gradient.cli package

Submodules

gradient.cli.cli module

gradient

gradient [OPTIONS] COMMAND [ARGS]...

apiKey

Save your api key

gradient apiKey [OPTIONS] [API_KEY]

Arguments

API_KEY

Optional argument

clusters

Manage clusters

gradient clusters [OPTIONS] COMMAND [ARGS]...
list

List your team clusters

gradient clusters list [OPTIONS]

Options

-l, --limit <cluster_limit>

Limit listed experiments per page

-o, --offset <cluster_offset>
--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

machineTypes

Manage machine types

gradient clusters machineTypes [OPTIONS] COMMAND [ARGS]...
list

List available machine types

gradient clusters machineTypes list [OPTIONS]

Options

--clusterId <cluster_id>

Filter machine types by cluster ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

datasets

Manage datasets

gradient datasets [OPTIONS] COMMAND [ARGS]...
create

Create dataset

gradient datasets create [OPTIONS]

Options

--name <name>

Required Dataset name

--storageProviderId <storage_provider_id>

Required Storage provider ID

--description <description>

Dataset description

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete dataset

gradient datasets delete [OPTIONS]

Options

--id <dataset_id>

Required Dataset ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show dataset details

gradient datasets details [OPTIONS]

Options

--id <dataset_id>

Required Dataset ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

files

Manage files

gradient datasets files [OPTIONS] COMMAND [ARGS]...
delete

Delete files

gradient datasets files delete [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--path <paths>

Sub-directory to delete

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

get

Get files

gradient datasets files get [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--source-path <source_paths>

File or directory to get

--target-path <target_path>

Required Target directory path

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List files

gradient datasets files list [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--path <path>

Sub-directory to list

--recursive <recursive>

Recursive list content

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

put

Put files

gradient datasets files put [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--source-path <source_paths>

Required File or directory to put

--target-path <target_path>

Target dataset file path

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List datasets

gradient datasets list [OPTIONS]

Options

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

tags

Manage tags

gradient datasets tags [OPTIONS] COMMAND [ARGS]...
delete

Delete dataset tag

gradient datasets tags delete [OPTIONS]

Options

--id <dataset_tag_id>

Required Dataset tag (ex: dsr8k5qzn401lb5:prod)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List dataset tags

gradient datasets tags list [OPTIONS]

Options

--id <dataset_id>

Required Dataset ID (ex: dsr8k5qzn401lb5)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

set

Set dataset tag

gradient datasets tags set [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--name <name>

Dataset tag name

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

update

Update dataset

gradient datasets update [OPTIONS]

Options

--id <dataset_id>

Required Dataset ID

--name <name>

Dataset name

--description <description>

Dataset description

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

versions

Manage versions

gradient datasets versions [OPTIONS] COMMAND [ARGS]...
commit

Commit dataset version

gradient datasets versions commit [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

create

Create dataset version

gradient datasets versions create [OPTIONS]

Options

--id <dataset_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--message <message>

Dataset version message

--source-path <source_paths>

Files to put into dataset version (setting this will commit the version)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete dataset version

gradient datasets versions delete [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show dataset version details

gradient datasets versions details [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List dataset versions

gradient datasets versions list [OPTIONS]

Options

--id <dataset_id>

Required Dataset ID (ex: dsr8k5qzn401lb5)

--isCommitted <is_committed>

Show filter by committed status

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

update

Update dataset version

gradient datasets versions update [OPTIONS]

Options

--id <dataset_version_id>

Required Dataset version ID (ex: dsr8k5qzn401lb5:klfoyy9)

--message <message>

Dataset version message

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

deployments

Manage Deployments

gradient deployments [OPTIONS] COMMAND [ARGS]...
create

Create a deployment

gradient deployments create [OPTIONS]

Options

--name <name>

Required Name

--projectId <project_id>

Required Project ID

--clusterId <cluster_id>

Cluster ID

--spec <spec_path>

Required Path to a deployment spec file

--apiKey <api_key>

API key to use this time only

delete

Delete a deployment

gradient deployments delete [OPTIONS]

Options

--apiKey <api_key>

API key to use this time only

--id <id>

Required ID

get

Get a deployment

gradient deployments get [OPTIONS]

Options

--apiKey <api_key>

API key to use this time only

--id <id>

Required ID

list

List deployments

gradient deployments list [OPTIONS]

Options

--apiKey <api_key>

API key to use this time only

update

Update a deployment

gradient deployments update [OPTIONS]

Options

--id <id>

Required ID

--name <name>

Name

--projectId <project_id>

Project ID

--clusterId <cluster_id>

Cluster ID

--spec <spec_path>

Path to a deployment spec file

--apiKey <api_key>

API key to use this time only

experiments

Manage experiments

gradient experiments [OPTIONS] COMMAND [ARGS]...
create

Create new experiment

gradient experiments create [OPTIONS] COMMAND [ARGS]...
multinode

Create multi node experiment

gradient experiments create multinode [OPTIONS]

Options

--name <name>

Name of new experiment

--ports <ports>

Port to use in new experiment

--workspace <workspace>

Path to workspace directory, archive, S3 or git repository

--workspaceRef <workspace_ref>

Git commit hash, branch name or tag

--workspaceUsername <username>

Workspace username

--workspacePassword <workspace_password>

Workspace password

--ignoreFiles <ignore_files>

Ignore certain files from uploading

--workingDirectory <working_directory>

Working directory for the experiment

--artifactDirectory <artifact_directory>

Artifacts directory

--clusterId <cluster ID>

Cluster ID

--experimentEnv <experiment_env>

Environment variables in a JSON

--projectId <project ID>

Required Project ID

--modelType <model type>

Model type

--modelPath <path>

Model path

--isPreemptible

Flag: is preemptible

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--experimentType <experiment_type_id>

Required Experiment Type

Options

GRPC | MPI

--workerContainer <container>

Required Worker container

--workerMachineType <machine type>

Required Worker machine type

--workerCommand <command>

Required Worker command

--workerCount <worker_count>

Required Worker count

--parameterServerContainer <container>

Parameter server container (GRPC only)

--parameterServerMachineType <machine type>

Parameter server machine type (GRPC only)

--parameterServerCommand <command>

Parameter server command (GRPC only)

--parameterServerCount <parameter_server_count>

Parameter server count (GRPC only)

--masterContainer <container>

Master container (MPI only)

--masterMachineType <machine type>

Master machine type (MPI only)

--masterCount <master_count>

Master count (MPI only)

--masterCommand <command>

Master command (MPI only)

--workerContainerUser <worker_container_user>

Worker container user

--workerRegistryUsername <worker_registry_username>

Worker container registry username

--workerRegistryPassword <password>

Worker registry password

--workerRegistryUrl <registry url>

Worker registry URL

--parameterServerContainerUser <parameter_server_container_user>

Parameter server container user

--parameterServerRegistryUsername <parameter_server_registry_username>

Parameter server registry username

--parameterServerRegistryPassword <password>

Parameter server registry password

--parameterServerRegistryUrl <registry url>

Parameter server registry URL

--masterContainerUser <master_container_user>

Master container user (MPI only)

--masterRegistryUsername <username>

Master registry username (MPI only)

--masterRegistryPassword <password>

Master registry password (MPI only)

--masterRegistryUrl <registry url>

Master registry URL (MPI only)

--datasetId <dateset id>

Dataset ID

--datasetUri <dateset uri>

Url to S3 bucket with dataset

--datasetName <dateset name>

Name of dataset

--datasetAwsAccessKeyId <AWS access key>

S3 bucket’s Access Key ID

--datasetAwsSecretAccessKey <dataset_secret_access_key_list>

S3 bucket’s Secret Access Key

--datasetAwsEndpoint <dataset_endpoint_list>

S3 endpoint URL

--datasetVersionId <version ID>

S3 dataset’s version ID

--datasetEtag <etag>

S3 dataset’s ETag

--datasetVolumeKind <dataset_volume_kind_list>

S3 dataset’s volume kind. If used, –datasetVolumeSize has to be set as well

Options

dynamic | shared

--datasetVolumeSize <dataset_volume_size_list>

S3 dataset’s volume size

--tensorboard

Creates new tensorboard for this experiment

--tensorboard_set <tensorboard ID>

Add to existing tensorboard

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

singlenode

Create single node experiment

gradient experiments create singlenode [OPTIONS]

Options

--name <name>

Name of new experiment

--ports <ports>

Port to use in new experiment

--workspace <workspace>

Path to workspace directory, archive, S3 or git repository

--workspaceRef <workspace_ref>

Git commit hash, branch name or tag

--workspaceUsername <username>

Workspace username

--workspacePassword <workspace_password>

Workspace password

--ignoreFiles <ignore_files>

Ignore certain files from uploading

--workingDirectory <working_directory>

Working directory for the experiment

--artifactDirectory <artifact_directory>

Artifacts directory

--clusterId <cluster ID>

Cluster ID

--experimentEnv <experiment_env>

Environment variables in a JSON

--projectId <project ID>

Required Project ID

--modelType <model type>

Model type

--modelPath <path>

Model path

--isPreemptible

Flag: is preemptible

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--container <container>

Required Container

--machineType <machine type>

Required Machine type

--command <command>

Required Container entrypoint command

--containerUser <container_user>

Container user

--registryUsername <registry_username>

Registry username

--registryPassword <password>

Registry password

--registryUrl <registry url>

Registry URL

--datasetId <dateset id>

Dataset ID

--datasetUri <dateset uri>

Url to S3 bucket with dataset

--datasetName <dateset name>

Name of dataset

--datasetAwsAccessKeyId <AWS access key>

S3 bucket’s Access Key ID

--datasetAwsSecretAccessKey <dataset_secret_access_key_list>

S3 bucket’s Secret Access Key

--datasetAwsEndpoint <dataset_endpoint_list>

S3 endpoint URL

--datasetVersionId <version ID>

S3 dataset’s version ID

--datasetEtag <etag>

S3 dataset’s ETag

--datasetVolumeKind <dataset_volume_kind_list>

S3 dataset’s volume kind. If used, –datasetVolumeSize has to be set as well

Options

dynamic | shared

--datasetVolumeSize <dataset_volume_size_list>

S3 dataset’s volume size

--tensorboard

Creates new tensorboard for this experiment

--tensorboard_set <tensorboard ID>

Add to existing tensorboard

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete an experiment

gradient experiments delete [OPTIONS]

Options

--id <id>

Required ID of the experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show detail of an experiment

gradient experiments details [OPTIONS]

Options

--id <id>

Required ID of the experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

hyperparameters

Manage hyperparameters

gradient experiments hyperparameters [OPTIONS] COMMAND [ARGS]...
create

Create hyperparameter

gradient experiments hyperparameters create [OPTIONS]

Options

--name <name>

Name of new experiment

--ports <ports>

Port to use in new experiment

--workspace <workspace>

Path to workspace directory, archive, S3 or git repository

--workspaceRef <workspace_ref>

Git commit hash, branch name or tag

--workspaceUsername <username>

Workspace username

--workspacePassword <workspace_password>

Workspace password

--ignoreFiles <ignore_files>

Ignore certain files from uploading

--workingDirectory <working_directory>

Working directory for the experiment

--artifactDirectory <artifact_directory>

Artifacts directory

--clusterId <cluster ID>

Cluster ID

--experimentEnv <experiment_env>

Environment variables in a JSON

--projectId <project ID>

Required Project ID

--modelType <model type>

Model type

--modelPath <path>

Model path

--isPreemptible

Flag: is preemptible

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--tuningCommand <tuning_command>

Required Tuning command

--workerContainer <worker_container>

Required Worker container

--workerContainerUser <worker_container_user>

Worker container user

--workerMachineType <worker_machine_type>

Required Worker machine type

--hyperparameterServerMachineType <hyperparameter_server_machine_type>

Hyperparameter Server machine type

--workerCommand <worker_command>

Required Worker command

--workerCount <worker_count>

Required Worker count

--workerUseDockerfile

Flag: use dockerfile

--workerDockerfilePath <dockerfile_path>

Path to

--workerRegistryUsername <worker_registry_username>

Worker registry username

--workerRegistryPassword <worker_registry_password>

Worker registry password

--hyperparameterServerRegistryUsername <hyperparameter_server_registry_username>

Hyperparameter server registry username

--hyperparameterServerRegistryPassword <hyperparameter_server_registry_password>

Hyperparameter server registry password

--hyperparameterServerContainer <hyperparameter_server_container>

Hyperparameter server container

--hyperparameterServerContainerUser <hyperparameter_server_container_user>

Hyperparameter server container user

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show details of hyperparameter

gradient experiments hyperparameters details [OPTIONS]

Options

--id <id_>

Required

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List hyperparameters

gradient experiments hyperparameters list [OPTIONS]

Options

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

run

Create and start hyperparameter tuning job

gradient experiments hyperparameters run [OPTIONS]

Options

--name <name>

Name of new experiment

--ports <ports>

Port to use in new experiment

--workspace <workspace>

Path to workspace directory, archive, S3 or git repository

--workspaceRef <workspace_ref>

Git commit hash, branch name or tag

--workspaceUsername <username>

Workspace username

--workspacePassword <workspace_password>

Workspace password

--ignoreFiles <ignore_files>

Ignore certain files from uploading

--workingDirectory <working_directory>

Working directory for the experiment

--artifactDirectory <artifact_directory>

Artifacts directory

--clusterId <cluster ID>

Cluster ID

--experimentEnv <experiment_env>

Environment variables in a JSON

--projectId <project ID>

Required Project ID

--modelType <model type>

Model type

--modelPath <path>

Model path

--isPreemptible

Flag: is preemptible

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--tuningCommand <tuning_command>

Required Tuning command

--workerContainer <worker_container>

Required Worker container

--workerContainerUser <worker_container_user>

Worker container user

--workerMachineType <worker_machine_type>

Required Worker machine type

--hyperparameterServerMachineType <hyperparameter_server_machine_type>

Hyperparameter Server machine type

--workerCommand <worker_command>

Required Worker command

--workerCount <worker_count>

Required Worker count

--workerUseDockerfile

Flag: use dockerfile

--workerDockerfilePath <dockerfile_path>

Path to

--workerRegistryUsername <worker_registry_username>

Worker registry username

--workerRegistryPassword <worker_registry_password>

Worker registry password

--hyperparameterServerRegistryUsername <hyperparameter_server_registry_username>

Hyperparameter server registry username

--hyperparameterServerRegistryPassword <hyperparameter_server_registry_password>

Hyperparameter server registry password

--hyperparameterServerContainer <hyperparameter_server_container>

Hyperparameter server container

--hyperparameterServerContainerUser <hyperparameter_server_container_user>

Hyperparameter server container user

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

start

Start hyperparameter tuning

gradient experiments hyperparameters start [OPTIONS]

Options

--id <id_>

Required

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

tags

Manage hyperparameter tags

gradient experiments hyperparameters tags [OPTIONS] COMMAND [ARGS]...
add

Add tags to hyperparameter

gradient experiments hyperparameters tags add [OPTIONS]

Options

--id <id>

Required ID of the hyperparameter

--tag <tags>

One or many tags that you want to add to hyperparameter

--tags <tags_comma>

Separated by comma tags that you want add to hyperparameter

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

remove

Remove tags from hyperparameter

gradient experiments hyperparameters tags remove [OPTIONS]

Options

--id <id>

Required ID of the hyperparameter

--tag <tags>

One or many tags that you want to remove from hyperparameter

--tags <tags_comma>

Separated by comma tags that you want to remove from hyperparameter

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List experiments

gradient experiments list [OPTIONS]

Options

-p, --projectId <project ID>

Filter by project IDs. Multiple use

--tag <tags>

Filter by tags. Multiple use

-l, --limit <exp_limit>

Limit listed experiments per page

-o, --offset <exp_offset>
--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

logs

List experiment logs

gradient experiments logs [OPTIONS]

Options

--id <experiment_id>

Required

--line <line>
--limit <limit>
--follow <follow>
--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

metrics

Read experiment metrics

gradient experiments metrics [OPTIONS] COMMAND [ARGS]...
get

Get experiment metrics. Shows CPU and RAM usage by default

gradient experiments metrics get [OPTIONS]

Options

--id <experiment_id>

Required ID of the experiment

--metric <metrics_list>

One or more metrics that you want to read: cpuPercentage, memoryUsage, gpuMemoryFree, gpuMemoryUsed, gpuPowerDraw, gpuTemp, gpuUtilization, gpuMemoryUtilization. Defaults to cpuPercentage and memoryUsage. To view available custom metrics, use command: gradient experiments metrics list

--interval <interval>

Interval

--start <start>

Timestamp of first time series metric to collect

--end <end>

Timestamp of last time series metric to collect

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List experiment metrics. Shows CPU and RAM usage by default

gradient experiments metrics list [OPTIONS]

Options

--id <experiment_id>

Required ID of the experiment

--interval <interval>

Interval

--start <start>

Timestamp of first time series metric to collect

--end <end>

Timestamp of last time series metric to collect

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

stream

Watch live experiment metrics. Shows CPU and RAM usage by default

gradient experiments metrics stream [OPTIONS]

Options

--id <experiment_id>

Required ID of the experiment

--metric <metrics_list>

One or more metrics that you want to read. Defaults to cpuPercentage and memoryUsage

Options

cpuPercentage | memoryUsage | gpuMemoryFree | gpuMemoryUsed | gpuPowerDraw | gpuTemp | gpuUtilization | gpuMemoryUtilization

--interval <interval>

Interval

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

run

Create and start new experiment

gradient experiments run [OPTIONS] COMMAND [ARGS]...
multinode

Create and start new multi node experiment

gradient experiments run multinode [OPTIONS]

Options

--name <name>

Name of new experiment

--ports <ports>

Port to use in new experiment

--workspace <workspace>

Path to workspace directory, archive, S3 or git repository

--workspaceRef <workspace_ref>

Git commit hash, branch name or tag

--workspaceUsername <username>

Workspace username

--workspacePassword <workspace_password>

Workspace password

--ignoreFiles <ignore_files>

Ignore certain files from uploading

--workingDirectory <working_directory>

Working directory for the experiment

--artifactDirectory <artifact_directory>

Artifacts directory

--clusterId <cluster ID>

Cluster ID

--experimentEnv <experiment_env>

Environment variables in a JSON

--projectId <project ID>

Required Project ID

--modelType <model type>

Model type

--modelPath <path>

Model path

--isPreemptible

Flag: is preemptible

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--experimentType <experiment_type_id>

Required Experiment Type

Options

GRPC | MPI

--workerContainer <container>

Required Worker container

--workerMachineType <machine type>

Required Worker machine type

--workerCommand <command>

Required Worker command

--workerCount <worker_count>

Required Worker count

--parameterServerContainer <container>

Parameter server container (GRPC only)

--parameterServerMachineType <machine type>

Parameter server machine type (GRPC only)

--parameterServerCommand <command>

Parameter server command (GRPC only)

--parameterServerCount <parameter_server_count>

Parameter server count (GRPC only)

--masterContainer <container>

Master container (MPI only)

--masterMachineType <machine type>

Master machine type (MPI only)

--masterCount <master_count>

Master count (MPI only)

--masterCommand <command>

Master command (MPI only)

--workerContainerUser <worker_container_user>

Worker container user

--workerRegistryUsername <worker_registry_username>

Worker container registry username

--workerRegistryPassword <password>

Worker registry password

--workerRegistryUrl <registry url>

Worker registry URL

--parameterServerContainerUser <parameter_server_container_user>

Parameter server container user

--parameterServerRegistryUsername <parameter_server_registry_username>

Parameter server registry username

--parameterServerRegistryPassword <password>

Parameter server registry password

--parameterServerRegistryUrl <registry url>

Parameter server registry URL

--masterContainerUser <master_container_user>

Master container user (MPI only)

--masterRegistryUsername <username>

Master registry username (MPI only)

--masterRegistryPassword <password>

Master registry password (MPI only)

--masterRegistryUrl <registry url>

Master registry URL (MPI only)

--no-logs

Don’t show logs. Only create, start and exit

--datasetId <dateset id>

Dataset ID

--datasetUri <dateset uri>

Url to S3 bucket with dataset

--datasetName <dateset name>

Name of dataset

--datasetAwsAccessKeyId <AWS access key>

S3 bucket’s Access Key ID

--datasetAwsSecretAccessKey <dataset_secret_access_key_list>

S3 bucket’s Secret Access Key

--datasetAwsEndpoint <dataset_endpoint_list>

S3 endpoint URL

--datasetVersionId <version ID>

S3 dataset’s version ID

--datasetEtag <etag>

S3 dataset’s ETag

--datasetVolumeKind <dataset_volume_kind_list>

S3 dataset’s volume kind. If used, –datasetVolumeSize has to be set as well

Options

dynamic | shared

--datasetVolumeSize <dataset_volume_size_list>

S3 dataset’s volume size

--tensorboard

Creates new tensorboard for this experiment

--tensorboard_set <tensorboard ID>

Add to existing tensorboard

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

singlenode

Create and start new single node experiment

gradient experiments run singlenode [OPTIONS]

Options

--name <name>

Name of new experiment

--ports <ports>

Port to use in new experiment

--workspace <workspace>

Path to workspace directory, archive, S3 or git repository

--workspaceRef <workspace_ref>

Git commit hash, branch name or tag

--workspaceUsername <username>

Workspace username

--workspacePassword <workspace_password>

Workspace password

--ignoreFiles <ignore_files>

Ignore certain files from uploading

--workingDirectory <working_directory>

Working directory for the experiment

--artifactDirectory <artifact_directory>

Artifacts directory

--clusterId <cluster ID>

Cluster ID

--experimentEnv <experiment_env>

Environment variables in a JSON

--projectId <project ID>

Required Project ID

--modelType <model type>

Model type

--modelPath <path>

Model path

--isPreemptible

Flag: is preemptible

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--container <container>

Required Container

--machineType <machine type>

Required Machine type

--command <command>

Required Container entrypoint command

--containerUser <container_user>

Container user

--registryUsername <registry_username>

Registry username

--registryPassword <password>

Registry password

--registryUrl <registry url>

Registry URL

--no-logs

Don’t show logs. Only create, start and exit

--datasetId <dateset id>

Dataset ID

--datasetUri <dateset uri>

Url to S3 bucket with dataset

--datasetName <dateset name>

Name of dataset

--datasetAwsAccessKeyId <AWS access key>

S3 bucket’s Access Key ID

--datasetAwsSecretAccessKey <dataset_secret_access_key_list>

S3 bucket’s Secret Access Key

--datasetAwsEndpoint <dataset_endpoint_list>

S3 endpoint URL

--datasetVersionId <version ID>

S3 dataset’s version ID

--datasetEtag <etag>

S3 dataset’s ETag

--datasetVolumeKind <dataset_volume_kind_list>

S3 dataset’s volume kind. If used, –datasetVolumeSize has to be set as well

Options

dynamic | shared

--datasetVolumeSize <dataset_volume_size_list>

S3 dataset’s volume size

--tensorboard

Creates new tensorboard for this experiment

--tensorboard_set <tensorboard ID>

Add to existing tensorboard

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

start

Start experiment

gradient experiments start [OPTIONS]

Options

--id <id>

Required ID of the experiment

--logs

Show logs

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

stop

Stop experiment

gradient experiments stop [OPTIONS]

Options

--id <id>

Required ID of the experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

tags

Manage tags for experiment

gradient experiments tags [OPTIONS] COMMAND [ARGS]...
add

Add tags to experiment

gradient experiments tags add [OPTIONS]

Options

--id <id>

Required ID of the experiment

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

remove

Remove tags from experiment

gradient experiments tags remove [OPTIONS]

Options

--id <id>

Required ID of the experiment

--tag <tags>

One or many tags that you want to remove from experiment

--tags <tags_comma>

Separated by comma tags that you want to remove from experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

jobs

Manage gradient jobs

gradient jobs [OPTIONS] COMMAND [ARGS]...
artifacts

Manage jobs’ artifacts

gradient jobs artifacts [OPTIONS] COMMAND [ARGS]...
destroy

Destroy job’s artifacts

gradient jobs artifacts destroy [OPTIONS]

Options

--id <job_id>

ID of the job

--files <files>
--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

download

List job’s artifacts

gradient jobs artifacts download [OPTIONS]

Options

--id <job_id>
--destinationDir <destination_directory>

Required

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

get

Get job’s artifacts

gradient jobs artifacts get [OPTIONS]

Options

--id <job_id>

ID of the job

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List job’s artifacts

gradient jobs artifacts list [OPTIONS]

Options

--id <job_id>

ID of the job

-s, --size

Show file size

-l, --links

Show file URL

--files <files>

Get only given file (use at the end * as a wildcard)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

create

Create job

gradient jobs create [OPTIONS]

Options

--name <name>

Job name

--machineType <machine_type>

Required Virtual machine type

--container <container>

Required Docker container

--command <command>

Job command/entrypoint

--ports <ports>

Mapped ports

--isPublic <is_public>

Flag: is job public

--workspace <workspace>

Path to workspace directory

--workingDirectory <working_directory>

Working directory for the experiment

--ignoreFiles <ignore_files>

Ignore certain files from uploading

--experimentId <experiment_id>

Experiment Id

--envVars <env_vars>

Environmental variables

--useDockerfile <use_dockerfile>

Flag: using Dockerfile

--isPreemptible <is_preemptible>

Flag: isPreemptible

--project <project>

Project name

--projectId <project_id>

Required Project ID

--startedByUserId <started_by_user_id>

User ID

--relDockerfilePath <rel_dockerfile_path>

Relative path to Dockerfile

--registryUsername <registry_username>

Docker registry username

--registryPassword <registry_password>

Docker registry password

--cluster <cluster>

Cluster name

--clusterId <cluster_id>

Cluster id

--nodeAttrs <node_attrs>

Cluster node details

--registryTarget <registry_target>

Docker registry target

--registryTargetUsername <registry_target_username>

Docker registry username

--registryTargetPassword <registry_target_password>

Docker registry password

--buildOnly

Determines whether to only build and not run image (default false)

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--dataset <datasets>

Separated by comma tags that you want add to experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete job

gradient jobs delete [OPTIONS]

Options

--id <job_id>

Required Delete job with given ID

--apiKey <api_key>

API key to use this time only

list

List jobs with optional filtering

gradient jobs list [OPTIONS]

Options

--project <project>

Use to filter jobs by project name

--projectId <project_id>

Use to filter jobs by project ID

--experimentId <experiment_id>

Use to filter jobs by experiment ID

--tag <tags>

Filter by tags. Multiple use

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

logs

List job logs

gradient jobs logs [OPTIONS]

Options

--id <job_id>

Required

--line <line>
--limit <limit>
--follow <follow>
--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

metrics

Read job metrics

gradient jobs metrics [OPTIONS] COMMAND [ARGS]...
get

Get job metrics. Shows CPU and RAM usage by default

gradient jobs metrics get [OPTIONS]

Options

--id <job_id>

Required ID of the job

--metric <metrics_list>

One or more metrics that you want to read: cpuPercentage, memoryUsage, gpuMemoryFree, gpuMemoryUsed, gpuPowerDraw, gpuTemp, gpuUtilization, gpuMemoryUtilization. Defaults to cpuPercentage and memoryUsage. To view available custom metrics, use command: gradient jobs metrics list

--interval <interval>

Interval

--start <start>

Timestamp of first time series metric to collect

--end <end>

Timestamp of last time series metric to collect

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List job metrics. Shows CPU and RAM usage by default

gradient jobs metrics list [OPTIONS]

Options

--id <job_id>

Required ID of the job

--interval <interval>

Interval

--start <start>

Timestamp of first time series metric to collect

--end <end>

Timestamp of last time series metric to collect

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

stream

Watch live job metrics. Shows CPU and RAM usage by default

gradient jobs metrics stream [OPTIONS]

Options

--id <job_id>

Required ID of the job

--metric <metrics_list>

One or more metrics that you want to read. Defaults to cpuPercentage and memoryUsage

Options

cpuPercentage | memoryUsage | gpuMemoryFree | gpuMemoryUsed | gpuPowerDraw | gpuTemp | gpuUtilization | gpuMemoryUtilization

--interval <interval>

Interval

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

stop

Stop running job

gradient jobs stop [OPTIONS]

Options

--id <job_id>

Required Stop job with given ID

--apiKey <api_key>

API key to use this time only

tags

Manage job tags

gradient jobs tags [OPTIONS] COMMAND [ARGS]...
add

Add tags to job

gradient jobs tags add [OPTIONS]

Options

--id <id>

Required ID of the job

--tag <tags>

One or many tags that you want to add to job

--tags <tags_comma>

Separated by comma tags that you want add to job

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

remove

Remove tags from job

gradient jobs tags remove [OPTIONS]

Options

--id <id>

Required ID of the job

--tag <tags>

One or many tags that you want to remove from job

--tags <tags_comma>

Separated by comma tags that you want to remove from job

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

logout

Log out / remove apiKey from config file

gradient logout [OPTIONS]

machines

Manage machines

gradient machines [OPTIONS] COMMAND [ARGS]...
availability

Get machine availability for the given region and machine type. Note: availability is only provided for the dedicated GPU machine types. Also, not all machine types are available in all regions

gradient machines availability [OPTIONS]

Options

--region <region>

Required Name of the region

Options

CA1 | NY2 | AMS1

--machineType <machine_type>

Required Machine type

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

create

Create a new Paperspace virtual machine. If you are using an individual account, you will be assigned as the owner of the machine. If you are a team administrator, you must specify the user that should be assigned to the machine, either by specifing a user id, or by providing an email address, password, first name and last name for the creation of a new user on the team.

gradient machines create [OPTIONS]

Options

--region <region>

Required Name of the region

Options

CA1 | NY2 | AMS1

--machineType <machine_type>

Required Machine type

--size <size>

Required Storage size for the machine in GB

--billingType <billing_type>

Required Either ‘monthly’ or ‘hourly’ billing

Options

hourly | monthly

--machineName <name>

Required A memorable name for this machine

--templateId <template_id>

Required Template id of the template to use for creating this machine

--assignPublicIp

Assign a new public ip address on machine creation. Cannot be used with dynamicPublicIp

--dynamicPublicIp

Assigns a new public ip address on machine start and releases it from the account on machine stop. Cannot be used with assignPublicIp

--networkId <network_id>

If creating on a specific network, specify its id

--teamId <team_id>

If creating the machine for a team, specify the team id

--userId <user_id>

If assigning to an existing user other than yourself, specify the user id (mutually exclusive with email, password, firstName, lastName)

--email <email>

If creating a new user for this machine, specify their email address (mutually exclusive with userId)

--password <password>

If creating a new user, specify their password (mutually exclusive with userId)

--firstName <first_name>

If creating a new user, specify their first name (mutually exclusive with userId)

--lastName <last_name>

If creating a new user, specify their last name (mutually exclusive with userId)

--notificationEmail <notification_email>

Send a notification to this email address when complete

--scriptId <script_id>

The script id of a script to be run on startup. See the Script Guide for more info on using scripts

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

destroy

Destroy the machine with the given id. When this action is performed, the machine is immediately shut down and marked for deletion from the datacenter. Any snapshots that were derived from the machine are also deleted. Access to the machine is terminated immediately and billing for the machine is prorated to the hour. This action can only be performed by the user who owns the machine, or in the case of a team, the team administrator.

gradient machines destroy [OPTIONS]

Options

--id <machine_id>

Required The id of the machine to destroy

--releasePublicIp

releases any assigned public ip address for the machine; defaults to false

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show machine information for the machine with the given id.

gradient machines details [OPTIONS]

Options

--id <machine_id>

Required Id of the machine to show

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List information about all machines available to either the current authenticated user or the team, if the user belongs to a team. The list method takes an optional first argument to limit the returned machine objects.

gradient machines list [OPTIONS]

Options

--params <params>

JSON used to filter machines. Use either this or a combination of following options

--id <id>

Optional machine id to match on

--name <name>

Filter by machine name

--os <os>

Filter by os used

--ram <ram>

Filter by machine RAM (in bytes)

--cpus <cpus>

Filter by CPU count

--gpu <gpu>

Filter by GPU type

--storageTotal <storage_total>

Filter by total storage

--storageUsed <storage_used>

Filter by storage used

--usageRate <usage_rate>

Filter by usage rate

--shutdownTimeoutInHours <shutdown_timeout_in_hours>

Filter by shutdown timeout

--performAutoSnapshot <perform_auto_snapshot>

Filter by performAutoSnapshot flag

--autoSnapshotFrequency <auto_snapshot_frequency>

Filter by autoSnapshotFrequency flag

Options

hour | day | week

--autoSnapshotSaveCount <auto_snapshot_save_count>

Filter by auto shapshots count

--agentType <agent_type>

Filter by agent type

--dtCreated <created_timestamp>

Filter by date created

--state <state>

Filter by state

--updatesPending <updates_pending>

Filter by updatesPending

--networkId <network_id>

Filter by network ID

--privateIpAddress <private_ip_address>

Filter by private IP address

--publicIpAddress <public_ip_address>

Filter by public IP address

--region <region>

Filter by region

Options

CA1 | NY2 | AMS1

--userId <user_id>

Filter by user ID

--teamId <team_id>

Filter by team ID

--dtLastRun <last_run_timestamp>

Filter by last run date

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

restart

Restart an individual machine. If the machine is already restarting, this action will request the machine be restarted again. This action can only be performed by the user who owns the machine

gradient machines restart [OPTIONS]

Options

--id <machine_id>

Required Id of the machine to restart

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

start

Start up an individual machine. If the machine is already started, this action is a no-op. If the machine is off, it will be booted up. This action can only be performed by the user who owns the machine

gradient machines start [OPTIONS]

Options

--id <machine_id>

Required Id of the machine to start

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

stop

Stop an individual machine. If the machine is already stopped or has been shut down, this action is a no-op. If the machine is running, it will be stopped and any users logged in will be immediately kicked out. This action can only be performed by the user who owns the machine

gradient machines stop [OPTIONS]

Options

--id <machine_id>

Required Id of the machine to start

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

tags

Manage machine tags

gradient machines tags [OPTIONS] COMMAND [ARGS]...
add

Add tags to machine

gradient machines tags add [OPTIONS]

Options

--id <id>

Required ID of the machine

--tag <tags>

One or many tags that you want to add to machine

--tags <tags_comma>

Separated by comma tags that you want add to machine

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

remove

Remove tags from machine

gradient machines tags remove [OPTIONS]

Options

--id <id>

Required ID of the machine

--tag <tags>

One or many tags that you want to remove from machine

--tags <tags_comma>

Separated by comma tags that you want to remove from machine

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

update

Update attributes of a machine

gradient machines update [OPTIONS]

Options

--id <machine_id>

Required Id of the machine to update

--machineName <name>

New name for the machine

--shutdownTimeoutInHours <shutdown_timeout_in_hours>

Number of hours before machine is shutdown if no one is logged in via the Paperspace client

--shutdownTimeoutForces <shutdown_timeout_forces>

Force shutdown at shutdown timeout, even if there is a Paperspace client connection

--performAutoSnapshot <perform_auto_snapshot>

Perform auto snapshots

--autoSnapshotFrequency <auto_snapshot_frequency>

One of ‘hour’, ‘day’, ‘week’, or null

Options

hour | day | week

--autoSnapshotSaveCount <auto_snapshot_save_count>

Number of snapshots to save

--dynamicPublicIp <dynamic_public_ip>

If true, assigns a new public ip address on machine start and releases it from the account on machine stop

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

utilization

Get machine utilization data for the machine with the given id. Machine upgrades are not represented in utilization data

gradient machines utilization [OPTIONS]

Options

--id <machine_id>

Required Id of the machine to start

--billingMonth <billing_month>

Required Month in YYYY-MM format

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

waitfor

Wait for the machine with the given id to enter a certain machine state. This action polls the server and returns only when we detect that the machine has transitioned into the given state.

gradient machines waitfor [OPTIONS]

Options

--id <machine_id>

Required Id of the machine to start

--state <state>

Required Name of the state to wait for

Options

off | serviceready | ready

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

models

Manage models

gradient models [OPTIONS] COMMAND [ARGS]...
create

Create a model from an url or dataset id

gradient models create [OPTIONS]

Options

--name <name>

Required Model name

--modelType <model_type>

Required Model type

Options

Tensorflow | ONNX | Custom

--datasetRef <dataset_ref>

Required Dataset ref to associate a model with

--projectId <project_id>

ID of a project

--modelSummary <model_summary>

Model summary

--notes <notes>

Additional notes

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete model

gradient models delete [OPTIONS]

Options

--id <model_id>

Required Model ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show model details

gradient models details [OPTIONS]

Options

--id <model_id>

Required Model ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

download

Download model files

gradient models download [OPTIONS]

Options

--id <model_id>

Required Model ID

--destinationDir <destination_directory>

Required Destination directory

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List models with optional filtering

gradient models list [OPTIONS]

Options

--experimentId <experiment_id>

Use to filter by experiment ID

--projectId <project_id>

Use to filter by project ID

--tag <tags>

Filter by tags. Multiple use

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

tags

Manage model tags

gradient models tags [OPTIONS] COMMAND [ARGS]...
add

Add tags to ml model

gradient models tags add [OPTIONS]

Options

--id <id>

Required ID of the model

--tag <tags>

One or many tags that you want to add to ml model

--tags <tags_comma>

Separated by comma tags that you want add to ml model

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

remove

Remove tags from ml model

gradient models tags remove [OPTIONS]

Options

--id <id>

Required ID of the model

--tag <tags>

One or many tags that you want to remove from ml model

--tags <tags_comma>

Separated by comma tags that you want to remove from ml model

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

upload

Upload a model file or directory

gradient models upload [OPTIONS] PATH

Options

--name <name>

Required Model name

--modelType <model_type>

Required Model type

Options

Tensorflow | ONNX | Custom

--projectId <project_id>

ID of a project

--clusterId <cluster_id>

ID of a cluster

--modelSummary <model_summary>

Model summary

--notes <notes>

Additional notes

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

Arguments

PATH

Required argument

notebooks

Manage notebooks

gradient notebooks [OPTIONS] COMMAND [ARGS]...
artifacts

Manage notebooks’ artifacts

gradient notebooks artifacts [OPTIONS] COMMAND [ARGS]...
list

List notebook’s artifacts

gradient notebooks artifacts list [OPTIONS]

Options

--id <notebook_id>

Required ID of the notebook

-s, --size

Show file size

-l, --links

Show file URL

--files <files>

Get only given file (use at the end * as a wildcard)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

create

Create new notebook

gradient notebooks create [OPTIONS]

Options

--machineType <machine_type>

Required Virtual Machine type label e.g. P5000

--container <container>

Required Container name

--projectId <project_id>

Required Project ID

--clusterId <cluster_id>

Cluster ID

--name <name>

Notebook name

--registryUsername <registry_username>

Registry username

--registryPassword <registry_password>

Registry password

--command <command>

Command (executed as /bin/sh -c ‘YOUR COMMAND’)

--containerUser <container_user>

Container user

--shutdownTimeout <shutdown_timeout>

Shutdown timeout in hours

--isPreemptible

Is preemptible

--isPublic

Is publically viewable

--environment <environment>

Environmental variables

--workspace <workspace>

S3 url or git repository. Directory uploads are not yet supported

--workspaceRef <workspace_ref>

Git commit hash, branch name or tag

--workspaceUsername <username>
--workspacePassword <workspace_password>

Workspace password

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete existing notebook

gradient notebooks delete [OPTIONS]

Options

--id <id_>

Notebook ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show notebook details

gradient notebooks details [OPTIONS]

Options

--id <id>

Required Notebook ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

fork

Fork existing notebook

gradient notebooks fork [OPTIONS]

Options

--id <id_>

Required Notebook ID

--projectId <project_id>

Required Project ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List notebooks

gradient notebooks list [OPTIONS]

Options

-l, --limit <n_limit>
-o, --offset <n_offset>
--tag <tags>

Filter by tags. Multiple use

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

logs

List notebook logs

gradient notebooks logs [OPTIONS]

Options

--id <notebook_id>

Required

--line <line>
--limit <limit>
--follow <follow>
--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

metrics

Read notebook metrics

gradient notebooks metrics [OPTIONS] COMMAND [ARGS]...
get

Get notebook metrics. Shows CPU and RAM usage by default

gradient notebooks metrics get [OPTIONS]

Options

--id <notebook_id>

Required ID of the notebook

--metric <metrics_list>

One or more metrics that you want to read: cpuPercentage, memoryUsage, gpuMemoryFree, gpuMemoryUsed, gpuPowerDraw, gpuTemp, gpuUtilization, gpuMemoryUtilization. Defaults to cpuPercentage and memoryUsage. To view available custom metrics, use command: gradient notebooks metrics list

--interval <interval>

Interval

--start <start>

Timestamp of first time series metric to collect

--end <end>

Timestamp of last time series metric to collect

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List notebook metrics. Shows CPU and RAM usage by default

gradient notebooks metrics list [OPTIONS]

Options

--id <notebook_id>

Required ID of the notebook

--interval <interval>

Interval

--start <start>

Timestamp of first time series metric to collect

--end <end>

Timestamp of last time series metric to collect

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

stream

Watch live notebook metrics. Shows CPU and RAM usage by default

gradient notebooks metrics stream [OPTIONS]

Options

--id <notebook_id>

Required ID of the notebook

--metric <metrics_list>

One or more metrics that you want to read. Defaults to cpuPercentage and memoryUsage

Options

cpuPercentage | memoryUsage | gpuMemoryFree | gpuMemoryUsed | gpuPowerDraw | gpuTemp | gpuUtilization | gpuMemoryUtilization

--interval <interval>

Interval

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

start

Start notebook

gradient notebooks start [OPTIONS]

Options

--id <id>

Required Notebook ID

--machineType <machine_type>

Required Virtual Machine type label e.g. P5000

--clusterId <cluster_id>

Cluster ID

--shutdownTimeout <shutdown_timeout>

Shutdown timeout in hours

--isPreemptible

Is preemptible

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

stop

Stop running notebook

gradient notebooks stop [OPTIONS]

Options

--id <notebook_id>

Required Stop notebook with given ID

--apiKey <api_key>

API key to use this time only

tags

Manage notebook tags

gradient notebooks tags [OPTIONS] COMMAND [ARGS]...
add

Add tags to notebook

gradient notebooks tags add [OPTIONS]

Options

--id <id>

Required ID of the notebook

--tag <tags>

One or many tags that you want to add to notebook

--tags <tags_comma>

Separated by comma tags that you want add to notebook

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

remove

Remove tags from notebook

gradient notebooks tags remove [OPTIONS]

Options

--id <id>

Required ID of the model

--tag <tags>

One or many tags that you want to remove from notebook

--tags <tags_comma>

Separated by comma tags that you want to remove from notebook

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

projects

Manage projects

gradient projects [OPTIONS] COMMAND [ARGS]...
create

Create project

gradient projects create [OPTIONS]

Options

--name <name>

Required Name of new project

--repositoryName <repository_name>

Name of the repository

--repositoryUrl <repository_url>

URL to the repository

--tag <tags>

One or many tags that you want to add to experiment

--tags <tags_comma>

Separated by comma tags that you want add to experiment

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete project and all its experiments

gradient projects delete [OPTIONS]

Options

--id <project_id>

Required ID of project

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show details of a project

gradient projects details [OPTIONS]

Options

--id <project_id>

Required Project ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List projects

gradient projects list [OPTIONS]

Options

--tag <tags>

Filter by tags. Multiple use

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

tags

Manage project tags

gradient projects tags [OPTIONS] COMMAND [ARGS]...
add

Add tags to project

gradient projects tags add [OPTIONS]

Options

--id <id>

Required ID of the project

--tag <tags>

One or many tags that you want to add to project

--tags <tags_comma>

Separated by comma tags that you want add to project

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

remove

Remove tags from project

gradient projects tags remove [OPTIONS]

Options

--id <id>

Required ID of the project

--tag <tags>

One or many tags that you want to remove from project

--tags <tags_comma>

Separated by comma tags that you want to remove from project

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

secrets

Manage secrets

gradient secrets [OPTIONS] COMMAND [ARGS]...
delete

Delete secret

gradient secrets delete [OPTIONS] [cluster|project|team]

Options

--id <entity_id>

Entity ID

--name <name>

Required Secret name

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

Arguments

ENTITY

Required argument

list

List secrets

gradient secrets list [OPTIONS] [cluster|project|team]

Options

--id <entity_id>

Entity ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

Arguments

ENTITY

Required argument

set

Set secret

gradient secrets set [OPTIONS] [cluster|project|team]

Options

--id <entity_id>

Entity ID

--name <name>

Required Secret name

--value <value>

Required Secret value

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

Arguments

ENTITY

Required argument

storageProviders

Manage storage providers

gradient storageProviders [OPTIONS] COMMAND [ARGS]...
create

Create storage providers

gradient storageProviders create [OPTIONS] COMMAND [ARGS]...
s3

Create S3 storage provider

gradient storageProviders create s3 [OPTIONS]

Options

--name <name>

Required Storage provider name

--bucket <bucket>

Required S3 bucket

--accessKey <access_key>

Required S3 access key ID

--secretAccessKey <secret_access_key>

Required S3 secret access key

--endpoint <endpoint>

S3 endpoint URL

--region <region>

S3 region

--signatureVersion <signature_version>

S3 signature version (ex: v4)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete storage provider

gradient storageProviders delete [OPTIONS]

Options

--id <id>

Required Storage provider ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show storage provider details

gradient storageProviders details [OPTIONS]

Options

--id <id>

Required Storage provider ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List storage providers

gradient storageProviders list [OPTIONS]

Options

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

update

Update storage providers

gradient storageProviders update [OPTIONS] COMMAND [ARGS]...
s3

Update S3 storage provider

gradient storageProviders update s3 [OPTIONS]

Options

--id <id>

Required Storage provider ID

--name <name>

Storage provider name

--bucket <bucket>

S3 bucket

--accessKey <access_key>

S3 access key ID

--secretAccessKey <secret_access_key>

S3 secret access key

--endpoint <endpoint>

S3 endpoint URL

--region <region>

S3 region

--signatureVersion <signature_version>

S3 signature version (ex: v4)

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

tensorboards

Manage tensorboards

gradient tensorboards [OPTIONS] COMMAND [ARGS]...
add-experiments

Update tensorboard experiments

gradient tensorboards add-experiments [OPTIONS]

Options

--id <tensorboard ID>

Required Tensorboard ID

--experiment <experiment ID>

Required One or more experiment IDs [–experiment id1 –experiment id2 …]

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

create

Create new tensorboard

gradient tensorboards create [OPTIONS]

Options

--experiment <experiment ID>

Required One or more experiment IDs [–experiment id1 –experiment id2 …]

--image <image>

Tensorboard Container Image, by default its set to tensorflow/tensorflow:latest

--username <username>

Basic Auth Username

--password <password>

Basic Auth Password

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

delete

Delete tensorboard

gradient tensorboards delete [OPTIONS]

Options

--id <tensorboard ID>

Required Tensorboard ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

details

Show details of a tensorboard

gradient tensorboards details [OPTIONS]

Options

--id <tensorboard ID>

Required Tensorboard ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

Show list of tensorboards

gradient tensorboards list [OPTIONS]

Options

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

remove-experiments

Update tensorboard experiments

gradient tensorboards remove-experiments [OPTIONS]

Options

--id <tensorboard ID>

Required Tensorboard ID

--experiment <experiment ID>

Required One or more experiment IDs [–experiment id1 –experiment id2 …]

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

version

Show the version and exit

gradient version [OPTIONS]

workflows

Manage workflows

gradient workflows [OPTIONS] COMMAND [ARGS]...
create

Create workflow

gradient workflows create [OPTIONS]

Options

--name <name>

Required Workflow name

--projectId <project_id>

Required Workflow name

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

get

Get workflow info

gradient workflows get [OPTIONS]

Options

--id <workflow_id>

Required Workflow Id

--show-runs

Fetch runs

--run <run>

Specify workload run

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

list

List workflows

gradient workflows list [OPTIONS]

Options

--projectId <project_id>

Project ID

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

logs

List logs for specific workflow

gradient workflows logs [OPTIONS]

Options

--id <workflow_id>

Required Workflow ID

--run <run>

Required Specify workload run

--logId <workflow_log_id>
--line <line>
--limit <limit>
--follow <follow>
--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file

run

Run workflow spec

gradient workflows run [OPTIONS]

Options

--id <workflow_id>

Required Workflow ID

--clusterId <cluster_id>

Cluster ID

--path <spec_path>

Required Path to spec

--inputPath <input_path>

Path to inputs

--apiKey <api_key>

API key to use this time only

--optionsFile <options_file>

Path to YAML file with predefined options

--createOptionsFile <createoptionsfile>

Generate template options file