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

fix: set vLLM dtype to auto to resolve weight casting corruption

Browse source
This commit is contained in:
SiriusClaw 2026-04-18 19:55:55 +00:00
parent bdf2e1d1a4
commit 576ba503b0
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
apps/base/customer1/openclaw/vllm-gemma.yaml
View file

@ -30,7 +30,7 @@ spec:
- "--model"
- "dphn/Dolphin-Mistral-24B-Venice-Edition" # Uncensored 24B (Native)
- "--dtype"
- "bfloat16"
- "auto"
- "--max-model-len"
- "8192"
- "--trust-remote-code"