mlx-examples/whisper/mlx_whisper/transcribe.py

541 lines
22 KiB
Python
Raw Normal View History

2023-12-01 03:08:53 +08:00
# Copyright © 2023 Apple Inc.
2023-11-30 00:17:26 +08:00
import sys
import warnings
from typing import List, Optional, Tuple, Union
2023-11-30 00:17:26 +08:00
import mlx.core as mx
import numpy as np
import tqdm
2023-11-30 00:17:26 +08:00
from .audio import (
FRAMES_PER_SECOND,
HOP_LENGTH,
N_FRAMES,
N_SAMPLES,
SAMPLE_RATE,
log_mel_spectrogram,
pad_or_trim,
)
from .decoding import DecodingOptions, DecodingResult
from .load_models import load_model
from .timing import add_word_timestamps
from .tokenizer import LANGUAGES, get_tokenizer
2023-11-30 00:17:26 +08:00
def _format_timestamp(seconds: float):
assert seconds >= 0, "non-negative timestamp expected"
milliseconds = round(seconds * 1000.0)
hours = milliseconds // 3_600_000
milliseconds -= hours * 3_600_000
minutes = milliseconds // 60_000
milliseconds -= minutes * 60_000
seconds = milliseconds // 1_000
milliseconds -= seconds * 1_000
hours_marker = f"{hours:02d}:" if hours > 0 else ""
return f"{hours_marker}{minutes:02d}:{seconds:02d}.{milliseconds:03d}"
def _get_end(segments: List[dict]) -> Optional[float]:
return next(
(w["end"] for s in reversed(segments) for w in reversed(s["words"])),
segments[-1]["end"] if segments else None,
)
2023-11-30 00:17:26 +08:00
class ModelHolder:
model = None
model_path = None
2023-11-30 00:17:26 +08:00
@classmethod
def get_model(cls, model_path: str, dtype: mx.Dtype):
if cls.model is None or model_path != cls.model_path:
cls.model = load_model(model_path, dtype=dtype)
cls.model_path = model_path
2023-11-30 00:17:26 +08:00
return cls.model
def transcribe(
audio: Union[str, np.ndarray, mx.array],
*,
path_or_hf_repo: str = "mlx-community/whisper-tiny",
2023-11-30 00:17:26 +08:00
verbose: Optional[bool] = None,
temperature: Union[float, Tuple[float, ...]] = (0.0, 0.2, 0.4, 0.6, 0.8, 1.0),
compression_ratio_threshold: Optional[float] = 2.4,
logprob_threshold: Optional[float] = -1.0,
no_speech_threshold: Optional[float] = 0.6,
condition_on_previous_text: bool = True,
initial_prompt: Optional[str] = None,
word_timestamps: bool = False,
2023-11-30 00:17:26 +08:00
prepend_punctuations: str = "\"'“¿([{-",
append_punctuations: str = "\"'.。,!?::”)]}、",
clip_timestamps: Union[str, List[float]] = "0",
hallucination_silence_threshold: Optional[float] = None,
2023-11-30 00:17:26 +08:00
**decode_options,
):
"""
Transcribe an audio file using Whisper
Parameters
----------
audio: Union[str, np.ndarray, mx.array]
The path to the audio file to open, or the audio waveform
path_or_hf_repo: str
The localpath to the Whisper model or HF Hub repo with the MLX converted weights.
2023-11-30 00:17:26 +08:00
verbose: bool
Whether to display the text being decoded to the console. If True, displays all the details,
If False, displays minimal details. If None, does not display anything
temperature: Union[float, Tuple[float, ...]]
Temperature for sampling. It can be a tuple of temperatures, which will be successively used
upon failures according to either `compression_ratio_threshold` or `logprob_threshold`.
compression_ratio_threshold: float
If the gzip compression ratio is above this value, treat as failed
logprob_threshold: float
If the average log probability over sampled tokens is below this value, treat as failed
no_speech_threshold: float
If the no_speech probability is higher than this value AND the average log probability
over sampled tokens is below `logprob_threshold`, consider the segment as silent
condition_on_previous_text: bool
if True, the previous output of the model is provided as a prompt for the next window;
disabling may make the text inconsistent across windows, but the model becomes less prone to
getting stuck in a failure loop, such as repetition looping or timestamps going out of sync.
word_timestamps: bool
Extract word-level timestamps using the cross-attention pattern and dynamic time warping,
and include the timestamps for each word in each segment.
prepend_punctuations: str
If word_timestamps is True, merge these punctuation symbols with the next word
append_punctuations: str
If word_timestamps is True, merge these punctuation symbols with the previous word
2023-11-30 00:17:26 +08:00
initial_prompt: Optional[str]
Optional text to provide as a prompt for the first window. This can be used to provide, or
"prompt-engineer" a context for transcription, e.g. custom vocabularies or proper nouns
to make it more likely to predict those word correctly.
decode_options: dict
Keyword arguments to construct `DecodingOptions` instances
clip_timestamps: Union[str, List[float]]
Comma-separated list start,end,start,end,... timestamps (in seconds) of clips to process.
The last end timestamp defaults to the end of the file.
hallucination_silence_threshold: Optional[float]
When word_timestamps is True, skip silent periods longer than this threshold (in seconds)
when a possible hallucination is detected
2023-11-30 00:17:26 +08:00
Returns
-------
A dictionary containing the resulting text ("text") and segment-level details ("segments"), and
the spoken language ("language"), which is detected when `decode_options["language"]` is None.
"""
2023-12-12 23:37:35 +08:00
dtype = mx.float16 if decode_options.get("fp16", True) else mx.float32
model = ModelHolder.get_model(path_or_hf_repo, dtype)
2023-11-30 00:17:26 +08:00
# Pad 30-seconds of silence to the input audio, for slicing
2023-12-13 00:26:52 +08:00
mel = log_mel_spectrogram(audio, n_mels=model.dims.n_mels, padding=N_SAMPLES)
2023-11-30 00:17:26 +08:00
content_frames = mel.shape[-2] - N_FRAMES
content_duration = float(content_frames * HOP_LENGTH / SAMPLE_RATE)
2023-11-30 00:17:26 +08:00
if verbose:
system_encoding = sys.getdefaultencoding()
if system_encoding != "utf-8":
make_safe = lambda x: x.encode(system_encoding, errors="replace").decode(
system_encoding
)
else:
make_safe = lambda x: x
if decode_options.get("language", None) is None:
if not model.is_multilingual:
decode_options["language"] = "en"
else:
if verbose:
print(
"Detecting language using up to the first 30 seconds. "
"Use the `language` decoding option to specify the language"
)
mel_segment = pad_or_trim(mel, N_FRAMES, axis=-2).astype(dtype)
_, probs = model.detect_language(mel_segment)
decode_options["language"] = max(probs, key=probs.get)
if verbose is not None:
print(
f"Detected language: {LANGUAGES[decode_options['language']].title()}"
)
language: str = decode_options["language"]
task: str = decode_options.get("task", "transcribe")
2023-12-13 00:26:52 +08:00
tokenizer = get_tokenizer(
model.is_multilingual,
num_languages=model.num_languages,
language=language,
task=task,
)
2023-11-30 00:17:26 +08:00
if isinstance(clip_timestamps, str):
clip_timestamps = [
float(ts) for ts in (clip_timestamps.split(",") if clip_timestamps else [])
]
seek_points: List[int] = [round(ts * FRAMES_PER_SECOND) for ts in clip_timestamps]
if len(seek_points) == 0:
seek_points.append(0)
if len(seek_points) % 2 == 1:
seek_points.append(content_frames)
seek_clips: List[Tuple[int, int]] = list(zip(seek_points[::2], seek_points[1::2]))
punctuation = "\"'“¿([{-\"'.。,!?::”)]}、"
if word_timestamps and task == "translate":
warnings.warn("Word-level timestamps on translations may not be reliable.")
2023-11-30 00:17:26 +08:00
def decode_with_fallback(segment: mx.array) -> DecodingResult:
temperatures = (
[temperature] if isinstance(temperature, (int, float)) else temperature
)
decode_result = None
for t in temperatures:
kwargs = {**decode_options}
if t > 0:
# disable beam_size and patience when t > 0
kwargs.pop("beam_size", None)
kwargs.pop("patience", None)
else:
# disable best_of when t == 0
kwargs.pop("best_of", None)
options = DecodingOptions(**kwargs, temperature=t)
decode_result = model.decode(segment, options)
needs_fallback = False
if (
compression_ratio_threshold is not None
and decode_result.compression_ratio > compression_ratio_threshold
):
needs_fallback = True # too repetitive
if (
logprob_threshold is not None
and decode_result.avg_logprob < logprob_threshold
):
needs_fallback = True # average log probability is too low
if (
no_speech_threshold is not None
and decode_result.no_speech_prob > no_speech_threshold
):
needs_fallback = False # silence
if not needs_fallback:
break
return decode_result
clip_idx = 0
seek = seek_clips[clip_idx][0]
2023-11-30 00:17:26 +08:00
input_stride = N_FRAMES // model.dims.n_audio_ctx # mel frames per output token: 2
time_precision = (
input_stride * HOP_LENGTH / SAMPLE_RATE
) # time per output token: 0.02 (seconds)
all_tokens = []
all_segments = []
prompt_reset_since = 0
if initial_prompt is not None:
initial_prompt_tokens = tokenizer.encode(" " + initial_prompt.strip())
all_tokens.extend(initial_prompt_tokens)
else:
initial_prompt_tokens = []
def new_segment(
*, start: float, end: float, tokens: mx.array, result: DecodingResult
):
tokens = tokens.tolist()
text_tokens = [token for token in tokens if token < tokenizer.eot]
return {
"seek": seek,
"start": start,
"end": end,
"text": tokenizer.decode(text_tokens),
"tokens": tokens,
"temperature": result.temperature,
"avg_logprob": result.avg_logprob,
"compression_ratio": result.compression_ratio,
"no_speech_prob": result.no_speech_prob,
}
# show the progress bar when verbose is False (if True, transcribed text will be printed)
with tqdm.tqdm(
total=content_frames, unit="frames", disable=verbose is not False
) as pbar:
last_speech_timestamp = 0.0
2024-03-31 04:13:58 +08:00
for seek_clip_start, seek_clip_end in seek_clips:
while seek < seek_clip_end:
time_offset = float(seek * HOP_LENGTH / SAMPLE_RATE)
window_end_time = float((seek + N_FRAMES) * HOP_LENGTH / SAMPLE_RATE)
segment_size = min(
N_FRAMES, content_frames - seek, seek_clip_end - seek
)
mel_segment = mel[seek : seek + segment_size]
segment_duration = segment_size * HOP_LENGTH / SAMPLE_RATE
mel_segment = pad_or_trim(mel_segment, N_FRAMES, axis=-2).astype(dtype)
decode_options["prompt"] = all_tokens[prompt_reset_since:]
result: DecodingResult = decode_with_fallback(mel_segment)
tokens = np.array(result.tokens)
if no_speech_threshold is not None:
# no voice activity check
should_skip = result.no_speech_prob > no_speech_threshold
if (
logprob_threshold is not None
and result.avg_logprob > logprob_threshold
):
# don't skip if the logprob is high enough, despite the no_speech_prob
should_skip = False
if should_skip:
seek += (
segment_size # fast-forward to the next segment boundary
)
continue
previous_seek = seek
current_segments = []
# anomalous words are very long/short/improbable
def word_anomaly_score(word: dict) -> float:
probability = word.get("probability", 0.0)
duration = word["end"] - word["start"]
score = 0.0
if probability < 0.15:
score += 1.0
if duration < 0.133:
score += (0.133 - duration) * 15
if duration > 2.0:
score += duration - 2.0
return score
def is_segment_anomaly(segment: Optional[dict]) -> bool:
if segment is None or not segment["words"]:
return False
words = [
w for w in segment["words"] if w["word"] not in punctuation
]
words = words[:8]
score = sum(word_anomaly_score(w) for w in words)
return score >= 3 or score + 0.01 >= len(words)
def next_words_segment(segments: List[dict]) -> Optional[dict]:
return next((s for s in segments if s["words"]), None)
timestamp_tokens = tokens >= tokenizer.timestamp_begin
single_timestamp_ending = timestamp_tokens[-2:].tolist() == [
False,
True,
]
consecutive = np.where(
np.logical_and(timestamp_tokens[:-1], timestamp_tokens[1:])
)[0]
consecutive += 1
if len(consecutive) > 0:
# if the output contains two consecutive timestamp tokens
slices = consecutive.tolist()
if single_timestamp_ending:
slices.append(len(tokens))
last_slice = 0
for current_slice in slices:
sliced_tokens = tokens[last_slice:current_slice]
start_timestamp_pos = (
sliced_tokens[0].item() - tokenizer.timestamp_begin
)
end_timestamp_pos = (
sliced_tokens[-1].item() - tokenizer.timestamp_begin
)
current_segments.append(
new_segment(
start=time_offset
+ start_timestamp_pos * time_precision,
end=time_offset + end_timestamp_pos * time_precision,
tokens=sliced_tokens,
result=result,
)
)
last_slice = current_slice
if single_timestamp_ending:
# single timestamp at the end means no speech after the last timestamp.
seek += segment_size
else:
# otherwise, ignore the unfinished segment and seek to the last timestamp
last_timestamp_pos = (
tokens[last_slice - 1].item() - tokenizer.timestamp_begin
)
seek += last_timestamp_pos * input_stride
else:
duration = segment_duration
timestamps = tokens[timestamp_tokens.nonzero()[0]]
if (
len(timestamps) > 0
and timestamps[-1].item() != tokenizer.timestamp_begin
):
# no consecutive timestamps but it has a timestamp; use the last one.
last_timestamp_pos = (
timestamps[-1].item() - tokenizer.timestamp_begin
)
duration = last_timestamp_pos * time_precision
2023-11-30 00:17:26 +08:00
current_segments.append(
new_segment(
2024-03-31 04:13:58 +08:00
start=time_offset,
end=time_offset + duration,
tokens=tokens,
2023-11-30 00:17:26 +08:00
result=result,
)
)
seek += segment_size
2024-03-31 04:13:58 +08:00
if word_timestamps:
add_word_timestamps(
segments=current_segments,
model=model,
tokenizer=tokenizer,
mel=mel_segment,
num_frames=segment_size,
prepend_punctuations=prepend_punctuations,
append_punctuations=append_punctuations,
last_speech_timestamp=last_speech_timestamp,
)
if not single_timestamp_ending:
last_word_end = _get_end(current_segments)
if last_word_end is not None and last_word_end > time_offset:
2024-03-31 04:13:58 +08:00
seek = round(last_word_end * FRAMES_PER_SECOND)
# skip silence before possible hallucinations
if hallucination_silence_threshold is not None:
threshold = hallucination_silence_threshold
if not single_timestamp_ending:
last_word_end = _get_end(current_segments)
if (
last_word_end is not None
and last_word_end > time_offset
):
remaining_duration = window_end_time - last_word_end
if remaining_duration > threshold:
seek = round(last_word_end * FRAMES_PER_SECOND)
else:
seek = previous_seek + segment_size
# if first segment might be a hallucination, skip leading silence
first_segment = next_words_segment(current_segments)
if first_segment is not None and is_segment_anomaly(
first_segment
):
gap = first_segment["start"] - time_offset
if gap > threshold:
seek = previous_seek + round(gap * FRAMES_PER_SECOND)
continue
# skip silence before any possible hallucination that is surrounded
# by silence or more hallucinations
hal_last_end = last_speech_timestamp
for si in range(len(current_segments)):
segment = current_segments[si]
if not segment["words"]:
continue
if is_segment_anomaly(segment):
next_segment = next_words_segment(
current_segments[si + 1 :]
)
2024-03-31 04:13:58 +08:00
if next_segment is not None:
hal_next_start = next_segment["words"][0]["start"]
else:
hal_next_start = time_offset + segment_duration
silence_before = (
segment["start"] - hal_last_end > threshold
or segment["start"] < threshold
or segment["start"] - time_offset < 2.0
)
silence_after = (
hal_next_start - segment["end"] > threshold
or is_segment_anomaly(next_segment)
or window_end_time - segment["end"] < 2.0
)
if silence_before and silence_after:
seek = round(
max(time_offset + 1, segment["start"])
* FRAMES_PER_SECOND
)
if content_duration - segment["end"] < threshold:
seek = content_frames
current_segments[si:] = []
break
hal_last_end = segment["end"]
2024-03-31 04:13:58 +08:00
last_word_end = _get_end(current_segments)
if last_word_end is not None:
last_speech_timestamp = last_word_end
if verbose:
for segment in current_segments:
start, end, text = (
segment["start"],
segment["end"],
segment["text"],
)
line = f"[{_format_timestamp(start)} --> {_format_timestamp(end)}] {text}"
print(make_safe(line))
# if a segment is instantaneous or does not contain text, clear it
for i, segment in enumerate(current_segments):
if (
segment["start"] == segment["end"]
or segment["text"].strip() == ""
):
segment["text"] = ""
segment["tokens"] = []
segment["words"] = []
all_segments.extend(
[
{"id": i, **segment}
for i, segment in enumerate(
current_segments, start=len(all_segments)
)
]
)
all_tokens.extend(
[
token
for segment in current_segments
for token in segment["tokens"]
]
)
2023-11-30 00:17:26 +08:00
2024-03-31 04:13:58 +08:00
if not condition_on_previous_text or result.temperature > 0.5:
# do not feed the prompt tokens if a high temperature was used
prompt_reset_since = len(all_tokens)
2023-11-30 00:17:26 +08:00
2024-03-31 04:13:58 +08:00
# update progress bar
pbar.update(min(content_frames, seek) - previous_seek)
2023-11-30 00:17:26 +08:00
return dict(
text=tokenizer.decode(all_tokens[len(initial_prompt_tokens) :]),
segments=all_segments,
language=language,
)