lcm

package

v0.0.0-...-8225f2e Latest Latest Go to latest Published: Feb 28, 2019 License: Apache-2.0 Imports: 42 Imported by: 1

Details

Valid go.mod file

The Go module system was introduced in Go 1.11 and is the official dependency management solution for Go.
Redistributable license

Redistributable licenses place minimal restrictions on how software can be used, modified, and redistributed.
Tagged version

Modules with tagged versions give importers more predictable builds.
Stable version

When a project reaches major version v1 it is considered stable.
Learn more about best practices

View Source

const PodLevelJobDir = "/job"

PodLevelJobDir represents the place to store the job state indicator files, as well as the $BREAK_FILE and $EXITCODE_FILE.

View Source

const PodLevelLogDir = PodLevelJobDir + "/logs"

PodLevelLogDir represents the place to store the per-learner logs.

View Source

var (
	//NativeFrameworks which support native distribution
	NativeFrameworks = []string{"tensorflow", "caffe2", "mxnet", "horovod", "pytorch"}
)

func GetVolumeClaim(volumeSize int64) (*v1core.PersistentVolumeClaim, error)

GetVolumeClaim returns a PersistentVolumeClaim struct for the given volume size (specified in bytes).

func InitLogger(trainingID string, userID string) *log.Entry

InitLogger ... initializes new logger with trainingID and userID

type Service interface {
	service.LifecycleManagerServer
	service.LifecycleHandler
	StopLCM()
}

Service LCM manages the lifecycle of the entire distributed deep learning job

func NewService() (Service, error)

NewService is a constructor to initialize LCM

type Training interface {
	Start() error
}

Training ...

func NewTraining(ctx context.Context, k8sClient kubernetes.Interface, req *service.JobDeploymentRequest, log *logger.LocLoggingEntry) Training

NewTraining ...