chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 11 Packages Projects Releases Wiki Activity

[metrics][EPLB]: Support selected count of physical experts on each GPU (#9825)

Browse Source
This commit is contained in:
Yingchun Lai
2025-10-28 20:56:19 -07:00
committed by GitHub
parent 0297773a2f
commit 5e36a0b455
3 changed files with 54 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/references/environment_variables.md
+9 -8
View File
@@ -6,14 +6,15 @@ SGLang supports various environment variables that can be used to configure its
## General Configuration
| Environment Variable | Description | Default Value |
| --- | --- | --- |
| `SGLANG_USE_MODELSCOPE` | Enable using models from ModelScope | `false` |
| `SGLANG_HOST_IP` | Host IP address for the server | `0.0.0.0` |
| `SGLANG_PORT` | Port for the server | auto-detected |
| `SGLANG_LOGGING_CONFIG_PATH` | Custom logging configuration path | Not set |
| `SGLANG_DISABLE_REQUEST_LOGGING` | Disable request logging | `false` |
| `SGLANG_HEALTH_CHECK_TIMEOUT` | Timeout for health check in seconds | `20` |
| Environment Variable | Description | Default Value |
|-------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------|---------------|
| `SGLANG_USE_MODELSCOPE` | Enable using models from ModelScope | `false` |
| `SGLANG_HOST_IP` | Host IP address for the server | `0.0.0.0` |
| `SGLANG_PORT` | Port for the server | auto-detected |
| `SGLANG_LOGGING_CONFIG_PATH` | Custom logging configuration path | Not set |
| `SGLANG_DISABLE_REQUEST_LOGGING` | Disable request logging | `false` |
| `SGLANG_HEALTH_CHECK_TIMEOUT` | Timeout for health check in seconds | `20` |
| `SGLANG_EPLB_HEATMAP_COLLECTION_INTERVAL` | The interval of passes to collect the metric of selected count of physical experts on each layer and GPU rank. 0 means disabled. | `0` |
## Performance Tuning