app.py

import logging
import os
from fastapi import FastAPI, HTTPException
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
from peft import PeftModel, PeftConfig
from mistral_common.tokens.tokenizers.mistral import MistralTokenizer

# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

# Initialize FastAPI app
app = FastAPI()

# Global variables for model, tokenizer, and pipeline
model = None
tokenizer = None
pipe = None

# Get the Hugging Face token from environment variable
hf_token = os.environ.get("HUGGINGFACE_TOKEN")
if not hf_token:
    raise ValueError("HUGGINGFACE_TOKEN environment variable is not set")

@app.on_event("startup")
async def load_model():
    global model, tokenizer, pipe
    
    try:
        logger.info("Loading PEFT configuration...")
        config = PeftConfig.from_pretrained("frankmorales2020/Mistral-7B-text-to-sql-flash-attention-2-dataeval", token=hf_token)
        
        logger.info("Loading base model...")
        base_model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3", token=hf_token)
        
        logger.info("Loading PEFT model...")
        model = PeftModel.from_pretrained(base_model, "frankmorales2020/Mistral-7B-text-to-sql-flash-attention-2-dataeval", token=hf_token)

        logger.info("Loading tokenizer...")
        tokenizer = MistralTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3", token=hf_token)

        logger.info("Creating pipeline...")
        pipe = pipeline("text2text-generation", model=model, tokenizer=tokenizer)
        
        logger.info("Model, tokenizer, and pipeline loaded successfully.")
    except ImportError as e:
        logger.error(f"Error importing required modules. Please check your installation: {e}")
        raise
    except Exception as e:
        logger.error(f"Error loading model or creating pipeline: {e}")
        raise

@app.get("/")
def home():
    return {"message": "Hello World"}

@app.get("/generate")
async def generate(text: str):
    if not pipe:
        raise HTTPException(status_code=503, detail="Model not loaded")
    
    try:
        output = pipe(text, max_length=100, num_return_sequences=1)
        return {"output": output[0]['generated_text']}
    except Exception as e:
        logger.error(f"Error during text generation: {e}")
        raise HTTPException(status_code=500, detail=f"Error during text generation: {str(e)}")

if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=7860)