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

chore: maximize vllm max_model_len on L4 to 32768

Browse source
This commit is contained in:
Sirius Claw 2026-04-20 16:14:27 +00:00
parent 399d1a58ed
commit fb951600d7
1 changed files with 2 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
apps/base/customer1/openclaw/vllm-l4.yaml
View file

@ -34,8 +34,8 @@ spec:
- --host=0.0.0.0 - --host=0.0.0.0
- --port=8000 - --port=8000
- --tensor-parallel-size=1 - --tensor-parallel-size=1
- --gpu-memory-utilization=0.90 - --gpu-memory-utilization=0.95
- --max-model-len=16384 - --max-model-len=32768
- --enable-auto-tool-choice - --enable-auto-tool-choice
- --trust-remote-code - --trust-remote-code
- --dtype=auto - --dtype=auto