Update app.py

app.py

@@ -2,18 +2,13 @@ import gradio as gr
 from huggingface_hub import InferenceClient
 import traceback
 
-# --- Helper: Gradio 6.5 Text Extractor ---
 def get_text(content):
-    """Extracts text safely from Gradio 6's list-of-dict message format."""
-    if isinstance(content, str):
-        return content
+    if isinstance(content, str): return content
     if isinstance(content, list):
         return "".join([block.get("text", "") for block in content if block.get("type") == "text"])
-    if isinstance(content, dict):
-        return content.get("text", str(content))
+    if isinstance(content, dict): return content.get("text", str(content))
     return str(content)
 
-# --- Core Function: The Chat Logic ---
 def respond(
     message,
     history: list[dict],
@@ -23,21 +18,19 @@ def respond(
     top_p,
     hf_token: gr.OAuthToken, 
 ):
-    # 1. Check Authentication
     if not hf_token or not hf_token.token:
-        yield "⚠️ Please **Login** using the button in the sidebar to access the @frusto360 AI."
+        yield "⚠️ Please **Login** in the sidebar to access @frusto360 AI."
         return
 
     try:
-        # 2. DIRECT ROUTING: Avoids the StopIteration error by bypassing the provider search
+        # ✅ NEW 2026 ROUTER URL
+        # We use the 'hf-inference' provider prefix on the new router domain
         MODEL_ID = "Frusto/llama-3.2-1b-frusto360-final"
-        # We use the direct inference URL as the base_url
-        client = InferenceClient(
-            base_url=f"https://api-inference.huggingface.co/models/{MODEL_ID}",
-            token=hf_token.token
-        )
+        API_URL = f"https://router.huggingface.co/hf-inference/models/{MODEL_ID}"
+        
+        client = InferenceClient(base_url=API_URL, token=hf_token.token)
 
-        # 3. Build Llama 3.2 Chat Template
+        # Build Prompt
         prompt = f"<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\n{system_message}<|eot_id|>"
         for msg in history:
             role = msg.get("role", "user")
@@ -46,36 +39,27 @@ def respond(
         
         prompt += f"<|start_header_id|>user<|end_header_id|>\n\n{get_text(message)}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
 
-        # 4. Stream the Response
         response = ""
-        # Note: 'model' parameter is OMITTED because it's already in the base_url
         for token in client.text_generation(
             prompt,
             max_new_tokens=max_tokens,
             stream=True,
             temperature=temperature,
             top_p=top_p,
-            stop=["<|eot_id|>", "<|start_header_id|>"]
+            stop=["<|eot_id|>"]
         ):
-            # Compatibility check for token format
             token_text = token if isinstance(token, str) else getattr(token, 'token', getattr(token, 'text', str(token)))
             response += token_text
             yield response
             
     except Exception as e:
-        error_msg = str(e)
-        if "503" in error_msg:
-            yield "⏳ **Model is starting up.** Hugging Face is loading the weights. Please try again in 30 seconds!"
-        elif "404" in error_msg:
-            yield f"❌ **Error 404:** Model not found or Inference API disabled on the model page."
-        else:
-            yield f"❌ **Error:** {error_msg}\n\n*Check the Space Logs for details.*"
+        yield f"❌ **Router Error:** {str(e)}\n\n*Note: Ensure 'Inference API' is enabled in your model settings.*"
 
-# --- Gradio UI Layout ---
+# UI Setup (Gradio 6.5)
 chatbot_interface = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="You are a helpful assistant developed by @frusto360.", label="System message"),
+        gr.Textbox(value="You are the @frusto360 AI.", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max tokens"),
         gr.Slider(minimum=0.1, maximum=2.0, 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"),
@@ -84,13 +68,9 @@ chatbot_interface = gr.ChatInterface(
 
 with gr.Blocks(fill_height=True) as demo:
     with gr.Sidebar():
-        gr.Markdown("## 🔐 @frusto360 Control")
+        gr.Markdown("## 🔐 @frusto360 Auth")
         gr.LoginButton()
-        gr.Markdown("---")
-        gr.Markdown("Created by [@frusto360](https://youtube.com/@frusto360)")
-    
     chatbot_interface.render()
 
 if __name__ == "__main__":
-    # Gradio 6.0+ prefers theme in launch()
     demo.launch(theme="glass")