[ROCm][CI] rm duplicate Distributed Torchrun ci test (#46729)
Signed-off-by: Divakar Verma <divakar.verma@amd.com>
Divakar Verma committed
Jun 25, 2026 at 10:58 UTC
cdfa2fd7e9eb26d911312989b6458d71f248608e
1 file changed
-24
.buildkite/test-amd.yaml
-24
@@ -221,30 +221,6 @@ steps:
221
- pytest -v -s distributed/test_shm_buffer.py
222
- pytest -v -s distributed/test_shm_storage.py
223
224
-- label: Distributed Torchrun + Shutdown Tests (2 GPUs) # TBD
225
- timeout_in_minutes: 180
226
- mirror_hardwares: [amdexperimental, amdproduction, amdgfx90anightly, amdmi250]
227
- agent_pool: mi250_2
228
- num_gpus: 2
229
- optional: true
230
- working_dir: "/vllm-workspace/tests"
231
- source_file_dependencies:
232
- - vllm/distributed/
233
- - vllm/engine/
234
- - vllm/executor/
235
- - vllm/worker/worker_base.py
236
- - vllm/v1/engine/
237
- - vllm/v1/worker/
238
- - tests/distributed/
239
- - tests/v1/shutdown
240
- - tests/v1/worker/test_worker_memory_snapshot.py
241
- - vllm/platforms/rocm.py
242
- commands:
243
- - VLLM_TEST_SAME_HOST=1 torchrun --nproc-per-node=4 distributed/test_same_node.py | grep 'Same node test passed'
244
- - VLLM_TEST_SAME_HOST=1 VLLM_TEST_WITH_DEFAULT_DEVICE_SET=1 torchrun --nproc-per-node=4 distributed/test_same_node.py | grep 'Same node test passed'
245
- - CUDA_VISIBLE_DEVICES=0,1 pytest -v -s v1/shutdown
246
- - pytest -v -s v1/worker/test_worker_memory_snapshot.py
247
-
224
- label: Pipeline + Context Parallelism (4 GPUs) # TBD
225
timeout_in_minutes: 180
226
mirror_hardwares: [amdexperimental, amdproduction, amdgfx90anightly, amdmi250]