Spaces:
Runtime error
Runtime error
Commit
·
172d443
1
Parent(s):
3c9ba70
Update main.py
Browse files
main.py
CHANGED
@@ -1,26 +1,5 @@
|
|
1 |
#uvicorn app:app --host 0.0.0.0 --port 8000 --reload
|
2 |
|
3 |
-
# from fastapi import FastAPI
|
4 |
-
# from transformers import pipeline
|
5 |
-
|
6 |
-
# pipe = pipeline("automatic-speech-recognition", model="Pranjal12345/whisper-small-ne-pranjal")
|
7 |
-
|
8 |
-
# audio_path = "/home/pranjal/Downloads/chinese_audio.mp3"
|
9 |
-
|
10 |
-
# with open("/home/pranjal/Downloads/chinese_audio.mp3", "rb") as audio_file:
|
11 |
-
# audio_data = audio_file.read()
|
12 |
-
|
13 |
-
# app = FastAPI()
|
14 |
-
|
15 |
-
# @app.get("/")
|
16 |
-
# def hello():
|
17 |
-
# output = pipe(input)
|
18 |
-
# return {"Output": output}
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
|
25 |
from fastapi import FastAPI
|
26 |
from transformers import WhisperProcessor, WhisperForConditionalGeneration
|
@@ -28,26 +7,19 @@ import librosa
|
|
28 |
|
29 |
app = FastAPI()
|
30 |
|
31 |
-
# Load model and processor
|
32 |
processor = WhisperProcessor.from_pretrained("openai/whisper-small")
|
33 |
model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-small")
|
34 |
model.config.forced_decoder_ids = None
|
35 |
|
36 |
-
# Path to your audio file
|
37 |
audio_file_path = "output.mp3"
|
38 |
|
39 |
-
# Read the audio file
|
40 |
audio_data, _ = librosa.load(audio_file_path, sr=16000)
|
41 |
|
42 |
@app.get("/")
|
43 |
def transcribe_audio():
|
44 |
-
# Process the audio data using the Whisper processor
|
45 |
input_features = processor(audio_data.tolist(), return_tensors="pt").input_features
|
46 |
|
47 |
-
# Generate transcription
|
48 |
predicted_ids = model.generate(input_features)
|
49 |
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
|
50 |
|
51 |
-
return {"transcription": transcription[0]}
|
52 |
-
|
53 |
-
|
|
|
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
|
|
|
7 |
|
8 |
app = FastAPI()
|
9 |
|
|
|
10 |
processor = WhisperProcessor.from_pretrained("openai/whisper-small")
|
11 |
model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-small")
|
12 |
model.config.forced_decoder_ids = None
|
13 |
|
|
|
14 |
audio_file_path = "output.mp3"
|
15 |
|
|
|
16 |
audio_data, _ = librosa.load(audio_file_path, sr=16000)
|
17 |
|
18 |
@app.get("/")
|
19 |
def transcribe_audio():
|
|
|
20 |
input_features = processor(audio_data.tolist(), return_tensors="pt").input_features
|
21 |
|
|
|
22 |
predicted_ids = model.generate(input_features)
|
23 |
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
|
24 |
|
25 |
+
return {"transcription": transcription[0]}
|
|
|
|