Switch to gradio API call

app.py

@@ -11,28 +11,18 @@ title="Whisper to Emotion"
 
 device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
 
-whisper_model = pipeline(
-    task="automatic-speech-recognition",
-    model="openai/whisper-large-v2",
-    chunk_length_s=30,
-    device=device,
-)
-
-all_special_ids = whisper_model.tokenizer.all_special_ids
-transcribe_token_id = all_special_ids[-5]
-translate_token_id = all_special_ids[-6]
+whisper = gr.Interface.load(name="spaces/sanchit-gandhi/whisper-large-v2")
 
 emotion_classifier = pipeline("text-classification",model='bhadresh-savani/distilbert-base-uncased-emotion')
 
 def translate_and_classify(audio):
-    task = "Transcribe in Spoken Language"
-    whisper_model.model.config.forced_decoder_ids = [[2, transcribe_token_id if task=="Transcribe in Spoken Language" else translate_token_id]]
-    text = whisper_model(audio)["text"]
     
-    emotion = emotion_classifier(text)
+    text_result = whisper(audio, None, "transcribe", fn_index=0)
+    
+    emotion = emotion_classifier(text_result)
     detected_emotion = emotion[0]["label"]
     print("Detected Emotion: ", detected_emotion)
-    return text, detected_emotion
+    return text_result, detected_emotion
 
 css = """
         .gradio-container {