import gradio as gr
from huggingface_hub import InferenceClient
import os

def respond(
    message,
    history: list[dict[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
    hf_token_string, 
):
    token = hf_token_string if hf_token_string else os.getenv("HF_TOKEN")
    
    if not token:
        yield "Error: No Token provided."
        return

    client = InferenceClient(token=token, model="meta-llama/Meta-Llama-3-8B-Instruct")

    messages = [{"role": "system", "content": system_message}]
    messages.extend(history)
    messages.append({"role": "user", "content": message})

    try:
        # We don't need a 'response' string variable here for the API
        for chunk in client.chat_completion(
            messages,
            max_tokens=max_tokens,
            stream=True,
            temperature=temperature,
            top_p=top_p,
        ):
            if len(chunk.choices) > 0:
                token_str = chunk.choices[0].delta.content
                if token_str:
                    # OPTIMIZATION: Yield ONLY the new token.
                    # This is what makes the API streaming "instant".
                    yield token_str
    except Exception as e:
        yield f"API Error: {str(e)}"

# The ChatInterface will now receive tokens one by one. 
# Note: In the Gradio UI, this might make tokens "replace" each other. 
# If you want the UI to still look normal while keeping the API fast,
# use the client-side logic below.
chatbot = gr.ChatInterface(
    respond,
    type="messages",
    additional_inputs=[
        gr.Textbox(value="You are a friendly Chatbot.", 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"),
        gr.Textbox(label="Hugging Face Token", type="password"), 
    ],
)

with gr.Blocks() as demo:
    with gr.Sidebar():
        gr.LoginButton()
    chatbot.render()

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