import gradio as gr
from huggingface_hub import InferenceClient

# Define a dictionary of pre-defined LLMs
# To add a new LLM:
# 1. Go to https://huggingface.co/models
# 2. Find an open-source LLM that supports the chat completion task
# 3. Copy the model's name (e.g., "mistralai/Mistral-7B-Instruct-v0.1")
# 4. Add it to this dictionary with a user-friendly name as the key
MODELS = {
    "Zephyr 7B Beta": "HuggingFaceH4/zephyr-7b-beta",
    "Mistral 7B Instruct": "mistralai/Mistral-7B-Instruct-v0.1",
    "Llama 2 7B": "meta-llama/Llama-2-7b-chat-hf",
    "FLAN-T5 XXL": "google/flan-t5-xxl",
    # Add more models here as needed
}

def respond(
    message, 
    history: list[tuple[str, str]], 
    model_name,
    system_message, 
    max_tokens, 
    temperature, 
    top_p,
):
    client = InferenceClient(MODELS[model_name])
    
    messages = [{"role": "system", "content": system_message}]
    for val in history:
        if val[0]:
            messages.append({"role": "user", "content": val[0]})
        if val[1]:
            messages.append({"role": "assistant", "content": val[1]})
    messages.append({"role": "user", "content": message})
    
    response = ""
    try:
        for message in client.chat_completion(
            messages, 
            max_tokens=max_tokens, 
            stream=True, 
            temperature=temperature, 
            top_p=top_p,
        ):
            token = message.choices[0].delta.content
            response += token
            yield response
    except Exception as e:
        yield f"Error: {str(e)}"

demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Dropdown(choices=list(MODELS.keys()), label="Select LLM", value=list(MODELS.keys())[0]),
        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 (nucleus sampling)"
        ),
    ],
)

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