add e2e test for vLLM with lora adapters

substratusai · Dec 13, 2024 · 4affddf · 4affddf
1 parent f0d7dd4
commit 4affddf
Show file tree

Hide file tree

Showing 4 changed files with 12 additions and 1 deletion.
diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml
@@ -33,6 +33,7 @@ jobs:
         - "autoscaler-restart"
         - "cache-shared-filesystem"
         - "engine-vllm-pvc"
+        - "engine-vllm-adapters"
     steps:
       - name: Checkout code
         uses: actions/checkout@v4

diff --git a/Makefile b/Makefile
@@ -95,6 +95,10 @@ test-e2e-cache-shared-filesystem: skaffold
 test-e2e-engine-vllm-pvc: skaffold
 	./test/e2e/run.sh engine-vllm-pvc --profile e2e-test-default
 
+.PHONY: test-e2e-engine-vllm-adapters
+test-e2e-engine-vllm-adapters: skaffold
+	./test/e2e/run.sh engine-vllm-adapters --profile e2e-test-default
+
 .PHONY: test-e2e-engine
 test-e2e-engine: skaffold
 	CACHE_PROFILE=$(CACHE_PROFILE) ./test/e2e/run.sh engine-$(ENGINE) --profile e2e-test-default

diff --git a/charts/models/values.yaml b/charts/models/values.yaml
@@ -304,6 +304,12 @@ catalog:
     url: "ollama://qwen2.5:7b"
     engine: OLlama
     resourceProfile: cpu:2
+  qwen2.5-0.5b-cpu:
+    enabled: false
+    features: ["TextGeneration"]
+    engine: VLLM
+    url: "hf://Qwen/Qwen2.5-0.5B-Instruct"
+    resourceProfile: cpu:1
   qwen2-500m-cpu:
     enabled: false
     features: ["TextGeneration"]

diff --git a/test/e2e-manual/gke-vllm-adapters/run.sh b/test/e2e-manual/gke-vllm-adapters/run.sh
@@ -16,4 +16,4 @@ curl -v http://localhost:8000/openai/v1/completions \
 # with adapter
 curl -v http://localhost:8000/openai/v1/completions \
   -H "Content-Type: application/json" \
-  -d '{"model": "tiny-llama/colorist", "prompt": "Who was the first president of the United States?", "max_tokens": 40}'
+  -d '{"model": "tiny-llama_colorist", "prompt": "Who was the first president of the United States?", "max_tokens": 40}'