import gradio as gr
from huggingface_hub import InferenceClient

client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")

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

    # Highlight the diagnosis in bold
    diagnosis_start = response.find("Предварительный диагноз:")
    if diagnosis_start != -1:
        diagnosis_end = response.find("\n", diagnosis_start)
        if diagnosis_end == -1:
            diagnosis_end = len(response)
        diagnosis = response[diagnosis_start:diagnosis_end]
        response = response[:diagnosis_start] + f"<b>{diagnosis}</b>" + response[diagnosis_end:]

    # Add an identification message to the response
    final_response = f"{response}\n\nСоздано больницей EMS штата Alta"
    yield final_response

demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(
            value="Привет! Я помощник врача в больнице EMS штата Alta! Опиши свои симптомы, и я поставлю предварительный диагноз.",
            label="Системное сообщение"
        ),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Максимальное количество новых токенов"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Температура"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (ядерное семплирование)",
        ),
    ],
)

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