tools/train_net.py

# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
r"""
Basic training script for PyTorch
"""

# Set up custom environment before nearly anything else is imported
# NOTE: this should be the first import (no not reorder)
from maskrcnn_benchmark.utils.env import setup_environment  # noqa F401 isort:skip

import argparse
import os

import torch
from maskrcnn_benchmark.config import cfg
from maskrcnn_benchmark.data import make_data_loader
from maskrcnn_benchmark.solver import make_lr_scheduler
from maskrcnn_benchmark.solver import make_optimizer
from maskrcnn_benchmark.engine.trainer import do_train
from maskrcnn_benchmark.modeling.detector import build_detection_model
from maskrcnn_benchmark.utils.checkpoint import DetectronCheckpointer
from maskrcnn_benchmark.utils.collect_env import collect_env_info
from maskrcnn_benchmark.utils.comm import synchronize, get_rank
from maskrcnn_benchmark.utils.imports import import_file
from maskrcnn_benchmark.utils.logging import setup_logger, Logger
from maskrcnn_benchmark.utils.miscellaneous import mkdir
# See if we can use apex.DistributedDataParallel instead of the torch default,
# and enable mixed-precision via apex.amp
try:
    from apex import amp
except ImportError:
    raise ImportError('Use APEX for multi-precision via apex.amp')

def train(cfg, local_rank, distributed):
    model = build_detection_model(cfg)
    device = torch.device(cfg.MODEL.DEVICE)
    model.to(device)

    optimizer = make_optimizer(cfg, model)
    scheduler = make_lr_scheduler(cfg, optimizer)
    
    # Initialize mixed-precision training
    use_mixed_precision = cfg.DTYPE == "float16"
    amp_opt_level = 'O1' if use_mixed_precision else 'O0'
    model, optimizer = amp.initialize(model, optimizer, opt_level=amp_opt_level)

    if distributed:
        model = torch.nn.parallel.DistributedDataParallel(
            model, device_ids=[local_rank], output_device=local_rank,
            # this should be removed if we update BatchNorm stats
            broadcast_buffers=False,
            # find_unused_parameters=True
        )

    arguments = {}
    arguments["iteration"] = 0

    output_dir = cfg.OUTPUT_DIR

    save_to_disk = get_rank() == 0
    checkpointer = DetectronCheckpointer(
        cfg, model, optimizer, scheduler, output_dir, save_to_disk
    )
    extra_checkpoint_data = checkpointer.load(cfg.MODEL.WEIGHT, resume=cfg.SOLVER.RESUME)
    if cfg.SOLVER.RESUME:
        arguments.update(extra_checkpoint_data)

    data_loader = make_data_loader(
        cfg,
        is_train=True,
        is_distributed=distributed,
        start_iter=arguments["iteration"],
    )

    checkpoint_period = cfg.SOLVER.CHECKPOINT_PERIOD
    tb_logger = Logger(cfg.OUTPUT_DIR, local_rank)
    do_train(
        model,
        data_loader,
        optimizer,
        scheduler,
        checkpointer,
        device,
        checkpoint_period,
        arguments,
        tb_logger,
        cfg,
        local_rank,
    )

    return model

def main():
    parser = argparse.ArgumentParser(description="PyTorch Object Detection Training")
    parser.add_argument(
        "--config-file",
        default="",
        metavar="FILE",
        help="path to config file",
        type=str,
    )
    parser.add_argument("--local_rank", type=int, default=0)
    parser.add_argument(
        "--skip-test",
        dest="skip_test",
        help="Do not test the final model",
        action="store_true",
    )
    parser.add_argument(
        "opts",
        help="Modify config options using the command-line",
        default=None,
        nargs=argparse.REMAINDER,
    )

    args = parser.parse_args()

    num_gpus = int(os.environ["WORLD_SIZE"]) if "WORLD_SIZE" in os.environ else 1
    args.distributed = num_gpus > 1

    if args.distributed:
        torch.cuda.set_device(args.local_rank)
        torch.distributed.init_process_group(
            backend="nccl", init_method="env://"
        )

    cfg.merge_from_file(args.config_file)
    cfg.merge_from_list(args.opts)
    cfg.freeze()

    output_dir = cfg.OUTPUT_DIR
    if output_dir:
        mkdir(output_dir)

    local_rank = get_rank()
    logger = setup_logger("maskrcnn_benchmark", output_dir, local_rank)
    if local_rank == 0:
        logger.info("Using {} GPUs".format(num_gpus))
        logger.info(args)

        logger.info("Collecting env info (might take some time)")
        logger.info("\n" + collect_env_info())

        logger.info("Loaded configuration file {}".format(args.config_file))
        with open(args.config_file, "r") as cf:
            config_str = "\n" + cf.read()
            logger.info(config_str)
        logger.info("Running with config:\n{}".format(cfg))

    model = train(cfg, args.local_rank, args.distributed)


if __name__ == "__main__":
    main()