app.py

import gradio as gr
from huggingface_hub import InferenceClient
import traceback

def get_text(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))
    return str(content)

def respond(
    message,
    history: list[dict],
    system_message,
    max_tokens,
    temperature,
    top_p,
    hf_token: gr.OAuthToken, 
):
    if not hf_token or not hf_token.token:
        yield "⚠️ Please **Login** in the sidebar to access @frusto360 AI."
        return

    try:
        # ✅ 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"
        API_URL = f"https://router.huggingface.co/hf-inference/models/{MODEL_ID}"
        
        client = InferenceClient(base_url=API_URL, token=hf_token.token)

        # 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")
            content = get_text(msg.get("content", ""))
            prompt += f"<|start_header_id|>{role}<|end_header_id|>\n\n{content}<|eot_id|>"
        
        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"

        response = ""
        for token in client.text_generation(
            prompt,
            max_new_tokens=max_tokens,
            stream=True,
            temperature=temperature,
            top_p=top_p,
            stop=["<|eot_id|>"]
        ):
            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:
        yield f"❌ **Router Error:** {str(e)}\n\n*Note: Ensure 'Inference API' is enabled in your model settings.*"

# UI Setup (Gradio 6.5)
chatbot_interface = gr.ChatInterface(
    respond,
    additional_inputs=[
        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"),
    ],
)

with gr.Blocks(fill_height=True) as demo:
    with gr.Sidebar():
        gr.Markdown("## 🔐 @frusto360 Auth")
        gr.LoginButton()
    chatbot_interface.render()

if __name__ == "__main__":
    demo.launch(theme="glass")