app.py2

from openai import OpenAI
import gradio as gr
import os

# Initialize the OpenAI Client with your API key and endpoint
api_key = os.environ.get("RUNPOD_API_KEY")  # Ensure your API key is correctly loaded from environment variables
client = OpenAI(
    api_key=api_key,
    base_url="https://api.runpod.ai/v2/vllm-k0g4c60zor9xuu/openai/v1",
)

def predict(message, history):
    # Format the history for OpenAI
    history_openai_format = []
    for human, assistant in history:
        history_openai_format.append({"role": "user", "content": human})
        history_openai_format.append({"role": "assistant", "content": assistant})
    history_openai_format.append({"role": "user", "content": message})
  
    response = client.chat.completions.create(
        model='ambrosfitz/llama-3-history',
        messages=history_openai_format,
        temperature=0,
        max_tokens=150,
        stream=False  # Set to False for simplicity in this example
    )

    # Ensure response handling is correct
    if response.choices and response.choices[0].message['content']:
        response_text = response.choices[0].message['content'].strip()
        history.append((message, response_text))  # Update history with the new Q&A pair
        return response_text
    else:
        return "No response generated."

demo = gr.ChatInterface(
    fn=predict,
    title="HistoryBot Chat",
    description="Interact with HistoryBot, a specialized assistant for American History. Ask any historical questions to get detailed and nuanced answers."
)

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