|
|
|
|
@ -1,6 +1,6 @@
|
|
|
|
|
# MIT License, Copyright (c) 2022 OpenAI.
|
|
|
|
|
# Copyright (c) 2022 PaddlePaddle Authors and . All Rights Reserved.
|
|
|
|
|
#
|
|
|
|
|
#
|
|
|
|
|
# Modified from OpenAI Whisper 2022 (https://github.com/openai/whisper/whisper)
|
|
|
|
|
import os
|
|
|
|
|
from dataclasses import dataclass
|
|
|
|
|
@ -265,7 +265,6 @@ class DecodingOptions:
|
|
|
|
|
task: str = "transcribe" # whether to perform X->X "transcribe" or X->English "translate"
|
|
|
|
|
language: Optional[
|
|
|
|
|
str] = None # language that the audio is in; uses detected language if None
|
|
|
|
|
|
|
|
|
|
# sampling-related options
|
|
|
|
|
temperature: float = 0.0
|
|
|
|
|
sample_len: Optional[int] = None # maximum number of tokens to sample
|
|
|
|
|
@ -361,10 +360,11 @@ class WhisperInference(Inference):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
@paddle.no_grad()
|
|
|
|
|
def detect_language(model: "Whisper",
|
|
|
|
|
mel: paddle.Tensor,
|
|
|
|
|
tokenizer: Tokenizer=None
|
|
|
|
|
) -> Tuple[paddle.Tensor, List[dict]]:
|
|
|
|
|
def detect_language(
|
|
|
|
|
model: "Whisper",
|
|
|
|
|
mel: paddle.Tensor,
|
|
|
|
|
resource_path: str,
|
|
|
|
|
tokenizer: Tokenizer=None) -> Tuple[paddle.Tensor, List[dict]]:
|
|
|
|
|
"""
|
|
|
|
|
Detect the spoken language in the audio, and return them as list of strings, along with the ids
|
|
|
|
|
of the most probable language tokens and the probability distribution over all language tokens.
|
|
|
|
|
@ -378,7 +378,8 @@ def detect_language(model: "Whisper",
|
|
|
|
|
list of dictionaries containing the probability distribution over all languages.
|
|
|
|
|
"""
|
|
|
|
|
if tokenizer is None:
|
|
|
|
|
tokenizer = get_tokenizer(model.is_multilingual)
|
|
|
|
|
tokenizer = get_tokenizer(
|
|
|
|
|
model.is_multilingual, resource_path=resource_path)
|
|
|
|
|
if tokenizer.language is None or tokenizer.language_token not in tokenizer.sot_sequence:
|
|
|
|
|
raise ValueError(
|
|
|
|
|
"This model doesn't have language tokens so it can't perform lang id"
|
|
|
|
|
@ -419,6 +420,7 @@ def detect_language(model: "Whisper",
|
|
|
|
|
def transcribe(
|
|
|
|
|
model: "Whisper",
|
|
|
|
|
mel: paddle.Tensor,
|
|
|
|
|
resource_path: str,
|
|
|
|
|
*,
|
|
|
|
|
verbose: Optional[bool]=None,
|
|
|
|
|
temperature: Union[float, Tuple[float, ...]]=(0.0, 0.2, 0.4, 0.6, 0.8,
|
|
|
|
|
@ -485,7 +487,7 @@ def transcribe(
|
|
|
|
|
"Detecting language using up to the first 30 seconds. Use `--language` to specify the language"
|
|
|
|
|
)
|
|
|
|
|
segment = pad_or_trim(mel, N_FRAMES)
|
|
|
|
|
_, probs = model.detect_language(segment)
|
|
|
|
|
_, probs = model.detect_language(segment, resource_path)
|
|
|
|
|
decode_options["language"] = max(probs, key=probs.get)
|
|
|
|
|
if verbose is not None:
|
|
|
|
|
print(
|
|
|
|
|
@ -495,7 +497,10 @@ def transcribe(
|
|
|
|
|
language = decode_options["language"]
|
|
|
|
|
task = decode_options.get("task", "transcribe")
|
|
|
|
|
tokenizer = get_tokenizer(
|
|
|
|
|
model.is_multilingual, language=language, task=task)
|
|
|
|
|
model.is_multilingual,
|
|
|
|
|
resource_path=resource_path,
|
|
|
|
|
language=language,
|
|
|
|
|
task=task)
|
|
|
|
|
|
|
|
|
|
def decode_with_fallback(segment: paddle.Tensor) -> DecodingResult:
|
|
|
|
|
temperatures = [temperature] if isinstance(temperature, (
|
|
|
|
|
@ -513,7 +518,7 @@ def transcribe(
|
|
|
|
|
kwargs.pop("best_of", None)
|
|
|
|
|
|
|
|
|
|
options = DecodingOptions(**kwargs, temperature=t)
|
|
|
|
|
decode_result = model.decode(segment, options)
|
|
|
|
|
decode_result = model.decode(segment, options, resource_path)
|
|
|
|
|
|
|
|
|
|
needs_fallback = False
|
|
|
|
|
if compression_ratio_threshold is not None and decode_result.compression_ratio > compression_ratio_threshold:
|
|
|
|
|
@ -979,14 +984,21 @@ class DecodingTask:
|
|
|
|
|
decoder: TokenDecoder
|
|
|
|
|
logit_filters: List[LogitFilter]
|
|
|
|
|
|
|
|
|
|
def __init__(self, model: "Whisper", options: DecodingOptions):
|
|
|
|
|
def __init__(self,
|
|
|
|
|
model: "Whisper",
|
|
|
|
|
options: DecodingOptions,
|
|
|
|
|
resource_path: str):
|
|
|
|
|
self.model = model
|
|
|
|
|
|
|
|
|
|
language = options.language or "en"
|
|
|
|
|
tokenizer = get_tokenizer(
|
|
|
|
|
model.is_multilingual, language=language, task=options.task)
|
|
|
|
|
model.is_multilingual,
|
|
|
|
|
resource_path=resource_path,
|
|
|
|
|
language=language,
|
|
|
|
|
task=options.task)
|
|
|
|
|
self.tokenizer: Tokenizer = tokenizer
|
|
|
|
|
self.options: DecodingOptions = self._verify_options(options)
|
|
|
|
|
self.resource_path: str = resource_path
|
|
|
|
|
|
|
|
|
|
self.beam_size: int = options.beam_size or options.best_of or 1
|
|
|
|
|
self.n_ctx: int = model.dims.n_text_ctx
|
|
|
|
|
@ -1112,13 +1124,14 @@ class DecodingTask:
|
|
|
|
|
|
|
|
|
|
def _detect_language(self,
|
|
|
|
|
audio_features: paddle.Tensor,
|
|
|
|
|
tokens: paddle.Tensor):
|
|
|
|
|
tokens: paddle.Tensor,
|
|
|
|
|
resource_path: str):
|
|
|
|
|
languages = [self.options.language] * audio_features.shape[0]
|
|
|
|
|
lang_probs = None
|
|
|
|
|
|
|
|
|
|
if self.options.language is None or self.options.task == "lang_id":
|
|
|
|
|
lang_tokens, lang_probs = self.model.detect_language(audio_features,
|
|
|
|
|
self.tokenizer)
|
|
|
|
|
lang_tokens, lang_probs = self.model.detect_language(
|
|
|
|
|
audio_features, self.tokenizer, self.resource_path)
|
|
|
|
|
languages = [max(probs, key=probs.get) for probs in lang_probs]
|
|
|
|
|
if self.options.language is None:
|
|
|
|
|
tokens[:, self.sot_index +
|
|
|
|
|
@ -1185,7 +1198,8 @@ class DecodingTask:
|
|
|
|
|
|
|
|
|
|
# detect language if requested, overwriting the language token
|
|
|
|
|
languages, language_probs = self._detect_language(
|
|
|
|
|
paddle.to_tensor(audio_features), paddle.to_tensor(tokens))
|
|
|
|
|
paddle.to_tensor(audio_features),
|
|
|
|
|
paddle.to_tensor(tokens), self.resource_path)
|
|
|
|
|
|
|
|
|
|
if self.options.task == "lang_id":
|
|
|
|
|
return [
|
|
|
|
|
@ -1254,10 +1268,11 @@ class DecodingTask:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
@paddle.no_grad()
|
|
|
|
|
def decode(model: "Whisper",
|
|
|
|
|
mel: paddle.Tensor,
|
|
|
|
|
options: DecodingOptions=DecodingOptions()
|
|
|
|
|
) -> Union[DecodingResult, List[DecodingResult]]:
|
|
|
|
|
def decode(
|
|
|
|
|
model: "Whisper",
|
|
|
|
|
mel: paddle.Tensor,
|
|
|
|
|
options: DecodingOptions=DecodingOptions(),
|
|
|
|
|
resource_path=str, ) -> Union[DecodingResult, List[DecodingResult]]:
|
|
|
|
|
"""
|
|
|
|
|
Performs decoding of 30-second audio segment(s), provided as Mel spectrogram(s).
|
|
|
|
|
|
|
|
|
|
@ -1281,7 +1296,7 @@ def decode(model: "Whisper",
|
|
|
|
|
if single:
|
|
|
|
|
mel = mel.unsqueeze(0)
|
|
|
|
|
|
|
|
|
|
result = DecodingTask(model, options).run(mel)
|
|
|
|
|
result = DecodingTask(model, options, resource_path).run(mel)
|
|
|
|
|
|
|
|
|
|
if single:
|
|
|
|
|
result = result[0]
|
|
|
|
|
@ -1407,7 +1422,7 @@ def hann_window(n_fft: int=N_FFT):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
@lru_cache(maxsize=None)
|
|
|
|
|
def mel_filters(device, n_mels: int=N_MELS) -> paddle.Tensor:
|
|
|
|
|
def mel_filters(resource_path: str, n_mels: int=N_MELS) -> paddle.Tensor:
|
|
|
|
|
"""
|
|
|
|
|
load the mel filterbank matrix for projecting STFT into a Mel spectrogram.
|
|
|
|
|
Allows decoupling librosa dependency; saved using:
|
|
|
|
|
@ -1418,14 +1433,13 @@ def mel_filters(device, n_mels: int=N_MELS) -> paddle.Tensor:
|
|
|
|
|
)
|
|
|
|
|
"""
|
|
|
|
|
assert n_mels == 80, f"Unsupported n_mels: {n_mels}"
|
|
|
|
|
with np.load(
|
|
|
|
|
os.path.join(
|
|
|
|
|
os.path.dirname(__file__), "assets", "mel_filters.npz")) as f:
|
|
|
|
|
with np.load(os.path.join(resource_path, "assets", "mel_filters.npz")) as f:
|
|
|
|
|
return paddle.to_tensor(f[f"mel_{n_mels}"])
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def log_mel_spectrogram(audio: Union[str, np.ndarray, paddle.Tensor],
|
|
|
|
|
n_mels: int=N_MELS):
|
|
|
|
|
n_mels: int=N_MELS,
|
|
|
|
|
resource_path: str=None):
|
|
|
|
|
"""
|
|
|
|
|
Compute the log-Mel spectrogram of
|
|
|
|
|
|
|
|
|
|
@ -1454,7 +1468,7 @@ def log_mel_spectrogram(audio: Union[str, np.ndarray, paddle.Tensor],
|
|
|
|
|
|
|
|
|
|
magnitudes = stft[:, :-1].abs()**2
|
|
|
|
|
|
|
|
|
|
filters = mel_filters(audio, n_mels)
|
|
|
|
|
filters = mel_filters(resource_path, n_mels)
|
|
|
|
|
mel_spec = filters @ magnitudes
|
|
|
|
|
mel_spec = paddle.to_tensor(mel_spec.numpy().tolist())
|
|
|
|
|
|
|
|
|
|
|
zxcd
2 years ago
committed by
GitHub