Source code for espnet2.train.trainer

"""Trainer module."""
import argparse
import dataclasses
import logging
import time
from contextlib import contextmanager
from dataclasses import is_dataclass
from pathlib import Path
from typing import Dict, Iterable, List, Optional, Sequence, Tuple, Union

import humanfriendly
import numpy as np
import torch
import torch.nn
import torch.optim
from packaging.version import parse as V
from typeguard import check_argument_types

from espnet2.iterators.abs_iter_factory import AbsIterFactory
from espnet2.main_funcs.average_nbest_models import average_nbest_models
from espnet2.main_funcs.calculate_all_attentions import calculate_all_attentions
from espnet2.schedulers.abs_scheduler import (
    AbsBatchStepScheduler,
    AbsEpochStepScheduler,
    AbsScheduler,
    AbsValEpochStepScheduler,
)
from espnet2.torch_utils.add_gradient_noise import add_gradient_noise
from espnet2.torch_utils.device_funcs import to_device
from espnet2.torch_utils.recursive_op import recursive_average
from espnet2.torch_utils.set_all_random_seed import set_all_random_seed
from espnet2.train.abs_espnet_model import AbsESPnetModel
from espnet2.train.distributed_utils import DistributedOption
from espnet2.train.reporter import Reporter, SubReporter
from espnet2.utils.build_dataclass import build_dataclass
from espnet2.utils.kwargs2args import kwargs2args

if torch.distributed.is_available():
    from torch.distributed import ReduceOp

autocast_args = dict()
if V(torch.__version__) >= V("1.6.0"):
    from torch.cuda.amp import GradScaler, autocast

    if (
        V(torch.__version__) >= V("1.10.0")
        and torch.cuda.is_available()
        and torch.cuda.is_bf16_supported()
    ):
        autocast_args = dict(dtype=torch.bfloat16)
else:
    # Nothing to do if torch<1.6.0
    @contextmanager
    def autocast(enabled=True):
        yield

    GradScaler = None

try:
    import fairscale
except ImportError:
    fairscale = None


[docs]@dataclasses.dataclass class TrainerOptions: ngpu: int resume: bool use_amp: bool train_dtype: str grad_noise: bool accum_grad: int grad_clip: float grad_clip_type: float log_interval: Optional[int] no_forward_run: bool use_matplotlib: bool use_tensorboard: bool use_wandb: bool output_dir: Union[Path, str] max_epoch: int seed: int sharded_ddp: bool patience: Optional[int] keep_nbest_models: Union[int, List[int]] nbest_averaging_interval: int early_stopping_criterion: Sequence[str] best_model_criterion: Sequence[Sequence[str]] val_scheduler_criterion: Sequence[str] unused_parameters: bool wandb_model_log_interval: int create_graph_in_tensorboard: bool
[docs]class Trainer: """Trainer having a optimizer. If you'd like to use multiple optimizers, then inherit this class and override the methods if necessary - at least "train_one_epoch()" >>> class TwoOptimizerTrainer(Trainer): ... @classmethod ... def add_arguments(cls, parser): ... ... ... ... @classmethod ... def train_one_epoch(cls, model, optimizers, ...): ... loss1 = model.model1(...) ... loss1.backward() ... optimizers[0].step() ... ... loss2 = model.model2(...) ... loss2.backward() ... optimizers[1].step() """ def __init__(self): raise RuntimeError("This class can't be instantiated.")
[docs] @classmethod def build_options(cls, args: argparse.Namespace) -> TrainerOptions: """Build options consumed by train(), eval(), and plot_attention()""" assert check_argument_types() return build_dataclass(TrainerOptions, args)
[docs] @classmethod def add_arguments(cls, parser: argparse.ArgumentParser): """Reserved for future development of another Trainer""" pass
[docs] @staticmethod def resume( checkpoint: Union[str, Path], model: torch.nn.Module, reporter: Reporter, optimizers: Sequence[torch.optim.Optimizer], schedulers: Sequence[Optional[AbsScheduler]], scaler: Optional[GradScaler], ngpu: int = 0, ): states = torch.load( checkpoint, map_location=f"cuda:{torch.cuda.current_device()}" if ngpu > 0 else "cpu", ) model.load_state_dict(states["model"]) reporter.load_state_dict(states["reporter"]) for optimizer, state in zip(optimizers, states["optimizers"]): optimizer.load_state_dict(state) for scheduler, state in zip(schedulers, states["schedulers"]): if scheduler is not None: scheduler.load_state_dict(state) if scaler is not None: if states["scaler"] is None: logging.warning("scaler state is not found") else: scaler.load_state_dict(states["scaler"]) logging.info(f"The training was resumed using {checkpoint}")
[docs] @classmethod def run( cls, model: AbsESPnetModel, optimizers: Sequence[torch.optim.Optimizer], schedulers: Sequence[Optional[AbsScheduler]], train_iter_factory: AbsIterFactory, valid_iter_factory: AbsIterFactory, plot_attention_iter_factory: Optional[AbsIterFactory], trainer_options, distributed_option: DistributedOption, ) -> None: """Perform training. This method performs the main process of training.""" assert check_argument_types() # NOTE(kamo): Don't check the type more strictly as far trainer_options assert is_dataclass(trainer_options), type(trainer_options) assert len(optimizers) == len(schedulers), (len(optimizers), len(schedulers)) if isinstance(trainer_options.keep_nbest_models, int): keep_nbest_models = [trainer_options.keep_nbest_models] else: if len(trainer_options.keep_nbest_models) == 0: logging.warning("No keep_nbest_models is given. Change to [1]") trainer_options.keep_nbest_models = [1] keep_nbest_models = trainer_options.keep_nbest_models output_dir = Path(trainer_options.output_dir) reporter = Reporter() if trainer_options.use_amp: if V(torch.__version__) < V("1.6.0"): raise RuntimeError( "Require torch>=1.6.0 for Automatic Mixed Precision" ) if trainer_options.sharded_ddp: if fairscale is None: raise RuntimeError( "Requiring fairscale. Do 'pip install fairscale'" ) scaler = fairscale.optim.grad_scaler.ShardedGradScaler() else: scaler = GradScaler() else: scaler = None if trainer_options.resume and (output_dir / "checkpoint.pth").exists(): cls.resume( checkpoint=output_dir / "checkpoint.pth", model=model, optimizers=optimizers, schedulers=schedulers, reporter=reporter, scaler=scaler, ngpu=trainer_options.ngpu, ) start_epoch = reporter.get_epoch() + 1 if start_epoch == trainer_options.max_epoch + 1: logging.warning( f"The training has already reached at max_epoch: {start_epoch}" ) if distributed_option.distributed: if trainer_options.sharded_ddp: dp_model = fairscale.nn.data_parallel.ShardedDataParallel( module=model, sharded_optimizer=optimizers, ) else: dp_model = torch.nn.parallel.DistributedDataParallel( model, device_ids=( # Perform multi-Process with multi-GPUs [torch.cuda.current_device()] if distributed_option.ngpu == 1 # Perform single-Process with multi-GPUs else None ), output_device=( torch.cuda.current_device() if distributed_option.ngpu == 1 else None ), find_unused_parameters=trainer_options.unused_parameters, ) elif distributed_option.ngpu > 1: dp_model = torch.nn.parallel.DataParallel( model, device_ids=list(range(distributed_option.ngpu)), ) else: # NOTE(kamo): DataParallel also should work with ngpu=1, # but for debuggability it's better to keep this block. dp_model = model if trainer_options.use_tensorboard and ( not distributed_option.distributed or distributed_option.dist_rank == 0 ): from torch.utils.tensorboard import SummaryWriter train_summary_writer = SummaryWriter( str(output_dir / "tensorboard" / "train") ) valid_summary_writer = SummaryWriter( str(output_dir / "tensorboard" / "valid") ) else: train_summary_writer = None start_time = time.perf_counter() for iepoch in range(start_epoch, trainer_options.max_epoch + 1): if iepoch != start_epoch: logging.info( "{}/{}epoch started. Estimated time to finish: {}".format( iepoch, trainer_options.max_epoch, humanfriendly.format_timespan( (time.perf_counter() - start_time) / (iepoch - start_epoch) * (trainer_options.max_epoch - iepoch + 1) ), ) ) else: logging.info(f"{iepoch}/{trainer_options.max_epoch}epoch started") set_all_random_seed(trainer_options.seed + iepoch) reporter.set_epoch(iepoch) # 1. Train and validation for one-epoch with reporter.observe("train") as sub_reporter: all_steps_are_invalid = cls.train_one_epoch( model=dp_model, optimizers=optimizers, schedulers=schedulers, iterator=train_iter_factory.build_iter(iepoch), reporter=sub_reporter, scaler=scaler, summary_writer=train_summary_writer, options=trainer_options, distributed_option=distributed_option, ) with reporter.observe("valid") as sub_reporter: cls.validate_one_epoch( model=dp_model, iterator=valid_iter_factory.build_iter(iepoch), reporter=sub_reporter, options=trainer_options, distributed_option=distributed_option, ) if not distributed_option.distributed or distributed_option.dist_rank == 0: # att_plot doesn't support distributed if plot_attention_iter_factory is not None: with reporter.observe("att_plot") as sub_reporter: cls.plot_attention( model=model, output_dir=output_dir / "att_ws", summary_writer=train_summary_writer, iterator=plot_attention_iter_factory.build_iter(iepoch), reporter=sub_reporter, options=trainer_options, ) # 2. LR Scheduler step for scheduler in schedulers: if isinstance(scheduler, AbsValEpochStepScheduler): scheduler.step( reporter.get_value(*trainer_options.val_scheduler_criterion) ) elif isinstance(scheduler, AbsEpochStepScheduler): scheduler.step() if trainer_options.sharded_ddp: for optimizer in optimizers: if isinstance(optimizer, fairscale.optim.oss.OSS): optimizer.consolidate_state_dict() if not distributed_option.distributed or distributed_option.dist_rank == 0: # 3. Report the results logging.info(reporter.log_message()) if trainer_options.use_matplotlib: reporter.matplotlib_plot(output_dir / "images") if train_summary_writer is not None: reporter.tensorboard_add_scalar(train_summary_writer, key1="train") reporter.tensorboard_add_scalar(valid_summary_writer, key1="valid") if trainer_options.use_wandb: reporter.wandb_log() # 4. Save/Update the checkpoint torch.save( { "model": model.state_dict(), "reporter": reporter.state_dict(), "optimizers": [o.state_dict() for o in optimizers], "schedulers": [ s.state_dict() if s is not None else None for s in schedulers ], "scaler": scaler.state_dict() if scaler is not None else None, }, output_dir / "checkpoint.pth", ) # 5. Save and log the model and update the link to the best model torch.save(model.state_dict(), output_dir / f"{iepoch}epoch.pth") # Creates a sym link latest.pth -> {iepoch}epoch.pth p = output_dir / "latest.pth" if p.is_symlink() or p.exists(): p.unlink() p.symlink_to(f"{iepoch}epoch.pth") _improved = [] for _phase, k, _mode in trainer_options.best_model_criterion: # e.g. _phase, k, _mode = "train", "loss", "min" if reporter.has(_phase, k): best_epoch = reporter.get_best_epoch(_phase, k, _mode) # Creates sym links if it's the best result if best_epoch == iepoch: p = output_dir / f"{_phase}.{k}.best.pth" if p.is_symlink() or p.exists(): p.unlink() p.symlink_to(f"{iepoch}epoch.pth") _improved.append(f"{_phase}.{k}") if len(_improved) == 0: logging.info("There are no improvements in this epoch") else: logging.info( "The best model has been updated: " + ", ".join(_improved) ) log_model = ( trainer_options.wandb_model_log_interval > 0 and iepoch % trainer_options.wandb_model_log_interval == 0 ) if log_model and trainer_options.use_wandb: import wandb logging.info("Logging Model on this epoch :::::") artifact = wandb.Artifact( name=f"model_{wandb.run.id}", type="model", metadata={"improved": _improved}, ) artifact.add_file(str(output_dir / f"{iepoch}epoch.pth")) aliases = [ f"epoch-{iepoch}", "best" if best_epoch == iepoch else "", ] wandb.log_artifact(artifact, aliases=aliases) # 6. Remove the model files excluding n-best epoch and latest epoch _removed = [] # Get the union set of the n-best among multiple criterion nbests = set().union( *[ set(reporter.sort_epochs(ph, k, m)[: max(keep_nbest_models)]) for ph, k, m in trainer_options.best_model_criterion if reporter.has(ph, k) ] ) # Generated n-best averaged model if ( trainer_options.nbest_averaging_interval > 0 and iepoch % trainer_options.nbest_averaging_interval == 0 ): average_nbest_models( reporter=reporter, output_dir=output_dir, best_model_criterion=trainer_options.best_model_criterion, nbest=keep_nbest_models, suffix=f"till{iepoch}epoch", ) for e in range(1, iepoch): p = output_dir / f"{e}epoch.pth" if p.exists() and e not in nbests: p.unlink() _removed.append(str(p)) if len(_removed) != 0: logging.info("The model files were removed: " + ", ".join(_removed)) # 7. If any updating haven't happened, stops the training if all_steps_are_invalid: logging.warning( "The gradients at all steps are invalid in this epoch. " f"Something seems wrong. This training was stopped at {iepoch}epoch" ) break # 8. Check early stopping if trainer_options.patience is not None: if reporter.check_early_stopping( trainer_options.patience, *trainer_options.early_stopping_criterion ): break else: logging.info( f"The training was finished at {trainer_options.max_epoch} epochs " ) # Generated n-best averaged model if not distributed_option.distributed or distributed_option.dist_rank == 0: average_nbest_models( reporter=reporter, output_dir=output_dir, best_model_criterion=trainer_options.best_model_criterion, nbest=keep_nbest_models, )
[docs] @classmethod def train_one_epoch( cls, model: torch.nn.Module, iterator: Iterable[Tuple[List[str], Dict[str, torch.Tensor]]], optimizers: Sequence[torch.optim.Optimizer], schedulers: Sequence[Optional[AbsScheduler]], scaler: Optional[GradScaler], reporter: SubReporter, summary_writer, options: TrainerOptions, distributed_option: DistributedOption, ) -> bool: assert check_argument_types() grad_noise = options.grad_noise accum_grad = options.accum_grad grad_clip = options.grad_clip grad_clip_type = options.grad_clip_type log_interval = options.log_interval no_forward_run = options.no_forward_run ngpu = options.ngpu use_wandb = options.use_wandb create_graph_in_tensorboard = options.create_graph_in_tensorboard distributed = distributed_option.distributed if log_interval is None: try: log_interval = max(len(iterator) // 20, 10) except TypeError: log_interval = 100 model.train() all_steps_are_invalid = True # [For distributed] Because iteration counts are not always equals between # processes, send stop-flag to the other processes if iterator is finished iterator_stop = torch.tensor(0).to("cuda" if ngpu > 0 else "cpu") start_time = time.perf_counter() for iiter, (utt_id, batch) in enumerate( reporter.measure_iter_time(iterator, "iter_time"), 1 ): assert isinstance(batch, dict), type(batch) if distributed: torch.distributed.all_reduce(iterator_stop, ReduceOp.SUM) if iterator_stop > 0: break batch["utt_id"] = utt_id batch = to_device(batch, "cuda" if ngpu > 0 else "cpu") if no_forward_run: all_steps_are_invalid = False continue if ( create_graph_in_tensorboard and iiter == 1 and summary_writer is not None ): if distributed: _model = getattr(model, "module") else: _model = model if _model is not None: try: _args = kwargs2args(_model.forward, batch) except (ValueError, TypeError): logging.warning( "inpect.signature() is failed for the model. " "The graph can't be added for tensorboard." ) else: try: summary_writer.add_graph( _model, _args, use_strict_trace=False ) except Exception: logging.warning( "summary_writer.add_graph() " "is failed for the model. " "The graph can't be added for tensorboard." ) del _args else: logging.warning( "model.module is not found (This should be a bug.)" ) del _model with autocast( scaler is not None, **autocast_args, ): with reporter.measure_time("forward_time"): retval = model(**batch) # Note(kamo): # Supporting two patterns for the returned value from the model # a. dict type if isinstance(retval, dict): loss = retval["loss"] stats = retval["stats"] weight = retval["weight"] optim_idx = retval.get("optim_idx") if optim_idx is not None and not isinstance(optim_idx, int): if not isinstance(optim_idx, torch.Tensor): raise RuntimeError( "optim_idx must be int or 1dim torch.Tensor, " f"but got {type(optim_idx)}" ) if optim_idx.dim() >= 2: raise RuntimeError( "optim_idx must be int or 1dim torch.Tensor, " f"but got {optim_idx.dim()}dim tensor" ) if optim_idx.dim() == 1: for v in optim_idx: if v != optim_idx[0]: raise RuntimeError( "optim_idx must be 1dim tensor " "having same values for all entries" ) optim_idx = optim_idx[0].item() else: optim_idx = optim_idx.item() # b. tuple or list type else: loss, stats, weight = retval optim_idx = None stats = {k: v for k, v in stats.items() if v is not None} if ngpu > 1 or distributed: # Apply weighted averaging for loss and stats loss = (loss * weight.type(loss.dtype)).sum() # if distributed, this method can also apply all_reduce() stats, weight = recursive_average(stats, weight, distributed) # Now weight is summation over all workers loss /= weight if distributed: # NOTE(kamo): Multiply world_size because DistributedDataParallel # automatically normalizes the gradient by world_size. loss *= torch.distributed.get_world_size() loss /= accum_grad reporter.register(stats, weight) with reporter.measure_time("backward_time"): if scaler is not None: # Scales loss. Calls backward() on scaled loss # to create scaled gradients. # Backward passes under autocast are not recommended. # Backward ops run in the same dtype autocast chose # for corresponding forward ops. scaler.scale(loss).backward() else: loss.backward() if iiter % accum_grad == 0: if scaler is not None: # Unscales the gradients of optimizer's assigned params in-place for iopt, optimizer in enumerate(optimizers): if optim_idx is not None and iopt != optim_idx: continue scaler.unscale_(optimizer) # gradient noise injection if grad_noise: add_gradient_noise( model, reporter.get_total_count(), duration=100, eta=1.0, scale_factor=0.55, ) # compute the gradient norm to check if it is normal or not grad_norm = torch.nn.utils.clip_grad_norm_( model.parameters(), max_norm=grad_clip, norm_type=grad_clip_type, ) # PyTorch<=1.4, clip_grad_norm_ returns float value if not isinstance(grad_norm, torch.Tensor): grad_norm = torch.tensor(grad_norm) if not torch.isfinite(grad_norm): logging.warning( f"The grad norm is {grad_norm}. Skipping updating the model." ) # Must invoke scaler.update() if unscale_() is used in the iteration # to avoid the following error: # RuntimeError: unscale_() has already been called # on this optimizer since the last update(). # Note that if the gradient has inf/nan values, # scaler.step skips optimizer.step(). if scaler is not None: for iopt, optimizer in enumerate(optimizers): if optim_idx is not None and iopt != optim_idx: continue scaler.step(optimizer) scaler.update() else: reporter.register( { "grad_norm": grad_norm, "clip": torch.where( grad_norm > grad_clip, grad_norm.new_tensor(100), grad_norm.new_tensor(0), ), "loss_scale": scaler.get_scale() if scaler else 1.0, } ) all_steps_are_invalid = False with reporter.measure_time("optim_step_time"): for iopt, (optimizer, scheduler) in enumerate( zip(optimizers, schedulers) ): if optim_idx is not None and iopt != optim_idx: continue if scaler is not None: # scaler.step() first unscales the gradients of # the optimizer's assigned params. scaler.step(optimizer) # Updates the scale for next iteration. scaler.update() else: optimizer.step() if isinstance(scheduler, AbsBatchStepScheduler): scheduler.step() for iopt, optimizer in enumerate(optimizers): if optim_idx is not None and iopt != optim_idx: continue optimizer.zero_grad() # Register lr and train/load time[sec/step], # where step refers to accum_grad * mini-batch reporter.register( dict( { f"optim{i}_lr{j}": pg["lr"] for i, optimizer in enumerate(optimizers) for j, pg in enumerate(optimizer.param_groups) if "lr" in pg }, train_time=time.perf_counter() - start_time, ), ) start_time = time.perf_counter() # NOTE(kamo): Call log_message() after next() reporter.next() if iiter % log_interval == 0: logging.info(reporter.log_message(-log_interval)) if summary_writer is not None: reporter.tensorboard_add_scalar(summary_writer, -log_interval) if use_wandb: reporter.wandb_log() else: if distributed: iterator_stop.fill_(1) torch.distributed.all_reduce(iterator_stop, ReduceOp.SUM) return all_steps_are_invalid
[docs] @classmethod @torch.no_grad() def validate_one_epoch( cls, model: torch.nn.Module, iterator: Iterable[Dict[str, torch.Tensor]], reporter: SubReporter, options: TrainerOptions, distributed_option: DistributedOption, ) -> None: assert check_argument_types() ngpu = options.ngpu no_forward_run = options.no_forward_run distributed = distributed_option.distributed model.eval() # [For distributed] Because iteration counts are not always equals between # processes, send stop-flag to the other processes if iterator is finished iterator_stop = torch.tensor(0).to("cuda" if ngpu > 0 else "cpu") for utt_id, batch in iterator: assert isinstance(batch, dict), type(batch) if distributed: torch.distributed.all_reduce(iterator_stop, ReduceOp.SUM) if iterator_stop > 0: break batch["utt_id"] = utt_id batch = to_device(batch, "cuda" if ngpu > 0 else "cpu") if no_forward_run: continue retval = model(**batch) if isinstance(retval, dict): stats = retval["stats"] weight = retval["weight"] else: _, stats, weight = retval if ngpu > 1 or distributed: # Apply weighted averaging for stats. # if distributed, this method can also apply all_reduce() stats, weight = recursive_average(stats, weight, distributed) reporter.register(stats, weight) reporter.next() else: if distributed: iterator_stop.fill_(1) torch.distributed.all_reduce(iterator_stop, ReduceOp.SUM)
[docs] @classmethod @torch.no_grad() def plot_attention( cls, model: torch.nn.Module, output_dir: Optional[Path], summary_writer, iterator: Iterable[Tuple[List[str], Dict[str, torch.Tensor]]], reporter: SubReporter, options: TrainerOptions, ) -> None: assert check_argument_types() import matplotlib ngpu = options.ngpu no_forward_run = options.no_forward_run matplotlib.use("Agg") import matplotlib.pyplot as plt from matplotlib.ticker import MaxNLocator model.eval() for ids, batch in iterator: assert isinstance(batch, dict), type(batch) assert len(next(iter(batch.values()))) == len(ids), ( len(next(iter(batch.values()))), len(ids), ) batch["utt_id"] = ids batch = to_device(batch, "cuda" if ngpu > 0 else "cpu") if no_forward_run: continue # 1. Forwarding model and gathering all attentions # calculate_all_attentions() uses single gpu only. att_dict = calculate_all_attentions(model, batch) # 2. Plot attentions: This part is slow due to matplotlib for k, att_list in att_dict.items(): assert len(att_list) == len(ids), (len(att_list), len(ids)) for id_, att_w in zip(ids, att_list): if isinstance(att_w, torch.Tensor): att_w = att_w.detach().cpu().numpy() if att_w.ndim == 2: att_w = att_w[None] elif att_w.ndim == 4: # In multispkr_asr model case, the dimension could be 4. att_w = np.concatenate( [att_w[i] for i in range(att_w.shape[0])], axis=0 ) elif att_w.ndim > 4 or att_w.ndim == 1: raise RuntimeError(f"Must be 2, 3 or 4 dimension: {att_w.ndim}") w, h = plt.figaspect(1.0 / len(att_w)) fig = plt.Figure(figsize=(w * 1.3, h * 1.3)) axes = fig.subplots(1, len(att_w)) if len(att_w) == 1: axes = [axes] for ax, aw in zip(axes, att_w): ax.imshow(aw.astype(np.float32), aspect="auto") ax.set_title(f"{k}_{id_}") ax.set_xlabel("Input") ax.set_ylabel("Output") ax.xaxis.set_major_locator(MaxNLocator(integer=True)) ax.yaxis.set_major_locator(MaxNLocator(integer=True)) if output_dir is not None: p = output_dir / id_ / f"{k}.{reporter.get_epoch()}ep.png" p.parent.mkdir(parents=True, exist_ok=True) fig.savefig(p) if summary_writer is not None: summary_writer.add_figure( f"{k}_{id_}", fig, reporter.get_epoch() ) if options.use_wandb: import wandb wandb.log({f"attention plot/{k}_{id_}": wandb.Image(fig)}) reporter.next()