from huggingface_hub import InferenceClient
import gradio as gr

client = InferenceClient("mistralai/Mistral-7B-Instruct-v0.3")


def format_prompt(message, history):
    prompt = "<s>"
    for user_prompt, bot_response in history:
        prompt += f"[INST] {user_prompt} [/INST] {bot_response}</s> "
    prompt += f"[INST] {message} [/INST]"
    return prompt


def generate(
    prompt, history, temperature=0.7, max_new_tokens=256, top_p=0.9, repetition_penalty=1.2,
):
    temperature = max(0.01, float(temperature))
    top_p = max(0.0, float(top_p))

    generate_kwargs = dict(
        temperature=temperature,
        max_new_tokens=max_new_tokens,
        top_p=top_p,
        repetition_penalty=repetition_penalty,
        do_sample=True,
        seed=42,
    )

    formatted_prompt = format_prompt(prompt, history)
    stream = client.text_generation(
        formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False
    )
    output = ""
    for response in stream:
        output += response.token.text
        yield output
    return output


additional_inputs = [
    gr.Slider(
        label="Temperature",
        value=0.7,
        minimum=0.1,
        maximum=1.0,
        step=0.05,
        interactive=True,
        info="Controls the creativity of the response. Higher values mean more randomness.",
    ),
    gr.Slider(
        label="Max Tokens",
        value=256,
        minimum=32,
        maximum=1024,
        step=32,
        interactive=True,
        info="The maximum number of tokens to generate.",
    ),
    gr.Slider(
        label="Top-p",
        value=0.9,
        minimum=0.1,
        maximum=1.0,
        step=0.05,
        interactive=True,
        info="Nucleus sampling. Lower values generate more deterministic responses.",
    ),
    gr.Slider(
        label="Repetition Penalty",
        value=1.2,
        minimum=1.0,
        maximum=2.0,
        step=0.1,
        interactive=True,
        info="Discourages repeated phrases in the output.",
    ),
]

css = """
  #chatbox {
    height: 600px;
    overflow: auto;
    border: none;
    box-shadow: 0px 4px 12px rgba(0, 0, 0, 0.1);
    border-radius: 8px;
  }
"""

with gr.Blocks(css=css) as demo:
    gr.HTML(
        """
        <div style="text-align: center; margin: 20px 0;">
            <h1 style="font-family: Arial, sans-serif; color: #333;">Mispel V1.0 Beta</h1>
            <p style="font-family: Arial, sans-serif; font-size: 16px; color: #555;">
                Powered by the <a href="https://medlarai.com" style="color: #007BFF; text-decoration: none;">MedlarAI</a> model.
                Your assistant for answering questions, generating text, and exploring ideas.
            </p>
        </div>
        """
    )
    # Replace gr.Box() with gr.Column()
    with gr.Column():
        gr.ChatInterface(
            generate,
            additional_inputs=additional_inputs,
            examples=[
                ["Explain quantum mechanics in simple terms."],
                ["What are the key principles of effective leadership?"],
                ["Write a short story about a time traveler."],
                ["Generate a Python function to calculate factorial using recursion."],
                ["What are the benefits of mindfulness meditation?"],
            ],
        )

demo.queue().launch(debug=True)