diff --git a/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml b/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml index 62b75d7..6135e98 100644 --- a/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml +++ b/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml @@ -41,7 +41,7 @@ spec: - --max-model-len=131072 - --enable-auto-tool-choice - --kv-cache-dtype=bfloat16 - - --max-num-batched-tokens=4096 + - --max-num-batched-tokens=16384 - --max-num-seqs=64 - --block-size=32 - --language-model-only