Experiment options

Web UI
CLI
Web UI

Experiment web UI Options List

Option

Description

Machine Type

This is the type of instance to run your Experiment's on. Many Experiments benefit from a machine with a GPU, but some can run just using a CPU.

Container

Experiments are run within a docker container. You can run a public or private container. Learn more here.

Workspace

The workspace is the collection of code that is run. The only option in the web UI is a Git repo (eg https://github.com/Paperspace/fast-style-transfer.git). The CLI offers more advanced options such as specifying a local directory, private Git repos, S3 buckets, and more.

Command

The command is the entry point to the container. This is the line of code that will kick off your experiment's job. It could be a bash script ./run.sh or python main.py as a few examples.

Ports

You have the option to attach a public IP automatically. Supports opening multiple ports simultaneously, separated by : . Learn more about opening ports here.

Custom Metrics

Enter a list of custom metrics to use with Gradient's statd client, such as percent_failure or percent_success.

CLI

Experiment CLI Options List

Option

Description

name

Experiment name

ports

An optional list of port mappings to open on the instance while the job is running. The port mappings are specified as 'XXXX:YYYY' where XXXX is an external port number and YYYY is an internal port number. Multiple port mappings can be provided as a comma separated list. Port numbers must be greater than 1023. Note: only /tcp protocol usage is supported. Learn more here.

workspace

Supports a path to a local directory, git repository, or s3 bucket.

workspaceRef

An option to specify a Git commit hash, branch name or tag. Learn more here.

workspaceArchive

Path to workspace archive. (Currently being deprecated in an upcoming version.)

workspaceUrl

Path to workspace archive. (Currently being deprecated in an upcoming version.)

workspaceUsername

Used to pass-in a username to a private repo.

workspacePassword

Used to pass-in a password to a private repo.

ignoreFiles

When using the directory upload option, this command can be used to ignore certain files from uploading.

workingDirectory

Working directory for the experiment

artifactDirectory

Where artifacts can be located.

clusterId

Cluster ID of the private Gradient cluster. Learn more here.

experimentEnv

Environmental variables in JSON String Format. Example:

{"HKS_EPOCHS":1,"HKS_MAX_EVALS":4,"DATASET_SIZE":100} Learn more here.

projectId

Your Project ID.

modelType

This option is used to specify the format of your model. Learn more here.

modelPath

This option is used to locate your model. Learn more here.

isPreemptible

Flag that specifies if the Experiment is preemptible.

container

A reference to a docker image in a public or private docker registry, or a container name provided by Paperspace. Docker image repository references must be in lowercase and may include a tag and a hostname prefix followed by a slash; if committed the hostname defaults to that of the public Docker Hub registry. An example docker image reference: docker.io/mynamespace/myimage:mytag. A container name may be mixed case. (Designated container names are currently only provided as part of various Gradient tutorials and samples.)

machineType

An optional machine type to run the job on: either 'GPU+', 'P4000', 'P5000', 'P6000', 'V100', 'K80', 'P100', or 'TPU'.

command

Experiment entrypoint command.

containerUser

Name of user in container.

registryUsername

Used to pass-in a username to a private container registry.

registryPassword

Used to pass-in a password to a private container registry.

registryUrl

Used to pass-in a path to a container in private container registry.

no-logs

Flag to prevent streaming logs in terminal.

datasetUri

Url to S3 bucket with dataset

datasetName

Name of dataset

datasetAwsAccessKeyId

S3 bucket's Access Key ID

datasetAwsSecretAccessKey

S3 bucket's Secret Access Key

datasetVersionId

S3 dataset's version ID

datasetEtag

S3 dataset's ETag

tensorboard

Add to existing TensorBoard. If no or many TensorBoard exists a new one will be created.

apiKey

API key to use this time only.

optionsFile

Path to YAML file with predefined options.

createOptionsFile

Generate template options file