From 068f0c856c04320fdc89268f81edf45df73265b2 Mon Sep 17 00:00:00 2001
From: ChanghoWoo <wooarchi@paasup.io>
Date: Thu, 27 Feb 2025 08:12:20 +0000
Subject: [PATCH] update model

---
 charts/vllm/custom-values.yaml | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/charts/vllm/custom-values.yaml b/charts/vllm/custom-values.yaml
index cefca80..7c795d4 100644
--- a/charts/vllm/custom-values.yaml
+++ b/charts/vllm/custom-values.yaml
@@ -1,5 +1,24 @@
 servingEngineSpec:
-  modelSpec: []
+  modelSpec:
+  - name: "llama3"
+    repository: "vllm/vllm-openai"
+    tag: "latest"
+    modelURL: "meta-llama/Llama-3.1-8B-Instruct"
+    replicaCount: 1
+
+    requestCPU: 10
+    requestMemory: "16Gi"
+    requestGPU: 1
+
+    pvcStorage: "50Gi"
+
+    vllmConfig:
+      enableChunkedPrefill: false
+      enablePrefixCaching: false
+      maxModelLen: 24576
+      dtype: "float16"
+      extraArgs: ["--disable-log-requests", "--gpu-memory-utilization", "0.8"]
+    hf_token: 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
   resources:
     requests:
       cpu: "4"
@@ -12,6 +31,7 @@ routerSpec:
   # -- The docker image of the router. The following values are defaults:
   repository: "lmcache/lmstack-router"
   tag: "latest"
+  enableRouter: true
   resources:
     requests:
       cpu: "2"