler  # noqa
from detectron2.solver.build import maybe_add_gradient_clipping
from detectron2.utils.logger import setup_logger

# MaskFormer
from mask2former import add_maskformer2_config
from ctvis import (add_ctvis_config,
                   YTVISDatasetMapper,
                   YTVISEvaluator,
                   RotateCOCOVideoDatasetMapper,
                   build_combined_loader,
                   build_detection_train_loader,
                   build_detection_test_loader)


class Trainer(DefaultTrainer):
    """
    Extension of the Trainer class adapted to CTVIS.
    """

    def __init__(self, cfg):
        """
        Args:
            cfg (CfgNode):
        """
        super(DefaultTrainer, self).__init__()
        logger = logging.getLogger("detectron2")
        if not logger.isEnabledFor(logging.INFO):  # setup_logger is not called for d2
            setup_logger()
        cfg = DefaultTrainer.auto_scale_workers(cfg, comm.get_world_size())

        # Assume these objects must be constructed in this order.
        model = self.build_model(cfg)
        optimizer = self.build_optimizer(cfg, model)
        data_loader = self.build_train_loader(cfg)

        model = create_ddp_model(model, broadcast_buffers=False, find_unused_parameters=cfg.FIND_UNUSED_PARAMETERS)
        self._trainer = (AMPTrainer if cfg.SOLVER.AMP.ENABLED else SimpleTrainer)(
            model, data_loader, optimizer
        )

        self.scheduler = self.build_lr_scheduler(cfg, optimizer)
        self.checkpointer = DetectionCheckpointer(
            # Assume you want to save checkpoints together with logs/statistics
            model,
            cfg.OUTPUT_DIR,
            trainer=weakref.proxy(self),
        )
        self.start_iter = 0
        self.max_iter = cfg.SOLVER.MAX_ITER
        self.cfg = cfg

        self.register_hooks(self.build_hooks())

    @classmethod
    def build_evaluator(cls, cfg, dataset_name, output_folder=None):
        """
            Create evaluator(s) for a given dataset.
            This uses the special metadata "evaluator_type" associated with each builtin dataset.
            For your own dataset, you can simply create an evaluator manually in your
            script and do not have to worry about the hacky if-else logic here.
        """
        if output_folder is None:
            output_folder = os.path.join(cfg.OUTPUT_DIR, "inference")
            os.makedirs(output_folder, exist_ok=True)
        evaluator_type = MetadataCatalog.get(dataset_name).evaluator_type
        if evaluator_type == "coco":
            raise NotImplementedError
        elif evaluator_type == "ytvis":
            DatasetEvaluator = YTVISEvaluator
        return DatasetEvaluator(dataset_name, cfg, True, output_folder)

    @classmethod
    def build_train_loader(cls, cfg):
        mappers = []
        # if dataset_name.startswith('coco'):
        for d_i, dataset_name in enumerate(cfg.DATASETS.TRAIN):
            if dataset_name.startswith('coco'):
                mappers.append(RotateCOCOVideoDatasetMapper(cfg, is_train=True, tgt_dataset_name=cfg.DATASETS.TRAIN[-1],
                                                            src_dataset_name=dataset_name))  # noqa
            elif dataset_name.startswith('ytvis') or dataset_name.startswith('ovis'):
                mappers.append(YTVISDatasetMapper(cfg, is_train=True))
            else:
                raise NotImplementedError

        assert len(mappers) > 0, "No dataset is chosen!"

        if len(mappers) == 1:
            mapper = mappers[0]
            return build_detection_train_loader(cfg, mapper=mapper, dataset_name=cfg.DATASETS.TRAIN[0])
        else:
            loaders = [
                build_detection_train_loader(cfg, mapper=mapper, dataset_name=dataset_name)
                for mapper, dataset_name in zip(mappers, cfg.DATASETS.TRAIN)
            ]
            combined_data_loader = build_combined_loader(cfg, loaders, cfg.DATASETS.DATASET_RATIO)
            return combined_data_loader

    @classmethod
    def build_test_loader(cls, cfg, dataset_name):
        dataset_name = cfg.DATASETS.TEST[0]
        if dataset_name.startswith('coco'):
            raise NotImplementedError
        elif dataset_name.startswith('ytvis') or dataset_name.startswith('ovis'):
            mapper = YTVISDatasetMapper(cfg, is_train=False)
        return build_detection_test_loader(cfg, dataset_name, mapper=mapper)

    @classmethod
    def build_lr_scheduler(cls, cfg, optimizer):
        """
        It now calls :func:`detectron2.solver.build_lr_scheduler`.
        Overwrite it if you'd like a different scheduler.
        """
        return build_lr_scheduler(cfg, optimizer)

    @classmethod
    def build_optimizer(cls, cfg, model):
        weight_decay_norm = cfg.SOLVER.WEIGHT_DECAY_NORM
        weight_decay_embed = cfg.SOLVER.WEIGHT_DECAY_EMBED

        defaults = {}
        defaults["lr"] = cfg.SOLVER.BASE_LR
        defaults["weight_decay"] = cfg.SOLVER.WEIGHT_DECAY

        norm_module_types = (
            torch.nn.BatchNorm1d,
            torch.nn.BatchNorm2d,
            torch.nn.BatchNorm3d,
            torch.nn.SyncBatchNorm,
            # NaiveSyncBatchNorm inherits from BatchNorm2d
            torch.nn.GroupNorm,
            torch.nn.InstanceNorm1d,
            torch.nn.InstanceNorm2d,
            torch.nn.InstanceNorm3d,
            torch.nn.LayerNorm,
            torch.nn.LocalResponseNorm,
        )

        params: List[Dict[str, Any]] = []
        memo: Set[torch.nn.parameter.Parameter] = set()
        for module_name, module in model.named_modules():
            for module_param_name, value in module.named_parameters(recurse=False):
                if not value.requires_grad:
                    continue
                # Avoid duplicating parameters
                if value in memo:
                    continue
                memo.add(value)

                hyperparams = copy.copy(defaults)
                if "backbone" in module_name:
                    hyperparams["lr"] = hyperparams["lr"] * cfg.SOLVER.BACKBONE_MULTIPLIER
                elif "reid_embed" not in module_name:
                    hyperparams["lr"] = hyperparams["lr"] * cfg.SOLVER.DETECTOR_MULTIPLIER
                else:
                    assert "reid_embed" in module_name, "Build optimizer failed!"

                if (
                        "relative_position_bias_table" in module_param_name
                        or "absolute_pos_embed" in module_param_name
                ):
                    print(module_param_name)
                    hyperparams["weight_decay"] = 0.0
                if isinstance(module, norm_module_types):
                    hyperparams["weight_decay"] = weight_decay_norm
                if isinstance(module, torch.nn.Embedding):
                    hyperparams["weight_decay"] = weight_decay_embed
                params.append({"params": [value], **hyperparams})

        def maybe_add_full_model_gradient_clipping(optim):
            # detectron2 doesn't have full model gradient clipping now
            clip_norm_val = cfg.SOLVER.CLIP_GRADIENTS.CLIP_VALUE
            enable = (
                    cfg.SOLVER.CLIP_GRADIENTS.ENABLED
                    and cfg.SOLVER.CLIP_GRADIENTS.CLIP_TYPE == "full_model"
                    and clip_norm_val > 0.0
            )

            class FullModelGradientClippingOptimizer(optim):
                def step(self, closure=None):
                    all_params = itertools.chain(*[x["params"] for x in self.param_groups])
                    torch.nn.utils.clip_grad_norm_(all_params, clip_norm_val)
                    super().step(closure=closure)

            return FullModelGradientClippingOptimizer if enable else optim

        optimizer_type = cfg.SOLVER.OPTIMIZER
        if optimizer_type == "SGD":
            optimizer = maybe_add_full_model_gradient_clipping(torch.optim.SGD)(
                params, cfg.SOLVER.BASE_LR, momentum=cfg.SOLVER.MOMENTUM
            )
        elif optimizer_type == "ADAMW":
            optimizer = maybe_add_full_model_gradient_clipping(torch.optim.AdamW)(
                params, cfg.SOLVER.BASE_LR
            )
        else:
            raise NotImplementedError(f"no optimizer type {optimizer_type}")
        if not cfg.SOLVER.CLIP_GRADIENTS.CLIP_TYPE == "full_model":
            optimizer = maybe_add_gradient_clipping(cfg, optimizer)
        return optimizer

    @classmethod
    def test(cls, cfg, model, evaluators=None):
        """
        Evaluate the given model. The given model is expected to already contain
        weights to evaluate.
        Args:
            cfg (CfgNode):
            model (nn.Module):
            evaluators (list[DatasetEvaluator] or None): if None, will call
                :meth:`build_evaluator`. Otherwise, must have the same length as
                ``cfg.DATASETS.TEST``.
        Returns:
            dict: a dict of result metrics
        """
        from torch.cuda.amp import autocast
        logger = logging.getLogger(__name__)
        if isinstance(evaluators, DatasetEvaluator):
            evaluators = [evaluators]
        if evaluators is not None:
            assert len(cfg.DATASETS.TEST) == len(evaluators), "{} != {}".format(
                len(cfg.DATASETS.TEST), len(evaluators)
            )

        results = OrderedDict()
        for idx, dataset_name in enumerate(cfg.DATASETS.TEST):
            data_loader = cls.build_test_loader(cfg, dataset_name)
            # When evaluators are passed in as arguments,
            # implicitly assume that evaluators can be created before data_loader.
            if evaluators is not None:
                evaluator = evaluators[idx]
            else:
                try:
                    evaluator = cls.build_evaluator(cfg, dataset_name)
                except NotImplementedError:
                    logger.warning(
                        "No evaluator found. Use `DefaultTrainer.test(evaluators=)`, "
                        "or implement its `build_evaluator` method."
                    )
                    results[dataset_name] = {}
                    continue
            with autocast():
                results_i = inference_on_dataset(model, data_loader, evaluator)
            results[dataset_name] = results_i
            if comm.is_main_process():
                assert isinstance(
                    results_i, dict
                ), "Evaluator must return a dict on the main process. Got {} instead.".format(
                    results_i
                )
                logger.info("Evaluation results for {} in csv format:".format(dataset_name))
                print_csv_format(results_i)

        if len(results) == 1:
            results = list(results.values())[0]
        return results


def setup(args):
    cfg = get_cfg()
    # for poly lr schedule
    add_deeplab_config(cfg)
    add_maskformer2_config(cfg)
    add_ctvis_config(cfg)
    cfg.merge_from_file(args.config_file)
    cfg.merge_from_list(args.opts)
    cfg.freeze()
    default_setup(cfg, args)
    # Setup logger for "mask_former" module
    setup_logger(name="mask2former")
    setup_logger(output=cfg.OUTPUT_DIR, distributed_rank=comm.get_rank(), name="ctvis")
    return cfg


def main(args):
    cfg = setup(args)

    if args.eval_only:
        model = Trainer.build_model(cfg)
        DetectionCheckpointer(model, save_dir=cfg.OUTPUT_DIR).resume_or_load(
            cfg.MODEL.WEIGHTS, resume=args.resume
        )
        res = Trainer.test(cfg, model)
        if cfg.TEST.AUG.ENABLED:
            raise NotImplementedError
        if comm.is_main_process():
            verify_results(cfg, res)
        return res

    trainer = Trainer(cfg)
    trainer.resume_or_load(resume=args.resume)
    return trainer.train()


if __name__ == "__main__":
    args = default_argument_parser().parse_args()
    print("Command Line Args:", args)

    launch(
        main,
        args.num_gpus,
        num_machines=args.num_machines,
        machine_rank=args.machine_rank,
        dist_url=args.dist_url,
        args=(args,))