EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

Browse Source
This commit is contained in:
Cheng Wan
2025-08-01 01:20:03 -07:00
committed by GitHub
parent c8d3a402c1
commit 6c88f6c8d9
38 changed files with 342 additions and 299 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
docs/backend/server_arguments.md
View File

@@ -212,8 +212,7 @@ Please consult the documentation below and [server_args.py](https://github.com/s
| Arguments | Description | Defaults |
|-----------|-------------|----------|
| `--ep-size` | The expert parallelism size. | 1 |
| `--enable-ep-moe` | Enabling expert parallelism for moe. The ep size is equal to the tp size. | False |
| `--enable-deepep-moe` | Enabling DeepEP MoE implementation for EP MoE. | False |
| `--moe-a2a-backend` | Select the backend for all-to-all communication for expert parallelism. | None |
| `--enable-flashinfer-cutlass-moe` | Enabling Flashinfer Cutlass MoE implementation for high throughput. | False |
| `--enable-flashinfer-trtllm-moe` | Enabling Flashinfer Trtllm MoE implementation for low latency. | False |
| `--deepep-mode` | Select the mode when enable DeepEP MoE, could be `normal`, `low_latency` or `auto`. Default is `auto`, which means `low_latency` for decode batch and `normal` for prefill batch. | auto |