Merge pull request #48 from sirius0xdev/feature/dolphin-awq-l4

Switch L4 vLLM to Dolphin Qwen2 7B AWQ for speed and uncensored compl…
2026-04-20 16:50:07 -04:00 · 2026-04-20 16:50:07 -04:00 · 887f91ac69
commit 887f91ac69
parent b4d4bd9841 e1cf661692
1 changed files with 2 additions and 1 deletions
--- a/apps/base/customer1/openclaw/vllm-l4.yaml
+++ b/apps/base/customer1/openclaw/vllm-l4.yaml
@ -30,7 +30,8 @@ spec:
            - name: HF_TOKEN
              value: ""
          args:
-            - --model=trohrbaugh/Qwen2.5-Coder-7B-Instruct-heretic
+            - --model=cognitivecomputations/dolphin-2.9.3-qwen2-7b-awq
+            - --quantization=awq
            - --host=0.0.0.0
            - --port=8000
            - --tensor-parallel-size=1