Use min new token ratio at start (#701)

2024-07-23 11:52:50 -07:00
parent 824a77d04d
commit 268684439b
2 changed files with 2 additions and 5 deletions
--- a/python/sglang/srt/server_args.py
+++ b/python/sglang/srt/server_args.py
@@ -29,7 +29,7 @@ class ServerArgs:
    max_prefill_tokens: Optional[int] = None
    max_running_requests: Optional[int] = None
    schedule_heuristic: str = "lpm"
-    schedule_conservativeness: float = 0.8
+    schedule_conservativeness: float = 1.0

    # Other runtime options
    tp_size: int = 1