import gradio as gr
from huggingface_hub import InferenceClient

#client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")
#client = InferenceClient("stanford-crfm/meditron-7b")

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 and len(val) == 2:
            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_chunk in client.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        delta = message_chunk.choices[0].delta
        if delta is None or delta.content is None:
            continue
        token = delta.content
        response += token
        yield response

# ✅ 5-line concise system prompt
default_system_prompt = (
    "You are a professional pharmacist who ONLY answers questions related to medications, including uses, dosages, side effects, interactions, and recommendations. "
    "If the user asks about anything NOT related to medications, politely reply that you can only help with medication-related questions and suggest they consult other resources. "
    "Always ask for the user's age before giving any dosage or advice. "
    "Include a clear disclaimer at the end: "
    "\"This information is for educational purposes only and does not replace professional medical advice. Please consult a licensed healthcare provider.\""
)

demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value=default_system_prompt, label="System message", lines=5),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.2, 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()