Kubeflow TensorFlow

Directory

Class	Description
`flytekitplugins.kftensorflow.task.Chief`
`flytekitplugins.kftensorflow.task.CleanPodPolicy`	CleanPodPolicy describes how to deal with pods when the job is finished.
`flytekitplugins.kftensorflow.task.Evaluator`
`flytekitplugins.kftensorflow.task.PS`
`flytekitplugins.kftensorflow.task.RestartPolicy`	RestartPolicy describes how the replicas should be restarted.
`flytekitplugins.kftensorflow.task.RunPolicy`	RunPolicy describes a set of policies to apply to the execution of a Kubeflow job.
`flytekitplugins.kftensorflow.task.TensorflowFunctionTask`	Plugin that submits a TFJob (see https://github.com/kubeflow/tf-operator) defined by the code within the _task_function to k8s cluster.
`flytekitplugins.kftensorflow.task.TfJob`	Configuration for an executable `TensorFlow Job`.
`flytekitplugins.kftensorflow.task.Worker`

Package	Description
`flytekitplugins.kftensorflow.task`	This Plugin adds the capability of running distributed tensorflow training to Flyte using backend plugins, natively on Kubernetes.