update transcribe inputs

app.py

@@ -28,9 +28,7 @@ pipe = pipeline(
 )
 
 
-def generate(user_prompt, temperature=0.1, max_new_tokens=128, top_p=0.95, repetition_penalty=1.0):
-    global instruct_history
-
+def generate(user_prompt, instruct_history, temperature=0.1, max_new_tokens=128, top_p=0.95, repetition_penalty=1.0):
     temperature = float(temperature)
     if temperature < 1e-2:
         temperature = 1e-2
@@ -54,21 +52,20 @@ def generate(user_prompt, temperature=0.1, max_new_tokens=128, top_p=0.95, repet
 
 
 @spaces.GPU(duration=60)
-def transcribe(audio):
-    global instruct_history, formatted_history
+def transcribe(audio, instruct_history=instruct_history, formatted_history=formatted_history):
     sr, y = audio
     y = y.astype(np.float32)
     y /= np.max(np.abs(y))
 
     inputs = pipe({"sampling_rate": sr, "raw": y})["text"]
 
-    formatted_history += f"""Human: {inputs}\n"""
+    formatted_history += f"""Human: {inputs}\n\n"""
 
-    llm_response = generate(inputs)
+    llm_response = generate(inputs, instruct_history)
 
     instruct_history += f""" {llm_response}</s>"""
 
-    formatted_history += f"""Friday: {llm_response}\n"""
+    formatted_history += f"""Friday: {llm_response}\n\n"""
 
     audio_response = gTTS(llm_response)
     audio_response.save("response.mp3")