[ci] test lint ci and lint tests dir (#28)

* [ci] add linting tests dir

* delete redundant tests

* [ci] change test gpu usage to 2

* [ci] update check ci

* [ci] fix init

* [ci] fix ray test running machine

* [ci] test self-hosted gpu

* [ci] fix yapf lint ci

* [ci] test checkout

* [ci] test origin

* [ci] fix ray test

* [ci] fix ray test

* [ci] add install verl in ray

.github/workflows/ray_test.yml

@@ -18,21 +18,15 @@ on:
 
 jobs:
   ray:
-    runs-on: self-host-gpu
-    strategy:
-      matrix:
-        python-version: ["3.12"]
+    runs-on: [self-hosted, gpu] # test if the enviroment is ready
     steps:
       - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
-      - name: Set up Python ${{ matrix.python-version }}
-        uses: actions/setup-python@0b93645e9fea7318ecaed2b359559ac225c90a2b # v5.3.0
         with:
-          python-version: ${{ matrix.python-version }}
-      # - name: Install dependencies
-      #   run: |
-      #     python -m pip install --upgrade pip
-      #     pip install yapf
-      #     pip install toml==0.10.2
+            fetch-depth: 0
+      - name: Install the current repository
+        run: |
+          pip install -e .
       - name: Running some ray test that only need 2 GPUs
         run: |
+          cd tests/ray
           pytest -s -x test_rvdz.py test_driverfunc_to_worker.py test_data_transfer.py test_colocated_workers.py test_check_worker_alive.py

.github/workflows/yapf_format.yml

@@ -24,8 +24,13 @@ jobs:
         python-version: ["3.12"]
     steps:
       - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
-        with:
-          ref: ${{ github.head_ref }}  # Checkout the branch associated with the pull request
+      # - name: checkout
+      #   run: |
+      #     commits=${{ github.event.pull_request.commits }}
+      #     if [[ -n "$commits" ]]; then
+      #       # Prepare enough depth for diffs with main
+      #       git fetch --depth="$(( commits + 1 ))"
+      #     fi
       - name: Set up Python ${{ matrix.python-version }}
         uses: actions/setup-python@0b93645e9fea7318ecaed2b359559ac225c90a2b # v5.3.0
         with:

tests/ray/check_worker_alive/main.py

@@ -42,7 +42,7 @@ def foo(self, wait_time):
 
     # test single-node-no-partition
     print(f"test single-node-no-partition")
-    resource_pool = RayResourcePool([8], use_gpu=True)
+    resource_pool = RayResourcePool([2], use_gpu=True)
     class_with_args = RayClassWithInitArgs(cls=TestActor)
 
     print("create worker group")

tests/ray/test_data_transfer.py

@@ -52,7 +52,7 @@ def do_nothing(self, data):
 def test_data_transfer():
     ray.init()
     # construct resource pool
-    resource_pool = RayResourcePool([8])
+    resource_pool = RayResourcePool([2])
     cls_with_init = RayClassWithInitArgs(cls=DummyWorker)
     # construct worker group
     wg = RayWorkerGroup(resource_pool, cls_with_init)

tests/ray/test_driverfunc_to_worker.py

@@ -56,7 +56,7 @@ def test():
     # construct model
     ray.init()
 
-    # create 8 workers, each hold a GPU
+    # create 2 workers, each hold a GPU
     resource_pool = RayResourcePool([2], use_gpu=True, name_prefix='a')
 
     class_with_args = RayClassWithInitArgs(cls=ModelActor)

tests/ray/test_rvdz.py

@@ -38,14 +38,14 @@ def test_rvdz():
     ray.init()
 
     group_name = "test_group"
-    world_size = 4
+    world_size = 2
 
     workers = [TestWorker.options(num_gpus=1).remote(rank, world_size, group_name) for rank in range(world_size)]
 
-    [worker.init.remote() for worker in workers]
+    ray.get([worker.init.remote() for worker in workers])
 
     ranks = ray.get([worker.test.remote() for worker in workers])
 
-    assert ranks == [0, 1, 2, 3], f"expecting [0, 1, 2, 3], got {ranks}"
+    assert ranks == [0, 1], f"expecting [0, 1], got {ranks}"
 
     ray.shutdown()