Clean up the usage of flashinfer (#610)

2024-07-12 13:00:03 -07:00
parent 519e20cfda
commit af4e7910e7
5 changed files with 46 additions and 75 deletions
--- a/python/sglang/srt/server_args.py
+++ b/python/sglang/srt/server_args.py
@@ -53,6 +53,7 @@ class ServerArgs:
    disable_flashinfer: bool = False
    disable_radix_cache: bool = False
    disable_regex_jump_forward: bool = False
+    disable_cuda_graph: bool = False
    disable_disk_cache: bool = False
    attention_reduce_in_fp32: bool = False
    enable_p2p_check: bool = False
@@ -294,6 +295,11 @@ class ServerArgs:
            action="store_true",
            help="Disable regex jump-forward",
        )
+        parser.add_argument(
+            "--disable-cuda-graph",
+            action="store_true",
+            help="Disable cuda graph.",
+        )
        parser.add_argument(
            "--disable-disk-cache",
            action="store_true",