Hugging Face's logo

Spaces:
Pranjal12345
/
Whisper_with_FastApi
Runtime error

App Files Community
Pranjal12345 commited on
Commit
5606356
•
1 Parent(s): b7f00c1

Update main.py

Browse files
Files changed (1) hide show
  1. main.py +1 -203
main.py CHANGED
@@ -1,204 +1,3 @@
1
- # # #uvicorn app:app --host 0.0.0.0 --port 8000 --reload
2
-
3
-
4
- # # # from fastapi import FastAPI
5
- # # # from transformers import WhisperProcessor, WhisperForConditionalGeneration
6
- # # # import librosa
7
- # # # import uvicorn
8
-
9
- # # # app = FastAPI()
10
-
11
- # # # processor = WhisperProcessor.from_pretrained("openai/whisper-small")
12
- # # # model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-small")
13
- # # # model.config.forced_decoder_ids = None
14
-
15
- # # # audio_file_path = "output.mp3"
16
-
17
- # # # audio_data, _ = librosa.load(audio_file_path, sr=16000)
18
-
19
- # # # @app.get("/")
20
- # # # def transcribe_audio():
21
- # # # input_features = processor(audio_data.tolist(), return_tensors="pt").input_features
22
-
23
- # # # predicted_ids = model.generate(input_features)
24
- # # # transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
25
-
26
- # # # return {"transcription": transcription[0]}
27
-
28
-
29
- # # # if __name__ == "__main__":
30
- # # # import uvicorn
31
- # # # uvicorn.run(app, host="0.0.0.0", port=8000)
32
-
33
-
34
- # # # if __name__=='__main__':
35
- # # # uvicorn.run('main:app', reload=True)
36
-
37
-
38
-
39
-
40
- # # #uvicorn app:app --host 0.0.0.0 --port 8000 --reload
41
- #
42
-
43
-
44
-
45
-
46
-
47
- # # # from fastapi import FastAPI
48
- # # # from transformers import WhisperProcessor, WhisperForConditionalGeneration
49
- # # # import librosa
50
- # # # import uvicorn
51
-
52
- # # # app = FastAPI()
53
-
54
- # # # # Load model and processor
55
- # # # processor = WhisperProcessor.from_pretrained("openai/whisper-small")
56
- # # # model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-small")
57
- # # # model.config.forced_decoder_ids = None
58
-
59
- # # # # Path to your audio file
60
- # # # audio_file_path = "/home/pranjal/Downloads/output.mp3"
61
-
62
- # # # # Read the audio file
63
- # # # audio_data, _ = librosa.load(audio_file_path, sr=16000)
64
-
65
- # # # @app.get("/")
66
- # # # def transcribe_audio():
67
- # # # # Process the audio data using the Whisper processor
68
- # # # input_features = processor(audio_data.tolist(), return_tensors="pt").input_features
69
-
70
- # # # # Generate transcription
71
- # # # predicted_ids = model.generate(input_features)
72
- # # # transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
73
-
74
- # # # return {"transcription": transcription[0]}
75
-
76
- # # # if __name__ == "__main__":
77
- # # # import uvicorn
78
- # # # uvicorn.run(app, host="0.0.0.0", port=8000)
79
-
80
-
81
- # # # if __name__=='__app__':
82
- # # # uvicorn.run('main:app', reload=True)
83
-
84
-
85
-
86
-
87
-
88
- # # from fastapi import FastAPI, UploadFile, File
89
- # # from transformers import WhisperProcessor, WhisperForConditionalGeneration
90
- # # import librosa
91
- # # from fastapi.responses import HTMLResponse
92
- # # import uvicorn
93
- # # import io
94
-
95
- # # app = FastAPI()
96
-
97
- # # # Load model and processor
98
- # # processor = WhisperProcessor.from_pretrained("openai/whisper-small")
99
- # # model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-small")
100
- # # model.config.forced_decoder_ids = None
101
-
102
- # # @app.get("/")
103
- # # def read_root():
104
- # # html_form = """
105
- # # <html>
106
- # # <body>
107
- # # <h2>ASR Transcription</h2>
108
- # # <form action="/transcribe" method="post" enctype="multipart/form-data">
109
- # # <label for="audio_file">Upload an audio file (MP3 or WAV):</label>
110
- # # <input type="file" id="audio_file" name="audio_file" accept=".mp3, .wav" required><br><br>
111
- # # <input type="submit" value="Transcribe">
112
- # # </form>
113
- # # </body>
114
- # # </html>
115
- # # """
116
- # # return HTMLResponse(content=html_form, status_code=200)
117
-
118
- # # @app.post("/transcribe")
119
- # # async def transcribe_audio(audio_file: UploadFile):
120
- # # try:
121
- # # # Read the uploaded audio file
122
- # # audio_data = await audio_file.read()
123
-
124
- # # # Process the audio data using the Whisper processor
125
- # # audio_data, _ = librosa.load(io.BytesIO(audio_data), sr=16000)
126
- # # input_features = processor(audio_data.tolist(), return_tensors="pt").input_features
127
-
128
- # # # Generate transcription
129
- # # predicted_ids = model.generate(input_features)
130
- # # transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
131
-
132
- # # return {"transcription": transcription[0]}
133
- # # except Exception as e:
134
- # # return {"error": str(e)}
135
-
136
- # # if __name__ == "__app__":
137
- # # uvicorn.run(app, host="0.0.0.0", port=8000, reload=True)
138
-
139
-
140
-
141
-
142
-
143
- # #uvicorn app:app --host 0.0.0.0 --port 8000 --reload
144
-
145
-
146
- # from fastapi import FastAPI, UploadFile, File
147
- # from transformers import WhisperProcessor, WhisperForConditionalGeneration
148
- # import librosa
149
- # from fastapi.responses import HTMLResponse
150
- # import uvicorn
151
- # import io
152
-
153
- # app = FastAPI()
154
-
155
- # # # Load model and processor
156
- # # processor = WhisperProcessor.from_pretrained("openai/whisper-medium")
157
- # # model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-medium")
158
- # # model.config.forced_decoder_ids = None
159
-
160
- # import whisper
161
- # model = whisper.load_model("small")
162
-
163
-
164
- # @app.get("/")
165
- # def read_root():
166
- # html_form = """
167
- # <html>
168
- # <body>
169
- # <h2>ASR Transcription</h2>
170
- # <form action="/transcribe" method="post" enctype="multipart/form-data">
171
- # <label for="audio_file">Upload an audio file (MP3 or WAV):</label>
172
- # <input type="file" id="audio_file" name="audio_file" accept=".mp3, .wav" required><br><br>
173
- # <input type="submit" value="Transcribe">
174
- # </form>
175
- # </body>
176
- # </html>
177
- # """
178
- # return HTMLResponse(content=html_form, status_code=200)
179
-
180
- # @app.post("/transcribe")
181
- # async def transcribe_audio(audio_file: UploadFile):
182
- # try:
183
- # # Read the uploaded audio file
184
- # audio_data = await audio_file.read()
185
-
186
- # # Process the audio data using the Whisper processor
187
- # # audio_data, _ = librosa.load(io.BytesIO(audio_data), sr=16000)
188
- # # input_features = processor(audio_data.tolist(), return_tensors="pt").input_features
189
-
190
- # # # Generate transcription
191
- # # predicted_ids = model.generate(input_features)
192
- # # transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
193
- # result = model.transcribe("/home/pranjal/Downloads/rt.mp3")
194
-
195
- # return {"transcription": result['text']}
196
- # except Exception as e:
197
- # return {"error": str(e)}
198
-
199
- # # if __name__ == "__app__":
200
- # # uvicorn.run(app, host="0.0.0.0", port=8000, reload=True)
201
-
202
  #uvicorn app:app --host 0.0.0.0 --port 8000 --reload
203
 
204
  from fastapi import FastAPI, UploadFile, File
@@ -261,5 +60,4 @@ async def transcribe_audio(audio_file: UploadFile):
261
  final_transcription = " ".join(transcription)
262
  final_transcription = remove_tags(final_transcription)
263
 
264
- return final_transcription
265
-
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  #uvicorn app:app --host 0.0.0.0 --port 8000 --reload
2
 
3
  from fastapi import FastAPI, UploadFile, File
 
60
  final_transcription = " ".join(transcription)
61
  final_transcription = remove_tags(final_transcription)
62
 
63
+ return final_transcription