Update app.py

app.py

@@ -26,7 +26,7 @@ feat_ext = WhisperFeatureExtractor.from_pretrained("rohitp1/kkkh_whisper_small_d
 
 p = pipeline('automatic-speech-recognition', model=model, tokenizer=tokenizer, feature_extractor=feat_ext)
 
-def transcribe(mic_input, upl_input):
+def transcribe(mic_input, upl_input, model_type):
     if mic_input:
         audio = mic_input
     else:
@@ -34,7 +34,7 @@ def transcribe(mic_input, upl_input):
     time.sleep(3)
     text = p(audio)["text"]
     # state = text + " "
-    return text
+    return text+" "+model_type
 
 
 
@@ -62,7 +62,7 @@ def transcribe(mic_input, upl_input):
 # demo.launch()
 
 def clear_inputs_and_outputs():
-    return [None, None, None]
+    return [None, None, None, None]
 
 # Main function
 if __name__ == "__main__":
@@ -84,10 +84,14 @@ if __name__ == "__main__":
                     source="upload", type="filepath", label="Upload a wav file"
                 )
 
+                with gr.Row():
+                    model_type = gr.inputs.Dropdown("gpt2", "distilgpt2"], type="text", label='Model Type')
+
                 with gr.Row():
                     clr_btn = gr.Button(value="Clear", variant="secondary")
                     prd_btn = gr.Button(value="Predict")
 
+
             # Outputs
             with gr.Column():
                 lbl_output = gr.Label(label="Top Predictions")
@@ -111,11 +115,11 @@ if __name__ == "__main__":
         clr_btn.click(
             fn=clear_inputs_and_outputs,
             inputs=[],
-            outputs=[mic_input, upl_input, lbl_output],
+            outputs=[mic_input, upl_input, model_type, lbl_output],
         )
         prd_btn.click(
             fn=transcribe,
-            inputs=[mic_input, upl_input],
+            inputs=[mic_input, upl_input, model_type],
             outputs=[lbl_output],
         )