icefall/egs/librispeech/ASR/tiny_transducer_ctc/train.py

#!/usr/bin/env python3
# Copyright    2021-2022  Xiaomi Corp.        (authors: Fangjun Kuang,
#                                                       Wei Kang,
#                                                       Mingshuang Luo,)
#                                                       Zengwei Yao)
#
# See ../../../../LICENSE for clarification regarding multiple authors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

"""
Usage:

cd egs/librispeech/ASR/
./prepare.sh

Run below if you want to use the phone lexicon instead of BPE:
python local/generate_unique_lexicon.py --lang-dir data/lang_phone 

"""
import argparse
import copy
import logging
import pprint
import warnings
from pathlib import Path
from shutil import copyfile
from typing import Any, Dict, Optional, Tuple, Union

import k2
import sentencepiece as spm
import torch
import torch.multiprocessing as mp
import torch.nn as nn
from asr_datamodule import LibriSpeechAsrDataModule
from decoder import Decoder
from encoder import Conv1dNet
from joiner import Joiner
from lhotse.cut import Cut
from lhotse.dataset.sampling.base import CutSampler
from lhotse.utils import fix_random_seed
from model import Transducer
from torch import Tensor
from torch.cuda.amp import GradScaler
from torch.nn.parallel import DistributedDataParallel as DDP
from torch.optim import AdamW
from torch.optim.lr_scheduler import StepLR
from torch.utils.tensorboard import SummaryWriter

from icefall import diagnostics, is_module_available
from icefall.checkpoint import load_checkpoint, remove_checkpoints
from icefall.checkpoint import save_checkpoint as save_checkpoint_impl
from icefall.checkpoint import (
    save_checkpoint_with_global_batch_idx,
    update_averaged_model,
)
from icefall.dist import cleanup_dist, setup_dist
from icefall.env import get_env_info
from icefall.err import raise_grad_scale_is_too_small_error
from icefall.hooks import register_inf_check_hooks
from icefall.lexicon import UniqLexicon
from icefall.utils import (
    AttributeDict,
    MetricsTracker,
    encode_supervisions,
    setup_logger,
    str2bool,
)

LRSchedulerType = torch.optim.lr_scheduler._LRScheduler


def set_batch_count(model: Union[nn.Module, DDP], batch_count: float) -> None:
    if isinstance(model, DDP):
        # get underlying nn.Module
        model = model.module
    for module in model.modules():
        if hasattr(module, "batch_count"):
            module.batch_count = batch_count


def add_model_arguments(parser: argparse.ArgumentParser):
    parser.add_argument(
        "--encoder-dim",
        type=int,
        default=256,
        help="Embedding dimension in the decoder model.",
    )

    parser.add_argument(
        "--decoder-dim",
        type=int,
        default=256,
        help="Embedding dimension in the decoder model.",
    )

    parser.add_argument(
        "--joiner-dim",
        type=int,
        default=256,
        help="""Dimension used in the joiner model.
        Outputs from the encoder and decoder model are projected
        to this dimension before adding.
        """,
    )

    parser.add_argument(
        "--conv-layers",
        type=int,
        default=10,
        help="""Number of convolution layers for the encoder.
        """,
    )

    parser.add_argument(
        "--channels",
        type=int,
        default=256,
        help="""Number of channels for the encoder.
        """,
    )

    parser.add_argument(
        "--skip-add",
        type=str2bool,
        default=False,
        help="""Use skip connection in the encoder.
        """,
    )


def get_parser():
    parser = argparse.ArgumentParser(
        formatter_class=argparse.ArgumentDefaultsHelpFormatter
    )

    parser.add_argument(
        "--exp-dir",
        type=str,
        default="tiny_transducer_ctc/exp",
        help="""The experiment dir.
        It specifies the directory where all training related
        files, e.g., checkpoints, log, etc, are saved
        """,
    )

    parser.add_argument(
        "--world-size",
        type=int,
        default=1,
        help="Number of GPUs for DDP training.",
    )

    parser.add_argument(
        "--master-port",
        type=int,
        default=12354,
        help="Master port to use for DDP training.",
    )

    parser.add_argument(
        "--tensorboard",
        type=str2bool,
        default=True,
        help="Should various information be logged in tensorboard.",
    )

    parser.add_argument(
        "--num-epochs",
        type=int,
        default=30,
        help="Number of epochs to train.",
    )

    parser.add_argument(
        "--start-epoch",
        type=int,
        default=1,
        help="""Resume training from this epoch. It should be positive.
        If larger than 1, it will load checkpoint from
        exp-dir/epoch-{start_epoch-1}.pt
        """,
    )

    parser.add_argument(
        "--start-batch",
        type=int,
        default=0,
        help="""If positive, --start-epoch is ignored and
        it loads the checkpoint from exp-dir/checkpoint-{start_batch}.pt
        """,
    )

    parser.add_argument(
        "--lang-dir",
        type=str,
        default="data/lang_bpe_500",
        help="""The lang dir
        It contains language related input files such as
        "lexicon.txt"
        """,
    )

    parser.add_argument(
        "--initial-lr",
        type=float,
        default=0.003,
        help="""The initial learning rate. This value should not need to be
        changed.""",
    )

    parser.add_argument(
        "--context-size",
        type=int,
        default=2,
        help="The context size in the decoder. 1 means bigram; " "2 means tri-gram",
    )

    parser.add_argument(
        "--prune-range",
        type=int,
        default=5,
        help="The prune range for rnnt loss, it means how many symbols(context)"
        "we are using to compute the loss",
    )

    parser.add_argument(
        "--lm-scale",
        type=float,
        default=0.25,
        help="The scale to smooth the loss with lm "
        "(output of prediction network) part.",
    )

    parser.add_argument(
        "--am-scale",
        type=float,
        default=0.0,
        help="The scale to smooth the loss with am (output of encoder network)" "part.",
    )

    parser.add_argument(
        "--simple-loss-scale",
        type=float,
        default=0.5,
        help="To get pruning ranges, we will calculate a simple version"
        "loss(joiner is just addition), this simple loss also uses for"
        "training (as a regularization item). We will scale the simple loss"
        "with this parameter before adding to the final loss.",
    )

    parser.add_argument(
        "--ctc-loss-scale",
        type=float,
        default=0.2,
        help="""Weight for CTC loss, between 0 and 1. 
        When set to 0, only transducer loss is used. 
        When set to 1, only CTC loss is used.""",
    )

    parser.add_argument(
        "--seed",
        type=int,
        default=42,
        help="The seed for random generators intended for reproducibility",
    )

    parser.add_argument(
        "--print-diagnostics",
        type=str2bool,
        default=False,
        help="Accumulate stats on activations, print them and exit.",
    )

    parser.add_argument(
        "--inf-check",
        type=str2bool,
        default=False,
        help="Add hooks to check for infinite module outputs and gradients.",
    )

    parser.add_argument(
        "--save-every-n",
        type=int,
        default=10000,
        help="""Save checkpoint after processing this number of batches"
        periodically. We save checkpoint to exp-dir/ whenever
        params.batch_idx_train % save_every_n == 0. The checkpoint filename
        has the form: f'exp-dir/checkpoint-{params.batch_idx_train}.pt'
        Note: It also saves checkpoint to `exp-dir/epoch-xxx.pt` at the
        end of each epoch where `xxx` is the epoch number counting from 0.
        """,
    )

    parser.add_argument(
        "--keep-last-k",
        type=int,
        default=5,
        help="""Only keep this number of checkpoints on disk.
        For instance, if it is 3, there are only 3 checkpoints
        in the exp-dir with filenames `checkpoint-xxx.pt`.
        It does not affect checkpoints with name `epoch-xxx.pt`.
        """,
    )

    parser.add_argument(
        "--average-period",
        type=int,
        default=200,
        help="""Update the averaged model, namely `model_avg`, after processing
        this number of batches. `model_avg` is a separate version of model,
        in which each floating-point parameter is the average of all the
        parameters from the start of training. Each time we take the average,
        we do: `model_avg = model * (average_period / batch_idx_train) +
            model_avg * ((batch_idx_train - average_period) / batch_idx_train)`.
        """,
    )

    parser.add_argument(
        "--use-fp16",
        type=str2bool,
        default=False,
        help="Whether to use half precision training.",
    )

    add_model_arguments(parser)

    return parser


def get_params() -> AttributeDict:
    """Return a dict containing training parameters.

    All training related parameters that are not passed from the commandline
    are saved in the variable `params`.

    Commandline options are merged into `params` after they are parsed, so
    you can also access them via `params`.

    Explanation of options saved in `params`:

        - best_train_loss: Best training loss so far. It is used to select
                           the model that has the lowest training loss. It is
                           updated during the training.

        - best_valid_loss: Best validation loss so far. It is used to select
                           the model that has the lowest validation loss. It is
                           updated during the training.

        - best_train_epoch: It is the epoch that has the best training loss.

        - best_valid_epoch: It is the epoch that has the best validation loss.

        - batch_idx_train: Used to writing statistics to tensorboard. It
                           contains number of batches trained so far across
                           epochs.

        - log_interval:  Print training loss if batch_idx % log_interval` is 0

        - reset_interval: Reset statistics if batch_idx % reset_interval is 0

        - valid_interval:  Run validation if batch_idx % valid_interval is 0

        - feature_dim: The model input dim. It has to match the one used
                       in computing features.

        - subsampling_factor:  The subsampling factor for the model.

        - num_decoder_layers: Number of decoder layer of transformer decoder.

        - warm_step: The warmup period that dictates the decay of the
              scale on "simple" (un-pruned) loss.
    """
    params = AttributeDict(
        {
            "best_train_loss": float("inf"),
            "best_valid_loss": float("inf"),
            "best_train_epoch": -1,
            "best_valid_epoch": -1,
            "batch_idx_train": 0,
            "log_interval": 500,
            "reset_interval": 200,
            "warm_step": 5000,
            "beam_size": 10,
            "use_double_scores": True,
            "env_info": get_env_info(),
            "feature_dim": 80,
            "subsampling_factor": 4,
            "use_dscnn": True,
            "activation": "doubleswish",
        }
    )

    return params


def get_encoder_model(params: AttributeDict) -> nn.Module:
    encoder = Conv1dNet(
        output_dim=params.encoder_dim,
        input_dim=params.feature_dim,
        conv_layers=params.conv_layers,
        channels=params.channels,
        subsampling_factor=params.subsampling_factor,
        skip_add=params.skip_add,
        dscnn=params.use_dscnn,
        activation=params.activation,
    )

    return encoder


def get_decoder_model(params: AttributeDict) -> nn.Module:
    decoder = Decoder(
        vocab_size=params.vocab_size,
        decoder_dim=params.decoder_dim,
        blank_id=params.blank_id,
        context_size=params.context_size,
    )
    return decoder


def get_joiner_model(params: AttributeDict) -> nn.Module:
    joiner = Joiner(
        encoder_dim=params.encoder_dim,
        decoder_dim=params.decoder_dim,
        joiner_dim=params.joiner_dim,
        vocab_size=params.vocab_size,
    )
    return joiner


def get_transducer_model(params: AttributeDict) -> Transducer:
    encoder = get_encoder_model(params)
    decoder = get_decoder_model(params)
    joiner = get_joiner_model(params)

    if is_module_available("thop"):
        if torch.cuda.is_available():
            device = torch.device("cuda:0")
        else:
            device = torch.device("cpu")
        # Assuming 10ms stride, 1000 frames is about 10 seconds.
        x = torch.zeros((1, 1000, params.feature_dim)).to(device)
        x_lens = torch.Tensor([1000]).int().to(device)
        from thop import clever_format, profile

        m = copy.deepcopy(encoder)
        m = m.to(device)
        ops, _ = clever_format(profile(m, (x, x_lens), verbose=False))
        logging.info(f"Encoder MAC ops for 10 seconds of audio is {ops}")
    else:
        logging.info("You can install thop to calculate the number of ops.")
        logging.info("Command: pip install thop")

    model = Transducer(
        encoder=encoder,
        decoder=decoder,
        joiner=joiner,
        encoder_dim=params.encoder_dim,
        decoder_dim=params.decoder_dim,
        joiner_dim=params.joiner_dim,
        vocab_size=params.vocab_size,
    )
    return model


def load_checkpoint_if_available(
    params: AttributeDict,
    model: nn.Module,
    model_avg: nn.Module = None,
    optimizer: Optional[torch.optim.Optimizer] = None,
    scheduler: Optional[LRSchedulerType] = None,
) -> Optional[Dict[str, Any]]:
    """Load checkpoint from file.

    If params.start_batch is positive, it will load the checkpoint from
    `params.exp_dir/checkpoint-{params.start_batch}.pt`. Otherwise, if
    params.start_epoch is larger than 1, it will load the checkpoint from
    `params.start_epoch - 1`.

    Apart from loading state dict for `model` and `optimizer` it also updates
    `best_train_epoch`, `best_train_loss`, `best_valid_epoch`,
    and `best_valid_loss` in `params`.

    Args:
      params:
        The return value of :func:`get_params`.
      model:
        The training model.
      model_avg:
        The stored model averaged from the start of training.
      optimizer:
        The optimizer that we are using.
      scheduler:
        The scheduler that we are using.
    Returns:
      Return a dict containing previously saved training info.
    """
    if params.start_batch > 0:
        filename = params.exp_dir / f"checkpoint-{params.start_batch}.pt"
    elif params.start_epoch > 1:
        filename = params.exp_dir / f"epoch-{params.start_epoch-1}.pt"
    else:
        return None

    assert filename.is_file(), f"{filename} does not exist!"

    saved_params = load_checkpoint(
        filename,
        model=model,
        model_avg=model_avg,
        optimizer=optimizer,
        scheduler=scheduler,
    )

    keys = [
        "best_train_epoch",
        "best_valid_epoch",
        "batch_idx_train",
        "best_train_loss",
        "best_valid_loss",
    ]
    for k in keys:
        params[k] = saved_params[k]

    if params.start_batch > 0:
        if "cur_epoch" in saved_params:
            params["start_epoch"] = saved_params["cur_epoch"]

        if "cur_batch_idx" in saved_params:
            params["cur_batch_idx"] = saved_params["cur_batch_idx"]

    return saved_params


def save_checkpoint(
    params: AttributeDict,
    model: Union[nn.Module, DDP],
    model_avg: Optional[nn.Module] = None,
    optimizer: Optional[torch.optim.Optimizer] = None,
    scheduler: Optional[LRSchedulerType] = None,
    sampler: Optional[CutSampler] = None,
    scaler: Optional[GradScaler] = None,
    rank: int = 0,
) -> None:
    """Save model, optimizer, scheduler and training stats to file.

    Args:
      params:
        It is returned by :func:`get_params`.
      model:
        The training model.
      model_avg:
        The stored model averaged from the start of training.
      optimizer:
        The optimizer used in the training.
      sampler:
       The sampler for the training dataset.
      scaler:
        The scaler used for mix precision training.
    """
    if rank != 0:
        return
    filename = params.exp_dir / f"epoch-{params.cur_epoch}.pt"
    save_checkpoint_impl(
        filename=filename,
        model=model,
        model_avg=model_avg,
        params=params,
        optimizer=optimizer,
        scheduler=scheduler,
        sampler=sampler,
        scaler=scaler,
        rank=rank,
    )

    if params.best_train_epoch == params.cur_epoch:
        best_train_filename = params.exp_dir / "best-train-loss.pt"
        copyfile(src=filename, dst=best_train_filename)

    if params.best_valid_epoch == params.cur_epoch:
        best_valid_filename = params.exp_dir / "best-valid-loss.pt"
        copyfile(src=filename, dst=best_valid_filename)


def compute_loss(
    params: AttributeDict,
    model: Union[nn.Module, DDP],
    sp: spm.SentencePieceProcessor,
    phone_lexicon: UniqLexicon,
    batch: dict,
    is_training: bool,
) -> Tuple[Tensor, MetricsTracker]:
    """
    Compute transducer loss given the model and its inputs.

    Args:
      params:
        Parameters for training. See :func:`get_params`.
      model:
        The model for training. It is an instance of Zipformer in our case.
      batch:
        A batch of data. See `lhotse.dataset.K2SpeechRecognitionDataset()`
        for the content in it.
      is_training:
        True for training. False for validation. When it is True, this
        function enables autograd during computation; when it is False, it
        disables autograd.
     warmup: a floating point value which increases throughout training;
        values >= 1.0 are fully warmed up and have all modules present.
    """
    device = model.device if isinstance(model, DDP) else next(model.parameters()).device
    feature = batch["inputs"]
    # at entry, feature is (N, T, C)
    assert feature.ndim == 3
    feature = feature.to(device)

    supervisions = batch["supervisions"]
    feature_lens = supervisions["num_frames"].to(device)

    batch_idx_train = params.batch_idx_train
    warm_step = params.warm_step

    texts = batch["supervisions"]["text"]
    if sp is not None:
        token_ids = sp.encode(texts, out_type=int)
        y = k2.RaggedTensor(token_ids).to(device)
    else:
        y = phone_lexicon.texts_to_token_ids(texts).to(device)
        token_ids = y.tolist()

    with torch.set_grad_enabled(is_training):
        simple_loss, pruned_loss, ctc_output = model(
            x=feature,
            x_lens=feature_lens,
            y=y,
            prune_range=params.prune_range,
            am_scale=params.am_scale,
            lm_scale=params.lm_scale,
        )

        s = params.simple_loss_scale
        # take down the scale on the simple loss from 1.0 at the start
        # to params.simple_loss scale by warm_step.
        simple_loss_scale = (
            s
            if batch_idx_train >= warm_step
            else 1.0 - (batch_idx_train / warm_step) * (1.0 - s)
        )
        pruned_loss_scale = (
            1.0
            if batch_idx_train >= warm_step
            else 0.1 + 0.9 * (batch_idx_train / warm_step)
        )

        loss = simple_loss_scale * simple_loss + pruned_loss_scale * pruned_loss

    # Compute ctc loss

    # NOTE: We need `encode_supervisions` to sort sequences with
    # different duration in decreasing order, required by
    # `k2.intersect_dense` called in `k2.ctc_loss`
    with warnings.catch_warnings():
        warnings.simplefilter("ignore")
        supervision_segments, token_ids = encode_supervisions(
            supervisions,
            subsampling_factor=params.subsampling_factor,
            token_ids=token_ids,
        )

    decoding_graph = k2.ctc_graph(token_ids, modified=False, device=device)
    dense_fsa_vec = k2.DenseFsaVec(
        ctc_output,
        supervision_segments,
        allow_truncate=params.subsampling_factor - 1,
    )

    ctc_loss = k2.ctc_loss(
        decoding_graph=decoding_graph,
        dense_fsa_vec=dense_fsa_vec,
        output_beam=params.beam_size,
        reduction="sum",
        use_double_scores=params.use_double_scores,
    )
    assert ctc_loss.requires_grad == is_training
    assert 0 <= params.ctc_loss_scale <= 1, "ctc_loss_scale must be between 0 and 1"
    loss = params.ctc_loss_scale * ctc_loss + (1 - params.ctc_loss_scale) * loss

    assert loss.requires_grad == is_training

    info = MetricsTracker()
    with warnings.catch_warnings():
        warnings.simplefilter("ignore")
        info["frames"] = (feature_lens // params.subsampling_factor).sum().item()

    # Note: We use reduction=sum while computing the loss.
    info["loss"] = loss.detach().cpu().item()
    info["simple_loss"] = simple_loss.detach().cpu().item()
    info["pruned_loss"] = pruned_loss.detach().cpu().item()
    info["ctc_loss"] = ctc_loss.detach().cpu().item()

    return loss, info


def compute_validation_loss(
    params: AttributeDict,
    model: Union[nn.Module, DDP],
    sp: spm.SentencePieceProcessor,
    phone_lexicon: UniqLexicon,
    valid_dl: torch.utils.data.DataLoader,
    world_size: int = 1,
) -> MetricsTracker:
    """Run the validation process."""
    model.eval()

    tot_loss = MetricsTracker()

    for batch_idx, batch in enumerate(valid_dl):
        loss, loss_info = compute_loss(
            params=params,
            model=model,
            sp=sp,
            phone_lexicon=phone_lexicon,
            batch=batch,
            is_training=False,
        )
        assert loss.requires_grad is False
        tot_loss = tot_loss + loss_info

    if world_size > 1:
        tot_loss.reduce(loss.device)

    loss_value = tot_loss["loss"] / tot_loss["frames"]
    if loss_value < params.best_valid_loss:
        params.best_valid_epoch = params.cur_epoch
        params.best_valid_loss = loss_value

    return tot_loss


def train_one_epoch(
    params: AttributeDict,
    model: Union[nn.Module, DDP],
    optimizer: torch.optim.Optimizer,
    scheduler: LRSchedulerType,
    sp: spm.SentencePieceProcessor,
    phone_lexicon: UniqLexicon,
    train_dl: torch.utils.data.DataLoader,
    valid_dl: torch.utils.data.DataLoader,
    scaler: GradScaler,
    model_avg: Optional[nn.Module] = None,
    tb_writer: Optional[SummaryWriter] = None,
    world_size: int = 1,
    rank: int = 0,
) -> None:
    """Train the model for one epoch.

    The training loss from the mean of all frames is saved in
    `params.train_loss`. It runs the validation process every
    `params.valid_interval` batches.

    Args:
      params:
        It is returned by :func:`get_params`.
      model:
        The model for training.
      optimizer:
        The optimizer we are using.
      scheduler:
        The learning rate scheduler, we call step() every step.
      train_dl:
        Dataloader for the training dataset.
      valid_dl:
        Dataloader for the validation dataset.
      scaler:
        The scaler used for mix precision training.
      model_avg:
        The stored model averaged from the start of training.
      tb_writer:
        Writer to write log messages to tensorboard.
      world_size:
        Number of nodes in DDP training. If it is 1, DDP is disabled.
      rank:
        The rank of the node in DDP training. If no DDP is used, it should
        be set to 0.
    """
    model.train()

    tot_loss = MetricsTracker()

    cur_batch_idx = params.get("cur_batch_idx", 0)

    for batch_idx, batch in enumerate(train_dl):
        if batch_idx < cur_batch_idx:
            continue
        cur_batch_idx = batch_idx

        params.batch_idx_train += 1
        batch_size = len(batch["supervisions"]["text"])

        try:
            with torch.cuda.amp.autocast(enabled=params.use_fp16):
                loss, loss_info = compute_loss(
                    params=params,
                    model=model,
                    sp=sp,
                    phone_lexicon=phone_lexicon,
                    batch=batch,
                    is_training=True,
                )
            # summary stats
            tot_loss = (tot_loss * (1 - 1 / params.reset_interval)) + loss_info

            # NOTE: We use reduction==sum and loss is computed over utterances
            # in the batch and there is no normalization to it so far.
            scaler.scale(loss).backward()
            set_batch_count(model, params.batch_idx_train)
            # scheduler.step_batch(params.batch_idx_train)

            scaler.step(optimizer)
            scaler.update()
            optimizer.zero_grad()
        except:  # noqa
            display_and_save_batch(
                batch, params=params, sp=sp, phone_lexicon=phone_lexicon
            )
            raise

        if params.print_diagnostics and batch_idx == 5:
            return

        if (
            rank == 0
            and params.batch_idx_train > 0
            and params.batch_idx_train % params.average_period == 0
        ):
            update_averaged_model(
                params=params,
                model_cur=model,
                model_avg=model_avg,
            )

        if (
            params.batch_idx_train > 0
            and params.batch_idx_train % params.save_every_n == 0
        ):
            params.cur_batch_idx = batch_idx
            save_checkpoint_with_global_batch_idx(
                out_dir=params.exp_dir,
                global_batch_idx=params.batch_idx_train,
                model=model,
                model_avg=model_avg,
                params=params,
                optimizer=optimizer,
                scheduler=scheduler,
                sampler=train_dl.sampler,
                scaler=scaler,
                rank=rank,
            )
            del params.cur_batch_idx
            remove_checkpoints(
                out_dir=params.exp_dir,
                topk=params.keep_last_k,
                rank=rank,
            )

        if batch_idx % 100 == 0 and params.use_fp16:
            # If the grad scale was less than 1, try increasing it.    The _growth_interval
            # of the grad scaler is configurable, but we can't configure it to have different
            # behavior depending on the current grad scale.
            cur_grad_scale = scaler._scale.item()
            if cur_grad_scale < 1.0 or (cur_grad_scale < 8.0 and batch_idx % 400 == 0):
                scaler.update(cur_grad_scale * 2.0)
            if cur_grad_scale < 0.01:
                logging.warning(f"Grad scale is small: {cur_grad_scale}")
            if cur_grad_scale < 1.0e-05:
                raise_grad_scale_is_too_small_error(cur_grad_scale)

        if batch_idx % params.log_interval == 0:
            cur_lr = scheduler.get_last_lr()[0]
            cur_grad_scale = scaler._scale.item() if params.use_fp16 else 1.0

            logging.info(
                f"Epoch {params.cur_epoch}, "
                f"batch {batch_idx}, loss[{loss_info}], "
                f"tot_loss[{tot_loss}], batch size: {batch_size}, "
                f"lr: {cur_lr:.2e}, "
                + (f"grad_scale: {scaler._scale.item()}" if params.use_fp16 else "")
            )

            if tb_writer is not None:
                tb_writer.add_scalar(
                    "train/learning_rate", cur_lr, params.batch_idx_train
                )

                loss_info.write_summary(
                    tb_writer, "train/current_", params.batch_idx_train
                )
                tot_loss.write_summary(tb_writer, "train/tot_", params.batch_idx_train)
                if params.use_fp16:
                    tb_writer.add_scalar(
                        "train/grad_scale", cur_grad_scale, params.batch_idx_train
                    )

    logging.info("Computing validation loss")
    valid_info = compute_validation_loss(
        params=params,
        model=model,
        sp=sp,
        phone_lexicon=phone_lexicon,
        valid_dl=valid_dl,
        world_size=world_size,
    )
    model.train()
    logging.info(f"Epoch {params.cur_epoch}, validation: {valid_info}")
    logging.info(
        f"Maximum memory allocated so far is {torch.cuda.max_memory_allocated()//1000000}MB"
    )
    if tb_writer is not None:
        valid_info.write_summary(tb_writer, "train/valid_", params.batch_idx_train)

    loss_value = tot_loss["loss"] / tot_loss["frames"]
    params.train_loss = loss_value
    if params.train_loss < params.best_train_loss:
        params.best_train_epoch = params.cur_epoch
        params.best_train_loss = params.train_loss


def run(rank, world_size, args):
    """
    Args:
      rank:
        It is a value between 0 and `world_size-1`, which is
        passed automatically by `mp.spawn()` in :func:`main`.
        The node with rank 0 is responsible for saving checkpoint.
      world_size:
        Number of GPUs for DDP training.
      args:
        The return value of get_parser().parse_args()
    """
    params = get_params()
    params.update(vars(args))
    if params.full_libri is False:
        params.valid_interval = 1600

    fix_random_seed(params.seed)
    if world_size > 1:
        setup_dist(rank, world_size, params.master_port)

    setup_logger(f"{params.exp_dir}/log/log-train")
    logging.info("Training started")

    if args.tensorboard and rank == 0:
        tb_writer = SummaryWriter(log_dir=f"{params.exp_dir}/tensorboard")
    else:
        tb_writer = None

    device = torch.device("cpu")
    if torch.cuda.is_available():
        device = torch.device("cuda", rank)
    logging.info(f"Device: {device}")

    if "lang_bpe" in str(params.lang_dir):
        sp = spm.SentencePieceProcessor()
        sp.load(params.lang_dir + "/bpe.model")
        # <blk> is defined in local/train_bpe_model.py
        params.blank_id = sp.piece_to_id("<blk>")
        params.vocab_size = sp.get_piece_size()
        phone_lexicon = None
    else:
        assert "lang_phone" in str(params.lang_dir)
        phone_lexicon = UniqLexicon(params.lang_dir)
        params.blank_id = 0
        params.vocab_size = max(phone_lexicon.tokens) + 1
        sp = None

    logging.info(pprint.pformat(params, indent=2))

    logging.info("About to create model")
    model = get_transducer_model(params)

    if rank == 0:
        num_param = sum([p.numel() for p in model.parameters()])
        enc_param = sum([p.numel() for p in model.encoder.parameters()])
        dec_param = sum([p.numel() for p in model.decoder.parameters()])
        join_param = sum([p.numel() for p in model.joiner.parameters()])
        ctc_param = sum([p.numel() for p in model.ctc_output.parameters()])

        logging.info(f"Number of model parameters: {num_param}")
        logging.info(f"Number of encoder parameters: {enc_param}")
        logging.info(f"Number of decoder parameters: {dec_param}")
        logging.info(f"Number of joiner parameters: {join_param}")
        logging.info(f"Number of ctc parameters: {ctc_param}")

    assert params.save_every_n >= params.average_period
    model_avg: Optional[nn.Module] = None
    if rank == 0:
        # model_avg is only used with rank 0
        model_avg = copy.deepcopy(model).to(torch.float64)

    assert params.start_epoch > 0, params.start_epoch
    checkpoints = load_checkpoint_if_available(
        params=params, model=model, model_avg=model_avg
    )

    model.to(device)
    if world_size > 1:
        logging.info("Using DDP")
        model = DDP(model, device_ids=[rank], find_unused_parameters=True)

    optimizer = AdamW(
        model.parameters(),
        lr=params.initial_lr,
        weight_decay=5e-4,
    )

    scheduler = StepLR(optimizer, step_size=2, gamma=0.8)

    if checkpoints and "optimizer" in checkpoints:
        logging.info("Loading optimizer state dict")
        optimizer.load_state_dict(checkpoints["optimizer"])

    if (
        checkpoints
        and "scheduler" in checkpoints
        and checkpoints["scheduler"] is not None
    ):
        logging.info("Loading scheduler state dict")
        scheduler.load_state_dict(checkpoints["scheduler"])

    if params.print_diagnostics:
        opts = diagnostics.TensorDiagnosticOptions(
            512
        )  # allow 4 megabytes per sub-module
        diagnostic = diagnostics.attach_diagnostics(model, opts)

    if params.inf_check:
        register_inf_check_hooks(model)

    librispeech = LibriSpeechAsrDataModule(args)

    if params.full_libri:
        train_cuts = librispeech.train_all_shuf_cuts()
    else:
        train_cuts = librispeech.train_clean_100_cuts()

    def remove_short_and_long_utt(c: Cut):
        # Keep only utterances with duration between 1 second and 20 seconds
        #
        # Caution: There is a reason to select 20.0 here. Please see
        # ../local/display_manifest_statistics.py
        #
        # You should use ../local/display_manifest_statistics.py to get
        # an utterance duration distribution for your dataset to select
        # the threshold
        return 1.0 <= c.duration <= 20.0

    train_cuts = train_cuts.filter(remove_short_and_long_utt)

    if params.start_batch > 0 and checkpoints and "sampler" in checkpoints:
        # We only load the sampler's state dict when it loads a checkpoint
        # saved in the middle of an epoch
        sampler_state_dict = checkpoints["sampler"]
    else:
        sampler_state_dict = None

    train_dl = librispeech.train_dataloaders(
        train_cuts, sampler_state_dict=sampler_state_dict
    )

    valid_cuts = librispeech.dev_clean_cuts()
    valid_cuts += librispeech.dev_other_cuts()
    valid_dl = librispeech.valid_dataloaders(valid_cuts)

    # if not params.print_diagnostics:
    #     scan_pessimistic_batches_for_oom(
    #         model=model,
    #         train_dl=train_dl,
    #         optimizer=optimizer,
    #         sp=sp,
    #         phone_lexicon=phone_lexicon,
    #         params=params,
    #     )

    scaler = GradScaler(enabled=params.use_fp16, init_scale=1.0)
    if checkpoints and "grad_scaler" in checkpoints:
        logging.info("Loading grad scaler state dict")
        scaler.load_state_dict(checkpoints["grad_scaler"])

    for epoch in range(params.start_epoch, params.num_epochs + 1):
        # scheduler.step_epoch(epoch - 1)
        fix_random_seed(params.seed + epoch - 1)
        train_dl.sampler.set_epoch(epoch - 1)

        if tb_writer is not None:
            tb_writer.add_scalar("train/epoch", epoch, params.batch_idx_train)

        params.cur_epoch = epoch

        train_one_epoch(
            params=params,
            model=model,
            model_avg=model_avg,
            optimizer=optimizer,
            scheduler=scheduler,
            sp=sp,
            phone_lexicon=phone_lexicon,
            train_dl=train_dl,
            valid_dl=valid_dl,
            scaler=scaler,
            tb_writer=tb_writer,
            world_size=world_size,
            rank=rank,
        )

        if params.print_diagnostics:
            diagnostic.print_diagnostics()
            break

        save_checkpoint(
            params=params,
            model=model,
            model_avg=model_avg,
            optimizer=optimizer,
            scheduler=scheduler,
            sampler=train_dl.sampler,
            scaler=scaler,
            rank=rank,
        )

    logging.info("Done!")

    if world_size > 1:
        torch.distributed.barrier()
        cleanup_dist()


def display_and_save_batch(
    batch: dict,
    params: AttributeDict,
    sp: spm.SentencePieceProcessor,
    phone_lexicon: UniqLexicon,
) -> None:
    """Display the batch statistics and save the batch into disk.

    Args:
      batch:
        A batch of data. See `lhotse.dataset.K2SpeechRecognitionDataset()`
        for the content in it.
      params:
        Parameters for training. See :func:`get_params`.
      sp:
        The BPE model.
    """
    from lhotse.utils import uuid4

    filename = f"{params.exp_dir}/batch-{uuid4()}.pt"
    logging.info(f"Saving batch to {filename}")
    torch.save(batch, filename)

    supervisions = batch["supervisions"]
    features = batch["inputs"]

    logging.info(f"features shape: {features.shape}")

    if sp is not None:
        y = sp.encode(supervisions["text"], out_type=int)
    else:
        y = phone_lexicon.texts_to_token_ids(supervisions["text"])
    num_tokens = sum(len(i) for i in y)
    logging.info(f"num tokens: {num_tokens}")


def scan_pessimistic_batches_for_oom(
    model: Union[nn.Module, DDP],
    train_dl: torch.utils.data.DataLoader,
    optimizer: torch.optim.Optimizer,
    sp: spm.SentencePieceProcessor,
    phone_lexicon: UniqLexicon,
    params: AttributeDict,
):
    from lhotse.dataset import find_pessimistic_batches

    logging.info(
        "Sanity check -- see if any of the batches in epoch 1 would cause OOM."
    )
    batches, crit_values = find_pessimistic_batches(train_dl.sampler)
    for criterion, cuts in batches.items():
        batch = train_dl.dataset[cuts]
        try:
            with torch.cuda.amp.autocast(enabled=params.use_fp16):
                loss, _ = compute_loss(
                    params=params,
                    model=model,
                    sp=sp,
                    phone_lexicon=phone_lexicon,
                    batch=batch,
                    is_training=True,
                )
            loss.backward()
            optimizer.zero_grad()
        except Exception as e:
            if "CUDA out of memory" in str(e):
                logging.error(
                    "Your GPU ran out of memory with the current "
                    "max_duration setting. We recommend decreasing "
                    "max_duration and trying again.\n"
                    f"Failing criterion: {criterion} "
                    f"(={crit_values[criterion]}) ..."
                )
            display_and_save_batch(
                batch, params=params, sp=sp, phone_lexicon=phone_lexicon
            )
            raise
        logging.info(
            f"Maximum memory allocated so far is {torch.cuda.max_memory_allocated()//1000000}MB"
        )


def main():
    parser = get_parser()
    LibriSpeechAsrDataModule.add_arguments(parser)
    args = parser.parse_args()
    args.exp_dir = Path(args.exp_dir)

    world_size = args.world_size
    assert world_size >= 1
    if world_size > 1:
        mp.spawn(run, args=(world_size, args), nprocs=world_size, join=True)
    else:
        run(rank=0, world_size=1, args=args)


torch.set_num_threads(1)
torch.set_num_interop_threads(1)

if __name__ == "__main__":
    main()