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 10, 19:00 UTC
[Bugfix] Fix EPLB state logging error (#31455)
Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Tyler Michael Smith committed
Jan 4, 2026 at 23:06 UTC
fe8a9fbd2e2b3c3443c599bae05c24d3501a901e
1 file changed
+1
-1
vllm/distributed/eplb/eplb_state.py
+1
-1
@@ -423,7 +423,7 @@ class EplbState:
423
# Set the policy based on the selected eplb algorithm type.
424
policy_type = self.parallel_config.eplb_config.policy
425
self.policy = EPLB_POLICIES[policy_type]
426
- logger.debug("Selected EPLB policy: %d", policy_type)
426
+ logger.debug("Selected EPLB policy: %s", policy_type)
427
if global_expert_load is not None:
428
ep_group = get_ep_group().device_group
429
assert global_expert_load.shape == (