Merge pull request #7 from sirius0xdev/feat/vllm-model-name-fix
fix: switch to official Qwen2.5 32B Instruct model
This commit is contained in:
commit
31c6e532c7
2 changed files with 5 additions and 2 deletions
|
|
@ -48,7 +48,10 @@ data:
|
|||
|
||||
"agents": {
|
||||
"defaults": {
|
||||
"model": { "primary": "openai/cognitivecomputations/dolphin-2.9.2-qwen1.5-32b" },
|
||||
"model": {
|
||||
"primary": "openai/Qwen/Qwen2.5-32B-Instruct",
|
||||
"fallbacks": ["google/gemini-3.1-pro-preview"]
|
||||
},
|
||||
"workspace": "~/.openclaw/workspace"
|
||||
},
|
||||
"list": [
|
||||
|
|
|
|||
|
|
@ -28,7 +28,7 @@ spec:
|
|||
command: ["python3", "-m", "vllm.entrypoints.openai.api_server"]
|
||||
args:
|
||||
- "--model"
|
||||
- "cognitivecomputations/dolphin-2.9.2-qwen1.5-32b" # Native 32B Uncensored (fits perfectly in 80GB VRAM)
|
||||
- "Qwen/Qwen2.5-32B-Instruct" # Official Qwen 2.5 32B Base Model
|
||||
- "--dtype"
|
||||
- "half"
|
||||
- "--max-model-len"
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue