chrissewell/whisperX
1
0
Fork 0
mirror of https://github.com/m-bain/whisperX.git synced 2025-07-01 18:17:27 -04:00
Code Issues Packages Projects Releases Wiki Activity

change regex sub to align_dictionary sub

Browse Source 
This allows the alignment method to be used for other languages -- depending on what characters are in the align_model dictionary.
This commit is contained in:
m-bain
2022-12-17 15:04:08 +00:00
committed by GitHub
parent 938341c05a
commit b47bb769a0
1 changed files with 1 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
whisperx/transcribe.py
View File

@ -7,7 +7,6 @@ import numpy as np
import torch import torch
import torchaudio import torchaudio
import tqdm import tqdm
import re
from .audio import SAMPLE_RATE, N_FRAMES, HOP_LENGTH, pad_or_trim, log_mel_spectrogram, load_audio from .audio import SAMPLE_RATE, N_FRAMES, HOP_LENGTH, pad_or_trim, log_mel_spectrogram, load_audio
from .alignment import get_trellis, backtrack, merge_repeats, merge_words from .alignment import get_trellis, backtrack, merge_repeats, merge_words
from .decoding import DecodingOptions, DecodingResult from .decoding import DecodingOptions, DecodingResult
@ -256,7 +255,6 @@ def align(
device: str, device: str,
extend_duration: float = 0.0, extend_duration: float = 0.0,
start_from_previous: bool = True, start_from_previous: bool = True,
resolution: str = "phrase",
): ):
print("Performing alignment...") print("Performing alignment...")
if not torch.is_tensor(audio): if not torch.is_tensor(audio):
@ -287,7 +285,7 @@ def align(
transcription = segment['text'].strip() transcription = segment['text'].strip()
t_words = transcription.split(' ') t_words = transcription.split(' ')
t_words_clean = [re.sub(r"[^a-zA-Z' ]", "", x) for x in t_words] t_words_clean = [''.join([w for w in word if w in model_dictionary.keys()]) for word in t_words]
t_words_nonempty = [x for x in t_words_clean if x != ""] t_words_nonempty = [x for x in t_words_clean if x != ""]
t_words_nonempty_idx = [x for x in range(len(t_words_clean)) if t_words_clean[x] != ""] t_words_nonempty_idx = [x for x in range(len(t_words_clean)) if t_words_clean[x] != ""]