v1

package
v0.0.0-...-ebf0d3c Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Oct 23, 2024 License: Apache-2.0 Imports: 15 Imported by: 1

Documentation

Overview

Package v1 is the v1 version of the API. +groupName=kubeflow.org

Package v1 contains API Schema definitions for the kubeflow.org v1 API group +kubebuilder:object:generate=true +groupName=kubeflow.org

Index

Constants

View Source 
const (
	// MPIJobDefaultPortName is name of the port used to communicate between Master and Workers.
	MPIJobDefaultPortName = "mpi-port"
	// MPIJobDefaultPort is default value of the port.
	MPIJobDefaultPort = 9999
	// MPIJobDefaultContainerName is the name of the MPIJob container.
	MPIJobDefaultContainerName = "mpi"
	// MPIJobDefaultRestartPolicy is default RestartPolicy for ReplicaSpec.
	MPIJobDefaultRestartPolicy = commonv1.RestartPolicyNever
	MPIJobKind                 = "MPIJob"
	// MPIJobPlural is the MPIJobPlural for TFJob.
	MPIJobPlural = "mpijobs"
	// MPIJobSingular is the singular for TFJob.
	MPIJobSingular = "mpijob"
	// MPIJobFrameworkName is the name of the ML Framework
	MPIJobFrameworkName = "mpi"
	// MPIJobReplicaTypeLauncher is the type for launcher replica.
	MPIJobReplicaTypeLauncher commonv1.ReplicaType = "Launcher"
	// MPIJobReplicaTypeWorker is the type for worker replicas.
	MPIJobReplicaTypeWorker commonv1.ReplicaType = "Worker"
)
View Source 
const (
	// MXJobDefaultPortName is name of the port used to communicate between scheduler and
	// servers & workers.
	MXJobDefaultPortName = "mxjob-port"
	// MXJobDefaultContainerName is the name of the MXJob container.
	MXJobDefaultContainerName = "mxnet"
	// MXJobDefaultPort is default value of the port.
	MXJobDefaultPort = 9091
	// MXJobDefaultRestartPolicy is default RestartPolicy for MXReplicaSpec.
	MXJobDefaultRestartPolicy = commonv1.RestartPolicyNever
	// MXJobKind is the kind name.
	MXJobKind = "MXJob"
	// MXJobPlural is the MXNetPlural for mxJob.
	MXJobPlural = "mxjobs"
	// MXJobSingular is the singular for mxJob.
	MXJobSingular = "mxjob"
	// MXJobFrameworkName is the name of the ML Framework
	MXJobFrameworkName = "mxnet"
	// MXJobReplicaTypeScheduler is the type for scheduler replica in MXNet.
	MXJobReplicaTypeScheduler commonv1.ReplicaType = "Scheduler"

	// MXJobReplicaTypeServer is the type for parameter servers of distributed MXNet.
	MXJobReplicaTypeServer commonv1.ReplicaType = "Server"

	// MXJobReplicaTypeWorker is the type for workers of distributed MXNet.
	// This is also used for non-distributed MXNet.
	MXJobReplicaTypeWorker commonv1.ReplicaType = "Worker"

	// MXJobReplicaTypeTunerTracker
	// This the auto-tuning tracker e.g. autotvm tracker, it will dispatch tuning task to TunerServer
	MXJobReplicaTypeTunerTracker commonv1.ReplicaType = "TunerTracker"

	// MXJobReplicaTypeTunerServer
	MXJobReplicaTypeTunerServer commonv1.ReplicaType = "TunerServer"

	// MXJobReplicaTypeTuner is the type for auto-tuning of distributed MXNet.
	// This is also used for non-distributed MXNet.
	MXJobReplicaTypeTuner commonv1.ReplicaType = "Tuner"
)
View Source 
const (
	// PaddleJobDefaultPortName is name of the port used to communicate between Master and
	// workers.
	PaddleJobDefaultPortName = "master"
	// PaddleJobDefaultContainerName is the name of the PaddleJob container.
	PaddleJobDefaultContainerName = "paddle"
	// PaddleJobDefaultPort is default value of the port.
	PaddleJobDefaultPort = 36543
	// PaddleJobDefaultRestartPolicy is default RestartPolicy for PaddleReplicaSpec.
	PaddleJobDefaultRestartPolicy = commonv1.RestartPolicyOnFailure
	// PaddleJobKind is the kind name.
	PaddleJobKind = "PaddleJob"
	// PaddleJobPlural is the PaddlePlural for paddleJob.
	PaddleJobPlural = "paddlejobs"
	// PaddleJobSingular is the singular for paddleJob.
	PaddleJobSingular = "paddlejob"
	// PaddleJobFrameworkName is the name of the ML Framework
	PaddleJobFrameworkName = "paddle"
	// PaddleJobReplicaTypeMaster is the type of Master of distributed Paddle
	PaddleJobReplicaTypeMaster commonv1.ReplicaType = "Master"
	// PaddleJobReplicaTypeWorker is the type for workers of distributed Paddle.
	PaddleJobReplicaTypeWorker commonv1.ReplicaType = "Worker"
)
View Source 
const (
	// PytorchJobDefaultPortName is name of the port used to communicate between Master and
	// workers.
	PytorchJobDefaultPortName = "pytorchjob-port"
	// PytorchJobDefaultContainerName is the name of the PyTorchJob container.
	PytorchJobDefaultContainerName = "pytorch"
	// PytorchJobDefaultPort is default value of the port.
	PytorchJobDefaultPort = 23456
	// PytorchJobDefaultRestartPolicy is default RestartPolicy for PyTorchReplicaSpec.
	PytorchJobDefaultRestartPolicy = commonv1.RestartPolicyOnFailure
	// PytorchJobKind is the kind name.
	PytorchJobKind = "PyTorchJob"
	// PytorchJobPlural is the PytorchPlural for pytorchJob.
	PytorchJobPlural = "pytorchjobs"
	// PytorchJobSingular is the singular for pytorchJob.
	PytorchJobSingular = "pytorchjob"
	// PytorchJobFrameworkName is the name of the ML Framework
	PytorchJobFrameworkName = "pytorch"
	// PyTorchJobReplicaTypeMaster is the type of Master of distributed PyTorch
	PyTorchJobReplicaTypeMaster commonv1.ReplicaType = "Master"
	// PyTorchJobReplicaTypeWorker is the type for workers of distributed PyTorch.
	PyTorchJobReplicaTypeWorker commonv1.ReplicaType = "Worker"
)
View Source 
const (
	// TFJobDefaultPortName is name of the port used to communicate between PS and
	// workers.
	TFJobDefaultPortName = "tfjob-port"
	// TFJobDefaultContainerName is the name of the TFJob container.
	TFJobDefaultContainerName = "tensorflow"
	// TFJobDefaultPort is default value of the port.
	TFJobDefaultPort = 2222
	// TFJobDefaultRestartPolicy is default RestartPolicy for TFReplicaSpec.
	TFJobDefaultRestartPolicy = commonv1.RestartPolicyNever
	// TFJobKind is the kind name.
	TFJobKind = "TFJob"
	// TFJobPlural is the TensorflowPlural for TFJob.
	TFJobPlural = "tfjobs"
	// TFJobSingular is the singular for TFJob.
	TFJobSingular = "tfjob"
	// TFJobFrameworkName is the name of the ML Framework
	TFJobFrameworkName = "tensorflow"
)
View Source 
const (
	// TFJobReplicaTypePS is the type for parameter servers of distributed TensorFlow.
	TFJobReplicaTypePS commonv1.ReplicaType = "PS"

	// TFJobReplicaTypeWorker is the type for workers of distributed TensorFlow.
	// This is also used for non-distributed TensorFlow.
	TFJobReplicaTypeWorker commonv1.ReplicaType = "Worker"

	// TFJobReplicaTypeChief is the type for chief worker of distributed TensorFlow.
	// If there is "chief" replica type, it's the "chief worker".
	// Else, worker:0 is the chief worker.
	TFJobReplicaTypeChief commonv1.ReplicaType = "Chief"

	// TFJobReplicaTypeMaster is the type for master worker of distributed TensorFlow.
	// This is similar to chief, and kept just for backwards compatibility.
	TFJobReplicaTypeMaster commonv1.ReplicaType = "Master"

	// TFJobReplicaTypeEval is the type for evaluation replica in TensorFlow.
	TFJobReplicaTypeEval commonv1.ReplicaType = "Evaluator"
)
View Source 
const (
	// XGBoostJobDefaultPortName is name of the port used to communicate between Master and Workers.
	XGBoostJobDefaultPortName = "xgboostjob-port"
	// XGBoostJobDefaultContainerName is the name of the XGBoostJob container.
	XGBoostJobDefaultContainerName = "xgboost"
	// XGBoostJobDefaultPort is default value of the port.
	XGBoostJobDefaultPort = 9999
	// XGBoostJobDefaultRestartPolicy is default RestartPolicy for XGBReplicaSpecs.
	XGBoostJobDefaultRestartPolicy = commonv1.RestartPolicyNever
	// XGBoostJobKind is the kind name.
	XGBoostJobKind = "XGBoostJob"
	// XGBoostJobPlural is the XGBoostJobPlural for XGBoostJob.
	XGBoostJobPlural = "xgboostjobs"
	// XGBoostJobSingular is the singular for XGBoostJob.
	XGBoostJobSingular = "xgboostjob"
	// XGBoostJobFrameworkName is the name of the ML Framework
	XGBoostJobFrameworkName = "xgboost"
	// XGBoostJobReplicaTypeMaster is the type for master replica.
	XGBoostJobReplicaTypeMaster commonv1.ReplicaType = "Master"
	// XGBoostJobReplicaTypeWorker is the type for worker replicas.
	XGBoostJobReplicaTypeWorker commonv1.ReplicaType = "Worker"
)

Variables

View Source 
var (
	// GroupVersion is group version used to register these objects
	GroupVersion = schema.GroupVersion{Group: "kubeflow.org", Version: "v1"}

	MPIJobSchemeGroupVersionKind = schema.GroupVersionKind{Group: "kubeflow.org", Version: "v1", Kind: MPIJobKind}

	// SchemeBuilder is used to add go types to the GroupVersionKind scheme
	SchemeBuilder = &scheme.Builder{GroupVersion: GroupVersion}

	// AddToScheme adds the types in this group-version to the given scheme.
	AddToScheme = SchemeBuilder.AddToScheme
)
View Source 
var SchemeGroupVersion = GroupVersion

SchemeGroupVersion is group version used to register these objects.

Functions

func GetOpenAPIDefinitions 

func GetOpenAPIDefinitions(ref common.ReferenceCallback) map[string]common.OpenAPIDefinition

func IsChieforMaster 

func IsChieforMaster(typ commonv1.ReplicaType) bool

IsChieforMaster returns true if the type is Master or Chief.

func IsEvaluator 

func IsEvaluator(typ commonv1.ReplicaType) bool

IsEvaluator returns true if the type is Evaluator.

func IsScheduler 

func IsScheduler(typ commonv1.ReplicaType) bool

IsScheduler returns true if the type is Scheduler.

func IsWorker 

func IsWorker(typ commonv1.ReplicaType) bool

IsWorker returns true if the type is Worker.

func RegisterDefaults 

func RegisterDefaults(scheme *runtime.Scheme) error

RegisterDefaults adds defaulters functions to the given scheme. Public to allow building arbitrary schemes. All generated defaulters are covering - they call all nested defaulters.

func Resource 

func Resource(resource string) schema.GroupResource

Resource takes an unqualified resource and returns a Group-qualified GroupResource.

func SetDefaults_MPIJob 

func SetDefaults_MPIJob(mpiJob *MPIJob)

func SetDefaults_MXJob 

func SetDefaults_MXJob(mxjob *MXJob)

SetDefaults_MXJob sets any unspecified values to defaults.

func SetDefaults_PaddleJob 

func SetDefaults_PaddleJob(job *PaddleJob)

SetDefaults_PaddleJob sets any unspecified values to defaults.

func SetDefaults_PyTorchJob 

func SetDefaults_PyTorchJob(job *PyTorchJob)

SetDefaults_PyTorchJob sets any unspecified values to defaults.

func SetDefaults_TFJob 

func SetDefaults_TFJob(tfJob *TFJob)

SetDefaults_TFJob sets any unspecified values to defaults.

func SetDefaults_XGBoostJob 

func SetDefaults_XGBoostJob(xgboostJob *XGBoostJob)

SetDefaults_XGBoostJob sets any unspecified values to defaults.

func SetObjectDefaults_MPIJob 

func SetObjectDefaults_MPIJob(in *MPIJob)

func SetObjectDefaults_MPIJobList 

func SetObjectDefaults_MPIJobList(in *MPIJobList)

func SetObjectDefaults_MXJob 

func SetObjectDefaults_MXJob(in *MXJob)

func SetObjectDefaults_MXJobList 

func SetObjectDefaults_MXJobList(in *MXJobList)

func SetObjectDefaults_PaddleJob 

func SetObjectDefaults_PaddleJob(in *PaddleJob)

func SetObjectDefaults_PaddleJobList 

func SetObjectDefaults_PaddleJobList(in *PaddleJobList)

func SetObjectDefaults_PyTorchJob 

func SetObjectDefaults_PyTorchJob(in *PyTorchJob)

func SetObjectDefaults_PyTorchJobList 

func SetObjectDefaults_PyTorchJobList(in *PyTorchJobList)

func SetObjectDefaults_TFJob 

func SetObjectDefaults_TFJob(in *TFJob)

func SetObjectDefaults_TFJobList 

func SetObjectDefaults_TFJobList(in *TFJobList)

func SetObjectDefaults_XGBoostJob 

func SetObjectDefaults_XGBoostJob(in *XGBoostJob)

func SetObjectDefaults_XGBoostJobList 

func SetObjectDefaults_XGBoostJobList(in *XGBoostJobList)

func ValidateV1MXJob 

func ValidateV1MXJob(mxJob *MXJob) error

ValidateV1MXJob checks that the kubeflowv1.MXJobSpec is valid.

func ValidateV1MpiJobSpec 

func ValidateV1MpiJobSpec(c *MPIJobSpec) error

func ValidateV1PaddleJob 

func ValidateV1PaddleJob(paddleJob *PaddleJob) error

func ValidateV1PyTorchJob 

func ValidateV1PyTorchJob(pytorchJob *PyTorchJob) error

func ValidateV1TFJob 

func ValidateV1TFJob(tfjob *TFJob) error

func ValidateV1XGBoostJob 

func ValidateV1XGBoostJob(xgboostJob *XGBoostJob) error

Types

type ElasticPolicy 

type ElasticPolicy struct {
	// minReplicas is the lower limit for the number of replicas to which the training job
	// can scale down.  It defaults to null.
	// +optional
	MinReplicas *int32 `json:"minReplicas,omitempty"`
	// upper limit for the number of pods that can be set by the autoscaler; cannot be smaller than MinReplicas, defaults to null.
	// +optional
	MaxReplicas *int32 `json:"maxReplicas,omitempty"`

	RDZVBackend *RDZVBackend `json:"rdzvBackend,omitempty"`
	RDZVPort    *int32       `json:"rdzvPort,omitempty"`
	RDZVHost    *string      `json:"rdzvHost,omitempty"`
	RDZVID      *string      `json:"rdzvId,omitempty"`
	// RDZVConf contains additional rendezvous configuration (<key1>=<value1>,<key2>=<value2>,...).
	RDZVConf []RDZVConf `json:"rdzvConf,omitempty"`
	// Start a local standalone rendezvous backend that is represented by a C10d TCP store
	// on port 29400. Useful when launching single-node, multi-worker job. If specified
	// --rdzv_backend, --rdzv_endpoint, --rdzv_id are auto-assigned; any explicitly set values
	// are ignored.
	Standalone *bool `json:"standalone,omitempty"`
	// Number of workers per node; supported values: [auto, cpu, gpu, int].
	NProcPerNode *int32 `json:"nProcPerNode,omitempty"`

	MaxRestarts *int32 `json:"maxRestarts,omitempty"`

	// Metrics contains the specifications which are used to calculate the
	// desired replica count (the maximum replica count across all metrics will
	// be used).  The desired replica count is calculated with multiplying the
	// ratio between the target value and the current value by the current
	// number of pods. Ergo, metrics used must decrease as the pod count is
	// increased, and vice-versa.  See the individual metric source types for
	// more information about how each type of metric must respond.
	// If not set, the HPA will not be created.
	// +optional
	Metrics []autoscalingv2.MetricSpec `json:"metrics,omitempty"`
}

func (*ElasticPolicy) DeepCopy 

func (in *ElasticPolicy) DeepCopy() *ElasticPolicy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ElasticPolicy.

func (*ElasticPolicy) DeepCopyInto 

func (in *ElasticPolicy) DeepCopyInto(out *ElasticPolicy)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type JobModeType 

type JobModeType string

JobModeType id the type for JobMode 

const (
	// Train Mode, in this mode requested MXReplicaSpecs need
	// has Server, Scheduler, Worker
	MXTrain JobModeType = "MXTrain"

	// Tune Mode, in this mode requested MXReplicaSpecs need
	// has Tuner
	MXTune JobModeType = "MXTune"
)

type MPIJob 

type MPIJob struct {
	metav1.TypeMeta   `json:",inline"`
	metav1.ObjectMeta `json:"metadata,omitempty"`
	Spec              MPIJobSpec         `json:"spec,omitempty"`
	Status            commonv1.JobStatus `json:"status,omitempty"`
}

func (*MPIJob) DeepCopy 

func (in *MPIJob) DeepCopy() *MPIJob

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MPIJob.

func (*MPIJob) DeepCopyInto 

func (in *MPIJob) DeepCopyInto(out *MPIJob)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*MPIJob) DeepCopyObject 

func (in *MPIJob) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type MPIJobList 

type MPIJobList struct {
	metav1.TypeMeta `json:",inline"`
	metav1.ListMeta `json:"metadata,omitempty"`
	Items           []MPIJob `json:"items"`
}

func (*MPIJobList) DeepCopy 

func (in *MPIJobList) DeepCopy() *MPIJobList

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MPIJobList.

func (*MPIJobList) DeepCopyInto 

func (in *MPIJobList) DeepCopyInto(out *MPIJobList)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*MPIJobList) DeepCopyObject 

func (in *MPIJobList) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type MPIJobSpec 

type MPIJobSpec struct {

	// Specifies the number of slots per worker used in hostfile.
	// Defaults to 1.
	// +optional
	SlotsPerWorker *int32 `json:"slotsPerWorker,omitempty"`

	// CleanPodPolicy defines the policy that whether to kill pods after the job completes.
	// Defaults to None.
	CleanPodPolicy *commonv1.CleanPodPolicy `json:"cleanPodPolicy,omitempty"`

	// `MPIReplicaSpecs` contains maps from `MPIReplicaType` to `ReplicaSpec` that
	// specify the MPI replicas to run.
	MPIReplicaSpecs map[commonv1.ReplicaType]*commonv1.ReplicaSpec `json:"mpiReplicaSpecs"`

	// MainContainer specifies name of the main container which
	// executes the MPI code.
	MainContainer string `json:"mainContainer,omitempty"`

	// `RunPolicy` encapsulates various runtime policies of the distributed training
	// job, for example how to clean up resources and how long the job can stay
	// active.
	RunPolicy commonv1.RunPolicy `json:"runPolicy,omitempty"`
}

func (*MPIJobSpec) DeepCopy 

func (in *MPIJobSpec) DeepCopy() *MPIJobSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MPIJobSpec.

func (*MPIJobSpec) DeepCopyInto 

func (in *MPIJobSpec) DeepCopyInto(out *MPIJobSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type MXJob 

type MXJob struct {
	metav1.TypeMeta   `json:",inline"`
	metav1.ObjectMeta `json:"metadata,omitempty"`

	Spec   MXJobSpec          `json:"spec,omitempty"`
	Status commonv1.JobStatus `json:"status,omitempty"`
}

MXJob is the Schema for the mxjobs API

func (*MXJob) DeepCopy 

func (in *MXJob) DeepCopy() *MXJob

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MXJob.

func (*MXJob) DeepCopyInto 

func (in *MXJob) DeepCopyInto(out *MXJob)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*MXJob) DeepCopyObject 

func (in *MXJob) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type MXJobList 

type MXJobList struct {
	metav1.TypeMeta `json:",inline"`
	metav1.ListMeta `json:"metadata,omitempty"`
	Items           []MXJob `json:"items"`
}

MXJobList contains a list of MXJob

func (*MXJobList) DeepCopy 

func (in *MXJobList) DeepCopy() *MXJobList

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MXJobList.

func (*MXJobList) DeepCopyInto 

func (in *MXJobList) DeepCopyInto(out *MXJobList)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*MXJobList) DeepCopyObject 

func (in *MXJobList) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type MXJobSpec 

type MXJobSpec struct {
	// RunPolicy encapsulates various runtime policies of the distributed training
	// job, for example how to clean up resources and how long the job can stay
	// active.
	//+kubebuilder:validation:Optional
	RunPolicy commonv1.RunPolicy `json:"runPolicy"`

	// JobMode specify the kind of MXjob to do. Different mode may have
	// different MXReplicaSpecs request
	JobMode JobModeType `json:"jobMode"`

	// MXReplicaSpecs is map of commonv1.ReplicaType and commonv1.ReplicaSpec
	// specifies the MX replicas to run.
	// For example,
	//   {
	//     "Scheduler": commonv1.ReplicaSpec,
	//     "Server": commonv1.ReplicaSpec,
	//     "Worker": commonv1.ReplicaSpec,
	//   }
	MXReplicaSpecs map[commonv1.ReplicaType]*commonv1.ReplicaSpec `json:"mxReplicaSpecs"`
}

MXJobSpec defines the desired state of MXJob

func (*MXJobSpec) DeepCopy 

func (in *MXJobSpec) DeepCopy() *MXJobSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MXJobSpec.

func (*MXJobSpec) DeepCopyInto 

func (in *MXJobSpec) DeepCopyInto(out *MXJobSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type MXJobStatus 

type MXJobStatus struct {
}

MXJobStatus defines the observed state of MXJob

func (*MXJobStatus) DeepCopy 

func (in *MXJobStatus) DeepCopy() *MXJobStatus

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MXJobStatus.

func (*MXJobStatus) DeepCopyInto 

func (in *MXJobStatus) DeepCopyInto(out *MXJobStatus)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type PaddleElasticPolicy 

type PaddleElasticPolicy struct {
	// minReplicas is the lower limit for the number of replicas to which the training job
	// can scale down.  It defaults to null.
	// +optional
	MinReplicas *int32 `json:"minReplicas,omitempty"`
	// upper limit for the number of pods that can be set by the autoscaler; cannot be smaller than MinReplicas, defaults to null.
	// +optional
	MaxReplicas *int32 `json:"maxReplicas,omitempty"`

	// MaxRestarts is the limit for restart times of pods in elastic mode.
	// +optional
	MaxRestarts *int32 `json:"maxRestarts,omitempty"`

	// Metrics contains the specifications which are used to calculate the
	// desired replica count (the maximum replica count across all metrics will
	// be used).  The desired replica count is calculated with multiplying the
	// ratio between the target value and the current value by the current
	// number of pods. Ergo, metrics used must decrease as the pod count is
	// increased, and vice-versa.  See the individual metric source types for
	// more information about how each type of metric must respond.
	// If not set, the HPA will not be created.
	// +optional
	Metrics []autoscalingv2.MetricSpec `json:"metrics,omitempty"`
}

func (*PaddleElasticPolicy) DeepCopy 

func (in *PaddleElasticPolicy) DeepCopy() *PaddleElasticPolicy

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PaddleElasticPolicy.

func (*PaddleElasticPolicy) DeepCopyInto 

func (in *PaddleElasticPolicy) DeepCopyInto(out *PaddleElasticPolicy)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type PaddleJob 

type PaddleJob struct {
	// Standard Kubernetes type metadata.
	metav1.TypeMeta `json:",inline"`

	metav1.ObjectMeta `json:"metadata,omitempty"`

	// Specification of the desired state of the PaddleJob.
	Spec PaddleJobSpec `json:"spec,omitempty"`

	// Most recently observed status of the PaddleJob.
	// Read-only (modified by the system).
	Status commonv1.JobStatus `json:"status,omitempty"`
}

PaddleJob Represents a PaddleJob resource.

func (*PaddleJob) DeepCopy 

func (in *PaddleJob) DeepCopy() *PaddleJob

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PaddleJob.

func (*PaddleJob) DeepCopyInto 

func (in *PaddleJob) DeepCopyInto(out *PaddleJob)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*PaddleJob) DeepCopyObject 

func (in *PaddleJob) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type PaddleJobList 

type PaddleJobList struct {
	// Standard type metadata.
	metav1.TypeMeta `json:",inline"`

	// Standard list metadata.
	metav1.ListMeta `json:"metadata,omitempty"`

	// List of PaddleJobs.
	Items []PaddleJob `json:"items"`
}

PaddleJobList is a list of PaddleJobs.

func (*PaddleJobList) DeepCopy 

func (in *PaddleJobList) DeepCopy() *PaddleJobList

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PaddleJobList.

func (*PaddleJobList) DeepCopyInto 

func (in *PaddleJobList) DeepCopyInto(out *PaddleJobList)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*PaddleJobList) DeepCopyObject 

func (in *PaddleJobList) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type PaddleJobSpec 

type PaddleJobSpec struct {
	// RunPolicy encapsulates various runtime policies of the distributed training
	// job, for example how to clean up resources and how long the job can stay
	// active.
	//+kubebuilder:validation:Optional
	RunPolicy commonv1.RunPolicy `json:"runPolicy"`

	// ElasticPolicy holds the elastic policy for paddle job.
	ElasticPolicy *PaddleElasticPolicy `json:"elasticPolicy,omitempty"`

	// A map of PaddleReplicaType (type) to ReplicaSpec (value). Specifies the Paddle cluster configuration.
	// For example,
	//   {
	//     "Master": PaddleReplicaSpec,
	//     "Worker": PaddleReplicaSpec,
	//   }
	PaddleReplicaSpecs map[commonv1.ReplicaType]*commonv1.ReplicaSpec `json:"paddleReplicaSpecs"`
}

PaddleJobSpec is a desired state description of the PaddleJob.

func (*PaddleJobSpec) DeepCopy 

func (in *PaddleJobSpec) DeepCopy() *PaddleJobSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PaddleJobSpec.

func (*PaddleJobSpec) DeepCopyInto 

func (in *PaddleJobSpec) DeepCopyInto(out *PaddleJobSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type PyTorchJob 

type PyTorchJob struct {
	// Standard Kubernetes type metadata.
	metav1.TypeMeta `json:",inline"`

	metav1.ObjectMeta `json:"metadata,omitempty"`

	// Specification of the desired state of the PyTorchJob.
	Spec PyTorchJobSpec `json:"spec,omitempty"`

	// Most recently observed status of the PyTorchJob.
	// Read-only (modified by the system).
	Status commonv1.JobStatus `json:"status,omitempty"`
}

PyTorchJob Represents a PyTorchJob resource.

func (*PyTorchJob) DeepCopy 

func (in *PyTorchJob) DeepCopy() *PyTorchJob

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PyTorchJob.

func (*PyTorchJob) DeepCopyInto 

func (in *PyTorchJob) DeepCopyInto(out *PyTorchJob)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*PyTorchJob) DeepCopyObject 

func (in *PyTorchJob) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type PyTorchJobList 

type PyTorchJobList struct {
	// Standard type metadata.
	metav1.TypeMeta `json:",inline"`

	// Standard list metadata.
	metav1.ListMeta `json:"metadata,omitempty"`

	// List of PyTorchJobs.
	Items []PyTorchJob `json:"items"`
}

PyTorchJobList is a list of PyTorchJobs.

func (*PyTorchJobList) DeepCopy 

func (in *PyTorchJobList) DeepCopy() *PyTorchJobList

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PyTorchJobList.

func (*PyTorchJobList) DeepCopyInto 

func (in *PyTorchJobList) DeepCopyInto(out *PyTorchJobList)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*PyTorchJobList) DeepCopyObject 

func (in *PyTorchJobList) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type PyTorchJobSpec 

type PyTorchJobSpec struct {
	// RunPolicy encapsulates various runtime policies of the distributed training
	// job, for example how to clean up resources and how long the job can stay
	// active.
	//+kubebuilder:validation:Optional
	RunPolicy commonv1.RunPolicy `json:"runPolicy"`

	ElasticPolicy *ElasticPolicy `json:"elasticPolicy,omitempty"`

	// A map of PyTorchReplicaType (type) to ReplicaSpec (value). Specifies the PyTorch cluster configuration.
	// For example,
	//   {
	//     "Master": PyTorchReplicaSpec,
	//     "Worker": PyTorchReplicaSpec,
	//   }
	PyTorchReplicaSpecs map[commonv1.ReplicaType]*commonv1.ReplicaSpec `json:"pytorchReplicaSpecs"`
}

PyTorchJobSpec is a desired state description of the PyTorchJob.

func (*PyTorchJobSpec) DeepCopy 

func (in *PyTorchJobSpec) DeepCopy() *PyTorchJobSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PyTorchJobSpec.

func (*PyTorchJobSpec) DeepCopyInto 

func (in *PyTorchJobSpec) DeepCopyInto(out *PyTorchJobSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type RDZVBackend 

type RDZVBackend string
const (
	// BackendC10D is the rendezvous backend type for C10d.
	BackendC10D RDZVBackend = "c10d"
	// BackendETCD is the rendezvous backend type for ETCD.
	BackendETCD RDZVBackend = "etcd"
	// BackendETCDV2 is the rendezvous backend type for ETCD v2.
	BackendETCDV2 RDZVBackend = "etcd-v2"
)

type RDZVConf 

type RDZVConf struct {
	Key   string `json:"key,omitempty"`
	Value string `json:"value,omitempty"`
}

func (*RDZVConf) DeepCopy 

func (in *RDZVConf) DeepCopy() *RDZVConf

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new RDZVConf.

func (*RDZVConf) DeepCopyInto 

func (in *RDZVConf) DeepCopyInto(out *RDZVConf)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type SuccessPolicy 

type SuccessPolicy string

SuccessPolicy is the success policy. 

const (
	SuccessPolicyDefault    SuccessPolicy = ""
	SuccessPolicyAllWorkers SuccessPolicy = "AllWorkers"
)

type TFJob 

type TFJob struct {
	// Standard Kubernetes type metadata.
	metav1.TypeMeta `json:",inline"`

	// +optional
	metav1.ObjectMeta `json:"metadata,omitempty"`

	// Specification of the desired state of the TFJob.
	// +optional
	Spec TFJobSpec `json:"spec,omitempty"`

	// Most recently observed status of the TFJob.
	// Populated by the system.
	// Read-only.
	// +optional
	Status commonv1.JobStatus `json:"status,omitempty"`
}

TFJob represents a TFJob resource.

func (*TFJob) DeepCopy 

func (in *TFJob) DeepCopy() *TFJob

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new TFJob.

func (*TFJob) DeepCopyInto 

func (in *TFJob) DeepCopyInto(out *TFJob)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*TFJob) DeepCopyObject 

func (in *TFJob) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type TFJobList 

type TFJobList struct {
	// Standard type metadata.
	metav1.TypeMeta `json:",inline"`

	// Standard list metadata.
	// +optional
	metav1.ListMeta `json:"metadata,omitempty"`

	// List of TFJobs.
	Items []TFJob `json:"items"`
}

TFJobList is a list of TFJobs.

func (*TFJobList) DeepCopy 

func (in *TFJobList) DeepCopy() *TFJobList

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new TFJobList.

func (*TFJobList) DeepCopyInto 

func (in *TFJobList) DeepCopyInto(out *TFJobList)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*TFJobList) DeepCopyObject 

func (in *TFJobList) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type TFJobSpec 

type TFJobSpec struct {
	// RunPolicy encapsulates various runtime policies of the distributed training
	// job, for example how to clean up resources and how long the job can stay
	// active.
	//+kubebuilder:validation:Optional
	RunPolicy commonv1.RunPolicy `json:"runPolicy"`

	// SuccessPolicy defines the policy to mark the TFJob as succeeded.
	// Default to "", using the default rules.
	// +optional
	SuccessPolicy *SuccessPolicy `json:"successPolicy,omitempty"`

	// A map of TFReplicaType (type) to ReplicaSpec (value). Specifies the TF cluster configuration.
	// For example,
	//   {
	//     "PS": ReplicaSpec,
	//     "Worker": ReplicaSpec,
	//   }
	TFReplicaSpecs map[commonv1.ReplicaType]*commonv1.ReplicaSpec `json:"tfReplicaSpecs"`

	// A switch to enable dynamic worker
	EnableDynamicWorker bool `json:"enableDynamicWorker,omitempty"`
}

TFJobSpec is a desired state description of the TFJob.

func (*TFJobSpec) DeepCopy 

func (in *TFJobSpec) DeepCopy() *TFJobSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new TFJobSpec.

func (*TFJobSpec) DeepCopyInto 

func (in *TFJobSpec) DeepCopyInto(out *TFJobSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

type XGBoostJob 

type XGBoostJob struct {
	metav1.TypeMeta   `json:",inline"`
	metav1.ObjectMeta `json:"metadata,omitempty"`

	Spec   XGBoostJobSpec     `json:"spec,omitempty"`
	Status commonv1.JobStatus `json:"status,omitempty"`
}

XGBoostJob is the Schema for the xgboostjobs API +k8s:openapi-gen=true

func (*XGBoostJob) DeepCopy 

func (in *XGBoostJob) DeepCopy() *XGBoostJob

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new XGBoostJob.

func (*XGBoostJob) DeepCopyInto 

func (in *XGBoostJob) DeepCopyInto(out *XGBoostJob)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*XGBoostJob) DeepCopyObject 

func (in *XGBoostJob) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type XGBoostJobList 

type XGBoostJobList struct {
	metav1.TypeMeta `json:",inline"`
	metav1.ListMeta `json:"metadata,omitempty"`
	Items           []XGBoostJob `json:"items"`
}

XGBoostJobList contains a list of XGBoostJob

func (*XGBoostJobList) DeepCopy 

func (in *XGBoostJobList) DeepCopy() *XGBoostJobList

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new XGBoostJobList.

func (*XGBoostJobList) DeepCopyInto 

func (in *XGBoostJobList) DeepCopyInto(out *XGBoostJobList)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

func (*XGBoostJobList) DeepCopyObject 

func (in *XGBoostJobList) DeepCopyObject() runtime.Object

DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.

type XGBoostJobSpec 

type XGBoostJobSpec struct {
	// INSERT ADDITIONAL SPEC FIELDS - desired state of cluster
	// Important: Run "make" to regenerate code after modifying this file
	//+kubebuilder:validation:Optional
	RunPolicy commonv1.RunPolicy `json:"runPolicy"`

	XGBReplicaSpecs map[commonv1.ReplicaType]*commonv1.ReplicaSpec `json:"xgbReplicaSpecs"`
}

XGBoostJobSpec defines the desired state of XGBoostJob

func (*XGBoostJobSpec) DeepCopy 

func (in *XGBoostJobSpec) DeepCopy() *XGBoostJobSpec

DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new XGBoostJobSpec.

func (*XGBoostJobSpec) DeepCopyInto 

func (in *XGBoostJobSpec) DeepCopyInto(out *XGBoostJobSpec)

DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.