From bcee730b1a9d25f0fd283a0ef6c19133ebeebf4f Mon Sep 17 00:00:00 2001 From: Sundri Lai Date: Thu, 1 Oct 2026 01:26:14 -0700 Subject: [PATCH] [Docs] Remove references to removed env vars (#59530) Signed-off-by: Sundri Lai --- docs/features/engram.md | 9 +++------ docs/serving/online_serving/renderer.md | 3 +-- 2 files changed, 4 insertions(+), 8 deletions(-) diff --git a/docs/features/engram.md b/docs/features/engram.md index 73dc05e9f1..023477b345 100644 --- a/docs/features/engram.md +++ b/docs/features/engram.md @@ -83,10 +83,7 @@ directly from the Triton kernel through unified virtual addressing (UVA), prefetching on a side CUDA stream so transfers overlap with the compute of preceding layers. The freed GPU memory goes to the KV cache. -`cpu_offload` defaults to `true` (on): it follows the -`VLLM_PLE_CPU_OFFLOAD` environment variable, which also defaults to `true`. -An explicit `--engram-config` value takes precedence over the environment -variable. +`cpu_offload` defaults to `true` (on) To keep the tables resident on GPU instead (e.g. when host memory is scarce): @@ -96,8 +93,8 @@ vllm serve deepseek-ai/DeepSeek-V4.1-Flash \ --engram-config.cpu_offload false ``` -or equivalently `VLLM_PLE_CPU_OFFLOAD=0`. CPU offload requires a GPU with UVA -support; vLLM fails fast when it is unavailable. +CPU offload requires a GPU with UVA support; +vLLM fails fast when it is unavailable. ## Data-parallel topologies (DeepSeek V4.1) diff --git a/docs/serving/online_serving/renderer.md b/docs/serving/online_serving/renderer.md index 9d494df22e..fc6579c258 100644 --- a/docs/serving/online_serving/renderer.md +++ b/docs/serving/online_serving/renderer.md @@ -51,8 +51,7 @@ and provision transport limits and memory accordingly. For example, start a standard inference server with scale-out endpoints enabled: ```bash -VLLM_ENABLE_SCALE_OUT_ENDPOINTS=1 \ - vllm serve meta-llama/Llama-3.1-8B-Instruct +vllm serve meta-llama/Llama-3.1-8B-Instruct --enable-scale-out ``` Send a Responses request and use `jq` to extract the `token_ids` field: