Update app.py

app.py

@@ -92,7 +92,7 @@ def initialize_model():
         MODEL_ID,
         torch_dtype=torch.float16,
         device_map="cuda",
-        attn_implementation="flash_attention_2",
+        # attn_implementation="flash_attention_2",
         trust_remote_code=True,
         quantization_config=quantization_config