[EPLB] Display the expert hotness comparison before and after eplb. (#6877)

### What this PR does / why we need it?
To intuitively show the effect of the eplb algorithm, we print the
expert heat before and after eplb.

### Does this PR introduce _any_ user-facing change?

### How was this patch tested?

![Snipaste_2026-02-28_17-23-42](https://github.com/user-attachments/assets/db1dadd1-cf96-44da-af34-57d41ccf412f)


- vLLM version: v0.16.0
- vLLM main:
15d76f74e2

Signed-off-by: shenchuxiaofugui <1311027364@qq.com>
This commit is contained in:
LI SHENGYONG
2026-03-06 09:53:29 +08:00
committed by GitHub
parent 18b52afe2b
commit ccd00798f3
3 changed files with 42 additions and 38 deletions

View File

@@ -66,7 +66,7 @@ vllm serve Qwen/Qwen3-235B-A22 \
--tensor-parallel-size 16 \
--enable-expert-parallel \
--additional-config '{
"expert_map_path": "/path/to/eplb.json"
"eplb_config": {"expert_map_path": "/path/to/eplb.json"}
}'
```