Update app.py

app.py

@@ -76,7 +76,7 @@ model = AutoModelForCausalLM.from_pretrained(
     "m-a-p/YuE-s1-7B-anneal-en-cot",
     torch_dtype=torch.float16,
     attn_implementation="flash_attention_2",
-    low_cpu_mem_usage=True,
+    # low_cpu_mem_usage=True,
 ).to(device)
 model.eval()