Update app.py

app.py

@@ -21,15 +21,15 @@ def load_audio(file: str, sr: int = SAMPLE_RATE):
 
     return np.frombuffer(out, np.int16).flatten().astype(np.float32) / 32768.0
 if 'processor' not in locals():
- processor = AutoProcessor.from_pretrained("openai/whisper-medium")
- model = AutoModelForSpeechSeq2Seq.from_pretrained("openai/whisper-medium")
+ processor = AutoProcessor.from_pretrained("openai/whisper-small")
+ model = AutoModelForSpeechSeq2Seq.from_pretrained("openai/whisper-small")
 
 
 
 pipe = pipeline('sentiment-analysis')
 #pipe2 =  pipeline(task="image-bestify", model="beihai/GFPGAN-V1.3-whole-image")
 text = st.text_area('entre com algum texto')
-st.title("Hot Dog? Or Not? ")
+st.title("Wav a ser transcrito ")
 
 wav_up = st.file_uploader("Upload a hot dog candidate image",type=['wav'])
 
@@ -52,8 +52,13 @@ if wav_up is not None:
  forced_decoder_ids = processor.get_decoder_prompt_ids(language = None, task = "transcribe")
 
  predicted_ids = model.generate(input_features, forced_decoder_ids = forced_decoder_ids)
- transcription = processor.batch_decode(predicted_ids, skip_special_tokens = True)
+ i=0
+ while i<2
+ 
+  
+  transcription = processor.batch_decode(predicted_ids, skip_special_tokens = True)
  st.json(transcription )
+ i=i+1
 print('3')
 if text:
  out=pipe(text)