Update app.py

app.py

@@ -6,7 +6,10 @@ import time
 
 from transformers import WhisperProcessor, WhisperForConditionalGeneration
 from datasets import load_dataset
-
+processor = WhisperProcessor.from_pretrained("openai/whisper-large-v2")
+model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-large-v2")
+model.config.forced_decoder_ids = None
+        
 # load model and processor
 #processor = WhisperProcessor.from_pretrained("openai/whisper-large-v2")
 #model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-large-v2")
@@ -44,9 +47,6 @@ def transcribe(language,audio, state=""):#language="Spanish",
     time.sleep(1)
     if language=="Multi":
         state=""
-        processor = WhisperProcessor.from_pretrained("openai/whisper-large-v2")
-        model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-large-v2")
-        model.config.forced_decoder_ids = None
         result = model.transcribe(audio)
         text = result["text"]#processor.batch_decode(predicted_ids, skip_special_tokens=False)
 
@@ -74,7 +74,7 @@ demo=gr.Interface(
 
     inputs=[
         #gr.Dropdown(["Spanish","Catalan","English", "French", "Japanese"],value="Spanish"),
-        gr.Dropdown(["Multi"],value="Multi"),
+        gr.Dropdown(["Multi","Spanish"],value="Multi"),
         
         #gr.Audio(source="microphone", type="filepath", streaming=True), 
         gr.inputs.Audio(source="microphone", type="filepath"),