diff --git a/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml b/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml index 2dc972f..c3d58d7 100644 --- 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=fp8 - --max-num-batched-tokens=65536 - - --max-num-seqs=128 + - --max-num-seqs=64 - --enable-chunked-prefill - --trust-remote-code - --dtype=auto