train_net.py

#!/usr/bin/env python3
# Copyright (c) Facebook, Inc. and its affiliates.

"""
DeepLab Training Script.

This script is a simplified version of the training script in detectron2/tools.
"""

import os
import torch
import itertools 
import copy
 
from typing import Any, Dict, List, Set

import detectron2.utils.comm as comm
import detectron2.data.transforms as T
from detectron2.checkpoint import DetectionCheckpointer
from detectron2.config import get_cfg
from detectron2.data import DatasetMapper, MetadataCatalog, build_detection_train_loader
from detectron2.engine import DefaultTrainer, default_argument_parser, default_setup, launch
from detectron2.evaluation import CityscapesSemSegEvaluator, DatasetEvaluators, SemSegEvaluator
from detectron2.solver.build import maybe_add_gradient_clipping
from detectron2.utils.logger import setup_logger
from detectron2.projects.point_rend import ColorAugSSDTransform
from detectron2.projects.deeplab import add_deeplab_config, build_lr_scheduler
from pna.config import add_pna_config
def build_sem_seg_train_aug(cfg):
    augs = [
        T.ResizeShortestEdge(
            cfg.INPUT.MIN_SIZE_TRAIN, cfg.INPUT.MAX_SIZE_TRAIN, cfg.INPUT.MIN_SIZE_TRAIN_SAMPLING
        )
    ] 
    if cfg.INPUT.CROP.ENABLED: 
        augs.append(
            T.RandomCrop_CategoryAreaConstraint(
                cfg.INPUT.CROP.TYPE,
                cfg.INPUT.CROP.SIZE,
                cfg.INPUT.CROP.SINGLE_CATEGORY_MAX_AREA,
                cfg.MODEL.SEM_SEG_HEAD.IGNORE_VALUE,
            )
        )
    if cfg.INPUT.COLOR_AUG_SSD:
        augs.append(ColorAugSSDTransform(img_format=cfg.INPUT.FORMAT))
    augs.append(T.RandomFlip())
    return augs
 
class Trainer(DefaultTrainer):
    """
    We use the "DefaultTrainer" which contains a number pre-defined logic for
    standard training workflow. They may not work for you, especially if you
    are working on a new research project. In that case you can use the cleaner
    "SimpleTrainer", or write your own training loop.
    """

    @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")
        evaluator_list = []
        evaluator_type = MetadataCatalog.get(dataset_name).evaluator_type
        if evaluator_type == "sem_seg":
            return SemSegEvaluator(
                dataset_name,
                distributed=True,
                output_dir=output_folder,
            )
        if evaluator_type == "cityscapes_sem_seg":
            return CityscapesSemSegEvaluator(dataset_name)
        if len(evaluator_list) == 0:
            raise NotImplementedError(
                "no Evaluator for the dataset {} with the type {}".format(
                    dataset_name, evaluator_type
                )
            )
        if len(evaluator_list) == 1:
            return evaluator_list[0]
        return DatasetEvaluators(evaluator_list)

    @classmethod
    def build_train_loader(cls, cfg):
        if "SemanticSegmentor" in cfg.MODEL.META_ARCHITECTURE:
            mapper = DatasetMapper(cfg, is_train=True, augmentations=build_sem_seg_train_aug(cfg))
        else:
            mapper = None
        return build_detection_train_loader(cfg, 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
                if (
                    "relative_position_bias_table" in module_param_name
                    or "absolute_pos_embed" in module_param_name
                    or "rpn" 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

 
def setup(args):
    """
    Create configs and perform basic setups.
    """
    cfg = get_cfg()
    add_deeplab_config(cfg)
    add_pna_config(cfg)
    cfg.merge_from_file(args.config_file)
    cfg.merge_from_list(args.opts)
    cfg.freeze()
    default_setup(cfg, args) 
    setup_logger(output=cfg.OUTPUT_DIR, distributed_rank=comm.get_rank(), name="PNA")
    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)
        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,),
    )