perf: let noromaid mixtral scale to zero

OpenRouterTeam · Apr 8, 2024 · b63c288 · b63c288
1 parent e0c3077
commit b63c288
Showing 1 changed file with 0 additions and 1 deletion.
diff --git a/modal/runner/containers/vllm_unified.py b/modal/runner/containers/vllm_unified.py
@@ -153,7 +153,6 @@ def __init__(self):
     gpu=modal.gpu.A100(count=1, memory=40),
     concurrent_inputs=4,
     max_containers=3,
-    keep_warm=1,
     quantization="GPTQ",
     dtype="float16",  # vLLM errors when using dtype="auto" with this model
 )