Buildkite¶

The GitHub webhook is configured to trigger the Buildkite pipeline. The current step configuration of the pipeline:

steps:
  - label: ":pipeline: Upload Pipeline"
    agents:
      queue: cpu
    priority: 200
    command: "bash .buildkite/scripts/bootstrap.sh"

Support Matrices¶

Besides continuous integration and continuous delivery, a major goal of our pipeline is to generate support matrices for our users for each release:

model support matrix (intended to replace this from the vllm upstream)
feature support matrix (intended to replace this from the vllm upstream)
kernel support matrix
kernel support matrix (microbenchmarks)
parallelism support matrix
quantization support matrix
rl support matrix

To support this requirement, each model and feature will go through a series of stages of testing, and the test results will be used to generate the support matrices automatically.

Adding a new model to CI¶

To add a new model to CI, model owners can use the prepared add_model_to_ci.py script. The script will populate a buildkite yaml config file in the .buildkite/models directory; config files under this directory will be integrated to our pipeline automatically.

Interactive Mode¶

The script will prompt you for inputs and generate a YAML in the appropriate directory based on your inputs.

python .buildkite/pipeline_generation/add_model_to_ci.py

Adding a new feature to CI¶

To add a new feature to CI, feature owners can use the prepared add_feature_to_ci.py script. The script will populate a buildkite yaml config file in the appropriate subdirectory (e.g., .buildkite/features, .buildkite/parallelism, etc.). These files will be integrated into our pipeline automatically.

Interactive Mode¶

The script will prompt you for inputs and generate a YAML in the appropriate directory based on your inputs.

python .buildkite/pipeline_generation/add_feature_to_ci.py