mirrors/icefall
1
0
Fork 0
mirror of https://github.com/k2-fsa/icefall.git synced 2025-08-08 09:32:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add LSTM for the multi-dataset setup. (#558)

Browse Source 
* Add LSTM for the multi-dataset setup.

* Add results

* fix style issues

* add missing file
This commit is contained in:
Fangjun Kuang 2022-09-16 18:40:25 +08:00 committed by GitHub
parent 145c44f710
commit 97b3fc53aa
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
24 changed files with 3513 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
.flake8
View File

@ -22,3 +22,11 @@ exclude =
**/data/**,
icefall/shared/make_kn_lm.py,
icefall/__init__.py
ignore =
# E203 white space before ":"
E203,
# W503 line break before binary operator
W503,
# E226 missing whitespace around arithmetic operator
E226,

1
egs/librispeech/ASR/README.md
View File

@ -26,6 +26,7 @@ The following table lists the differences among them.
| `conv_emformer_transducer_stateless` | ConvEmformer | Embedding + Conv1d | Using ConvEmformer for streaming ASR + mechanisms in reworked model |
| `conv_emformer_transducer_stateless2` | ConvEmformer | Embedding + Conv1d | Using ConvEmformer with simplified memory for streaming ASR + mechanisms in reworked model |
| `lstm_transducer_stateless` | LSTM | Embedding + Conv1d | Using LSTM with mechanisms in reworked model |
| `lstm_transducer_stateless2` | LSTM | Embedding + Conv1d | Using LSTM with mechanisms in reworked model + gigaspeech (multi-dataset setup) |
The decoder in `transducer_stateless` is modified from the paper
[Rnn-Transducer with Stateless Prediction Network](https://ieeexplore.ieee.org/document/9054419/).

70
egs/librispeech/ASR/RESULTS.md
View File

@ -1,5 +1,75 @@
## Results
#### LibriSpeech BPE training results (Pruned Stateless LSTM RNN-T + multi-dataset)
[lstm_transducer_stateless2](./lstm_transducer_stateless2)
See <https://github.com/k2-fsa/icefall/pull/558> for more details.
The WERs are:
| | test-clean | test-other | comment |
|-------------------------------------|------------|------------|-------------------------|
| greedy search (max sym per frame 1) | 2.78 | 7.36 | --iter 468000 --avg 16 |
| modified_beam_search | 2.73 | 7.15 | --iter 468000 --avg 16 |
| fast_beam_search | 2.76 | 7.31 | --iter 468000 --avg 16 |
| greedy search (max sym per frame 1) | 2.77 | 7.35 | --iter 472000 --avg 18 |
| modified_beam_search | 2.75 | 7.08 | --iter 472000 --avg 18 |
| fast_beam_search | 2.77 | 7.29 | --iter 472000 --avg 18 |
The training command is:
```bash
#!/usr/bin/env bash
export CUDA_VISIBLE_DEVICES="0,1,2,3,4,5,6,7"
./lstm_transducer_stateless2/train.py \
--world-size 8 \
--num-epochs 35 \
--start-epoch 1 \
--full-libri 1 \
--exp-dir lstm_transducer_stateless2/exp \
--max-duration 500 \
--use-fp16 0 \
--lr-epochs 10 \
--num-workers 2 \
--giga-prob 0.9
```
**Note**: It was killed manually after getting `epoch-18.pt`. Also, we resumed
training after getting `epoch-9.pt`.
The tensorboard log can be found at
<https://tensorboard.dev/experiment/1ziQ2LFmQY2mt4dlUr5dyA/>
The decoding command is
```bash
for m in greedy_search fast_beam_search modified_beam_search; do
for iter in 472000; do
for avg in 8 10 12 14 16 18; do
./lstm_transducer_stateless2/decode.py \
--iter $iter \
--avg $avg \
--exp-dir lstm_transducer_stateless2/exp \
--max-duration 600 \
--num-encoder-layers 12 \
--rnn-hidden-size 1024 \
--decoding-method $m \
--use-averaged-model True \
--beam 4 \
--max-contexts 4 \
--max-states 8 \
--beam-size 4
done
done
done
```
Pretrained models, training logs, decoding logs, and decoding results
are available at
<https://huggingface.co/csukuangfj/icefall-asr-librispeech-lstm-transducer-stateless2-2022-09-03>
#### LibriSpeech BPE training results (Pruned Stateless LSTM RNN-T)
[lstm_transducer_stateless](./lstm_transducer_stateless)

2
egs/librispeech/ASR/lstm_transducer_stateless/lstm.py
View File

@ -773,7 +773,7 @@ class RandomCombine(nn.Module):
"""
logprobs = (
torch.randn(num_frames, self.num_inputs, dtype=dtype, device=device)
* self.stddev
* self.stddev # noqa
)
logprobs[:, -1] += self.final_log_weight
return logprobs.softmax(dim=1)

0
egs/librispeech/ASR/lstm_transducer_stateless2/__init__.py Normal file
View File

1
egs/librispeech/ASR/lstm_transducer_stateless2/asr_datamodule.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless3/asr_datamodule.py

1
egs/librispeech/ASR/lstm_transducer_stateless2/beam_search.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless2/beam_search.py

827
egs/librispeech/ASR/lstm_transducer_stateless2/decode.py Executable file
View File

@ -0,0 +1,827 @@
#!/usr/bin/env python3
#
# Copyright 2021-2022 Xiaomi Corporation (Author: Fangjun Kuang,
# 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:
(1) greedy search
./lstm_transducer_stateless2/decode.py \
--epoch 35 \
--avg 15 \
--exp-dir ./lstm_transducer_stateless2/exp \
--max-duration 600 \
--decoding-method greedy_search
(2) beam search (not recommended)
./lstm_transducer_stateless2/decode.py \
--epoch 35 \
--avg 15 \
--exp-dir ./lstm_transducer_stateless2/exp \
--max-duration 600 \
--decoding-method beam_search \
--beam-size 4
(3) modified beam search
./lstm_transducer_stateless2/decode.py \
--epoch 35 \
--avg 15 \
--exp-dir ./lstm_transducer_stateless2/exp \
--max-duration 600 \
--decoding-method modified_beam_search \
--beam-size 4
(4) fast beam search (one best)
./lstm_transducer_stateless2/decode.py \
--epoch 35 \
--avg 15 \
--exp-dir ./lstm_transducer_stateless2/exp \
--max-duration 600 \
--decoding-method fast_beam_search \
--beam 20.0 \
--max-contexts 8 \
--max-states 64
(5) fast beam search (nbest)
./lstm_transducer_stateless2/decode.py \
--epoch 30 \
--avg 15 \
--exp-dir ./pruned_transducer_stateless3/exp \
--max-duration 600 \
--decoding-method fast_beam_search_nbest \
--beam 20.0 \
--max-contexts 8 \
--max-states 64 \
--num-paths 200 \
--nbest-scale 0.5
(6) fast beam search (nbest oracle WER)
./lstm_transducer_stateless2/decode.py \
--epoch 35 \
--avg 15 \
--exp-dir ./lstm_transducer_stateless2/exp \
--max-duration 600 \
--decoding-method fast_beam_search_nbest_oracle \
--beam 20.0 \
--max-contexts 8 \
--max-states 64 \
--num-paths 200 \
--nbest-scale 0.5
(7) fast beam search (with LG)
./lstm_transducer_stateless2/decode.py \
--epoch 35 \
--avg 15 \
--exp-dir ./lstm_transducer_stateless2/exp \
--max-duration 600 \
--decoding-method fast_beam_search_nbest_LG \
--beam 20.0 \
--max-contexts 8 \
--max-states 64
"""
import argparse
import logging
import math
from collections import defaultdict
from pathlib import Path
from typing import Dict, List, Optional, Tuple
import k2
import sentencepiece as spm
import torch
import torch.nn as nn
from asr_datamodule import AsrDataModule
from beam_search import (
beam_search,
fast_beam_search_nbest,
fast_beam_search_nbest_LG,
fast_beam_search_nbest_oracle,
fast_beam_search_one_best,
greedy_search,
greedy_search_batch,
modified_beam_search,
)
from librispeech import LibriSpeech
from train import add_model_arguments, get_params, get_transducer_model
from icefall.checkpoint import (
average_checkpoints,
average_checkpoints_with_averaged_model,
find_checkpoints,
load_checkpoint,
)
from icefall.lexicon import Lexicon
from icefall.utils import (
AttributeDict,
setup_logger,
store_transcripts,
str2bool,
write_error_stats,
)
LOG_EPS = math.log(1e-10)
def get_parser():
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter
)
parser.add_argument(
"--epoch",
type=int,
default=30,
help="""It specifies the checkpoint to use for decoding.
Note: Epoch counts from 1.
You can specify --avg to use more checkpoints for model averaging.""",
)
parser.add_argument(
"--iter",
type=int,
default=0,
help="""If positive, --epoch is ignored and it
will use the checkpoint exp_dir/checkpoint-iter.pt.
You can specify --avg to use more checkpoints for model averaging.
""",
)
parser.add_argument(
"--avg",
type=int,
default=15,
help="Number of checkpoints to average. Automatically select "
"consecutive checkpoints before the checkpoint specified by "
"'--epoch' and '--iter'",
)
parser.add_argument(
"--use-averaged-model",
type=str2bool,
default=True,
help="Whether to load averaged model. Currently it only supports "
"using --epoch. If True, it would decode with the averaged model "
"over the epoch range from `epoch-avg` (excluded) to `epoch`."
"Actually only the models with epoch number of `epoch-avg` and "
"`epoch` are loaded for averaging. ",
)
parser.add_argument(
"--exp-dir",
type=str,
default="lstm_transducer_stateless2/exp",
help="The experiment dir",
)
parser.add_argument(
"--bpe-model",
type=str,
default="data/lang_bpe_500/bpe.model",
help="Path to the BPE model",
)
parser.add_argument(
"--lang-dir",
type=Path,
default="data/lang_bpe_500",
help="The lang dir containing word table and LG graph",
)
parser.add_argument(
"--decoding-method",
type=str,
default="greedy_search",
help="""Possible values are:
- greedy_search
- beam_search
- modified_beam_search
- fast_beam_search
- fast_beam_search_nbest
- fast_beam_search_nbest_oracle
- fast_beam_search_nbest_LG
If you use fast_beam_search_nbest_LG, you have to specify
`--lang-dir`, which should contain `LG.pt`.
""",
)
parser.add_argument(
"--beam-size",
type=int,
default=4,
help="""An integer indicating how many candidates we will keep for each
frame. Used only when --decoding-method is beam_search or
modified_beam_search.""",
)
parser.add_argument(
"--beam",
type=float,
default=20.0,
help="""A floating point value to calculate the cutoff score during beam
search (i.e., `cutoff = max-score - beam`), which is the same as the
`beam` in Kaldi.
Used only when --decoding-method is fast_beam_search,
fast_beam_search_nbest, fast_beam_search_nbest_LG,
and fast_beam_search_nbest_oracle
""",
)
parser.add_argument(
"--ngram-lm-scale",
type=float,
default=0.01,
help="""
Used only when --decoding_method is fast_beam_search_nbest_LG.
It specifies the scale for n-gram LM scores.
""",
)
parser.add_argument(
"--max-contexts",
type=int,
default=8,
help="""Used only when --decoding-method is
fast_beam_search, fast_beam_search_nbest, fast_beam_search_nbest_LG,
and fast_beam_search_nbest_oracle""",
)
parser.add_argument(
"--max-states",
type=int,
default=64,
help="""Used only when --decoding-method is
fast_beam_search, fast_beam_search_nbest, fast_beam_search_nbest_LG,
and fast_beam_search_nbest_oracle""",
)
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(
"--max-sym-per-frame",
type=int,
default=1,
help="""Maximum number of symbols per frame.
Used only when --decoding_method is greedy_search""",
)
parser.add_argument(
"--num-paths",
type=int,
default=200,
help="""Number of paths for nbest decoding.
Used only when the decoding method is fast_beam_search_nbest,
fast_beam_search_nbest_LG, and fast_beam_search_nbest_oracle""",
)
parser.add_argument(
"--nbest-scale",
type=float,
default=0.5,
help="""Scale applied to lattice scores when computing nbest paths.
Used only when the decoding method is fast_beam_search_nbest,
fast_beam_search_nbest_LG, and fast_beam_search_nbest_oracle""",
)
add_model_arguments(parser)
return parser
def decode_one_batch(
params: AttributeDict,
model: nn.Module,
sp: spm.SentencePieceProcessor,
batch: dict,
word_table: Optional[k2.SymbolTable] = None,
decoding_graph: Optional[k2.Fsa] = None,
) -> Dict[str, List[List[str]]]:
"""Decode one batch and return the result in a dict. The dict has the
following format:
- key: It indicates the setting used for decoding. For example,
if greedy_search is used, it would be "greedy_search"
If beam search with a beam size of 7 is used, it would be
"beam_7"
- value: It contains the decoding result. `len(value)` equals to
batch size. `value[i]` is the decoding result for the i-th
utterance in the given batch.
Args:
params:
It's the return value of :func:`get_params`.
model:
The neural model.
sp:
The BPE model.
batch:
It is the return value from iterating
`lhotse.dataset.K2SpeechRecognitionDataset`. See its documentation
for the format of the `batch`.
word_table:
The word symbol table.
decoding_graph:
The decoding graph. Can be either a `k2.trivial_graph` or LG, Used
only when --decoding_method is fast_beam_search, fast_beam_search_nbest,
fast_beam_search_nbest_oracle, and fast_beam_search_nbest_LG.
Returns:
Return the decoding result. See above description for the format of
the returned dict.
"""
device = next(model.parameters()).device
feature = batch["inputs"]
assert feature.ndim == 3
feature = feature.to(device)
# at entry, feature is (N, T, C)
supervisions = batch["supervisions"]
feature_lens = supervisions["num_frames"].to(device)
# tail padding here to alleviate the tail deletion problem
num_tail_padded_frames = 35
feature = torch.nn.functional.pad(
feature,
(0, 0, 0, num_tail_padded_frames),
mode="constant",
value=LOG_EPS,
)
feature_lens += num_tail_padded_frames
encoder_out, encoder_out_lens, _ = model.encoder(
x=feature, x_lens=feature_lens
)
hyps = []
if params.decoding_method == "fast_beam_search":
hyp_tokens = fast_beam_search_one_best(
model=model,
decoding_graph=decoding_graph,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
beam=params.beam,
max_contexts=params.max_contexts,
max_states=params.max_states,
)
for hyp in sp.decode(hyp_tokens):
hyps.append(hyp.split())
elif params.decoding_method == "fast_beam_search_nbest_LG":
hyp_tokens = fast_beam_search_nbest_LG(
model=model,
decoding_graph=decoding_graph,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
beam=params.beam,
max_contexts=params.max_contexts,
max_states=params.max_states,
num_paths=params.num_paths,
nbest_scale=params.nbest_scale,
)
for hyp in hyp_tokens:
hyps.append([word_table[i] for i in hyp])
elif params.decoding_method == "fast_beam_search_nbest":
hyp_tokens = fast_beam_search_nbest(
model=model,
decoding_graph=decoding_graph,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
beam=params.beam,
max_contexts=params.max_contexts,
max_states=params.max_states,
num_paths=params.num_paths,
nbest_scale=params.nbest_scale,
)
for hyp in sp.decode(hyp_tokens):
hyps.append(hyp.split())
elif params.decoding_method == "fast_beam_search_nbest_oracle":
hyp_tokens = fast_beam_search_nbest_oracle(
model=model,
decoding_graph=decoding_graph,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
beam=params.beam,
max_contexts=params.max_contexts,
max_states=params.max_states,
num_paths=params.num_paths,
ref_texts=sp.encode(supervisions["text"]),
nbest_scale=params.nbest_scale,
)
for hyp in sp.decode(hyp_tokens):
hyps.append(hyp.split())
elif (
params.decoding_method == "greedy_search"
and params.max_sym_per_frame == 1
):
hyp_tokens = greedy_search_batch(
model=model,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
)
for hyp in sp.decode(hyp_tokens):
hyps.append(hyp.split())
elif params.decoding_method == "modified_beam_search":
hyp_tokens = modified_beam_search(
model=model,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
beam=params.beam_size,
)
for hyp in sp.decode(hyp_tokens):
hyps.append(hyp.split())
else:
batch_size = encoder_out.size(0)
for i in range(batch_size):
# fmt: off
encoder_out_i = encoder_out[i:i + 1, :encoder_out_lens[i]]
# fmt: on
if params.decoding_method == "greedy_search":
hyp = greedy_search(
model=model,
encoder_out=encoder_out_i,
max_sym_per_frame=params.max_sym_per_frame,
)
elif params.decoding_method == "beam_search":
hyp = beam_search(
model=model,
encoder_out=encoder_out_i,
beam=params.beam_size,
)
else:
raise ValueError(
f"Unsupported decoding method: {params.decoding_method}"
)
hyps.append(sp.decode(hyp).split())
if params.decoding_method == "greedy_search":
return {"greedy_search": hyps}
elif "fast_beam_search" in params.decoding_method:
key = f"beam_{params.beam}_"
key += f"max_contexts_{params.max_contexts}_"
key += f"max_states_{params.max_states}"
if "nbest" in params.decoding_method:
key += f"_num_paths_{params.num_paths}_"
key += f"nbest_scale_{params.nbest_scale}"
if "LG" in params.decoding_method:
key += f"_ngram_lm_scale_{params.ngram_lm_scale}"
return {key: hyps}
else:
return {f"beam_size_{params.beam_size}": hyps}
def decode_dataset(
dl: torch.utils.data.DataLoader,
params: AttributeDict,
model: nn.Module,
sp: spm.SentencePieceProcessor,
word_table: Optional[k2.SymbolTable] = None,
decoding_graph: Optional[k2.Fsa] = None,
) -> Dict[str, List[Tuple[List[str], List[str]]]]:
"""Decode dataset.
Args:
dl:
PyTorch's dataloader containing the dataset to decode.
params:
It is returned by :func:`get_params`.
model:
The neural model.
sp:
The BPE model.
word_table:
The word symbol table.
decoding_graph:
The decoding graph. Can be either a `k2.trivial_graph` or LG, Used
only when --decoding_method is fast_beam_search, fast_beam_search_nbest,
fast_beam_search_nbest_oracle, and fast_beam_search_nbest_LG.
Returns:
Return a dict, whose key may be "greedy_search" if greedy search
is used, or it may be "beam_7" if beam size of 7 is used.
Its value is a list of tuples. Each tuple contains two elements:
The first is the reference transcript, and the second is the
predicted result.
"""
num_cuts = 0
try:
num_batches = len(dl)
except TypeError:
num_batches = "?"
if params.decoding_method == "greedy_search":
log_interval = 50
else:
log_interval = 20
results = defaultdict(list)
for batch_idx, batch in enumerate(dl):
texts = batch["supervisions"]["text"]
cut_ids = [cut.id for cut in batch["supervisions"]["cut"]]
hyps_dict = decode_one_batch(
params=params,
model=model,
sp=sp,
decoding_graph=decoding_graph,
word_table=word_table,
batch=batch,
)
for name, hyps in hyps_dict.items():
this_batch = []
assert len(hyps) == len(texts)
for cut_id, hyp_words, ref_text in zip(cut_ids, hyps, texts):
ref_words = ref_text.split()
this_batch.append((cut_id, ref_words, hyp_words))
results[name].extend(this_batch)
num_cuts += len(texts)
if batch_idx % log_interval == 0:
batch_str = f"{batch_idx}/{num_batches}"
logging.info(
f"batch {batch_str}, cuts processed until now is {num_cuts}"
)
return results
def save_results(
params: AttributeDict,
test_set_name: str,
results_dict: Dict[str, List[Tuple[List[int], List[int]]]],
):
test_set_wers = dict()
for key, results in results_dict.items():
recog_path = (
params.res_dir / f"recogs-{test_set_name}-{key}-{params.suffix}.txt"
)
results = sorted(results)
store_transcripts(filename=recog_path, texts=results)
logging.info(f"The transcripts are stored in {recog_path}")
# The following prints out WERs, per-word error statistics and aligned
# ref/hyp pairs.
errs_filename = (
params.res_dir / f"errs-{test_set_name}-{key}-{params.suffix}.txt"
)
with open(errs_filename, "w") as f:
wer = write_error_stats(
f, f"{test_set_name}-{key}", results, enable_log=True
)
test_set_wers[key] = wer
logging.info("Wrote detailed error stats to {}".format(errs_filename))
test_set_wers = sorted(test_set_wers.items(), key=lambda x: x[1])
errs_info = (
params.res_dir
/ f"wer-summary-{test_set_name}-{key}-{params.suffix}.txt"
)
with open(errs_info, "w") as f:
print("settings\tWER", file=f)
for key, val in test_set_wers:
print("{}\t{}".format(key, val), file=f)
s = "\nFor {}, WER of different settings are:\n".format(test_set_name)
note = "\tbest for {}".format(test_set_name)
for key, val in test_set_wers:
s += "{}\t{}{}\n".format(key, val, note)
note = ""
logging.info(s)
@torch.no_grad()
def main():
parser = get_parser()
AsrDataModule.add_arguments(parser)
args = parser.parse_args()
args.exp_dir = Path(args.exp_dir)
params = get_params()
params.update(vars(args))
assert params.decoding_method in (
"greedy_search",
"beam_search",
"fast_beam_search",
"fast_beam_search_nbest",
"fast_beam_search_nbest_LG",
"fast_beam_search_nbest_oracle",
"modified_beam_search",
)
params.res_dir = params.exp_dir / params.decoding_method
if params.iter > 0:
params.suffix = f"iter-{params.iter}-avg-{params.avg}"
else:
params.suffix = f"epoch-{params.epoch}-avg-{params.avg}"
if "fast_beam_search" in params.decoding_method:
params.suffix += f"-beam-{params.beam}"
params.suffix += f"-max-contexts-{params.max_contexts}"
params.suffix += f"-max-states-{params.max_states}"
if "nbest" in params.decoding_method:
params.suffix += f"-nbest-scale-{params.nbest_scale}"
params.suffix += f"-num-paths-{params.num_paths}"
if "LG" in params.decoding_method:
params.suffix += f"-ngram-lm-scale-{params.ngram_lm_scale}"
elif "beam_search" in params.decoding_method:
params.suffix += (
f"-{params.decoding_method}-beam-size-{params.beam_size}"
)
else:
params.suffix += f"-context-{params.context_size}"
params.suffix += f"-max-sym-per-frame-{params.max_sym_per_frame}"
if params.use_averaged_model:
params.suffix += "-use-averaged-model"
setup_logger(f"{params.res_dir}/log-decode-{params.suffix}")
logging.info("Decoding started")
device = torch.device("cpu")
if torch.cuda.is_available():
device = torch.device("cuda", 0)
logging.info(f"Device: {device}")
sp = spm.SentencePieceProcessor()
sp.load(params.bpe_model)
# <blk> and <unk> are defined in local/train_bpe_model.py
params.blank_id = sp.piece_to_id("<blk>")
params.unk_id = sp.piece_to_id("<unk>")
params.vocab_size = sp.get_piece_size()
logging.info(params)
logging.info("About to create model")
model = get_transducer_model(params, enable_giga=False)
if not params.use_averaged_model:
if params.iter > 0:
filenames = find_checkpoints(
params.exp_dir, iteration=-params.iter
)[: params.avg]
if len(filenames) == 0:
raise ValueError(
f"No checkpoints found for"
f" --iter {params.iter}, --avg {params.avg}"
)
elif len(filenames) < params.avg:
raise ValueError(
f"Not enough checkpoints ({len(filenames)}) found for"
f" --iter {params.iter}, --avg {params.avg}"
)
logging.info(f"averaging {filenames}")
model.to(device)
model.load_state_dict(
average_checkpoints(filenames, device=device), strict=False
)
elif params.avg == 1:
load_checkpoint(f"{params.exp_dir}/epoch-{params.epoch}.pt", model)
else:
start = params.epoch - params.avg + 1
filenames = []
for i in range(start, params.epoch + 1):
if i >= 1:
filenames.append(f"{params.exp_dir}/epoch-{i}.pt")
logging.info(f"averaging {filenames}")
model.to(device)
model.load_state_dict(
average_checkpoints(filenames, device=device),
strict=False,
)
else:
if params.iter > 0:
filenames = find_checkpoints(
params.exp_dir, iteration=-params.iter
)[: params.avg + 1]
if len(filenames) == 0:
raise ValueError(
f"No checkpoints found for"
f" --iter {params.iter}, --avg {params.avg}"
)
elif len(filenames) < params.avg + 1:
raise ValueError(
f"Not enough checkpoints ({len(filenames)}) found for"
f" --iter {params.iter}, --avg {params.avg}"
)
filename_start = filenames[-1]
filename_end = filenames[0]
logging.info(
"Calculating the averaged model over iteration checkpoints"
f" from {filename_start} (excluded) to {filename_end}"
)
model.to(device)
model.load_state_dict(
average_checkpoints_with_averaged_model(
filename_start=filename_start,
filename_end=filename_end,
device=device,
),
strict=False,
)
else:
assert params.avg > 0, params.avg
start = params.epoch - params.avg
assert start >= 1, start
filename_start = f"{params.exp_dir}/epoch-{start}.pt"
filename_end = f"{params.exp_dir}/epoch-{params.epoch}.pt"
logging.info(
f"Calculating the averaged model over epoch range from "
f"{start} (excluded) to {params.epoch}"
)
model.to(device)
model.load_state_dict(
average_checkpoints_with_averaged_model(
filename_start=filename_start,
filename_end=filename_end,
device=device,
),
strict=False,
)
model.to(device)
model.eval()
if "fast_beam_search" in params.decoding_method:
if params.decoding_method == "fast_beam_search_nbest_LG":
lexicon = Lexicon(params.lang_dir)
word_table = lexicon.word_table
lg_filename = params.lang_dir / "LG.pt"
logging.info(f"Loading {lg_filename}")
decoding_graph = k2.Fsa.from_dict(
torch.load(lg_filename, map_location=device)
)
decoding_graph.scores *= params.ngram_lm_scale
else:
word_table = None
decoding_graph = k2.trivial_graph(
params.vocab_size - 1, device=device
)
else:
decoding_graph = None
word_table = None
num_param = sum([p.numel() for p in model.parameters()])
logging.info(f"Number of model parameters: {num_param}")
# we need cut ids to display recognition results.
args.return_cuts = True
asr_datamodule = AsrDataModule(args)
librispeech = LibriSpeech(manifest_dir=args.manifest_dir)
test_clean_cuts = librispeech.test_clean_cuts()
test_other_cuts = librispeech.test_other_cuts()
test_clean_dl = asr_datamodule.test_dataloaders(test_clean_cuts)
test_other_dl = asr_datamodule.test_dataloaders(test_other_cuts)
test_sets = ["test-clean", "test-other"]
test_dl = [test_clean_dl, test_other_dl]
for test_set, test_dl in zip(test_sets, test_dl):
results_dict = decode_dataset(
dl=test_dl,
params=params,
model=model,
sp=sp,
word_table=word_table,
decoding_graph=decoding_graph,
)
save_results(
params=params,
test_set_name=test_set,
results_dict=results_dict,
)
logging.info("Done!")
if __name__ == "__main__":
main()

1
egs/librispeech/ASR/lstm_transducer_stateless2/decoder.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless2/decoder.py

1
egs/librispeech/ASR/lstm_transducer_stateless2/encoder_interface.py Symbolic link
View File

@ -0,0 +1 @@
../transducer_stateless/encoder_interface.py

400
egs/librispeech/ASR/lstm_transducer_stateless2/export.py Executable file
View File

@ -0,0 +1,400 @@
#!/usr/bin/env python3
# flake8: noqa
#
# Copyright 2021-2022 Xiaomi Corporation (Author: Fangjun Kuang, 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.
# This script converts several saved checkpoints
# to a single one using model averaging.
"""
Usage:
(1) Export to torchscript model using torch.jit.trace()
./lstm_transducer_stateless2/export.py \
--exp-dir ./lstm_transducer_stateless2/exp \
--bpe-model data/lang_bpe_500/bpe.model \
--epoch 35 \
--avg 10 \
--jit-trace 1
It will generate 3 files: `encoder_jit_trace.pt`,
`decoder_jit_trace.pt`, and `joiner_jit_trace.pt`.
(2) Export `model.state_dict()`
./lstm_transducer_stateless2/export.py \
--exp-dir ./lstm_transducer_stateless2/exp \
--bpe-model data/lang_bpe_500/bpe.model \
--epoch 35 \
--avg 10
It will generate a file `pretrained.pt` in the given `exp_dir`. You can later
load it by `icefall.checkpoint.load_checkpoint()`.
To use the generated file with `lstm_transducer_stateless2/decode.py`,
you can do:
cd /path/to/exp_dir
ln -s pretrained.pt epoch-9999.pt
cd /path/to/egs/librispeech/ASR
./lstm_transducer_stateless2/decode.py \
--exp-dir ./lstm_transducer_stateless2/exp \
--epoch 9999 \
--avg 1 \
--max-duration 600 \
--decoding-method greedy_search \
--bpe-model data/lang_bpe_500/bpe.model
Check ./pretrained.py for its usage.
Note: If you don't want to train a model from scratch, we have
provided one for you. You can get it at
https://huggingface.co/csukuangfj/icefall-asr-librispeech-lstm-transducer-stateless2-2022-09-03
with the following commands:
sudo apt-get install git-lfs
git lfs install
git clone https://huggingface.co/csukuangfj/icefall-asr-librispeech-lstm-transducer-stateless2-2022-09-03
# You will find the pre-trained models in icefall-asr-librispeech-lstm-transducer-stateless2-2022-09-03/exp
"""
import argparse
import logging
from pathlib import Path
import sentencepiece as spm
import torch
import torch.nn as nn
from scaling_converter import convert_scaled_to_non_scaled
from train import add_model_arguments, get_params, get_transducer_model
from icefall.checkpoint import (
average_checkpoints,
average_checkpoints_with_averaged_model,
find_checkpoints,
load_checkpoint,
)
from icefall.utils import str2bool
def get_parser():
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter
)
parser.add_argument(
"--epoch",
type=int,
default=28,
help="""It specifies the checkpoint to use for averaging.
Note: Epoch counts from 0.
You can specify --avg to use more checkpoints for model averaging.""",
)
parser.add_argument(
"--iter",
type=int,
default=0,
help="""If positive, --epoch is ignored and it
will use the checkpoint exp_dir/checkpoint-iter.pt.
You can specify --avg to use more checkpoints for model averaging.
""",
)
parser.add_argument(
"--avg",
type=int,
default=15,
help="Number of checkpoints to average. Automatically select "
"consecutive checkpoints before the checkpoint specified by "
"'--epoch' and '--iter'",
)
parser.add_argument(
"--use-averaged-model",
type=str2bool,
default=True,
help="Whether to load averaged model. Currently it only supports "
"using --epoch. If True, it would decode with the averaged model "
"over the epoch range from `epoch-avg` (excluded) to `epoch`."
"Actually only the models with epoch number of `epoch-avg` and "
"`epoch` are loaded for averaging. ",
)
parser.add_argument(
"--exp-dir",
type=str,
default="pruned_transducer_stateless3/exp",
help="""It specifies the directory where all training related
files, e.g., checkpoints, log, etc, are saved
""",
)
parser.add_argument(
"--bpe-model",
type=str,
default="data/lang_bpe_500/bpe.model",
help="Path to the BPE model",
)
parser.add_argument(
"--jit-trace",
type=str2bool,
default=False,
help="""True to save a model after applying torch.jit.trace.
It will generate 3 files:
- encoder_jit_trace.pt
- decoder_jit_trace.pt
- joiner_jit_trace.pt
Check ./jit_pretrained.py for how to use them.
""",
)
parser.add_argument(
"--context-size",
type=int,
default=2,
help="The context size in the decoder. 1 means bigram; "
"2 means tri-gram",
)
add_model_arguments(parser)
return parser
def export_encoder_model_jit_trace(
encoder_model: nn.Module,
encoder_filename: str,
) -> None:
"""Export the given encoder model with torch.jit.trace()
Note: The warmup argument is fixed to 1.
Args:
encoder_model:
The input encoder model
encoder_filename:
The filename to save the exported model.
"""
x = torch.zeros(1, 100, 80, dtype=torch.float32)
x_lens = torch.tensor([100], dtype=torch.int64)
states = encoder_model.get_init_states()
traced_model = torch.jit.trace(encoder_model, (x, x_lens, states))
traced_model.save(encoder_filename)
logging.info(f"Saved to {encoder_filename}")
def export_decoder_model_jit_trace(
decoder_model: nn.Module,
decoder_filename: str,
) -> None:
"""Export the given decoder model with torch.jit.trace()
Note: The argument need_pad is fixed to False.
Args:
decoder_model:
The input decoder model
decoder_filename:
The filename to save the exported model.
"""
y = torch.zeros(10, decoder_model.context_size, dtype=torch.int64)
need_pad = torch.tensor([False])
traced_model = torch.jit.trace(decoder_model, (y, need_pad))
traced_model.save(decoder_filename)
logging.info(f"Saved to {decoder_filename}")
def export_joiner_model_jit_trace(
joiner_model: nn.Module,
joiner_filename: str,
) -> None:
"""Export the given joiner model with torch.jit.trace()
Note: The argument project_input is fixed to True. A user should not
project the encoder_out/decoder_out by himself/herself. The exported joiner
will do that for the user.
Args:
joiner_model:
The input joiner model
joiner_filename:
The filename to save the exported model.
"""
encoder_out_dim = joiner_model.encoder_proj.weight.shape[1]
decoder_out_dim = joiner_model.decoder_proj.weight.shape[1]
encoder_out = torch.rand(1, encoder_out_dim, dtype=torch.float32)
decoder_out = torch.rand(1, decoder_out_dim, dtype=torch.float32)
traced_model = torch.jit.trace(joiner_model, (encoder_out, decoder_out))
traced_model.save(joiner_filename)
logging.info(f"Saved to {joiner_filename}")
@torch.no_grad()
def main():
args = get_parser().parse_args()
args.exp_dir = Path(args.exp_dir)
params = get_params()
params.update(vars(args))
device = torch.device("cpu")
if torch.cuda.is_available():
device = torch.device("cuda", 0)
logging.info(f"device: {device}")
sp = spm.SentencePieceProcessor()
sp.load(params.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()
logging.info(params)
logging.info("About to create model")
model = get_transducer_model(params, enable_giga=False)
num_param = sum([p.numel() for p in model.parameters()])
logging.info(f"Number of model parameters: {num_param}")
if not params.use_averaged_model:
if params.iter > 0:
filenames = find_checkpoints(
params.exp_dir, iteration=-params.iter
)[: params.avg]
if len(filenames) == 0:
raise ValueError(
f"No checkpoints found for"
f" --iter {params.iter}, --avg {params.avg}"
)
elif len(filenames) < params.avg:
raise ValueError(
f"Not enough checkpoints ({len(filenames)}) found for"
f" --iter {params.iter}, --avg {params.avg}"
)
logging.info(f"averaging {filenames}")
model.to(device)
model.load_state_dict(
average_checkpoints(filenames, device=device),
strict=False,
)
elif params.avg == 1:
load_checkpoint(f"{params.exp_dir}/epoch-{params.epoch}.pt", model)
else:
start = params.epoch - params.avg + 1
filenames = []
for i in range(start, params.epoch + 1):
if i >= 1:
filenames.append(f"{params.exp_dir}/epoch-{i}.pt")
logging.info(f"averaging {filenames}")
model.to(device)
model.load_state_dict(
average_checkpoints(filenames, device=device),
strict=False,
)
else:
if params.iter > 0:
filenames = find_checkpoints(
params.exp_dir, iteration=-params.iter
)[: params.avg + 1]
if len(filenames) == 0:
raise ValueError(
f"No checkpoints found for"
f" --iter {params.iter}, --avg {params.avg}"
)
elif len(filenames) < params.avg + 1:
raise ValueError(
f"Not enough checkpoints ({len(filenames)}) found for"
f" --iter {params.iter}, --avg {params.avg}"
)
filename_start = filenames[-1]
filename_end = filenames[0]
logging.info(
"Calculating the averaged model over iteration checkpoints"
f" from {filename_start} (excluded) to {filename_end}"
)
model.to(device)
model.load_state_dict(
average_checkpoints_with_averaged_model(
filename_start=filename_start,
filename_end=filename_end,
device=device,
),
strict=False,
)
else:
assert params.avg > 0, params.avg
start = params.epoch - params.avg
assert start >= 1, start
filename_start = f"{params.exp_dir}/epoch-{start}.pt"
filename_end = f"{params.exp_dir}/epoch-{params.epoch}.pt"
logging.info(
f"Calculating the averaged model over epoch range from "
f"{start} (excluded) to {params.epoch}"
)
model.to(device)
model.load_state_dict(
average_checkpoints_with_averaged_model(
filename_start=filename_start,
filename_end=filename_end,
device=device,
),
strict=False,
)
model.to("cpu")
model.eval()
if params.jit_trace is True:
convert_scaled_to_non_scaled(model, inplace=True)
logging.info("Using torch.jit.trace()")
encoder_filename = params.exp_dir / "encoder_jit_trace.pt"
export_encoder_model_jit_trace(model.encoder, encoder_filename)
decoder_filename = params.exp_dir / "decoder_jit_trace.pt"
export_decoder_model_jit_trace(model.decoder, decoder_filename)
joiner_filename = params.exp_dir / "joiner_jit_trace.pt"
export_joiner_model_jit_trace(model.joiner, joiner_filename)
else:
logging.info("Not using torchscript")
# Save it using a format so that it can be loaded
# by :func:`load_checkpoint`
filename = params.exp_dir / "pretrained.pt"
torch.save({"model": model.state_dict()}, str(filename))
logging.info(f"Saved to {filename}")
if __name__ == "__main__":
formatter = (
"%(asctime)s %(levelname)s [%(filename)s:%(lineno)d] %(message)s"
)
logging.basicConfig(format=formatter, level=logging.INFO)
main()

1
egs/librispeech/ASR/lstm_transducer_stateless2/gigaspeech.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless3/gigaspeech.py

323
egs/librispeech/ASR/lstm_transducer_stateless2/jit_pretrained.py Executable file
View File

@ -0,0 +1,323 @@
#!/usr/bin/env python3
# flake8: noqa
# Copyright 2022 Xiaomi Corp. (authors: Fangjun Kuang, 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.
"""
This script loads torchscript models, either exported by `torch.jit.trace()`
or by `torch.jit.script()`, and uses them to decode waves.
You can use the following command to get the exported models:
./lstm_transducer_stateless2/export.py \
--exp-dir ./lstm_transducer_stateless2/exp \
--bpe-model data/lang_bpe_500/bpe.model \
--epoch 20 \
--avg 10 \
--jit-trace 1
Usage of this script:
./lstm_transducer_stateless2/jit_pretrained.py \
--encoder-model-filename ./lstm_transducer_stateless2/exp/encoder_jit_trace.pt \
--decoder-model-filename ./lstm_transducer_stateless2/exp/decoder_jit_trace.pt \
--joiner-model-filename ./lstm_transducer_stateless2/exp/joiner_jit_trace.pt \
--bpe-model ./data/lang_bpe_500/bpe.model \
/path/to/foo.wav \
/path/to/bar.wav
"""
import argparse
import logging
import math
from typing import List
import kaldifeat
import sentencepiece as spm
import torch
import torchaudio
from torch.nn.utils.rnn import pad_sequence
def get_parser():
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter
)
parser.add_argument(
"--encoder-model-filename",
type=str,
required=True,
help="Path to the encoder torchscript model. ",
)
parser.add_argument(
"--decoder-model-filename",
type=str,
required=True,
help="Path to the decoder torchscript model. ",
)
parser.add_argument(
"--joiner-model-filename",
type=str,
required=True,
help="Path to the joiner torchscript model. ",
)
parser.add_argument(
"--bpe-model",
type=str,
help="""Path to bpe.model.""",
)
parser.add_argument(
"sound_files",
type=str,
nargs="+",
help="The input sound file(s) to transcribe. "
"Supported formats are those supported by torchaudio.load(). "
"For example, wav and flac are supported. "
"The sample rate has to be 16kHz.",
)
parser.add_argument(
"--sample-rate",
type=int,
default=16000,
help="The sample rate of the input sound file",
)
parser.add_argument(
"--context-size",
type=int,
default=2,
help="Context size of the decoder model",
)
return parser
def read_sound_files(
filenames: List[str], expected_sample_rate: float
) -> List[torch.Tensor]:
"""Read a list of sound files into a list 1-D float32 torch tensors.
Args:
filenames:
A list of sound filenames.
expected_sample_rate:
The expected sample rate of the sound files.
Returns:
Return a list of 1-D float32 torch tensors.
"""
ans = []
for f in filenames:
wave, sample_rate = torchaudio.load(f)
assert sample_rate == expected_sample_rate, (
f"expected sample rate: {expected_sample_rate}. "
f"Given: {sample_rate}"
)
# We use only the first channel
ans.append(wave[0])
return ans
def greedy_search(
decoder: torch.jit.ScriptModule,
joiner: torch.jit.ScriptModule,
encoder_out: torch.Tensor,
encoder_out_lens: torch.Tensor,
context_size: int,
) -> List[List[int]]:
"""Greedy search in batch mode. It hardcodes --max-sym-per-frame=1.
Args:
decoder:
The decoder model.
joiner:
The joiner model.
encoder_out:
A 3-D tensor of shape (N, T, C)
encoder_out_lens:
A 1-D tensor of shape (N,).
context_size:
The context size of the decoder model.
Returns:
Return the decoded results for each utterance.
"""
assert encoder_out.ndim == 3
assert encoder_out.size(0) >= 1, encoder_out.size(0)
packed_encoder_out = torch.nn.utils.rnn.pack_padded_sequence(
input=encoder_out,
lengths=encoder_out_lens.cpu(),
batch_first=True,
enforce_sorted=False,
)
device = encoder_out.device
blank_id = 0 # hard-code to 0
batch_size_list = packed_encoder_out.batch_sizes.tolist()
N = encoder_out.size(0)
assert torch.all(encoder_out_lens > 0), encoder_out_lens
assert N == batch_size_list[0], (N, batch_size_list)
hyps = [[blank_id] * context_size for _ in range(N)]
decoder_input = torch.tensor(
hyps,
device=device,
dtype=torch.int64,
) # (N, context_size)
decoder_out = decoder(
decoder_input,
need_pad=torch.tensor([False]),
).squeeze(1)
offset = 0
for batch_size in batch_size_list:
start = offset
end = offset + batch_size
current_encoder_out = packed_encoder_out.data[start:end]
current_encoder_out = current_encoder_out
# current_encoder_out's shape: (batch_size, encoder_out_dim)
offset = end
decoder_out = decoder_out[:batch_size]
logits = joiner(
current_encoder_out,
decoder_out,
)
# logits'shape (batch_size, vocab_size)
assert logits.ndim == 2, logits.shape
y = logits.argmax(dim=1).tolist()
emitted = False
for i, v in enumerate(y):
if v != blank_id:
hyps[i].append(v)
emitted = True
if emitted:
# update decoder output
decoder_input = [h[-context_size:] for h in hyps[:batch_size]]
decoder_input = torch.tensor(
decoder_input,
device=device,
dtype=torch.int64,
)
decoder_out = decoder(
decoder_input,
need_pad=torch.tensor([False]),
)
decoder_out = decoder_out.squeeze(1)
sorted_ans = [h[context_size:] for h in hyps]
ans = []
unsorted_indices = packed_encoder_out.unsorted_indices.tolist()
for i in range(N):
ans.append(sorted_ans[unsorted_indices[i]])
return ans
@torch.no_grad()
def main():
parser = get_parser()
args = parser.parse_args()
logging.info(vars(args))
device = torch.device("cpu")
if torch.cuda.is_available():
device = torch.device("cuda", 0)
logging.info(f"device: {device}")
encoder = torch.jit.load(args.encoder_model_filename)
decoder = torch.jit.load(args.decoder_model_filename)
joiner = torch.jit.load(args.joiner_model_filename)
encoder.eval()
decoder.eval()
joiner.eval()
encoder.to(device)
decoder.to(device)
joiner.to(device)
sp = spm.SentencePieceProcessor()
sp.load(args.bpe_model)
logging.info("Constructing Fbank computer")
opts = kaldifeat.FbankOptions()
opts.device = device
opts.frame_opts.dither = 0
opts.frame_opts.snip_edges = False
opts.frame_opts.samp_freq = args.sample_rate
opts.mel_opts.num_bins = 80
fbank = kaldifeat.Fbank(opts)
logging.info(f"Reading sound files: {args.sound_files}")
waves = read_sound_files(
filenames=args.sound_files,
expected_sample_rate=args.sample_rate,
)
waves = [w.to(device) for w in waves]
logging.info("Decoding started")
features = fbank(waves)
feature_lengths = [f.size(0) for f in features]
features = pad_sequence(
features,
batch_first=True,
padding_value=math.log(1e-10),
)
feature_lengths = torch.tensor(feature_lengths, device=device)
states = encoder.get_init_states(batch_size=features.size(0), device=device)
encoder_out, encoder_out_lens, _ = encoder(
x=features,
x_lens=feature_lengths,
states=states,
)
hyps = greedy_search(
decoder=decoder,
joiner=joiner,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
context_size=args.context_size,
)
s = "\n"
for filename, hyp in zip(args.sound_files, hyps):
words = sp.decode(hyp)
s += f"{filename}:\n{words}\n\n"
logging.info(s)
logging.info("Decoding Done")
if __name__ == "__main__":
formatter = (
"%(asctime)s %(levelname)s [%(filename)s:%(lineno)d] %(message)s"
)
logging.basicConfig(format=formatter, level=logging.INFO)
main()

1
egs/librispeech/ASR/lstm_transducer_stateless2/joiner.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless2/joiner.py

1
egs/librispeech/ASR/lstm_transducer_stateless2/librispeech.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless3/librispeech.py

1
egs/librispeech/ASR/lstm_transducer_stateless2/lstm.py Symbolic link
View File

@ -0,0 +1 @@
../lstm_transducer_stateless/lstm.py

241
egs/librispeech/ASR/lstm_transducer_stateless2/model.py Normal file
View File

@ -0,0 +1,241 @@
# Copyright 2021 Xiaomi Corp. (authors: Fangjun Kuang, Wei Kang)
#
# 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.
from typing import Optional, Tuple
import k2
import torch
import torch.nn as nn
from encoder_interface import EncoderInterface
from scaling import ScaledLinear
from icefall.utils import add_sos
class Transducer(nn.Module):
"""It implements https://arxiv.org/pdf/1211.3711.pdf
"Sequence Transduction with Recurrent Neural Networks"
"""
def __init__(
self,
encoder: EncoderInterface,
decoder: nn.Module,
joiner: nn.Module,
encoder_dim: int,
decoder_dim: int,
joiner_dim: int,
vocab_size: int,
decoder_giga: Optional[nn.Module] = None,
joiner_giga: Optional[nn.Module] = None,
):
"""
Args:
encoder:
It is the transcription network in the paper. Its accepts
two inputs: `x` of (N, T, encoder_dim) and `x_lens` of shape (N,).
It returns two tensors: `logits` of shape (N, T, encoder_dm) and
`logit_lens` of shape (N,).
decoder:
It is the prediction network in the paper. Its input shape
is (N, U) and its output shape is (N, U, decoder_dim).
It should contain one attribute: `blank_id`.
joiner:
It has two inputs with shapes: (N, T, encoder_dim) and
(N, U, decoder_dim).
Its output shape is (N, T, U, vocab_size). Note that its output
contains unnormalized probs, i.e., not processed by log-softmax.
encoder_dim:
Output dimension of the encoder network.
decoder_dim:
Output dimension of the decoder network.
joiner_dim:
Input dimension of the joiner network.
vocab_size:
Output dimension of the joiner network.
decoder_giga:
Optional. The decoder network for the GigaSpeech dataset.
joiner_giga:
Optional. The joiner network for the GigaSpeech dataset.
"""
super().__init__()
assert isinstance(encoder, EncoderInterface), type(encoder)
assert hasattr(decoder, "blank_id")
self.encoder = encoder
self.decoder = decoder
self.joiner = joiner
self.decoder_giga = decoder_giga
self.joiner_giga = joiner_giga
self.simple_am_proj = ScaledLinear(
encoder_dim, vocab_size, initial_speed=0.5
)
self.simple_lm_proj = ScaledLinear(decoder_dim, vocab_size)
if decoder_giga is not None:
self.simple_am_proj_giga = ScaledLinear(
encoder_dim, vocab_size, initial_speed=0.5
)
self.simple_lm_proj_giga = ScaledLinear(decoder_dim, vocab_size)
def forward(
self,
x: torch.Tensor,
x_lens: torch.Tensor,
y: k2.RaggedTensor,
libri: bool = True,
prune_range: int = 5,
am_scale: float = 0.0,
lm_scale: float = 0.0,
warmup: float = 1.0,
reduction: str = "sum",
) -> Tuple[torch.Tensor, torch.Tensor]:
"""
Args:
x:
A 3-D tensor of shape (N, T, C).
x_lens:
A 1-D tensor of shape (N,). It contains the number of frames in `x`
before padding.
y:
A ragged tensor with 2 axes [utt][label]. It contains labels of each
utterance.
libri:
True to use the decoder and joiner for the LibriSpeech dataset.
False to use the decoder and joiner for the GigaSpeech dataset.
prune_range:
The prune range for rnnt loss, it means how many symbols(context)
we are considering for each frame to compute the loss.
am_scale:
The scale to smooth the loss with am (output of encoder network)
part
lm_scale:
The scale to smooth the loss with lm (output of predictor network)
part
warmup:
A value warmup >= 0 that determines which modules are active, values
warmup > 1 "are fully warmed up" and all modules will be active.
reduction:
"sum" to sum the losses over all utterances in the batch.
"none" to return the loss in a 1-D tensor for each utterance
in the batch.
Returns:
Return the transducer loss.
Note:
Regarding am_scale & lm_scale, it will make the loss-function one of
the form:
lm_scale * lm_probs + am_scale * am_probs +
(1-lm_scale-am_scale) * combined_probs
"""
assert reduction in ("sum", "none"), reduction
assert x.ndim == 3, x.shape
assert x_lens.ndim == 1, x_lens.shape
assert y.num_axes == 2, y.num_axes
assert x.size(0) == x_lens.size(0) == y.dim0
encoder_out, x_lens, _ = self.encoder(x, x_lens, warmup=warmup)
assert torch.all(x_lens > 0)
if libri:
decoder = self.decoder
simple_lm_proj = self.simple_lm_proj
simple_am_proj = self.simple_am_proj
joiner = self.joiner
else:
decoder = self.decoder_giga
simple_lm_proj = self.simple_lm_proj_giga
simple_am_proj = self.simple_am_proj_giga
joiner = self.joiner_giga
# Now for the decoder, i.e., the prediction network
row_splits = y.shape.row_splits(1)
y_lens = row_splits[1:] - row_splits[:-1]
blank_id = decoder.blank_id
sos_y = add_sos(y, sos_id=blank_id)
# sos_y_padded: [B, S + 1], start with SOS.
sos_y_padded = sos_y.pad(mode="constant", padding_value=blank_id)
# decoder_out: [B, S + 1, decoder_dim]
decoder_out = decoder(sos_y_padded)
# Note: y does not start with SOS
# y_padded : [B, S]
y_padded = y.pad(mode="constant", padding_value=0)
y_padded = y_padded.to(torch.int64)
boundary = torch.zeros(
(x.size(0), 4), dtype=torch.int64, device=x.device
)
boundary[:, 2] = y_lens
boundary[:, 3] = x_lens
lm = simple_lm_proj(decoder_out)
am = simple_am_proj(encoder_out)
with torch.cuda.amp.autocast(enabled=False):
simple_loss, (px_grad, py_grad) = k2.rnnt_loss_smoothed(
lm=lm.float(),
am=am.float(),
symbols=y_padded,
termination_symbol=blank_id,
lm_only_scale=lm_scale,
am_only_scale=am_scale,
boundary=boundary,
reduction=reduction,
return_grad=True,
)
# ranges : [B, T, prune_range]
ranges = k2.get_rnnt_prune_ranges(
px_grad=px_grad,
py_grad=py_grad,
boundary=boundary,
s_range=prune_range,
)
# am_pruned : [B, T, prune_range, encoder_dim]
# lm_pruned : [B, T, prune_range, decoder_dim]
am_pruned, lm_pruned = k2.do_rnnt_pruning(
am=joiner.encoder_proj(encoder_out),
lm=joiner.decoder_proj(decoder_out),
ranges=ranges,
)
# logits : [B, T, prune_range, vocab_size]
# project_input=False since we applied the decoder's input projections
# prior to do_rnnt_pruning (this is an optimization for speed).
logits = joiner(am_pruned, lm_pruned, project_input=False)
with torch.cuda.amp.autocast(enabled=False):
pruned_loss = k2.rnnt_loss_pruned(
logits=logits.float(),
symbols=y_padded,
ranges=ranges,
termination_symbol=blank_id,
boundary=boundary,
reduction=reduction,
)
return (simple_loss, pruned_loss)

1
egs/librispeech/ASR/lstm_transducer_stateless2/optim.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless2/optim.py

355
egs/librispeech/ASR/lstm_transducer_stateless2/pretrained.py Executable file
View File

@ -0,0 +1,355 @@
#!/usr/bin/env python3
# Copyright 2021 Xiaomi Corp. (authors: Fangjun Kuang, 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:
(1) greedy search
./lstm_transducer_stateless2/pretrained.py \
--checkpoint ./lstm_transducer_stateless2/exp/pretrained.pt \
--bpe-model ./data/lang_bpe_500/bpe.model \
--method greedy_search \
/path/to/foo.wav \
/path/to/bar.wav
(2) beam search
./lstm_transducer_stateless2/pretrained.py \
--checkpoint ./lstm_transducer_stateless2/exp/pretrained.pt \
--bpe-model ./data/lang_bpe_500/bpe.model \
--method beam_search \
--beam-size 4 \
/path/to/foo.wav \
/path/to/bar.wav
(3) modified beam search
./lstm_transducer_stateless2/pretrained.py \
--checkpoint ./lstm_transducer_stateless2/exp/pretrained.pt \
--bpe-model ./data/lang_bpe_500/bpe.model \
--method modified_beam_search \
--beam-size 4 \
/path/to/foo.wav \
/path/to/bar.wav
(4) fast beam search
./lstm_transducer_stateless2/pretrained.py \
--checkpoint ./lstm_transducer_stateless2/exp/pretrained.pt \
--bpe-model ./data/lang_bpe_500/bpe.model \
--method fast_beam_search \
--beam-size 4 \
/path/to/foo.wav \
/path/to/bar.wav
You can also use `./lstm_transducer_stateless2/exp/epoch-xx.pt`.
Note: ./lstm_transducer_stateless2/exp/pretrained.pt is generated by
./lstm_transducer_stateless2/export.py
You can find pretrained models by visiting
https://huggingface.co/csukuangfj/icefall-asr-librispeech-lstm-transducer-stateless2-2022-09-03/tree/main/exp
"""
import argparse
import logging
import math
from typing import List
import k2
import kaldifeat
import sentencepiece as spm
import torch
import torchaudio
from beam_search import (
beam_search,
fast_beam_search_one_best,
greedy_search,
greedy_search_batch,
modified_beam_search,
)
from torch.nn.utils.rnn import pad_sequence
from train import add_model_arguments, get_params, get_transducer_model
def get_parser():
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter
)
parser.add_argument(
"--checkpoint",
type=str,
required=True,
help="Path to the checkpoint. "
"The checkpoint is assumed to be saved by "
"icefall.checkpoint.save_checkpoint().",
)
parser.add_argument(
"--bpe-model",
type=str,
help="""Path to bpe.model.""",
)
parser.add_argument(
"--method",
type=str,
default="greedy_search",
help="""Possible values are:
- greedy_search
- beam_search
- modified_beam_search
- fast_beam_search
""",
)
parser.add_argument(
"sound_files",
type=str,
nargs="+",
help="The input sound file(s) to transcribe. "
"Supported formats are those supported by torchaudio.load(). "
"For example, wav and flac are supported. "
"The sample rate has to be 16kHz.",
)
parser.add_argument(
"--sample-rate",
type=int,
default=16000,
help="The sample rate of the input sound file",
)
parser.add_argument(
"--beam-size",
type=int,
default=4,
help="""An integer indicating how many candidates we will keep for each
frame. Used only when --method is beam_search or
modified_beam_search.""",
)
parser.add_argument(
"--beam",
type=float,
default=4,
help="""A floating point value to calculate the cutoff score during beam
search (i.e., `cutoff = max-score - beam`), which is the same as the
`beam` in Kaldi.
Used only when --method is fast_beam_search""",
)
parser.add_argument(
"--max-contexts",
type=int,
default=4,
help="""Used only when --method is fast_beam_search""",
)
parser.add_argument(
"--max-states",
type=int,
default=8,
help="""Used only when --method is fast_beam_search""",
)
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(
"--max-sym-per-frame",
type=int,
default=1,
help="""Maximum number of symbols per frame. Used only when
--method is greedy_search.
""",
)
add_model_arguments(parser)
return parser
def read_sound_files(
filenames: List[str], expected_sample_rate: float
) -> List[torch.Tensor]:
"""Read a list of sound files into a list 1-D float32 torch tensors.
Args:
filenames:
A list of sound filenames.
expected_sample_rate:
The expected sample rate of the sound files.
Returns:
Return a list of 1-D float32 torch tensors.
"""
ans = []
for f in filenames:
wave, sample_rate = torchaudio.load(f)
assert sample_rate == expected_sample_rate, (
f"expected sample rate: {expected_sample_rate}. "
f"Given: {sample_rate}"
)
# We use only the first channel
ans.append(wave[0])
return ans
@torch.no_grad()
def main():
parser = get_parser()
args = parser.parse_args()
params = get_params()
params.update(vars(args))
sp = spm.SentencePieceProcessor()
sp.load(params.bpe_model)
# <blk> is defined in local/train_bpe_model.py
params.blank_id = sp.piece_to_id("<blk>")
params.unk_id = sp.piece_to_id("<unk>")
params.vocab_size = sp.get_piece_size()
logging.info(f"{params}")
device = torch.device("cpu")
if torch.cuda.is_available():
device = torch.device("cuda", 0)
logging.info(f"device: {device}")
logging.info("Creating model")
model = get_transducer_model(params, enable_giga=False)
num_param = sum([p.numel() for p in model.parameters()])
logging.info(f"Number of model parameters: {num_param}")
checkpoint = torch.load(args.checkpoint, map_location="cpu")
model.load_state_dict(checkpoint["model"], strict=False)
model.to(device)
model.eval()
model.device = device
logging.info("Constructing Fbank computer")
opts = kaldifeat.FbankOptions()
opts.device = device
opts.frame_opts.dither = 0
opts.frame_opts.snip_edges = False
opts.frame_opts.samp_freq = params.sample_rate
opts.mel_opts.num_bins = params.feature_dim
fbank = kaldifeat.Fbank(opts)
logging.info(f"Reading sound files: {params.sound_files}")
waves = read_sound_files(
filenames=params.sound_files, expected_sample_rate=params.sample_rate
)
waves = [w.to(device) for w in waves]
logging.info("Decoding started")
features = fbank(waves)
feature_lengths = [f.size(0) for f in features]
features = pad_sequence(
features, batch_first=True, padding_value=math.log(1e-10)
)
feature_lengths = torch.tensor(feature_lengths, device=device)
encoder_out, encoder_out_lens, _ = model.encoder(
x=features, x_lens=feature_lengths
)
num_waves = encoder_out.size(0)
hyps = []
msg = f"Using {params.method}"
if params.method == "beam_search":
msg += f" with beam size {params.beam_size}"
logging.info(msg)
if params.method == "fast_beam_search":
decoding_graph = k2.trivial_graph(params.vocab_size - 1, device=device)
hyp_tokens = fast_beam_search_one_best(
model=model,
decoding_graph=decoding_graph,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
beam=params.beam,
max_contexts=params.max_contexts,
max_states=params.max_states,
)
for hyp in sp.decode(hyp_tokens):
hyps.append(hyp.split())
elif params.method == "modified_beam_search":
hyp_tokens = modified_beam_search(
model=model,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
beam=params.beam_size,
)
for hyp in sp.decode(hyp_tokens):
hyps.append(hyp.split())
elif params.method == "greedy_search" and params.max_sym_per_frame == 1:
hyp_tokens = greedy_search_batch(
model=model,
encoder_out=encoder_out,
encoder_out_lens=encoder_out_lens,
)
for hyp in sp.decode(hyp_tokens):
hyps.append(hyp.split())
else:
for i in range(num_waves):
# fmt: off
encoder_out_i = encoder_out[i:i+1, :encoder_out_lens[i]]
# fmt: on
if params.method == "greedy_search":
hyp = greedy_search(
model=model,
encoder_out=encoder_out_i,
max_sym_per_frame=params.max_sym_per_frame,
)
elif params.method == "beam_search":
hyp = beam_search(
model=model,
encoder_out=encoder_out_i,
beam=params.beam_size,
)
else:
raise ValueError(f"Unsupported method: {params.method}")
hyps.append(sp.decode(hyp).split())
s = "\n"
for filename, hyp in zip(params.sound_files, hyps):
words = " ".join(hyp)
s += f"{filename}:\n{words}\n\n"
logging.info(s)
logging.info("Decoding Done")
if __name__ == "__main__":
formatter = (
"%(asctime)s %(levelname)s [%(filename)s:%(lineno)d] %(message)s"
)
logging.basicConfig(format=formatter, level=logging.INFO)
main()

1
egs/librispeech/ASR/lstm_transducer_stateless2/scaling.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless2/scaling.py

1
egs/librispeech/ASR/lstm_transducer_stateless2/scaling_converter.py Symbolic link
View File

@ -0,0 +1 @@
../pruned_transducer_stateless3/scaling_converter.py

1273
egs/librispeech/ASR/lstm_transducer_stateless2/train.py Executable file
View File

File diff suppressed because it is too large Load Diff

4
egs/librispeech/ASR/pruned_transducer_stateless3/scaling_converter.py
View File

@ -60,7 +60,7 @@ def scaled_linear_to_linear(scaled_linear: ScaledLinear) -> nn.Linear:
linear = torch.nn.Linear(
in_features=scaled_linear.in_features,
out_features=scaled_linear.out_features,
bias=True, # otherwise, it throws errors when converting to PNNX format.
bias=True, # otherwise, it throws errors when converting to PNNX format
# device=weight.device, # Pytorch version before v1.9.0 does not has
# this argument. Comment out for now, we will
# see if it will raise error for versions
@ -206,7 +206,7 @@ def scaled_lstm_to_lstm(scaled_lstm: ScaledLSTM) -> nn.LSTM:
return lstm
# Copied from https://pytorch.org/docs/1.9.0/_modules/torch/nn/modules/module.html#Module.get_submodule
# Copied from https://pytorch.org/docs/1.9.0/_modules/torch/nn/modules/module.html#Module.get_submodule # noqa
# get_submodule was added to nn.Module at v1.9.0
def get_submodule(model, target):
if target == "":

2
egs/librispeech/ASR/pruned_transducer_stateless3/train.py
View File

@ -39,7 +39,7 @@ cd egs/librispeech/ASR/
--world-size 4 \
--num-epochs 30 \
--start-epoch 0 \
--use_fp16 1 \
--use-fp16 1 \
--exp-dir pruned_transducer_stateless3/exp \
--full-libri 1 \
--max-duration 550