Update rtx6000-vllm.yaml

2026-05-22 23:39:27 -04:00 · 2026-05-22 23:39:27 -04:00 · fa6dd3c8a6
commit fa6dd3c8a6
parent bfd9bbc3b6
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=bfloat16
            - --max-num-batched-tokens=65536
-            - --max-num-seqs=64
+            - --max-num-seqs=32
            - --block-size=32 
            - --language-model-only
            - --enable-chunked-prefill