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

[1/N] Introduce Mooncake Backend and Mooncake EP to Support Elastic EP (#10423)

Browse Source 
Co-authored-by: Hank Han <hanhan7630@outlook.com>
Co-authored-by: Shangming Cai <csmthu@gmail.com>
This commit is contained in:
Xun Sun
2025-10-15 10:40:54 +08:00
committed by GitHub
parent 74737b2863
commit a40229f6f8
13 changed files with 798 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
docs/advanced_features/server_arguments.md
View File

@@ -134,6 +134,8 @@ Please consult the documentation below and [server_args.py](https://github.com/s
| Arguments | Description | Defaults |
|-----------|-------------|----------|
| `--device` | The device to use ('cuda', 'xpu', 'hpu', 'npu', 'cpu'). Defaults to auto-detection if not specified. | None |
| `--elastic-ep-backend` | Select the collective communication backend for elastic EP. Currently supports 'mooncake'. | None |
| `--mooncake-ib-device` | The InfiniBand devices for Mooncake Backend, accepts multiple comma-separated devices. Default is None, which triggers automatic device detection when Mooncake Backend is enabled. | None |
| `--tp-size` | The tensor parallelism size. | 1 |
| `--pp-size` | The pipeline parallelism size. | 1 |
| `--pp-max-micro-batch-size` | The maximum micro batch size in pipeline parallelism. | None |
@@ -246,7 +248,7 @@ Please consult the documentation below and [server_args.py](https://github.com/s
| Arguments | Description | Defaults |
|-----------|-------------|----------|
| `--ep-size` | The expert parallelism size. | 1 |
| `--moe-a2a-backend` | Select the backend for all-to-all communication for expert parallelism. | none |
| `--moe-a2a-backend` | Select the backend for all-to-all communication for expert parallelism, could be `deepep` or `mooncake`. | none |
| `--moe-runner-backend` | Select the runner backend for MoE. | auto |
| `--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 |
| `--ep-num-redundant-experts` | Allocate this number of redundant experts in expert parallel. | 0 |