app.py

import json

import gradio as gr
import requests
import uvicorn
from fastapi import FastAPI
from huggingface_hub import InferenceClient
from starlette.responses import StreamingResponse, JSONResponse

"""
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("microsoft/Phi-3-mini-4k-instruct")


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

"""
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 friendly Chatbot.", label="System message"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (nucleus sampling)",
        ),
    ],
)


app = FastAPI()

@app.head("/ask")
def ask_head():
    return StreamingResponse("", media_type="application/json")

@app.get("/ask")
def ask_get(message: str = "", system_message: str = "You are a friendly Chatbot.", max_tokens: int = 512, temperature: float = 0.7, top_p: float = 0.95):
    predict_response = requests.post('http://localhost:7860/call/chat', json={'data': [message, [], system_message, max_tokens, temperature, top_p]}).json()
    if "event_id" not in predict_response:
        return predict_response
    out = requests.get(f'http://localhost:7860/call/chat/{predict_response["event_id"]}').text
    return JSONResponse([json.loads(out.rsplit("event: complete\ndata: ", 1)[-1])[0].strip()])

if __name__ == "__main__":
    app = gr.mount_gradio_app(app, demo, path="/")
    uvicorn.run(app, host="0.0.0.0", port=7860)