sirius/gcloud-lab
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

numseq lower

Browse source
This commit is contained in:
sirius0xdev 2026-05-10 02:20:42 +00:00
parent a718acffa6
commit ad30b9524f
1 changed files with 2 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml
View file

@ -37,12 +37,12 @@ spec:
- --port=8000 - --port=8000
- --tensor-parallel-size=1 - --tensor-parallel-size=1
- --tokenizer-mode=hf - --tokenizer-mode=hf
- --gpu-memory-utilization=0.90 - --gpu-memory-utilization=0.95
- --max-model-len=65536 - --max-model-len=65536
- --enable-auto-tool-choice - --enable-auto-tool-choice
- --kv-cache-dtype=fp8 - --kv-cache-dtype=fp8
- --max-num-batched-tokens=65536 - --max-num-batched-tokens=65536
- --max-num-seqs=64 - --max-num-seqs=32
- --enable-chunked-prefill - --enable-chunked-prefill
- --trust-remote-code - --trust-remote-code
- --dtype=auto - --dtype=auto