Spaces:
Runtime error
Runtime error
| from fastapi import FastAPI | |
| from transformers import pipeline | |
| import torch | |
| from pydantic import BaseModel | |
| import os | |
| import numpy as np # Explicit numpy import | |
| from transformers import AutoModelForSequenceClassification, AutoTokenizer, pipeline | |
| import torch | |
| # Fix numpy initialization | |
| np.zeros(1) # Force numpy load before model | |
| app = FastAPI() | |
| # Disable xformers if needed | |
| torch.backends.cuda.enable_flash_sdp(False) | |
| torch.backends.cuda.enable_mem_efficient_sdp(False) | |
| # Configure paths and device | |
| os.environ["TRANSFORMERS_CACHE"] = "/tmp/huggingface" | |
| os.makedirs(os.environ["TRANSFORMERS_CACHE"], exist_ok=True) | |
| def readiness_check(): | |
| return {"status": "ready"} | |
| model_name = "win2win/3-epochs-classifier-ver2" | |
| try: | |
| model = AutoModelForSequenceClassification.from_pretrained(model_name) | |
| tokenizer = AutoTokenizer.from_pretrained(model_name) | |
| # Create pipeline with explicit classes | |
| classifier = pipeline( | |
| "text-classification", | |
| model=model, | |
| tokenizer=tokenizer, | |
| device="cuda" if torch.cuda.is_available() else "cpu" | |
| ) | |
| print("Model loaded successfully!") | |
| except Exception as e: | |
| print(f"Error loading model: {str(e)}") | |
| raise | |
| class Request(BaseModel): | |
| text: str | |
| async def predict(request: Request): | |
| return model(request.text) | |
| if __name__ == "__main__": | |
| import uvicorn | |
| uvicorn.run(app, host="0.0.0.0", port=8000) |