Update main.py

main.py

@@ -4,6 +4,10 @@ from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import StreamingResponse
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import torch
+import os
+os.environ["TRANSFORMERS_CACHE"] = "/app/hf_cache"
+
+
 
 # Load Qwen model and tokenizer (once)
 model_name = "Qwen/Qwen2.5-0.5B-Instruct"