# Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
#
# 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 __future__ import annotations
import math
import os
from dataclasses import dataclass, field
from typing import List, Optional, Tuple, Union
import torch
from nemo.collections.asr.parts.context_biasing import BoostingTreeModelConfig, GPUBoostingTreeModel
from nemo.collections.asr.parts.submodules.ctc_batched_beam_decoding import BatchedBeamCTCComputer
from nemo.collections.asr.parts.submodules.ngram_lm import DEFAULT_TOKEN_OFFSET, NGramGPULanguageModel
from nemo.collections.asr.parts.submodules.wfst_decoder import RivaDecoderConfig, WfstNbestHypothesis
from nemo.collections.asr.parts.utils import rnnt_utils
from nemo.collections.common.parts.optional_cuda_graphs import WithOptionalCudaGraphs
from nemo.collections.common.tokenizers.tokenizer_spec import TokenizerSpec
from nemo.core.classes import Typing, typecheck
from nemo.core.neural_types import HypothesisType, LengthsType, LogprobsType, NeuralType
from nemo.utils import logging
def pack_hypotheses(
hypotheses: List[rnnt_utils.NBestHypotheses],
logitlen: torch.Tensor,
) -> List[rnnt_utils.NBestHypotheses]:
if logitlen is not None:
if hasattr(logitlen, 'cpu'):
logitlen_cpu = logitlen.to('cpu')
else:
logitlen_cpu = logitlen
for idx, hyp in enumerate(hypotheses): # type: rnnt_utils.NBestHypotheses
for candidate_idx, cand in enumerate(hyp.n_best_hypotheses):
cand.y_sequence = torch.tensor(cand.y_sequence, dtype=torch.long)
if logitlen is not None:
cand.length = logitlen_cpu[idx]
if cand.dec_state is not None:
cand.dec_state = _states_to_device(cand.dec_state)
return hypotheses
def pack_wfst_hypotheses(
hypotheses: List['WfstNbestHypothesis'],
logits: torch.Tensor,
logitlen: torch.Tensor,
) -> List[rnnt_utils.NBestHypotheses]:
logitlen_cpu = logitlen.to('cpu')
new_hypotheses = []
for idx, nbest_hyp in enumerate(hypotheses): # type: WfstNbestHypothesis
new_hyp = []
y_sequence = logits[idx, : logitlen[idx]].to('cpu')
length = logitlen_cpu[idx]
for candidate_idx, cand in enumerate(nbest_hyp):
cand_hyp = rnnt_utils.Hypothesis(
y_sequence=[],
score=cand.score,
text=" ".join(cand.words),
timestamp=list(cand.timesteps),
alignments=list(cand.alignment),
)
cand_hyp.y_sequence = y_sequence
if logitlen is not None:
cand_hyp.length = length
new_hyp.append(cand_hyp)
new_hypotheses.append(rnnt_utils.NBestHypotheses(new_hyp))
return new_hypotheses
def _states_to_device(dec_state, device='cpu'):
if torch.is_tensor(dec_state):
dec_state = dec_state.to(device)
elif isinstance(dec_state, (list, tuple)):
dec_state = tuple(_states_to_device(dec_i, device) for dec_i in dec_state)
return dec_state
class AbstractBeamCTCInfer(Typing):
"""A beam CTC decoder.
Provides a common abstraction for sample level beam decoding.
Args:
blank_id: int, index of the blank token. Can be 0 or len(vocabulary).
beam_size: int, size of the beam used in the underlying beam search engine.
"""
@property
def input_types(self):
"""Returns definitions of module input ports."""
return {
"decoder_output": NeuralType(('B', 'T', 'D'), LogprobsType()),
"decoder_lengths": NeuralType(tuple('B'), LengthsType()),
}
@property
def output_types(self):
"""Returns definitions of module output ports."""
return {"predictions": [NeuralType(elements_type=HypothesisType())]}
def __init__(self, blank_id: int, beam_size: int):
self.blank_id = blank_id
if beam_size < 1:
raise ValueError("Beam search size cannot be less than 1!")
self.beam_size = beam_size
# Variables set by corresponding setter methods
self.vocab = None
self.decoding_type = None
self.tokenizer = None
# Utility maps for vocabulary
self.vocab_index_map = None
self.index_vocab_map = None
# Internal variable, used to prevent double reduction of consecutive tokens (ctc collapse)
self.override_fold_consecutive_value = None
def set_vocabulary(self, vocab: List[str]):
"""
Set the vocabulary of the decoding framework.
Args:
vocab: List of str. Each token corresponds to its location in the vocabulary emitted by the model.
Note that this vocabulary must NOT contain the "BLANK" token.
"""
self.vocab = vocab
self.vocab_index_map = {v: i for i, v in enumerate(vocab)}
self.index_vocab_map = {i: v for i, v in enumerate(vocab)}
def set_decoding_type(self, decoding_type: str):
"""
Sets the decoding type of the framework. Can support either char or subword models.
Args:
decoding_type: Str corresponding to decoding type. Only supports "char" and "subword".
"""
decoding_type = decoding_type.lower()
supported_types = ['char', 'subword']
if decoding_type not in supported_types:
raise ValueError(
f"Unsupported decoding type. Supported types = {supported_types}.\n" f"Given = {decoding_type}"
)
self.decoding_type = decoding_type
def set_tokenizer(self, tokenizer: TokenizerSpec):
"""
Set the tokenizer of the decoding framework.
Args:
tokenizer: NeMo tokenizer object, which inherits from TokenizerSpec.
"""
self.tokenizer = tokenizer
@typecheck()
def forward(
self,
decoder_output: torch.Tensor,
decoder_lengths: torch.Tensor,
) -> Tuple[List[Union[rnnt_utils.Hypothesis, rnnt_utils.NBestHypotheses]]]:
"""Returns a list of hypotheses given an input batch of the encoder hidden embedding.
Output token is generated auto-repressively.
Args:
decoder_output: A tensor of size (batch, timesteps, features) or (batch, timesteps) (each timestep is a label).
decoder_lengths: list of int representing the length of each sequence
output sequence.
Returns:
packed list containing batch number of sentences (Hypotheses).
"""
raise NotImplementedError()
def __call__(self, *args, **kwargs):
return self.forward(*args, **kwargs)
[docs]
class BeamCTCInfer(AbstractBeamCTCInfer):
"""A beam CTC decoder.
Provides a common abstraction for sample level and batch level greedy decoding.
Args:
blank_index: int index of the blank token. Can be 0 or len(vocabulary).
preserve_alignments: Bool flag which preserves the history of logprobs generated during
decoding (sample / batched). When set to true, the Hypothesis will contain
the non-null value for `logprobs` in it. Here, `logprobs` is a torch.Tensors.
compute_timestamps: A bool flag, which determines whether to compute the character/subword, or
word based timestamp mapping the output log-probabilities to discrite intervals of timestamps.
The timestamps will be available in the returned Hypothesis.timestep as a dictionary.
"""
def __init__(
self,
blank_id: int,
beam_size: int,
search_type: str = "default",
return_best_hypothesis: bool = True,
preserve_alignments: bool = False,
compute_timestamps: bool = False,
ngram_lm_alpha: float = 0.3,
beam_beta: float = 0.0,
ngram_lm_model: str = None,
flashlight_cfg: Optional['FlashlightConfig'] = None,
pyctcdecode_cfg: Optional['PyCTCDecodeConfig'] = None,
):
super().__init__(blank_id=blank_id, beam_size=beam_size)
self.search_type = search_type
self.return_best_hypothesis = return_best_hypothesis
self.preserve_alignments = preserve_alignments
self.compute_timestamps = compute_timestamps
if self.compute_timestamps:
raise ValueError("Currently this flag is not supported for beam search algorithms.")
self.vocab = None # This must be set by specific method by user before calling forward() !
if search_type == "default" or search_type == "nemo":
self.search_algorithm = self.default_beam_search
elif search_type == "pyctcdecode":
self.search_algorithm = self._pyctcdecode_beam_search
elif search_type == "flashlight":
self.search_algorithm = self.flashlight_beam_search
else:
raise NotImplementedError(
f"The search type ({search_type}) supplied is not supported!\n"
f"Please use one of : (default, nemo, pyctcdecode)"
)
# Log the beam search algorithm
logging.info(f"Beam search algorithm: {search_type}")
self.ngram_lm_alpha = ngram_lm_alpha
self.beam_beta = beam_beta
# Default beam search args
self.ngram_lm_model = ngram_lm_model
# PyCTCDecode params
if pyctcdecode_cfg is None:
pyctcdecode_cfg = PyCTCDecodeConfig()
self.pyctcdecode_cfg = pyctcdecode_cfg # type: PyCTCDecodeConfig
if flashlight_cfg is None:
flashlight_cfg = FlashlightConfig()
self.flashlight_cfg = flashlight_cfg
# Default beam search scorer functions
self.default_beam_scorer = None
self.pyctcdecode_beam_scorer = None
self.flashlight_beam_scorer = None
self.token_offset = 0
[docs]
@typecheck()
def forward(
self,
decoder_output: torch.Tensor,
decoder_lengths: torch.Tensor,
) -> Tuple[List[Union[rnnt_utils.Hypothesis, rnnt_utils.NBestHypotheses]]]:
"""Returns a list of hypotheses given an input batch of the encoder hidden embedding.
Output token is generated auto-repressively.
Args:
decoder_output: A tensor of size (batch, timesteps, features).
decoder_lengths: list of int representing the length of each sequence
output sequence.
Returns:
packed list containing batch number of sentences (Hypotheses).
"""
if self.vocab is None:
raise RuntimeError("Please set the vocabulary with `set_vocabulary()` before calling this function.")
if self.decoding_type is None:
raise ValueError("Please set the decoding type with `set_decoding_type()` before calling this function.")
with torch.no_grad(), torch.inference_mode():
# Process each sequence independently
prediction_tensor = decoder_output
if prediction_tensor.ndim != 3:
raise ValueError(
f"`decoder_output` must be a tensor of shape [B, T, V] (log probs, float). "
f"Provided shape = {prediction_tensor.shape}"
)
# determine type of input - logprobs or labels
out_len = decoder_lengths if decoder_lengths is not None else None
hypotheses = self.search_algorithm(prediction_tensor, out_len)
# Pack results into Hypotheses
packed_result = pack_hypotheses(hypotheses, decoder_lengths)
# Pack the result
if self.return_best_hypothesis and isinstance(packed_result[0], rnnt_utils.NBestHypotheses):
packed_result = [res.n_best_hypotheses[0] for res in packed_result] # type: Hypothesis
return (packed_result,)
[docs]
@torch.no_grad()
def default_beam_search(
self, x: torch.Tensor, out_len: torch.Tensor
) -> List[Union[rnnt_utils.Hypothesis, rnnt_utils.NBestHypotheses]]:
"""
Open Seq2Seq Beam Search Algorithm (DeepSpeed)
Args:
x: Tensor of shape [B, T, V+1], where B is the batch size, T is the maximum sequence length,
and V is the vocabulary size. The tensor contains log-probabilities.
out_len: Tensor of shape [B], contains lengths of each sequence in the batch.
Returns:
A list of NBestHypotheses objects, one for each sequence in the batch.
"""
if self.compute_timestamps:
raise ValueError(
f"Beam Search with strategy `{self.search_type}` does not support time stamp calculation!"
)
if self.default_beam_scorer is None:
# Check for filepath
if self.ngram_lm_model is None or not os.path.exists(self.ngram_lm_model):
raise FileNotFoundError(
f"KenLM binary file not found at : {self.ngram_lm_model}. "
f"Please set a valid path in the decoding config."
)
# perform token offset for subword models
if self.decoding_type == 'subword':
vocab = [chr(idx + self.token_offset) for idx in range(len(self.vocab))]
else:
# char models
vocab = self.vocab
# Must import at runtime to avoid circular dependency due to module level import.
from nemo.collections.asr.modules.beam_search_decoder import BeamSearchDecoderWithLM
self.default_beam_scorer = BeamSearchDecoderWithLM(
vocab=vocab,
lm_path=self.ngram_lm_model,
beam_width=self.beam_size,
alpha=self.ngram_lm_alpha,
beta=self.beam_beta,
num_cpus=max(1, os.cpu_count()),
input_tensor=False,
)
x = x.to('cpu')
with typecheck.disable_checks():
data = [x[sample_id, : out_len[sample_id], :].softmax(dim=-1) for sample_id in range(len(x))]
beams_batch = self.default_beam_scorer.forward(log_probs=data, log_probs_length=None)
# For each sample in the batch
nbest_hypotheses = []
for beams_idx, beams in enumerate(beams_batch):
# For each beam candidate / hypothesis in each sample
hypotheses = []
for candidate_idx, candidate in enumerate(beams):
hypothesis = rnnt_utils.Hypothesis(
score=0.0, y_sequence=[], dec_state=None, timestamp=[], last_token=None
)
# For subword encoding, NeMo will double encode the subword (multiple tokens) into a
# singular unicode id. In doing so, we preserve the semantic of the unicode token, and
# compress the size of the final KenLM ARPA / Binary file.
# In order to do double encoding, we shift the subword by some token offset.
# This step is ignored for character based models.
if self.decoding_type == 'subword':
pred_token_ids = [ord(c) - self.token_offset for c in candidate[1]]
else:
# Char models
pred_token_ids = [self.vocab_index_map[c] for c in candidate[1]]
# We preserve the token ids and the score for this hypothesis
hypothesis.y_sequence = pred_token_ids
hypothesis.score = candidate[0]
# If alignment must be preserved, we preserve a view of the output logprobs.
# Note this view is shared amongst all beams within the sample, be sure to clone it if you
# require specific processing for each sample in the beam.
# This is done to preserve memory.
if self.preserve_alignments:
hypothesis.alignments = x[beams_idx][: out_len[beams_idx]]
hypotheses.append(hypothesis)
# Wrap the result in NBestHypothesis.
hypotheses = rnnt_utils.NBestHypotheses(hypotheses)
nbest_hypotheses.append(hypotheses)
return nbest_hypotheses
@torch.no_grad()
def _pyctcdecode_beam_search(
self, x: torch.Tensor, out_len: torch.Tensor
) -> List[Union[rnnt_utils.Hypothesis, rnnt_utils.NBestHypotheses]]:
"""
PyCTCDecode Beam Search Algorithm. Should support Char and Subword models.
Args:
x: Tensor of shape [B, T, V+1], where B is the batch size, T is the maximum sequence length,
and V is the vocabulary size. The tensor contains log-probabilities.
out_len: Tensor of shape [B], contains lengths of each sequence in the batch.
Returns:
A list of NBestHypotheses objects, one for each sequence in the batch.
"""
if self.compute_timestamps:
raise ValueError(
f"Beam Search with strategy `{self.search_type}` does not support time stamp calculation!"
)
try:
import pyctcdecode
except (ImportError, ModuleNotFoundError):
raise ImportError(
"Could not load `pyctcdecode` library. Please install it from pip using :\n"
"pip install --upgrade pyctcdecode"
)
if self.pyctcdecode_beam_scorer is None:
self.pyctcdecode_beam_scorer = pyctcdecode.build_ctcdecoder(
labels=self.vocab, kenlm_model_path=self.ngram_lm_model, alpha=self.ngram_lm_alpha, beta=self.beam_beta
) # type: pyctcdecode.BeamSearchDecoderCTC
x = x.to('cpu').numpy()
with typecheck.disable_checks():
beams_batch = []
for sample_id in range(len(x)):
logprobs = x[sample_id, : out_len[sample_id], :]
result = self.pyctcdecode_beam_scorer.decode_beams(
logprobs,
beam_width=self.beam_size,
beam_prune_logp=self.pyctcdecode_cfg.beam_prune_logp,
token_min_logp=self.pyctcdecode_cfg.token_min_logp,
prune_history=self.pyctcdecode_cfg.prune_history,
hotwords=self.pyctcdecode_cfg.hotwords,
hotword_weight=self.pyctcdecode_cfg.hotword_weight,
lm_start_state=None,
) # Output format: text, last_lm_state, text_frames, logit_score, lm_score
beams_batch.append(result)
nbest_hypotheses = []
for beams_idx, beams in enumerate(beams_batch):
hypotheses = []
for candidate_idx, candidate in enumerate(beams):
# Candidate = (text, last_lm_state, text_frames, logit_score, lm_score)
hypothesis = rnnt_utils.Hypothesis(
score=0.0, y_sequence=[], dec_state=None, timestamp=[], last_token=None
)
# TODO: Requires token ids to be returned rather than text.
if self.decoding_type == 'subword':
if self.tokenizer is None:
raise ValueError("Tokenizer must be provided for subword decoding. Use set_tokenizer().")
pred_token_ids = self.tokenizer.text_to_ids(candidate[0])
else:
if self.vocab is None:
raise ValueError("Vocab must be provided for character decoding. Use set_vocab().")
chars = list(candidate[0])
pred_token_ids = [self.vocab_index_map[c] for c in chars]
hypothesis.y_sequence = pred_token_ids
hypothesis.text = candidate[0] # text
hypothesis.score = candidate[4] # score
# Inject word level timestamps
hypothesis.timestamp = candidate[2] # text_frames
if self.preserve_alignments:
hypothesis.alignments = torch.from_numpy(x[beams_idx][: out_len[beams_idx]])
hypotheses.append(hypothesis)
hypotheses = rnnt_utils.NBestHypotheses(hypotheses)
nbest_hypotheses.append(hypotheses)
return nbest_hypotheses
[docs]
@torch.no_grad()
def flashlight_beam_search(
self, x: torch.Tensor, out_len: torch.Tensor
) -> List[Union[rnnt_utils.Hypothesis, rnnt_utils.NBestHypotheses]]:
"""
Flashlight Beam Search Algorithm. Should support Char and Subword models.
Args:
x: Tensor of shape [B, T, V+1], where B is the batch size, T is the maximum sequence length,
and V is the vocabulary size. The tensor contains log-probabilities.
out_len: Tensor of shape [B], contains lengths of each sequence in the batch.
Returns:
A list of NBestHypotheses objects, one for each sequence in the batch.
"""
if self.compute_timestamps:
raise ValueError(
f"Beam Search with strategy `{self.search_type}` does not support time stamp calculation!"
)
if self.flashlight_beam_scorer is None:
# Check for filepath
if self.ngram_lm_model is None or not os.path.exists(self.ngram_lm_model):
raise FileNotFoundError(
f"KenLM binary file not found at : {self.ngram_lm_model}. "
"Please set a valid path in the decoding config."
)
# perform token offset for subword models
# if self.decoding_type == 'subword':
# vocab = [chr(idx + self.token_offset) for idx in range(len(self.vocab))]
# else:
# # char models
# vocab = self.vocab
# Must import at runtime to avoid circular dependency due to module level import.
from nemo.collections.asr.modules.flashlight_decoder import FlashLightKenLMBeamSearchDecoder
self.flashlight_beam_scorer = FlashLightKenLMBeamSearchDecoder(
lm_path=self.ngram_lm_model,
vocabulary=self.vocab,
tokenizer=self.tokenizer,
lexicon_path=self.flashlight_cfg.lexicon_path,
boost_path=self.flashlight_cfg.boost_path,
beam_size=self.beam_size,
beam_size_token=self.flashlight_cfg.beam_size_token,
beam_threshold=self.flashlight_cfg.beam_threshold,
lm_weight=self.ngram_lm_alpha,
word_score=self.beam_beta,
unk_weight=self.flashlight_cfg.unk_weight,
sil_weight=self.flashlight_cfg.sil_weight,
)
x = x.to('cpu')
with typecheck.disable_checks():
beams_batch = self.flashlight_beam_scorer.forward(log_probs=x)
# For each sample in the batch
nbest_hypotheses = []
for beams_idx, beams in enumerate(beams_batch):
# For each beam candidate / hypothesis in each sample
hypotheses = []
for candidate_idx, candidate in enumerate(beams):
hypothesis = rnnt_utils.Hypothesis(
score=0.0, y_sequence=[], dec_state=None, timestamp=[], last_token=None
)
# We preserve the token ids and the score for this hypothesis
hypothesis.y_sequence = candidate['tokens'].tolist()
hypothesis.score = candidate['score']
# If alignment must be preserved, we preserve a view of the output logprobs.
# Note this view is shared amongst all beams within the sample, be sure to clone it if you
# require specific processing for each sample in the beam.
# This is done to preserve memory.
if self.preserve_alignments:
hypothesis.alignments = x[beams_idx][: out_len[beams_idx]]
hypotheses.append(hypothesis)
# Wrap the result in NBestHypothesis.
hypotheses = rnnt_utils.NBestHypotheses(hypotheses)
nbest_hypotheses.append(hypotheses)
return nbest_hypotheses
[docs]
def set_decoding_type(self, decoding_type: str):
super().set_decoding_type(decoding_type)
# Please check train_kenlm.py in scripts/asr_language_modeling/ to find out why we need
# TOKEN_OFFSET for BPE-based models
if self.decoding_type == 'subword':
self.token_offset = DEFAULT_TOKEN_OFFSET
class WfstCTCInfer(AbstractBeamCTCInfer):
"""A WFST-based beam CTC decoder.
Provides a common abstraction for sample level and batch level beam decoding.
Args:
TBD
"""
def __init__(
self,
blank_id: int,
beam_size: int,
search_type: str = "riva",
return_best_hypothesis: bool = True,
preserve_alignments: bool = False,
compute_timestamps: bool = False,
decoding_mode: str = 'nbest', # 'nbest', 'mbr' ('mbr' works only for search_type == 'riva' and beam_size == 1)
open_vocabulary_decoding: bool = False,
beam_width: float = 10.0,
lm_weight: float = 1.0,
device: str = "cuda",
arpa_lm_path: str = None,
wfst_lm_path: str = None,
riva_decoding_cfg: Optional['RivaDecoderConfig'] = None,
):
super().__init__(blank_id=blank_id, beam_size=beam_size)
self.search_type = search_type
self.return_best_hypothesis = return_best_hypothesis
self.preserve_alignments = preserve_alignments
self.compute_timestamps = compute_timestamps
self.decoding_algorithm = None
if search_type in ("default", "riva"):
self.decoding_algorithm = self._riva_decoding
# Log the WFST search_type
logging.info(f"WFST beam search search_type: {search_type}")
self.search_type = search_type
if beam_size > 1 and decoding_mode != 'nbest':
logging.warning(
f"`beam_size` > 1 is supported only for `decoding_mode` == `nbest`\n"
f"(provided: `{decoding_mode}`).\n"
f"`beam_size` rewritten as 1"
)
self.beam_size = 1
self.decoding_mode = decoding_mode
self.open_vocabulary_decoding = open_vocabulary_decoding
self._tokenword_disambig_id = -1
self.beam_width = beam_width
self.lm_weight = lm_weight
self.device = device
# Default beam search args
self.arpa_lm_path = arpa_lm_path
self.wfst_lm_path = wfst_lm_path
self.riva_decoding_cfg = riva_decoding_cfg
# Default beam search scorer functions
self.riva_decoder = None
@typecheck()
def forward(
self,
decoder_output: torch.Tensor,
decoder_lengths: torch.Tensor,
) -> Tuple[List[Union[rnnt_utils.Hypothesis, rnnt_utils.NBestHypotheses]]]:
"""Returns a list of hypotheses given an input batch of the encoder hidden embedding.
Output token is generated auto-repressively.
Args:
decoder_output: A tensor of size (batch, timesteps, features).
decoder_lengths: list of int representing the length of each sequence
output sequence.
Returns:
packed list containing batch number of sentences (Hypotheses).
"""
if self.vocab is None:
raise RuntimeError("Please set the vocabulary with `set_vocabulary()` before calling this function.")
if self.decoding_type != 'subword':
raise ValueError(
f"`decoding_type` other than `subword` is not supported. Provided: `{self.decoding_type}`"
)
elif self.tokenizer is None:
raise ValueError("Tokenizer must be provided for subword decoding. Use set_tokenizer().")
if self.decoding_algorithm is None:
raise NotImplementedError(
f"The decoding search_type ({self.search_type}) supplied is not supported!\n"
f"Please use one of : (default, riva)"
)
with torch.no_grad(), torch.inference_mode():
# Process each sequence independently
prediction_tensor = decoder_output
if prediction_tensor.ndim != 3:
raise ValueError(
f"`decoder_output` must be a tensor of shape [B, T, V] (log probs, float). "
f"Provided shape = {prediction_tensor.shape}"
)
hypotheses = self.decoding_algorithm(prediction_tensor, decoder_lengths)
# Pack results into Hypotheses
packed_result = pack_wfst_hypotheses(hypotheses, prediction_tensor, decoder_lengths)
# Pack the result
if self.return_best_hypothesis and isinstance(packed_result[0], rnnt_utils.NBestHypotheses):
packed_result = [res.n_best_hypotheses[0] for res in packed_result] # type: Hypothesis
return (packed_result,)
def _prepare_decoding_lm_wfst(self) -> Union[str, 'kaldifst.StdFst']: # noqa: F821
"""TBD"""
arpa_lm_path_exists = self.arpa_lm_path is not None and os.path.exists(self.arpa_lm_path)
wfst_lm_path_exists = self.wfst_lm_path is not None and os.path.exists(self.wfst_lm_path)
lm_fst = None
if wfst_lm_path_exists:
if self.search_type == "riva" and not self.wfst_lm_path.endswith(".fst"):
raise ValueError(
f"Search type `riva` expects WFSTs in the `.fst` format. Provided: `{self.wfst_lm_path}`"
)
if arpa_lm_path_exists:
logging.warning(
"Both `arpa_lm_path` and `wfst_lm_path` are provided and not empty. The latter will be used."
)
lm_fst = self.wfst_lm_path
elif not arpa_lm_path_exists:
raise FileNotFoundError(
f"Arpa LM file not found at `{self.arpa_lm_path}` and WFST LM is not found at `{self.wfst_lm_path}`.\n"
f"Please set a valid path in the decoding config for at least one of those."
)
else:
logging.warning(
f"Since WFST LM is not found at `{self.wfst_lm_path}`, "
f"it will be made from the Arpa LM at `{self.arpa_lm_path}`.\n"
f"This procedure will take some time."
)
if self.wfst_lm_path is not None:
logging.info(f"WFST LM will be buffered at `{self.wfst_lm_path}`.")
write_tlg_path = self.wfst_lm_path
else:
logging.warning("Consider providing a write-permitted `wfst_lm_path` for WFST LM buffering.")
write_tlg_path = None
ctc_topology = "default" # there is no way to indicate the need of other topologies
target = "kaldi"
from nemo.collections.asr.parts.utils.wfst_utils import mkgraph_ctc_ov
lm_fst, tokenword_disambig_id = mkgraph_ctc_ov(
tokenizer=self.tokenizer,
lm_path=self.arpa_lm_path,
topology_name=ctc_topology,
write_tlg_path=write_tlg_path,
open_vocabulary=self.open_vocabulary_decoding,
target=target,
)
self._tokenword_disambig_id = tokenword_disambig_id
return lm_fst
@torch.no_grad()
def _riva_decoding(self, x: torch.Tensor, out_len: torch.Tensor) -> List['WfstNbestHypothesis']:
"""
Riva Asrlib WFST decoder Algorithm.
Args:
x: Tensor of shape [B, T, V+1], where B is the batch size, T is the maximum sequence length,
and V is the vocabulary size. The tensor contains log-probabilities.
out_len: Tensor of shape [B], contains lengths of each sequence in the batch.
Returns:
A list of WfstNbestHypothesis objects, one for each sequence in the batch.
"""
if self.riva_decoder is None:
lm_fst = self._prepare_decoding_lm_wfst()
if self.open_vocabulary_decoding and self._tokenword_disambig_id == -1:
# trying to extract tokenword_disambig_id from the lm_fst
if isinstance(lm_fst, str):
# use importer instead of direct import to possibly get an installation message
from nemo.collections.asr.parts.utils.wfst_utils import kaldifst_importer
kaldifst = kaldifst_importer()
lm_fst = kaldifst.StdVectorFst.read(self.wfst_lm_path)
tokenword_disambig_id = lm_fst.output_symbols.find("#1")
if tokenword_disambig_id == -1:
raise ValueError(
"Cannot determine `tokenword_disambig_id` "
"which is required if `open_vocabulary_decoding` == True"
)
self._tokenword_disambig_id = tokenword_disambig_id
if not self.device.startswith("cuda"):
raise ValueError(f"Riva decoder does not support non-cuda device. Provided: `{self.device}`")
from nemo.collections.asr.parts.submodules.wfst_decoder import RivaGpuWfstDecoder
self.riva_decoder = RivaGpuWfstDecoder(
lm_fst=lm_fst,
decoding_mode=self.decoding_mode,
beam_size=self.beam_width,
config=self.riva_decoding_cfg,
tokenword_disambig_id=self._tokenword_disambig_id,
lm_weight=self.lm_weight,
nbest_size=self.beam_size,
)
return self.riva_decoder.decode(x.to(device=self.device), out_len.to(device=self.device))
class BeamBatchedCTCInfer(AbstractBeamCTCInfer, WithOptionalCudaGraphs):
"""
A batched beam CTC decoder.
Args:
blank_index: int index of the blank token. Can be 0 or len(vocabulary).
beam_size: int size of the beam.
return_best_hypothesis: When set to True (default), returns a single Hypothesis.
When set to False, returns a NBestHypotheses container, which contains a list of Hypothesis.
preserve_alignments: Bool flag which preserves the history of logprobs generated during
decoding (sample / batched). When set to true, the Hypothesis will contain
the non-null value for `logprobs` in it. Here, `logprobs` is a torch.Tensors.
compute_timestamps: A bool flag, which determines whether to compute the character/subword, or
word based timestamp mapping the output log-probabilities to discrite intervals of timestamps.
The timestamps will be available in the returned Hypothesis.timestep as a dictionary.
ngram_lm_alpha: float, the language model weight.
beam_beta: float, the word insertion weight.
beam_threshold: float, the beam pruning threshold.
ngram_lm_model: str, the path to the ngram model.
boosting_tree: BoostingTreeModelConfig, the boosting tree model config.
boosting_tree_alpha: float, the boosting tree alpha.
allow_cuda_graphs: bool, whether to allow cuda graphs for the beam search algorithm.
"""
def __init__(
self,
blank_index: int,
beam_size: int,
return_best_hypothesis: bool = True,
preserve_alignments: bool = False,
compute_timestamps: bool = False,
ngram_lm_alpha: float = 1.0,
beam_beta: float = 0.0,
beam_threshold: float = 20.0,
ngram_lm_model: str = None,
boosting_tree: BoostingTreeModelConfig = None,
boosting_tree_alpha: float = 0.0,
allow_cuda_graphs: bool = True,
tokenizer: TokenizerSpec = None,
):
super().__init__(blank_id=blank_index, beam_size=beam_size)
self.return_best_hypothesis = return_best_hypothesis
self.preserve_alignments = preserve_alignments
self.compute_timestamps = compute_timestamps
self.allow_cuda_graphs = allow_cuda_graphs
if self.compute_timestamps:
raise ValueError("`Compute timestamps` is not supported for batched beam search.")
if self.preserve_alignments:
raise ValueError("`Preserve alignments` is not supported for batched beam search.")
self.beam_beta = beam_beta
self.beam_threshold = beam_threshold
# load fusion models from paths (ngram_lm_model and boosting_tree_model)
fusion_models, fusion_models_alpha = [], []
if ngram_lm_model is not None:
assert blank_index != 0, "Blank should not be the first token in the vocabulary"
fusion_models.append(NGramGPULanguageModel.from_file(lm_path=ngram_lm_model, vocab_size=blank_index))
fusion_models_alpha.append(ngram_lm_alpha)
if boosting_tree and not BoostingTreeModelConfig.is_empty(boosting_tree):
assert blank_index != 0, "Blank should not be the first token in the vocabulary"
fusion_models.append(GPUBoostingTreeModel.from_config(boosting_tree, tokenizer=tokenizer))
fusion_models_alpha.append(boosting_tree_alpha)
if not fusion_models:
fusion_models = None
fusion_models_alpha = None
# # Default beam search args
self.search_algorithm = BatchedBeamCTCComputer(
blank_index=blank_index,
beam_size=beam_size,
return_best_hypothesis=return_best_hypothesis,
preserve_alignments=preserve_alignments,
compute_timestamps=compute_timestamps,
fusion_models=fusion_models,
fusion_models_alpha=fusion_models_alpha,
beam_beta=beam_beta,
beam_threshold=beam_threshold,
allow_cuda_graphs=allow_cuda_graphs,
)
def disable_cuda_graphs(self) -> bool:
"""Disable CUDA graphs (e.g., for decoding in training)"""
if isinstance(self.search_algorithm, WithOptionalCudaGraphs):
return self.search_algorithm.disable_cuda_graphs()
return False
def maybe_enable_cuda_graphs(self) -> bool:
"""Enable CUDA graphs (if allowed)"""
if isinstance(self.search_algorithm, WithOptionalCudaGraphs):
return self.search_algorithm.maybe_enable_cuda_graphs()
return False
@typecheck()
def forward(
self,
decoder_output: torch.Tensor,
decoder_lengths: torch.Tensor,
) -> Tuple[List[Union[rnnt_utils.Hypothesis, rnnt_utils.NBestHypotheses]]]:
"""Returns a list of hypotheses given an input batch of the encoder hidden embedding.
Output token is generated auto-repressively.
Args:
decoder_output: A tensor of size (batch, timesteps, features).
decoder_lengths: list of int representing the length of each sequence
output sequence.
Returns:
packed list containing batch number of sentences (Hypotheses).
"""
with torch.no_grad(), torch.inference_mode():
if decoder_output.ndim != 3:
raise ValueError(
f"`decoder_output` must be a tensor of shape [B, T, V] (log probs, float). "
f"Provided shape = {decoder_output.shape}"
)
batched_beam_hyps = self.search_algorithm(decoder_output, decoder_lengths)
batch_size = decoder_lengths.shape[0]
if self.return_best_hypothesis:
hyps = batched_beam_hyps.to_hyps_list(score_norm=False)[:batch_size]
else:
hyps = batched_beam_hyps.to_nbest_hyps_list(score_norm=False)[:batch_size]
return (hyps,)
@dataclass
class PyCTCDecodeConfig:
# These arguments cannot be imported from pyctcdecode (optional dependency)
# Therefore we copy the values explicitly
# Taken from pyctcdecode.constant
beam_prune_logp: float = -10.0
token_min_logp: float = -5.0
prune_history: bool = False
hotwords: Optional[List[str]] = None
hotword_weight: float = 10.0
@dataclass
class FlashlightConfig:
lexicon_path: Optional[str] = None
boost_path: Optional[str] = None
beam_size_token: int = 16
beam_threshold: float = 20.0
unk_weight: float = -math.inf
sil_weight: float = 0.0
@dataclass
class BeamCTCInferConfig:
beam_size: int
search_type: str = 'default'
preserve_alignments: bool = False
compute_timestamps: bool = False
return_best_hypothesis: bool = True
allow_cuda_graphs: bool = True
beam_alpha: Optional[float] = None # Deprecated
beam_beta: float = 1.0
beam_threshold: float = 20.0
kenlm_path: Optional[str] = None # Deprecated, default should be None
ngram_lm_alpha: Optional[float] = 1.0
ngram_lm_model: Optional[str] = None
boosting_tree: BoostingTreeModelConfig = field(default_factory=BoostingTreeModelConfig)
boosting_tree_alpha: Optional[float] = 0.0
flashlight_cfg: Optional[FlashlightConfig] = field(default_factory=lambda: FlashlightConfig())
pyctcdecode_cfg: Optional[PyCTCDecodeConfig] = field(default_factory=lambda: PyCTCDecodeConfig())
@dataclass
class WfstCTCInferConfig:
beam_size: int
search_type: str = "riva"
return_best_hypothesis: bool = True
preserve_alignments: bool = False
compute_timestamps: bool = False
decoding_mode: str = 'nbest' # 'nbest', 'mbr' ('mbr' works only for search_type == 'riva' and beam_size == 1)
open_vocabulary_decoding: bool = False
beam_width: float = 10.0
lm_weight: float = 1.0
device: str = "cuda"
arpa_lm_path: Optional[str] = None
wfst_lm_path: Optional[str] = None
riva_decoding_cfg: Optional['RivaDecoderConfig'] = field(default_factory=lambda: RivaDecoderConfig())
