app.py

import gradio as gr
from huggingface_hub import InferenceClient
import os
from rag import local_retriever, global_retriever

"""
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
"""


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


def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    search_strategy,
    top_p,
):
    if search_strategy == "Global":
        return global_retriever(message, 2, "multiple paragraphs")
    else:
        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=2048,
            stream=True,
            temperature=1.0,
            top_p=top_p,
        ):
            token = message.choices[0].delta.content

            response += token

        return response


"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""
demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(
            value="You are a medical assistant Chatbot. For any query that you don't know, you will say 'I don't know'. You will answer with the given information:",
            label="System message",
        ),
        gr.Dropdown(choices=["Local", "Global"], label="Select search strategy"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (nucleus sampling)",
        ),
    ],
)


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