Update app.py

app.py

@@ -1,7 +1,6 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
 
-
 def respond(
     message,
     history: list[dict[str, str]],
@@ -9,61 +8,67 @@ def respond(
     max_tokens,
     temperature,
     top_p,
-    hf_token: gr.OAuthToken,
+    hf_token: gr.OAuthToken, # Gradio injects this if logged in
 ):
-    """
-    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="Frusto/llama-3.2-1b-frusto360-final")
-
-    messages = [{"role": "system", "content": system_message}]
+    # Ensure token exists (User must click Login)
+    if not hf_token or not hf_token.token:
+        yield "⚠️ Please login using the button in the sidebar to access the @frusto360 AI."
+        return
 
-    messages.extend(history)
+    # Use direct InferenceClient (more stable for custom models)
+    client = InferenceClient(model="Frusto/llama-3.2-1b-frusto360-final", token=hf_token.token)
 
-    messages.append({"role": "user", "content": message})
+    # 1. Manually build the Llama 3.2 Chat Template
+    prompt = f"<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\n{system_message}<|eot_id|>"
+    
+    for msg in history:
+        role = msg['role']
+        content = msg['content']
+        prompt += f"<|start_header_id|>{role}<|end_header_id|>\n\n{content}<|eot_id|>"
+    
+    # Add current user message
+    prompt += f"<|start_header_id|>user<|end_header_id|>\n\n{message}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
 
     response = ""
+    try:
+        # 2. Use text_generation instead of chat_completion
+        for token in client.text_generation(
+            prompt,
+            max_new_tokens=max_tokens,
+            stream=True,
+            temperature=temperature,
+            top_p=top_p,
+            stop=["<|eot_id|>"] # Stop generating at the end-of-turn token
+        ):
+            response += token
+            yield response
+            
+    except Exception as e:
+        error_msg = str(e)
+        if "503" in error_msg:
+            yield "⏳ Model is waking up... please wait 60 seconds and try again."
+        else:
+            yield f"❌ Error: {error_msg}"
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        choices = message.choices
-        token = ""
-        if len(choices) and choices[0].delta.content:
-            token = choices[0].delta.content
-
-        response += token
-        yield response
-
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# --- UI Setup ---
 chatbot = gr.ChatInterface(
     respond,
+    type="messages", # Ensures history is a list of dictionaries
     additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
+        gr.Textbox(value="You are the @frusto360 AI assistant. Created by @frusto360.", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p"),
     ],
 )
 
-with gr.Blocks() as demo:
+with gr.Blocks(theme="glass") as demo:
     with gr.Sidebar():
+        gr.Markdown("### 🔐 Authentication")
         gr.LoginButton()
+        gr.Markdown("Login to use your Hugging Face account permissions.")
+    
     chatbot.render()
 
-
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()