Update app.py

app.py

@@ -17,7 +17,8 @@ tokenizer = AutoTokenizer.from_pretrained(model_name)
 print("Loading model...")
 model = AutoModelForCausalLM.from_pretrained(
     model_name,
-    dtype=torch.float32
+    dtype=torch.float32,
+    low_cpu_mem_usage=True
 ).to("cpu")
 
 print("Model loaded successfully!")
@@ -122,4 +123,4 @@ async def chat(req: ChatRequest):
 
 # Required for HuggingFace Spaces
 if __name__ == "__main__":
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+    uvicorn.run(app, host="0.0.0.0", port=7860, workers=3)