Update app.py

app.py

@@ -54,9 +54,6 @@ import numpy as np
 # Load model directly
 from transformers import AutoProcessor, AutoModelForPreTraining
 
-processor = AutoProcessor.from_pretrained("facebook/wav2vec2-base")
-model = AutoModelForPreTraining.from_pretrained("facebook/wav2vec2-base")
-
 def transcribe_audio(audio_bytes):
  #   processor = AutoProcessor.from_pretrained("facebook/s2t-small-librispeech-asr")
   #  model = AutoModelForSpeechSeq2Seq.from_pretrained("facebook/s2t-small-librispeech-asr")
@@ -64,20 +61,14 @@ def transcribe_audio(audio_bytes):
     model = AutoModelForPreTraining.from_pretrained("facebook/wav2vec2-base")
     # Convert audio bytes to numpy array
     audio_array = np.frombuffer(audio_bytes, dtype=np.int16)
-
     # Normalize audio array
     audio_tensor = torch.tensor(audio_array, dtype=torch.float64) / 32768.0
-
     # Provide inputs to the processor
-    #inputs = processor(audio=audio_tensor, sampling_rate=16000, return_tensors="pt")
     input_features = processor(audio_tensor, sampling_rate=16000, return_tensors="pt").input_features
-
    # generate token ids
     predicted_ids = model.generate(input_features)
     # decode token ids to text
     transcription = processor.batch_decode(predicted_ids, skip_special_tokens=False)
-
-    transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
     return transcription
 
 # Streamlit app