A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 12, 07:15 UTC
Fix AttributeError about _use_fi_prefill (#29734)
Signed-off-by: Huamin Li <3ericli@gmail.com>
Huamin Li committed
Nov 29, 2025 at 22:04 UTC
82c795d6f28ee365bfa822f30612e5da35c93fc0
1 file changed
+1
-1
vllm/v1/attention/backends/mla/common.py
+1
-1
@@ -2061,7 +2061,7 @@ class MLACommonImpl(MLACommonBaseImpl[M], Generic[M]):
2061
attn_out,
2062
lse,
2063
get_dcp_group(),
2064
- is_lse_base_on_e=not self._use_fi_prefill,
2064
+ is_lse_base_on_e=not getattr(self, "_use_fi_prefill", False),
2065
)
2066
2067
# v_up projection