app.py

import gradio as gr
from huggingface_hub import InferenceClient

"""
For more information on `huggingface_hub` Inference API support, check:
https://huggingface.co/docs/huggingface_hub/en/guides/inference
"""

# Initialize the Inference API Client with your model
client = InferenceClient("one1cat/FineTunes_LLM_CFR_49")


def respond(message, history, system_message, max_tokens, temperature, top_p):
    """
    Generates responses using the fine-tuned CFR 49 model.
    """

    # Format prompt
    prompt = f"{system_message}\n\nUser: {message}\n\nAssistant:"

    # Generate response
    response = ""

    try:
        for token in client.text_generation(
            prompt,
            max_new_tokens=max_tokens,
            temperature=temperature,
            top_p=top_p,
            stream=True,  # Enables token streaming
        ):
            response += token
            yield response

    except Exception as e:
        yield f"Error: {str(e)}"


# Gradio Chat Interface
demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value="You are an AI trained on CFR 49 regulations.", 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)"),
    ],
)

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