feat(tests): Move to k8s cluster for CI/CD

huggingface · Jun 25, 2024 · dbe261a · dbe261a
1 parent 6ae816e
commit dbe261a
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 73 deletions.
diff --git a/.github/workflows/pr_functional_tests.yml b/.github/workflows/pr_functional_tests.yml
diff --git a/.github/workflows/pr_integration_tests.yml → .github/workflows/pr_tests.yml b/.github/workflows/pr_integration_tests.yml → .github/workflows/pr_tests.yml
@@ -14,7 +14,6 @@ concurrency:
 
 env:
   OPTIMUM_NVIDIA_IS_CI: ON
-  RUN_CPU_ONLY: OFF
   RUN_NIGHTLY: OFF
   RUN_SLOW: ON
   HF_HUB_ENABLE_HF_TRANSFER: ON
@@ -27,10 +26,11 @@ jobs:
         config:
           - name: GPU-enabled Optimum-Nvidia Integration Test Suite
             image: nvidia/cuda:12.1.0-devel-ubuntu22.04
-        cuda_arch: [sm_86, sm_89]
+        gpu_target: ["nvidia-multi-gpu-l4-runners", "nvidia-multi-gpu-a10-runners"]
 
     name: ${{ matrix.config.name }}
-    runs-on: [ci, nvidia-gpu, multi-gpu, "${{ matrix.cuda_arch }}"]
+    runs-on:
+      group: "${{matrix.gpu_target}}"
 
     container:
       image: ${{ matrix.config.image }}

diff --git a/src/optimum/nvidia/__init__.py b/src/optimum/nvidia/__init__.py
@@ -21,5 +21,5 @@
 from .models import AutoModelForCausalLM
 from .optimizations import IntoModelOptQuantizeConfig
 
-# from .pipelines import pipeline
+from .pipelines import pipeline
 from .version import VERSION, __version__