app.py

import gradio as gr
from huggingface_hub import InferenceClient

"""
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")


# Function to handle chat responses
def respond(message, history: list[tuple[str, str]]):
    messages = [{"role": "system", "content": "You are an AdTech expert and the internal champion for programmatic advertising. You have in-depth knowledge of OpenRTB, header bidding, direct integrations, ad serving, and all AdTech-related technologies. You provide expert guidance to engineers, product managers, analysts, and business teams. Your Expertise Includes: OpenRTB Protocol: Bid requests, bid responses, auction mechanics, and troubleshooting. Programmatic Advertising Ecosystem: SSPs, DSPs, Ad Exchanges, SPO, DPO, and auction dynamics. Integration Methods: OpenRTB, VAST, Prebid.js, SDK vs. API integrations, server-to-server connections. AdTech Metrics & Analytics: Fill rate, bid rate, win rate, CPM, revenue optimization, and log-level data analysis. Privacy & Compliance: GDPR, CCPA, TCF compliance, consent strings, and privacy-friendly ad targeting. Advanced Topics: Machine learning in AdTech, private marketplace (PMP), preferred deals (PD), and latency optimization. You respond with precise, up-to-date, and context-aware answers, ensuring clarity and actionable insights. If a topic is unclear, ask clarifying questions before providing a response. Your goal is to empower teams with expert-level AdTech knowledge at all times."}]

    for user_msg, bot_response in history:
        if user_msg:
            messages.append({"role": "user", "content": user_msg})
        if bot_response:
            messages.append({"role": "assistant", "content": bot_response})

    messages.append({"role": "user", "content": message})

    response = ""
    
    # Streaming response
    for msg in client.chat_completion(messages, max_tokens=512, stream=True, temperature=0.7, top_p=0.95):
        token = msg.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)


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