app.py

import gradio as gr
import requests
import json
import logging
import time
from requests.exceptions import RequestException

# Set up logging
logging.basicConfig(level=logging.DEBUG, format='%(asctime)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)

# Hugging Face API settings
API_URL = "https://api-inference.huggingface.co/models/mattshumer/Reflection-Llama-3.1-70B"

def query(payload, max_retries=3, delay=5):
    for attempt in range(max_retries):
        try:
            logger.info(f"Sending payload to API (attempt {attempt + 1}/{max_retries}): {payload}")
            response = requests.post(API_URL, json=payload, timeout=60)
            logger.info(f"Received response with status code: {response.status_code}")
            
            if response.status_code == 200:
                return response.json()
            elif response.status_code == 503:
                logger.warning("Model is loading. Retrying...")
                time.sleep(delay)
            else:
                logger.error(f"API request failed with status code {response.status_code}: {response.text}")
                return {"error": f"API request failed with status code {response.status_code}"}
        
        except RequestException as e:
            logger.error(f"Request failed: {str(e)}")
            if attempt < max_retries - 1:
                logger.info(f"Retrying in {delay} seconds...")
                time.sleep(delay)
            else:
                return {"error": f"Failed to connect after {max_retries} attempts: {str(e)}"}

    return {"error": "Maximum retries reached"}

def generate_text(prompt):
    logger.info(f"Received prompt: {prompt}")
    
    try:
        payload = {
            "inputs": prompt,
            "parameters": {
                "max_new_tokens": 100,
                "temperature": 0.7,
                "top_p": 0.95,
                "do_sample": True
            }
        }
        logger.info("Calling Hugging Face Inference API for text generation...")
        response = query(payload)
        
        logger.info(f"Raw response from API: {json.dumps(response, indent=2)}")
        
        if "error" in response:
            return f"Error: {response['error']}"
        
        if isinstance(response, list) and len(response) > 0:
            generated_text = response[0].get('generated_text', '')
            logger.info(f"Processed response: {generated_text[:100]}...")  # Log first 100 chars
            return generated_text
        else:
            error_msg = f"Unexpected response format: {response}"
            logger.error(error_msg)
            return error_msg

    except Exception as e:
        error_msg = f"Error generating text: {str(e)}"
        logger.exception(error_msg)
        return error_msg

iface = gr.Interface(
    fn=generate_text,
    inputs=gr.Textbox(lines=5, label="Enter your prompt"),
    outputs=gr.Textbox(label="Generated Response"),
    title="Reflection Llama 3.1 70B Demo",
    description="Enter a prompt to generate text using the Reflection Llama 3.1 70B model. Please be patient, as the model may take some time to respond."
)

if __name__ == "__main__":
    logger.info("Starting Gradio interface...")
    iface.launch()
    logger.info("Gradio interface launched.")