From e1cf661692dfe19678d200ed6cad2c5cd224eaff Mon Sep 17 00:00:00 2001 From: Sirius Claw Date: Mon, 20 Apr 2026 20:47:30 +0000 Subject: [PATCH] Switch L4 vLLM to Dolphin Qwen2 7B AWQ for speed and uncensored compliance --- apps/base/customer1/openclaw/vllm-l4.yaml | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/apps/base/customer1/openclaw/vllm-l4.yaml b/apps/base/customer1/openclaw/vllm-l4.yaml index 6c0a00c..aadbc63 100644 --- 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