Update app.py

app.py

@@ -1,10 +1,9 @@
 import torch
 import torchaudio
 from torchaudio.transforms import Resample
-#from transformers import Speech2TextProcessor, Speech2TextForConditionalGeneration
+from transformers import Speech2TextProcessor, Speech2TextForConditionalGeneration
 import streamlit as st
 from audio_recorder_streamlit import audio_recorder
-from transformers import AutoProcessor, AutoModelForPreTraining
 
 def preprocess_audio(audio_bytes, sample_rate=16000):
     # Load audio and convert to mono if necessary
@@ -20,11 +19,11 @@ def preprocess_audio(audio_bytes, sample_rate=16000):
     return waveform
 
 def transcribe_audio(audio_bytes):
-    #model = Speech2TextForConditionalGeneration.from_pretrained("facebook/wav2vec2-large-100k-voxpopuli")
-    #processor = Speech2TextProcessor.from_pretrained("facebook/wav2vec2-large-100k-voxpopuli")
+    model = Speech2TextForConditionalGeneration.from_pretrained("facebook/s2t-small-librispeech-asr")
+    processor = Speech2TextProcessor.from_pretrained("facebook/s2t-small-librispeech-asr")
     # Load model directly
-    processor = AutoProcessor.from_pretrained("facebook/wav2vec2-large-100k-voxpopuli")
-    model = AutoModelForPreTraining.from_pretrained("facebook/wav2vec2-large-100k-voxpopuli")
+    #processor = AutoProcessor.from_pretrained("facebook/wav2vec2-large-100k-voxpopuli")
+    #model = AutoModelForPreTraining.from_pretrained("facebook/wav2vec2-large-100k-voxpopuli")
     # Preprocess audio
     input_features = preprocess_audio(audio_bytes)