remove node label selector

substratusai · Nov 1, 2024 · 5e820ae · 5e820ae
1 parent bd26bcd
commit 5e820ae
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 5 deletions.
diff --git a/charts/kubeai/values.yaml b/charts/kubeai/values.yaml
@@ -109,8 +109,6 @@ resourceProfiles:
         operator: "Equal"
         value: "present"
         effect: "NoSchedule"
-    nodeSelector:
-      nvidia.com/gpu.family: "hopper"
   nvidia-gpu-a100-80gb:
     imageName: "nvidia-gpu"
     limits:

diff --git a/manifests/models/llama-3.1-70b-instruct-awq-int4-gh200.yaml b/manifests/models/llama-3.1-70b-instruct-awq-int4-gh200.yaml
@@ -8,10 +8,10 @@ spec:
   url: hf://hugging-quants/Meta-Llama-3.1-70B-Instruct-AWQ-INT4
   engine: VLLM
   args:
-    - --max-model-len=16384
-    - --max-num-batched-token=16384
+    - --max-model-len=4000
+    - --max-num-batched-token=4096
     - --enable-prefix-caching
     - --disable-log-requests
-  targetRequests: 50
+  targetRequests: 4000
   minReplicas: 1
   resourceProfile: nvidia-gpu-gh200:1