41 lines
1.0 KiB
YAML
41 lines
1.0 KiB
YAML
group: Expert Parallelism
|
|
depends_on:
|
|
- image-build
|
|
steps:
|
|
- label: EPLB Algorithm
|
|
timeout_in_minutes: 15
|
|
device: h200_18gb
|
|
working_dir: "/vllm-workspace/tests"
|
|
source_file_dependencies:
|
|
- vllm/distributed/eplb
|
|
- tests/distributed/test_eplb_algo.py
|
|
- tests/distributed/test_eplb_utils.py
|
|
commands:
|
|
- pytest -v -s distributed/test_eplb_algo.py
|
|
- pytest -v -s distributed/test_eplb_utils.py
|
|
|
|
- label: EPLB Execution # 17min
|
|
timeout_in_minutes: 27
|
|
working_dir: "/vllm-workspace/tests"
|
|
num_devices: 4
|
|
source_file_dependencies:
|
|
- vllm/distributed/eplb
|
|
- tests/distributed/test_eplb_execute.py
|
|
commands:
|
|
- pytest -v -s distributed/test_eplb_execute.py
|
|
- pytest -v -s distributed/test_eplb_spec_decode.py
|
|
|
|
- label: Elastic EP Scaling Test
|
|
timeout_in_minutes: 20
|
|
device: h100
|
|
working_dir: "/vllm-workspace/tests"
|
|
num_devices: 4
|
|
source_file_dependencies:
|
|
- vllm/distributed/
|
|
- vllm/engine/
|
|
- vllm/executor/
|
|
- vllm/compilation/
|
|
- tests/distributed/
|
|
commands:
|
|
- pytest -v -s distributed/test_elastic_ep.py
|