From fee37d9e8da009534765768d0a09224914ad8f61 Mon Sep 17 00:00:00 2001 From: Baizhou Zhang Date: Wed, 7 May 2025 09:04:22 -0700 Subject: [PATCH] [Doc]Fix description for dp_size argument (#6063) --- docs/backend/server_arguments.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/backend/server_arguments.md b/docs/backend/server_arguments.md index 76206aaee..d99d9756b 100644 --- a/docs/backend/server_arguments.md +++ b/docs/backend/server_arguments.md @@ -100,7 +100,7 @@ Please consult the documentation below and [server_args.py](https://github.com/s | Arguments | Description | Defaults | |-----------|-------------|---------| -| `dp_size` | Will be deprecated. The number of data-parallel copies of the model. [SGLang router](../router/router.md) is recommended instead of the current naive data parallel. | `1` | +| `dp_size` | For non-DeepSeek models, this is the the number of data-parallel copies of the model. For DeepSeek models, this is the group size of [data parallel attention](https://docs.sglang.ai/references/deepseek.html#data-parallelism-attention) on DeepSeek models. | `1` | | `load_balance_method` | Will be deprecated. Load balancing strategy for data parallel requests. | `"round_robin"` | ### Expert parallelism