From bd09ea4958365c737be5c6496afe3995ce03c0b9 Mon Sep 17 00:00:00 2001
From: sirius0xdev <sirius0xdev@proton.me>
Date: Tue, 21 Apr 2026 03:03:59 +0000
Subject: [PATCH] fix oom

---
 apps/base/customer1/openclaw/vllm-l4.yaml | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/apps/base/customer1/openclaw/vllm-l4.yaml b/apps/base/customer1/openclaw/vllm-l4.yaml
index 5e28bd9..89a7152 100644
--- a/apps/base/customer1/openclaw/vllm-l4.yaml
+++ b/apps/base/customer1/openclaw/vllm-l4.yaml
@@ -38,7 +38,10 @@ spec:
             - --port=8000
             - --tensor-parallel-size=1
             - --gpu-memory-utilization=0.95
-            - --max-model-len=32768
+            - --max-model-len=16384
+            - --max-num-batched-tokens=2048
+            - --max-num-seqs=128
+            - --enable-chunked-prefill
             - --enable-auto-tool-choice 
             - --trust-remote-code
             - --dtype=auto