Update rtx6000-vllm.yaml

2026-05-24 15:58:35 -04:00 · 2026-05-24 15:58:35 -04:00 · 5bf0601e8d
commit 5bf0601e8d
parent 8aeaf2100f
1 changed files with 1 additions and 1 deletions
--- a/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml
+++ b/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml
@ -42,7 +42,7 @@ spec:
            - --enable-auto-tool-choice 
            - --kv-cache-dtype=auto
            - --max-num-batched-tokens=16384
-            - --max-num-seqs=4
+            - --max-num-seqs=12
            - --block-size=32 
            - --language-model-only
            - --enable-chunked-prefill