app.py

import gradio as gr
from huggingface_hub import InferenceClient

# Create an InferenceClient to interact with the model
client = InferenceClient("meta-llama/Llama-3.2-3B-Instruct")

# Define the function to generate a response
def respond(message, history: list[tuple[str, str]], system_message, max_tokens, temperature, top_p):
    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 = ""
    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

# Custom CSS for styling
css = """
body {
    font-family: 'Arial', sans-serif;
    background-color: #f8f9fa; /* Light background */
    color: #333;
}
.gr-button {
    background-color: #0b2545 !important;
    color: white !important;
    border: none !important;
    border-radius: 25px !important;
    padding: 8px 20px !important;
    font-size: 14px;
    font-weight: bold;
    cursor: pointer;
}
.gr-button:hover {
    background-color: #0a1b35 !important;
}
.search-box {
    border-radius: 20px;
    border: 1px solid #ccc;
    padding: 10px;
    width: 100%;
    font-size: 16px;
    background-color: #ffffff;
}
"""

# Main function to create the interface
with gr.Blocks(css=css) as demo:
    gr.Markdown("<h1 style='text-align: center;'>Health Assistant GPT</h1>")
    gr.Markdown("<h3 style='text-align: center;'>What do you want to know about health and wellness?</h3>")

    # Sidebar
    with gr.Sidebar():
        gr.Markdown("### Settings")
        system_message = gr.Textbox(
            value="You are a virtual health assistant designed to provide accurate and reliable information related to health, wellness, and medical topics. Your primary goal is to assist users with their health-related queries, offer general guidance, and suggest when to consult a licensed medical professional. If a user asks a question that is unrelated to health, wellness, or medical topics, respond politely but firmly.",
            label="System message",
            visible=False
        )
        max_tokens = gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens", visible=False)
        temperature = gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature", visible=False)
        top_p = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)", visible=False)

    # Main content
    with gr.Row():
        with gr.Column(scale=7):
            gr.Markdown("### Ask a health-related question:")
            search_input = gr.Textbox(label="Search Input", placeholder="Type your health-related question here...", lines=1)
            submit_button = gr.Button("Generate Response")
            output = gr.Markdown()

        with gr.Column(scale=3):
            gr.Markdown("### Upload a relevant file (Optional):")
            uploaded_file = gr.File(label="Upload PDF")

    # Button click action to trigger response generation
    submit_button.click(
        fn=respond,
        inputs=[search_input, [], system_message, max_tokens, temperature, top_p],  # Empty history for fresh chat
        outputs=output
    )

demo.launch()