airflow.providers.google.cloud.triggers.mlengine¶

类¶

MLEngineStartTrainingJobTrigger

MLEngineStartTrainingJobTrigger 在 trigger worker 上运行以执行启动训练作业操作。

模块内容¶

class airflow.providers.google.cloud.triggers.mlengine.MLEngineStartTrainingJobTrigger(conn_id, job_id, region, poll_interval=4.0, package_uris=None, training_python_module=None, training_args=None, runtime_version=None, python_version=None, job_dir=None, project_id=PROVIDE_PROJECT_ID, labels=None, gcp_conn_id='google_cloud_default', impersonation_chain=None)[source]¶

基类: airflow.triggers.base.BaseTrigger

MLEngineStartTrainingJobTrigger 在 trigger worker 上运行以执行启动训练作业操作。

参数:

conn_id (str) – 引用 google cloud 连接 id
job_id (str) – 作业的 ID。其后缀将为作业配置的哈希值
project_id (str) – 作业正在运行的 Google Cloud 项目
poll_interval (float) – 检查状态的轮询周期，单位为秒

conn_id[source]¶

job_id[source]¶

project_id = None[source]¶

region[source]¶

poll_interval = 4.0[source]¶

runtime_version = None[source]¶

python_version = None[source]¶

job_dir = None[source]¶

package_uris = None[source]¶

training_python_module = None[source]¶

training_args = None[source]¶

labels = None[source]¶

gcp_conn_id = 'google_cloud_default'[source]¶

impersonation_chain = None[source]¶

serialize()[source]¶

序列化 MLEngineStartTrainingJobTrigger 参数和类路径。

async run()[source]¶

获取当前作业执行状态并生成一个 TriggerEvent。