Update app.py

app.py

@@ -45,9 +45,9 @@ def inference(audio):
     result = whisper.decode(model, mel, options)
     
     # Added prompt below
-    prompt = "The following is a transcript of someone talking, please predict what they will say next. \n"
+    input_prompt = "The following is a transcript of someone talking, please predict what they will say next. \n"
     ### code 
-    input_ids = tokenizer(result.text, return_tensors="pt").input_ids
+    input_ids = tokenizer(input_prompt, result.text, return_tensors="pt").input_ids
     print("inputs ", input_ids)
 
     # prompt length