app.py

import json
import gradio as gr
from huggingface_hub import InferenceClient
import os

import requests
from rag import run_rag
"""
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 chat(
    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]})
    message =run_rag(message, history)
 
    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 += str(token)
 
        yield response

"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""
chatbot = gr.Chatbot(
    label="Retrieval Augmented Generation News & Finance",
    # avatar_images=[None, BOT_AVATAR],
    show_copy_button=True, 
    likeable=True, 
    layout="bubble")
theme = gr.themes.Base(
    font=[gr.themes.GoogleFont('Libre Franklin'), gr.themes.GoogleFont('Public Sans'), 'system-ui', 'sans-serif'],
)
EXAMPLES = [
    [ "Tell me about the latest news in the world ?"],
    [ "Tell me about the increase in the price of Bitcoin ?"],
    [ "Tell me about the actual situation in Ukraine ?"],
    [ "Tell me about current situation in palestine ?"],
]
max_new_tokens = gr.Slider(
    minimum=1,
    maximum=2048,
    value=512,
    step=1,
    interactive=True,
    label="Max new tokens",
)
temperature = gr.Slider(
    minimum=0.1,
    maximum=0.9,
    value=0.6,
    step=0.1,
    visible=True,
    interactive=True,
    label="Temperature",
    info="Higher values will produce more diverse outputs.",
)
top_p = gr.Slider(
    minimum=0.1,
    maximum=1,
    value=0.9,
    step=0.05,
    visible=True,
    interactive=True,
    label="Top-p (nucleus sampling)",
    info="Higher values is equivalent to sampling more low-probability tokens.",
)

with gr.Blocks(
    fill_height=True,
    css=""".gradio-container .avatar-container {height: 40px width: 40px !important;} #duplicate-button {margin: auto; color: white; background: #f1a139; border-radius: 100vh; margin-top: 2px; margin-bottom: 2px;}""",
) as main:
    gr.ChatInterface(
        chat,
        chatbot=chatbot,
        title="Retrieval Augmented Generation (RAG) Chatbot",
        description="A chatbot that uses a RAG model to generate responses based on the input query.",
        examples=EXAMPLES,
        theme=theme,
        fill_height=True,
        multimodal=True,
        additional_inputs=[
            max_new_tokens,
            temperature,
            top_p,
        ],
    )
with gr.Blocks(theme=theme, css="footer {visibility: hidden}textbox{resize:none}", title="RAG") as demo:
    gr.TabbedInterface([main] , tab_names=["Chatbot"]  )
demo.launch()