vllm/.buildkite/test-template.j2

57 lines
1.8 KiB
Django/Jinja

{% set docker_image = "us-central1-docker.pkg.dev/vllm-405802/vllm-ci-test-repo/vllm-test:$BUILDKITE_COMMIT" %}
{% set default_num_gpu = 1 %}
{% set default_working_dir = "/vllm-workspace/tests" %}
steps:
- label: ":docker: build image"
commands:
- "docker build --build-arg max_jobs=16 --tag {{ docker_image }} --target test --progress plain ."
- "docker push {{ docker_image }}"
env:
DOCKER_BUILDKIT: "1"
retry:
automatic:
- exit_status: -1 # Agent was lost
limit: 5
- wait
{% for step in steps %}
- label: "{{ step.label }}"
agents:
queue: kubernetes
soft_fail: {{ step.soft_fail or false }}
retry:
automatic:
- exit_status: -1 # Agent was lost
limit: 5
plugins:
- kubernetes:
podSpec:
volumes:
- name: dshm
emptyDir:
medium: Memory
containers:
- image: "{{ docker_image }}"
command: ["bash"]
args:
- '-c'
- "'cd {{ (step.working_dir or default_working_dir) | safe }} && {{ step.command or (step.commands | join(' && ')) | safe }}'"
{% if not step.no_gpu %}
resources:
requests:
nvidia.com/gpu: "{{ step.num_gpus or default_num_gpu }}"
limits:
nvidia.com/gpu: "{{ step.num_gpus or default_num_gpu }}"
{% endif %}
env:
- name: HF_TOKEN
valueFrom:
secretKeyRef:
name: hf-token-secret
key: token
volumeMounts:
- mountPath: /dev/shm
name: dshm
{% endfor %}