Update app.py

app.py

@@ -47,7 +47,7 @@ def inference(audio):
     # Added prompt two 2 lines below
     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(prompt, result.text, return_tensors="pt").input_ids
     print("inputs ", input_ids)
 
     # prompt length
@@ -87,7 +87,7 @@ def inference(audio):
     # print audio data as text
     # print(result.text)
     # prompt
-    getText = generator(prompt, result.text, max_new_tokens=10, num_return_sequences=5)
+    getText = generator(result.text, max_new_tokens=10, num_return_sequences=5)
     # pprint("getText: ", getText)
     # pprint("text.result: ", result.text)
     # result.text