default use of ck fa

vllm/envs.py

@@ -130,7 +130,7 @@ environment_variables: Dict[str, Callable[[], Any]] = {
 
     # flag to control if vllm should use triton flash attention
     "VLLM_USE_TRITON_FLASH_ATTN":
-    lambda: (os.environ.get("VLLM_USE_TRITON_FLASH_ATTN", "True").lower() in
+    lambda: (os.environ.get("VLLM_USE_TRITON_FLASH_ATTN", "False").lower() in
              ("true", "1")),
 
     # local rank of the process in the distributed setting, used to determine