train.py

#!/usr/bin/env python3 -u
# Copyright (c) Facebook, Inc. and its affiliates.
#
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.
"""
Train a new model on one or across multiple GPUs.
"""

import logging
import math
import os
import sys

import numpy as np
import torch
from fairseq import (
    checkpoint_utils,
    distributed_utils,
    options,
    quantization_utils,
    tasks,
    utils,
)
from fairseq.data import iterators
from fairseq.logging import meters, metrics, progress_bar
from fairseq.model_parallel.megatron_trainer import MegatronTrainer
from fairseq.trainer import Trainer

logging.basicConfig(
    format='%(asctime)s | %(levelname)s | %(name)s | %(message)s',
    datefmt='%Y-%m-%d %H:%M:%S',
    level=logging.INFO,
    stream=sys.stdout,
)
logger = logging.getLogger('fairseq_cli.train')


def main(args, init_distributed=False):
    utils.import_user_module(args)

    assert (
            args.max_tokens is not None or getattr(args, "max_sentences", None) is not None
    ), "Must specify batch size either with --max-tokens or --max-sentences"

    metrics.reset()

    np.random.seed(args.seed)
    utils.set_torch_seed(args.seed)

    if distributed_utils.is_master(args):
        checkpoint_utils.verify_checkpoint_directory(args.save_dir)

    # Print args
    logger.info(args)

    # Setup task, e.g., translation, language modeling, etc.
    task = tasks.setup_task(args)

    # Load valid dataset (we load training data below, based on the latest checkpoint)
    for valid_sub_split in args.valid_subset.split(','):
        task.load_dataset(valid_sub_split, combine=False, epoch=1)

    # Build model and criterion
    model = task.build_model(args)
    criterion = task.build_criterion(args)
    logger.info(model)
    logger.info("task: {} ({})".format(args.task, task.__class__.__name__))
    logger.info("model: {} ({})".format(args.arch, model.__class__.__name__))
    logger.info("criterion: {} ({})".format(args.criterion, criterion.__class__.__name__))
    logger.info('num. model params: {} (num. trained: {})'.format(
        sum(p.numel() for p in model.parameters()),
        sum(p.numel() for p in model.parameters() if p.requires_grad),
    ))

    # (optionally) Configure quantization
    if args.quantization_config_path is not None:
        quantizer = quantization_utils.Quantizer(
            config_path=args.quantization_config_path,
            max_epoch=args.max_epoch,
            max_update=args.max_update,
        )
    else:
        quantizer = None

    # Build trainer
    if args.model_parallel_size == 1:
        trainer = Trainer(args, task, model, criterion, quantizer)
    else:
        trainer = MegatronTrainer(args, task, model, criterion)

    logger.info('training on {} devices (GPUs/TPUs)'.format(args.distributed_world_size))
    logger.info('max tokens per GPU = {} and max sentences per GPU = {}'.format(
        args.max_tokens,
        getattr(args, "max_sentences", None),
    ))

    # Load the latest checkpoint if one is available and restore the
    # corresponding train iterator
    extra_state, epoch_itr = checkpoint_utils.load_checkpoint(args, trainer)

    # Train until the learning rate gets too small
    max_epoch = args.max_epoch or math.inf
    lr = trainer.get_lr()
    train_meter = meters.StopwatchMeter()
    train_meter.start()
    while (
            lr > args.min_lr
            and epoch_itr.next_epoch_idx <= max_epoch
    ):
        # train for one epoch
        valid_losses, should_stop = train(args, trainer, task, epoch_itr)
        if should_stop:
            break

        # only use first validation loss to update the learning rate
        lr = trainer.lr_step(epoch_itr.epoch, valid_losses[0])
        epoch_itr = trainer.get_train_iterator(
            epoch_itr.next_epoch_idx,
            # sharded data: get train iterator for next epoch
            load_dataset=(os.pathsep in getattr(args, 'data', '')),
        )
        if hasattr(model, "epoch_strategy") and epoch_itr.next_epoch_idx % getattr(model.args, "epoch_every", 1) == 0:
            model.epoch_strategy()

    train_meter.stop()
    logger.info('done training in {:.1f} seconds'.format(train_meter.sum))


def should_stop_early(args, valid_loss):
    # skip check if no validation was done in the current epoch
    if valid_loss is None:
        return False
    if args.patience <= 0:
        return False

    def is_better(a, b):
        return a > b if args.maximize_best_checkpoint_metric else a < b

    prev_best = getattr(should_stop_early, 'best', None)
    if prev_best is None or is_better(valid_loss, prev_best):
        should_stop_early.best = valid_loss
        should_stop_early.num_runs = 0
        return False
    else:
        should_stop_early.num_runs += 1
        if should_stop_early.num_runs >= args.patience:
            logger.info('early stop since valid performance hasn\'t improved for last {} runs'.format(args.patience))
            return True
        else:
            return False


@metrics.aggregate('train')
def train(args, trainer, task, epoch_itr):
    """Train the model for one epoch and return validation losses."""
    # Initialize data iterator
    itr = epoch_itr.next_epoch_itr(
        fix_batches_to_gpus=args.fix_batches_to_gpus,
        shuffle=(epoch_itr.next_epoch_idx > args.curriculum),
    )
    update_freq = (
        args.update_freq[epoch_itr.epoch - 1]
        if epoch_itr.epoch <= len(args.update_freq)
        else args.update_freq[-1]
    )
    itr = iterators.GroupedIterator(itr, update_freq)
    if getattr(args, "tpu", False):
        itr = utils.tpu_data_loader(itr)
    progress = progress_bar.progress_bar(
        itr,
        log_format=args.log_format,
        log_interval=args.log_interval,
        epoch=epoch_itr.epoch,
        tensorboard_logdir=(
            args.tensorboard_logdir if distributed_utils.is_master(args) else None
        ),
        default_log_format=('tqdm' if not args.no_progress_bar else 'simple'),
    )

    trainer.begin_epoch(epoch_itr.epoch)

    valid_subsets = args.valid_subset.split(',')
    should_stop = False
    num_updates = trainer.get_num_updates()
    for i, samples in enumerate(progress):
        with metrics.aggregate("train_inner"), torch.autograd.profiler.record_function(
                "train_step-%d" % i
        ):
            log_output = trainer.train_step(samples)

        # log mid-epoch stats
        if log_output is not None:
            num_updates = trainer.get_num_updates()
            if num_updates % args.log_interval == 0:
                stats = get_training_stats(metrics.get_smoothed_values('train_inner'))
                progress.log(stats, tag='train_inner', step=num_updates)

                # reset mid-epoch stats after each log interval
                # the end-of-epoch stats will still be preserved
                metrics.reset_meters('train_inner')

        # end_of_epoch = not itr.has_next()
        end_of_epoch = not itr.has_next() and not args.no_epoch_checkpoints  # remove checkpoint updates
        valid_losses, should_stop = validate_and_save(
            args, trainer, task, epoch_itr, valid_subsets, end_of_epoch
        )

        if should_stop:
            break

    # log end-of-epoch stats
    stats = get_training_stats(metrics.get_smoothed_values('train'))
    progress.print(stats, tag='train', step=num_updates)

    # reset epoch-level meters
    metrics.reset_meters('train')
    return valid_losses, should_stop


def validate_and_save(args, trainer, task, epoch_itr, valid_subsets, end_of_epoch):
    num_updates = trainer.get_num_updates()
    do_save = (
                      args.save_interval_updates > 0
                      and num_updates > 0
                      and num_updates % args.save_interval_updates == 0
                      and num_updates >= args.validate_after_updates
              ) or (end_of_epoch and epoch_itr.epoch % args.save_interval == 0)
    do_validate = (
                          (not end_of_epoch and do_save)  # validate during mid-epoch saves
                          or (end_of_epoch and epoch_itr.epoch % args.validate_interval == 0)
                          or (args.validate_interval_updates > 0 and num_updates % args.validate_interval_updates == 0)
                  ) and not args.disable_validation and num_updates > args.validate_after_updates

    # Validate
    valid_losses = [None]
    if do_validate:
        valid_losses = validate(args, trainer, task, epoch_itr, valid_subsets)

    # Stopping conditions
    max_update = args.max_update or math.inf
    should_stop = (
            should_stop_early(args, valid_losses[0])
            or trainer.get_num_updates() >= max_update
            or (
                    args.stop_time_hours
                    and trainer.cumulative_training_time() / (60 * 60) > args.stop_time_hours
            )
    )

    # Save checkpoint
    if do_save or should_stop:
        checkpoint_utils.save_checkpoint(args, trainer, epoch_itr, valid_losses[0])
        logger.info("save checkpoint done")

    return valid_losses, should_stop


def get_training_stats(stats):
    stats['wall'] = round(metrics.get_meter('default', 'wall').elapsed_time, 0)
    return stats


def validate(args, trainer, task, epoch_itr, subsets):
    """Evaluate the model on the validation set(s) and return the losses."""

    if args.fixed_validation_seed is not None:
        # set fixed seed for every validation
        utils.set_torch_seed(args.fixed_validation_seed)

    valid_losses = []
    for subset in subsets:
        # logger.info('begin validation on "{}" subset'.format(subset))

        # Initialize data iterator
        itr = trainer.get_valid_iterator(subset).next_epoch_itr(shuffle=False)
        if getattr(args, "tpu", False):
            itr = utils.tpu_data_loader(itr)
        progress = progress_bar.progress_bar(
            itr,
            log_format=args.log_format,
            log_interval=args.log_interval,
            epoch=epoch_itr.epoch,
            prefix=f"valid on '{subset}' subset",
            tensorboard_logdir=(
                args.tensorboard_logdir if distributed_utils.is_master(args) else None
            ),
            default_log_format=('tqdm' if not args.no_progress_bar else 'simple'),
        )

        # create a new root metrics aggregator so validation metrics
        # don't pollute other aggregators (e.g., train meters)
        with metrics.aggregate(new_root=True) as agg:
            for sample in progress:
                trainer.valid_step(sample)

        # log validation stats
        stats = get_valid_stats(args, trainer, agg.get_smoothed_values())
        progress.print(stats, tag=subset, step=trainer.get_num_updates())

        valid_losses.append(stats[args.best_checkpoint_metric])
    return valid_losses


def get_valid_stats(args, trainer, stats):
    stats['num_updates'] = trainer.get_num_updates()
    if hasattr(checkpoint_utils.save_checkpoint, 'best'):
        key = 'best_{0}'.format(args.best_checkpoint_metric)
        best_function = max if args.maximize_best_checkpoint_metric else min
        stats[key] = best_function(
            checkpoint_utils.save_checkpoint.best,
            stats[args.best_checkpoint_metric],
        )
    return stats


def cli_main(modify_parser=None):
    parser = options.get_training_parser()
    options.add_generation_args(parser)
    add_finetuning_args(group=parser)
    args = options.parse_args_and_arch(parser, modify_parser=modify_parser)

    if args.profile:
        with torch.cuda.profiler.profile():
            with torch.autograd.profiler.emit_nvtx():
                distributed_utils.call_main(args, main)
    else:
        distributed_utils.call_main(args, main)


def add_finetuning_args(group):
    group.add_argument("--finetune", action="store_true")
    group.add_argument("--finetune-length-pred", action="store_true")
    group.add_argument("--finetune-vq-pred", action="store_true")
    group.add_argument("--retain-vq-code", action="store_true")


if __name__ == '__main__':
    cli_main()