From 853ad9cbf5c838b30bc221b5851276cfdcf77625 Mon Sep 17 00:00:00 2001
From: Peter Pan <Peter.Pan@daocloud.io>
Date: Wed, 5 Feb 2025 15:58:26 +0800
Subject: [PATCH] docs: add vLLM new feature(>-v0.7.1) supporting DeepSeek-R1

refer to https://github.com/vllm-project/vllm/releases/tag/v0.7.1
---
 README.md | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/README.md b/README.md
index d9d4ccf..49310e0 100644
--- a/README.md
+++ b/README.md
@@ -191,6 +191,8 @@ For instance, you can easily start a service using [vLLM](https://github.com/vll
 ```shell
 vllm serve deepseek-ai/DeepSeek-R1-Distill-Qwen-32B --tensor-parallel-size 2 --max-model-len 32768 --enforce-eager
 ```
+With `vllm` >= v0.7.1, you can additionally append `--enable-reasoning --reasoning-parser deepseek_r1` to above shell command, for reasoning output parsing.
+
 
 You can also easily start a service using [SGLang](https://github.com/sgl-project/sglang)