diff --git a/python/sglang/srt/layers/attention/flashinfer_backend.py b/python/sglang/srt/layers/attention/flashinfer_backend.py index b524a9afcb85..11f81721970e 100644 --- a/python/sglang/srt/layers/attention/flashinfer_backend.py +++ b/python/sglang/srt/layers/attention/flashinfer_backend.py @@ -281,6 +281,7 @@ def fast_prefill_plan( fixed_split_size if fixed_split_size is not None else -1, False, # disable_split_kv 0, # num_colocated_ctas + 0, # uniform_q_len ] self._plan_info = self._cached_module.plan(*args)