Update app.py

app.py

@@ -5,12 +5,12 @@ import httpx
 import gradio as gr
 import torch
 from fastapi import FastAPI, HTTPException, Request
-from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM, set_default_dtype
+from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
 import uvicorn
 import asyncio
 
-# ✅ Use float16 to reduce memory usage (for Hugging Face Spaces)
-set_default_dtype(torch.float16)
+# ✅ Use float16 to reduce memory usage
+torch.set_default_dtype(torch.float16)
 
 # ✅ Hugging Face API Token
 HF_TOKEN = os.getenv("HF_TOKEN")