Spaces:
Build error
Build error
Update stt.py
Browse files
stt.py
CHANGED
@@ -1,8 +1,13 @@
|
|
|
|
1 |
import os
|
2 |
import torch
|
3 |
import torchaudio
|
4 |
import spaces
|
|
|
|
|
|
|
5 |
from transformers import WhisperProcessor, WhisperForConditionalGeneration
|
|
|
6 |
|
7 |
# Create directories
|
8 |
os.makedirs("transcriptions", exist_ok=True)
|
@@ -20,63 +25,79 @@ WHISPER_MODEL_SIZES = {
|
|
20 |
'large': 'openai/whisper-large-v3',
|
21 |
}
|
22 |
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
|
|
|
|
|
|
|
|
29 |
# Get model identifier
|
30 |
-
model_id = WHISPER_MODEL_SIZES.get(model_size.lower(), WHISPER_MODEL_SIZES['base'])
|
31 |
|
32 |
-
# Load model and processor
|
33 |
if whisper_model is None or whisper_processor is None or (whisper_model and whisper_model.config._name_or_path != model_id):
|
34 |
-
print(f"Loading Whisper {model_size} model...")
|
35 |
whisper_processor = WhisperProcessor.from_pretrained(model_id)
|
36 |
whisper_model = WhisperForConditionalGeneration.from_pretrained(model_id)
|
37 |
print(f"Model loaded on device: {whisper_model.device}")
|
|
|
|
|
|
|
|
|
|
|
38 |
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
if speech_array.shape[0] > 1:
|
44 |
-
speech_array = torch.mean(speech_array, dim=0, keepdim=True)
|
45 |
-
|
46 |
-
# Resample to 16kHz if needed
|
47 |
-
if sample_rate != 16000:
|
48 |
-
resampler = torchaudio.transforms.Resample(sample_rate, 16000)
|
49 |
-
speech_array = resampler(speech_array)
|
50 |
-
|
51 |
-
# Prepare inputs for the model
|
52 |
-
input_features = whisper_processor(
|
53 |
-
speech_array.squeeze().numpy(),
|
54 |
-
sampling_rate=16000,
|
55 |
-
return_tensors="pt"
|
56 |
-
).input_features
|
57 |
-
|
58 |
-
# Generate transcription
|
59 |
-
generation_kwargs = {}
|
60 |
-
|
61 |
-
if language:
|
62 |
-
forced_decoder_ids = whisper_processor.get_decoder_prompt_ids(language=language, task="transcribe")
|
63 |
-
generation_kwargs["forced_decoder_ids"] = forced_decoder_ids
|
64 |
-
|
65 |
-
# Run the model
|
66 |
-
with torch.no_grad():
|
67 |
-
predicted_ids = whisper_model.generate(input_features, **generation_kwargs)
|
68 |
|
69 |
-
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
-
|
74 |
-
|
75 |
-
|
76 |
-
|
77 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
78 |
|
79 |
-
#
|
80 |
async def transcribe_audio(audio_file_path, model_size="base", language="en"):
|
81 |
-
|
82 |
-
|
|
|
|
|
|
|
|
1 |
+
# stt.py
|
2 |
import os
|
3 |
import torch
|
4 |
import torchaudio
|
5 |
import spaces
|
6 |
+
import numpy as np
|
7 |
+
from typing import Tuple
|
8 |
+
from numpy.typing import NDArray
|
9 |
from transformers import WhisperProcessor, WhisperForConditionalGeneration
|
10 |
+
import tempfile
|
11 |
|
12 |
# Create directories
|
13 |
os.makedirs("transcriptions", exist_ok=True)
|
|
|
25 |
'large': 'openai/whisper-large-v3',
|
26 |
}
|
27 |
|
28 |
+
class WhisperSTTModel:
|
29 |
+
def __init__(self, model_size="base", language="en"):
|
30 |
+
self.model_size = model_size
|
31 |
+
self.language = language
|
32 |
+
self._initialize_model()
|
33 |
+
|
34 |
+
@spaces.GPU
|
35 |
+
def _initialize_model(self):
|
36 |
+
global whisper_model, whisper_processor
|
37 |
+
|
38 |
# Get model identifier
|
39 |
+
model_id = WHISPER_MODEL_SIZES.get(self.model_size.lower(), WHISPER_MODEL_SIZES['base'])
|
40 |
|
41 |
+
# Load model and processor if not already loaded
|
42 |
if whisper_model is None or whisper_processor is None or (whisper_model and whisper_model.config._name_or_path != model_id):
|
43 |
+
print(f"Loading Whisper {self.model_size} model...")
|
44 |
whisper_processor = WhisperProcessor.from_pretrained(model_id)
|
45 |
whisper_model = WhisperForConditionalGeneration.from_pretrained(model_id)
|
46 |
print(f"Model loaded on device: {whisper_model.device}")
|
47 |
+
|
48 |
+
@spaces.GPU
|
49 |
+
def stt(self, audio: Tuple[int, NDArray[np.float32]]) -> str:
|
50 |
+
"""Transcribe audio to text following the STTModel protocol"""
|
51 |
+
sample_rate, audio_array = audio
|
52 |
|
53 |
+
try:
|
54 |
+
# Convert to mono if needed
|
55 |
+
if len(audio_array.shape) > 1 and audio_array.shape[0] > 1:
|
56 |
+
audio_array = np.mean(audio_array, axis=0)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
|
58 |
+
# Convert numpy array to torch tensor
|
59 |
+
speech_array = torch.tensor(audio_array).unsqueeze(0)
|
60 |
+
|
61 |
+
# Resample to 16kHz if needed
|
62 |
+
if sample_rate != 16000:
|
63 |
+
resampler = torchaudio.transforms.Resample(sample_rate, 16000)
|
64 |
+
speech_array = resampler(speech_array)
|
65 |
+
|
66 |
+
# Prepare inputs for the model
|
67 |
+
input_features = whisper_processor(
|
68 |
+
speech_array.squeeze().numpy(),
|
69 |
+
sampling_rate=16000,
|
70 |
+
return_tensors="pt"
|
71 |
+
).input_features
|
72 |
+
|
73 |
+
# Generate transcription
|
74 |
+
generation_kwargs = {}
|
75 |
+
|
76 |
+
if self.language:
|
77 |
+
forced_decoder_ids = whisper_processor.get_decoder_prompt_ids(language=self.language, task="transcribe")
|
78 |
+
generation_kwargs["forced_decoder_ids"] = forced_decoder_ids
|
79 |
+
|
80 |
+
# Run the model
|
81 |
+
with torch.no_grad():
|
82 |
+
predicted_ids = whisper_model.generate(input_features, **generation_kwargs)
|
83 |
+
|
84 |
+
# Decode the output
|
85 |
+
transcription = whisper_processor.batch_decode(predicted_ids, skip_special_tokens=True)
|
86 |
+
|
87 |
+
# Return the transcribed text
|
88 |
+
return transcription[0]
|
89 |
+
|
90 |
+
except Exception as e:
|
91 |
+
print(f"Error during transcription: {str(e)}")
|
92 |
+
return ""
|
93 |
+
|
94 |
+
# Create a singleton instance for easy import
|
95 |
+
whisper_stt = WhisperSTTModel(model_size="base", language="en")
|
96 |
|
97 |
+
# Legacy function for backward compatibility
|
98 |
async def transcribe_audio(audio_file_path, model_size="base", language="en"):
|
99 |
+
"""For compatibility with older code"""
|
100 |
+
# Load audio from file
|
101 |
+
speech_array, sample_rate = torchaudio.load(audio_file_path)
|
102 |
+
# Use the new model to transcribe
|
103 |
+
return whisper_stt.stt((sample_rate, speech_array.squeeze().numpy()))
|