xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-09 13:15:48 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Docs]add eplb_config param use docs (#24213)

Browse Source 
Signed-off-by: rongfu.leng <rongfu.leng@daocloud.io>
This commit is contained in:
rongfu.leng 2025-09-09 00:36:57 +08:00 committed by GitHub
parent 55be93baf5
commit c44797a4d6
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 27 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
docs/serving/expert_parallel_deployment.md
View File

@ -123,12 +123,33 @@ When enabled, vLLM collects load statistics with every forward pass and periodic
### EPLB Parameters
Configure EPLB with the `--eplb-config` argument, which accepts a JSON string. The available keys and their descriptions are:
| Parameter | Description | Default |
|-----------|-------------|---------|
| `--eplb-window-size` | Number of engine steps to track for rebalancing decisions | - |
| `--eplb-step-interval` | Frequency of rebalancing (every N engine steps) | - |
| `--eplb-log-balancedness` | Log balancedness metrics (avg tokens per expert ÷ max tokens per expert) | `false` |
| `--num-redundant-experts` | Additional global experts per EP rank beyond equal distribution | `0` |
| `window_size`| Number of engine steps to track for rebalancing decisions | 1000 |
| `step_interval`| Frequency of rebalancing (every N engine steps) | 3000 |
| `log_balancedness` | Log balancedness metrics (avg tokens per expert ÷ max tokens per expert) | `false` |
| `num_redundant_experts` | Additional global experts per EP rank beyond equal distribution | `0` |
For example:
```bash
vllm serve Qwen/Qwen3-30B-A3B \
--enable-eplb \
--eplb-config '{"window_size":1000,"step_interval":3000,"num_redundant_experts":2,"log_balancedness":true}'
```
??? tip "Prefer individual arguments instead of JSON?"
```bash
vllm serve Qwen/Qwen3-30B-A3B \
--enable-eplb \
--eplb-config.window_size 1000 \
--eplb-config.step_interval 3000 \
--eplb-config.num_redundant_experts 2 \
--eplb-config.log_balancedness true
```
### Expert Distribution Formula
@ -146,12 +167,10 @@ VLLM_ALL2ALL_BACKEND=pplx VLLM_USE_DEEP_GEMM=1 vllm serve deepseek-ai/DeepSeek-V
--data-parallel-size 8 \ # Data parallelism
--enable-expert-parallel \ # Enable EP
--enable-eplb \ # Enable load balancer
--eplb-log-balancedness \ # Log balancing metrics
--eplb-window-size 1000 \ # Track last 1000 engine steps
--eplb-step-interval 3000 # Rebalance every 3000 steps
--eplb-config '{"window_size":1000,"step_interval":3000,"num_redundant_experts":2,"log_balancedness":true}'
```
For multi-node deployment, add these EPLB flags to each node's command. We recommend setting `--num-redundant-experts` to 32 in large scale use cases so the most popular experts are always available.
For multi-node deployment, add these EPLB flags to each node's command. We recommend setting `--eplb-config '{"num_redundant_experts":32}'` to 32 in large scale use cases so the most popular experts are always available.
## Disaggregated Serving (Prefill/Decode Split)