claude suggestions

app.py

@@ -1,5 +1,6 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
+from typing import Optional
 
 interviewer_system_prompt = """
 # Task
@@ -114,7 +115,7 @@ with gr.Blocks() as demo:
     def user_submit(user_message, history):
         return "", history + [{"role": "user", "content": user_message}]
     
-    def bot_respond(history, system_msg, max_tok, temp, top_p_val, hf_token):
+    def bot_respond(history, system_msg, max_tok, temp, top_p_val, hf_token: Optional[gr.OAuthToken] = None):
         if not history or history[-1]["role"] != "user":
             return history
         
@@ -126,7 +127,7 @@ with gr.Blocks() as demo:
             bot_response = response
             yield history + [{"role": "assistant", "content": bot_response}]
     
-    def handle_auto_respond(history, max_tok, temp, top_p_val, hf_token):
+    def handle_auto_respond(history, max_tok, temp, top_p_val, hf_token: Optional[gr.OAuthToken] = None):
         if not history:
             return history
         
@@ -137,7 +138,7 @@ with gr.Blocks() as demo:
         updated_history = history + [{"role": "user", "content": auto_response}]
         return updated_history
     
-    def bot_respond_after_auto(history, system_msg, max_tok, temp, top_p_val, hf_token):
+    def bot_respond_after_auto(history, system_msg, max_tok, temp, top_p_val, hf_token: Optional[gr.OAuthToken] = None):
         # This is called after auto-respond to get the interviewer's response
         if not history or history[-1]["role"] != "user":
             return history
@@ -158,7 +159,7 @@ with gr.Blocks() as demo:
         queue=False
     ).then(
         bot_respond,
-        [chatbot, system_prompt, max_tokens, temperature, top_p, gr.OAuthToken()],
+        [chatbot, system_prompt, max_tokens, temperature, top_p],
         chatbot
     )
     
@@ -169,19 +170,19 @@ with gr.Blocks() as demo:
         queue=False
     ).then(
         bot_respond,
-        [chatbot, system_prompt, max_tokens, temperature, top_p, gr.OAuthToken()],
+        [chatbot, system_prompt, max_tokens, temperature, top_p],
         chatbot
     )
     
     # Wire up the auto-respond button
     auto_respond_btn.click(
         handle_auto_respond,
-        [chatbot, max_tokens, temperature, top_p, gr.OAuthToken()],
+        [chatbot, max_tokens, temperature, top_p],
         chatbot,
         queue=False
     ).then(
         bot_respond_after_auto,
-        [chatbot, system_prompt, max_tokens, temperature, top_p, gr.OAuthToken()],
+        [chatbot, system_prompt, max_tokens, temperature, top_p],
         chatbot
     )
 
@@ -193,12 +194,13 @@ def respond(
     max_tokens,
     temperature,
     top_p,
-    hf_token: gr.OAuthToken,
+    hf_token: Optional[gr.OAuthToken] = None,
 ):
     """
     For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
     """
-    client = InferenceClient(token=hf_token.token, model="CohereForAI/aya-23-35B")
+    token = hf_token.token if hf_token else None
+    client = InferenceClient(token=token, model="CohereForAI/aya-23-35B")
 
     messages = [{"role": "system", "content": system_message}]
 
@@ -229,7 +231,7 @@ def generate_auto_response(
     max_tokens,
     temperature,
     top_p,
-    hf_token: gr.OAuthToken,
+    hf_token: Optional[gr.OAuthToken] = None,
 ):
     """
     Generates an automatic candidate response based on the interviewer's last question.
@@ -243,7 +245,8 @@ def generate_auto_response(
     if not last_message or last_message.get("role") != "assistant":
         return "Thank you for your question. I'm happy to answer."
     
-    client = InferenceClient(token=hf_token.token, model="CohereForAI/aya-23-35B")
+    token = hf_token.token if hf_token else None
+    client = InferenceClient(token=token, model="CohereForAI/aya-23-35B")
     
     # Create a system prompt for the candidate