Update app.py

app.py

@@ -117,8 +117,10 @@ try:
     model = AutoModelForCausalLM.from_pretrained(
         base_model_path,
         torch_dtype=torch.float16,
-        trust_remote_code=True,
-        device_map="auto"
+        device_map="cpu",          # Explicitly set CPU
+        load_in_8bit=True          # Enable int8 quantization
+        # trust_remote_code=True,
+        # device_map="auto"
     )
 
     # Load tokenizer