AI Studio SDK v2

Getting Started

The AI Studio SDK was developed in Python and is designed to help data scientists to interact with AI Studio from their code, experiments and models. Through the SDK, you can create experiments, manage models, automate your machine learning pipeline and more.

The topics in this page:

• Prerequisites
• Download and Install the AI Studio SDK
  • Install options
• Authenticating the AI Studio SDK
  • Inside a AI Studio job scope
  • Authenticate using a local configuration file
  • Authenticate using environment variables
• User Operations
  • Verify you are logged in
  • Set the default organization
• Project Operations
  • Create a new project
  • Get the project's object:
  • List all the projects in the organization:
  • Delete a project:
  • Project File Operations
  • Update the project's settings
• Datasource Operations
  • Create a Datasource
  • Get a Datasource
  • Access Datasource attributes
  • List all existing Datasources
  • Delete a Datasource
  • Adding and Removing Collaborators for a Private Datasource
  • Datasource File Operations
  • Datasources Additions to Jobs
• Dataset Operations
  • Create a Dataset
  • Dataset ID
  • Get a Dataset
  • Access Dataset attributes
  • List all existing Datasets
  • Delete a Dataset
  • Dataset File Operations
  • Dataset Commits
  • Checking Dataset Commit Indexing Status
  • Dataset Caching
  • Dataset Queries
• Workspaces operations:
  • Workspace ID
  • Fetch the workspace object
  • Create a new workspace
  • Access workspace attributes
  • Sync workspace's file changes
  • Stop a running workspace
  • Start a stopped workspace
  • List all of the workspaces
  • Delete workspaces
  • Operate a Tensorboard
• Experiment Operations
  • Experiment slug
  • Fetch the experiment object
  • Create a new experiment
  • Create a grid of experiments
  • Get an existing experiment
  • list all experiment in the project
  • Delete experiment
  • Stop a running experiment
  • Get experiment's system utilization
  • Track an experiment manually
  • Examples
  • Metadata operations on experiments
  • Get the experiment's last 40 logs:
  • Upload artifacts
  • Fetch a list of experiment's artifacts:
  • Download the experiment's artifacts
  • Merge experiment's commit into project files
  • Download the experiment's artifacts
  • Merge experiment's commit into project files
  • Create a Tag
  • Charts
  • Operate a Tensorboard
• Flow Operations
  • Create a Flow
  • Access Flow attributes
  • Flow slug
  • Get a Flow
  • List Flows
  • Run a Flow
  • Update Flow
  • Delete Flow
  • Schedule a Flow
  • Trigger webhook
  • Toggle dataset update trigger
  • Flow versions
• Endpoint Operations
  • Create Endpoint
  • Endpoint slug
  • Get Endpoint object
  • List Endpoints
  • Stop running Endpoints
  • Start a stopped Endpoint
  • Delete Endpoints
  • Endpoint Attributes
  • Update The Endpoint's version
  • Update the Endpoint's replica set
  • Rollback version
  • Get sample code
  • Send prediction
  • Log metrics
  • Poll charts
  • Add logs to a running endpoint
  • Set feedback loop
  • Add Continual Learning rules
  • Control batch Endpoint
• Webapps Operations
  • Create a webapp
  • Webapp's slug
  • Get webapp object
  • List all the webapps in the project
  • Delete webapp
  • Stop running webapp
  • Share and Permissions
• Resource Operations
  • Connect your existing Kubernetes cluster
  • Build managed EKS cluster (Supported on MetaCloud)
  • Get an existing resource
  • Update an existing resource
  • Delete a resource
• Templates Operations
  • Create a new template
  • Get an existing template
  • List all existing templates
  • Update an existing template
  • Delete an existing template
• Registry commands
  • Create a registry
  • Get a registry
  • List all registries
  • Update a registry
  • Delete a registry
• Image commands
  • Create an image
  • Get an image
  • List all images
  • Delete an image

Prerequisites

In order to run the pip commands, Python (version 3.6 or later) should be installed on the system.

Download and Install the AI Studio SDK

To install, open up your terminal/command prompt and run the following command:

pip3 install cnvrgv2

In order to use cnvrgv2 on windows, the use of Windows Subsystem for Linux (WSL) is required. First, install WSL with following commands:

## Update the WSL:
wsl --update
## List the WSL distributions:
wsl --list -o
## Set the default WSL version:
wsl --set-default-version 1
## Install Ubuntu:
wsl --install -d Ubuntu
## Boot Ubuntu:
wsl -d Ubuntu

Then, install python:

## Install required software:
sudo apt install software-properties-common
## Add Python PPA:
sudo add-apt-repository ppa:deadsnakes/ppa
## Update the system:
sudo apt update
## Install Python 3.10:
sudo apt install python3.10
## Install pip for Python 3:
sudo apt install python3-pip

Finally, install cnvrgv2:

pip3 install cnvrgv2

Install options

When on self-hosted AI Studio environemnt, you can specify an option for cnvrgv2, to fit the object storage you intend to work with in your AI Studio environment.

For Metacloud, we'll use the default installation without any option.

Add the options to the install command as needed, you can add multiple options by separating with comma:

pip install "cnvrgv2[options]"

available options are:

• azure - Install packages relevant for Azure storage client
• google - Install packages relevant for GCP storage client
• python3.6 - Install specific dependencies for python version 3.6

SDK Operations

Authenticating the AI Studio SDK

You can can authenticate when credentials as parameters:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg(domain="https://app.cnvrg.io",
              email="Johndoe@acme.com",
              password="123123",
              )

Login command parameters

Parameter	Value Type	Description	required	default
domain	Text	AI Studio domain/url starting from http/s, excluding the organization slug e.g. https://app.cnvrgdomain.com/	Yes
email	Text	User email	Yes
password	Text	authenticate using password	Password or token are required
token	Text	authenticate using the API Token	Password or token are required
organization	Text	AI Studio organization, this can be used when the user is a member of multiple organizations	No	First organization the user was added to

For Metacloud and self-hosted AI Studio environments with SSO authentication, The user's API Token must be used instead of the Password, with the token parameter. the token can be retrieved from the user settings page, under API Token field:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg(domain="https://app.domain.metacloud.cnvrg.io",
              email="Johndoe@acme.com",
              token="YOUR API KEY")

NOTE

As a security measure, please do not put your credentials into your code.

Inside a AI Studio job scope

The AI Studio SDK will already be initialized and authenticated with cnvrgv2 using the account that is logged in. You can start using AI Studio SDK functions immediately by running:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()

Authenticate using a local configuration file

You can authenticate to the AI Studio SDK by creating a configuration file in your working directory

• In your working directory create a directory called .cnvrg

  mkdir .cnvrg
  

• Inside the directory .cnvrg create a configuration file named cnvrg.config
• Edit the file and insert the following:

  check_certificate: <false/true>
  domain: <cnvrg_full_domain>
  keep_duration_days: null
  organization: <organiztion_name>
  token: <user_access_token>
  user: <user_email>
  version: null
  

• Once you finish editing, save the file Now you can simply run the following in your code and it will log you in automatically:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()

Authenticate using environment variables

You can authenticate to the AI Studio SDK by setting the following environment variable:

• CNVRG_TOKEN: Your API token ( You can find it in your user settings page)
• CNVRG_URL: Your cnvrg url that you use to view AI Studio through the browser, for example: https://app.prod.cnvrg.io
• CNVRG_USER: Your email that you use to log in to AI Studio
• CNVRG_ORGANIZATION: The organization name you use Once you set those environment variables, you can simply run the following and it will log you in automatically:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()

NOTE

The following documentation assume you have successfully logged in to the SDK and loaded the AI Studio object

User Operations

Verify you are logged in

To get the logged in user object you can simply run:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
user = cnvrg.me()

Once you have the user object you can get the user fields like: email, username, organizations, git_access_token, name, time_zone and more.

For example:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
user = cnvrg.me()
email = user.email

Set the default organization

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cnvrg.set_organization("my-org")

NOTE

As a default, your current organization will be used

Project Operations

Create a new project

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.create("myproject")

Get the project's object:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")

Once you have the project object you can get the project fields like: title, slug, git_url, git_branch, last_commit, num_files and more.

For example:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
title = myproj.title

NOTE

You can also reference the current project from within a job scope:

from cnvrgv2 import Project
p = Project()

List all the projects in the organization:

List all projects that the current user is allowed to view

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.list()

To order the projects list by created_at run the following:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.list(sort="-created_at")

TIP

sort the list by: -key -> DESC | key -> ASC

Delete a project:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
myproj.delete()

Project File Operations

Clone the project to the current working directory

myproj.clone()

Available Parameters:

Parameter	type	description	required	default
commit	string	Sha1 of the commit to clone	No	latest
override	boolean	re-clone even if the project already cloned	No	False
current_dir	boolean	Whether to clone to the current dir or not	No	clone into a directory created with the project name
fullpath_filter	string	Filter on path of file by part of the path	No	None
threads	integer	number of threads that will be used in order to clone the project	No	40

Download project's latest commit

In order to fetch latest project commit to your local directory, use following command:

myproj.download()

WARNING

The Project must be cloned first

upload all file changes from local project directory

In order to upload all file changes to your cnvrg project, use following command:

import os
from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
myproj.clone()
os.chdir("myproject")
myproj.upload()

Available Parameters:

Parameter	type	description	required	default
message	string	The commit message	No	""
output_dir	string	will only sync the files in the output_dir specified	No	file changes from current dir
git_diff	boolean	upload files from git diff output in addition to the given paths	No	Only upload files mentioned in the paths

WARNING

• This command must be used from a directory linked to a AI Studio project
• Folder/file names must not contain special characters, which are - +?!@#$%&^*(){}[]

Upload specific files to a project

myproj.put_files(paths=['/files_dir/file1.txt', '/files_dir/file2.txt'])

Available Parameters:

Parameter	type	description	required	default
paths	List	The list of file paths that will be uploaded to the project	Yes
message	string	The commit message	No	""
override	boolean	Upload every file specified, even if it already exists in the latest commit	No	False
force	boolean	Create a new commit with only the files included in the current upload	No	False
fullpath_filter	string	Filter on path of file by part of the path	No	None
git_diff	boolean	upload files from git diff output in addition to the given paths	No	False (Only upload files mentioned in the paths)

NOTE

• Files will get uploaded with their relative path as specified in the put_files command for example: myproj.put_files(paths=['/files_dir/file1.txt', '/files_dir/file2.txt']) file1.txt and file2.txt will be under files_dir directory in the Storage.
• Reference to parent paths ('../../path') is not supported

WARNING

Folder/file names must not contain special characters, which are - +?!@#$%&^*(){}[]

Sync project

Sync local file changes into the project using following command:

myproj.sync()

Available Parameters:

Parameter	type	description	required	default
message	string	The commit message	No	""
output_dir	string	will only sync the files in the output_dir specified	No	file changes from current dir
git_diff	boolean	upload files from git diff output in addition to the given paths	No	Only upload files mentioned in the paths

Remove files from a project

You can remove files from the project using following command:

myproj.remove_files(paths='*', 
          message='This will delete everything!')

Available Parameters:

Parameter	type	description	required	default
paths	List	The list of file/folder paths that will be deleted (regex and wilfcard are allowed)	Yes
message	string	The commit message	No	""

List the project's content

You can list all of the files and folders that are in the project:

myproj.list_files()
myproj.list_folders(commit_sha1='xxxxxxxxx')

Available Parameters:

Parameter	type	description	required	default
commit_sha1	string	Sha1 string of the commit to list the files from	No	None
sort	string	Key to sort the list by (-key -> DESC / key -> ASC)	No	"-id"
query_raw	string	query to be filtered with wild card (for example: 'filename*' or '*.png')	No	None

List project's commits

You can list all of the files and folders that are in the project:

myproj.list_commits(sort="-key -> DESC")

Update the project's settings

You can change any of the project's settings by passing them as keyword arguments. For example, to change the project title

myproj.settings.update(title='NewProjectTitle')

To update project's collaborators

myproj.settings.update(
    privacy='private',
    collaborators=['person1@email.com', 'person2@email.com']
)

Available Parameters:

Parameter	type	description
title	string	The name of the project
default_image	string	The name of the image to set to be the project's default image
default_computes	List	The list of the project's default compute template names
privacy	string	The project's privacy set to either 'private' or 'public'
mount_folders	List	Paths to be mounted to the docker container
env_variables	List	KEY=VALUE pairs to be exported as environment variables to each job
check_stuckiness	boolean	Idle experiment behaviour (Stop (False) or Restart (True))
max_restarts	int	When "check_stuckiness" is True this sets how many times to repeatedly restart a single experiment each time it idles
stuck_time	int	The duration (in minutes) that an experiment must be idle for before it is stopped or restarted
autosync	boolean	Whether or not to preform periodic automatic sync
sync_time	int	The interval (in minutes) between each automatic sync of jobs
collaborators	List	The complete and updated list of collaborators' email addresses on the project
command_to_execute	string	The project's default command to execute when starting a new job
run_tensorboard_by_default	boolean	Whether or not to run Tensorboard by default with each launched experiment
run_jupyter_by_default	boolean	Whether or not to run Jupyter by default with each launched experiment
requirements_path	string	The default path to the requirements.txt file that will run with every job
is_git	boolean	Whether if the project is linked to a git repo or not
git_repo	string	The address of the git repo
git_branch	string	The default branch
private_repo	boolean	Whether the repo is private or not
email_on_success	boolean	If email should be sent when the experiment finishes successfully
email_on_error	boolean	If email should be sent when the experiment finishes with an error

Setup Git Integrations in project settings

For a public git repository

myproj.settings.update(is_git=True, git_repo="MyGitRepo", git_branch="MyBranch")

For a private git repository using Oauth Token First make sure that the git Oauth Token is saved in your profile and then run

myproj.settings.update(is_git=True, git_repo="PrivateGitRepo", git_branch="MyBranch", private_repo=True)

To disable git integrations

myproj.settings.update(is_git=False)

Configure secrets

In order to edit an existing secret from your project's settings, or add a secret, use following command:

myproj.settings.update_secret(secret_key="KEY", secret_value="VALUE")

In order to delete secrets from your project's settings, use following command:

## Delete single secret
myproj.settings.delete_secret("KEY")

## Delete multiple secrets
myproj.settings.delete_secret(["KEY1","KEY2"])

Datasource Operations

Create a Datasource

You can create a new Datasource in AI Studio:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
dsource = cnvrg.datasource.create(name="mybucket",storage_type=StorageTypes.S3)

Available Parameters:

Parameter	type	description	required
storage_type	string	The storage type (s3/Minio)	Yes
name	string	The name of the new Datasource	Yes
bucket_name	string	The name of the bucket	Yes
path	string	Path in the bucket	No
endpoint	string	Endpoint URL	No
region	string	The region determines where your data is physically stored	Yes
descripton	string	Typically includes information about the purpose and use of the bucket, along with any relevant configuration details	Yes
credentials	dict	Access key ID and Secret access key. In this format(for s3 compatibles): {'access_key_id': '***', "secret_access_key": "***"}	Yes
collaborators	list	The complete and updated list of collaborators email addresses for this Datasource	No
public	bool	Permissions: public or private	No

Note

• Only an admin can create datasources.
• By default, datasource is created as private.
• For Minio, endpoint_url is mandatory

Get a Datasource

To fetch a Dataset from AI Studio you can use:

dsource = cnvrg.datasources.get(slug='datasource_slug')

Note

A non-admin user can only perform Get requests on Datasources they are explicitly assigned to, as well as on public Datasources accessible to all users.

Access Datasource attributes

You can access the Datasource attributes by using regular dot notation:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
dsource = cnvrg.datasources.get(slug='dsource_slug')
dsource.slug
dsource.collaborators
dsource.storage_type

Available attributes:

Attribute	type	description
slug	string	The unique slug value of the Datasource
storage_type	string	The bucket storage type for the Datasource
name	string	The name of the Datasource
admin	string	The admin of the Datasource (the creator)
path	string	The path in the storage bucket
endpoint	string	Endpoint URL
region	int	The region determines where your data is physically stored
public	bool	Permissions: public or private
collaborators	list	The complete and updated list of collaborators email addresses for this Datasource

Note

• The AI Studio platform does not return credentials in response to GET requests for security and privacy reasons.
• GET requests with a faulty slug should throw an error (pattern = r'^[a-z0-9-]+$').

List all existing Datasources

You can list all the Datasources in the current organization:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
datasources = e2e_client.datasources.list(filter=filter, sort="id")
for data_source in datasources:
    print(data_source.name)

• Optional Filtering and Sorting
  In the context of dataset retrieval, both filtering and sorting are optional parameters that can enhance data management.
  Filtering allows users to narrow down results based on specific criteria. It operates similarly to dataset or project filters.
  For example: You can filter by storage type: StorageTypes.S3, StorageTypes.MINIO etc:

s3_filter = {'operator': "AND",
             'conditions':[{'key': "type", 'operator': "is", 'value': StorageTypes.S3}]}

TIP

Helper function: list_count returns the total count of Datasources

cnvrg.datasources.list_count(filter=filter, sort="id")

Delete a Datasource

To delete a Datasource call the delete() function on it's instance:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
dsource = cnvrg.datasources.get("slug")
dsource.delete()

Adding and Removing Collaborators for a Private Datasource

• Adding a Collaborator:

dsource.add_collaborator('abc@efg.com')

• Removing a Collaborator:

dsource.remove_collaborator('abc@efg.com')

While adding/removing Collaborators. An error will be received in the following cases:

1. If user does not exist in the Organization.
2. If user is Admin (it can’t be removed from Datasource).
3. If user is not Admin and not assigned.
4. Authorization error to perform action if current user is not Admin.

Datasource File Operations

• For the following operations, if a path is set on the Datasource, the ALWAYS path will be added to the request.

List Files

Lists the objects in the bucket path and returns a list of object keys:

for file_name in datasource.list_objects():
  print(file_name)

Note

If the Datasource has a path, files will be listed for the given path only.

Files Downloader

This is a helper function which downloads page by page for better performance.

downloader = ds.get_files_downloader(page_size=page_size, max_workers=max_workers, destination_folder=dest_dir)
for page in downloader.page_iterator:
    downloader.download_objects(page)

• This will download page by page to the destination_dir.
• If no destination is sent, it will download to the current directory.
• Page size default is 100

Downloading single file

Downloads a single file from the bucket.

download_file(file_path, destination_path)

• file_path: The path of the file you wish to download from the bucket. ( This is the actual file path in the bucket)
• destination_path: The destination file path where the file will be downloaded. This might be a relative or absolute path. In any case, all the folders in the tree should be existing folders (if the path in a/b/c.txt -> a/ and b/ should be existing folders)

Note

If the datasource has a path in backet, the file_path is appended to the path.

Example: For instance, if the datasource has ‘path1’ then : datasource.download_file("path_to/file.txt", "destination/directory/file_name.txt") this will download the following path from the bucket: "path1/path_to/file.txt"

WARNING

When downloading a file within a job (workspace or experiment), ensure that you save the datasource files to a designated folder /data. This precaution prevents the unintended upload of these files to project files during periodic synchronization or end-of-job synchronization processes.

File Upload

Upload a single file to the bucket.

upload_file(file_path, destination_path)

• file_path: The path of the file you wish to upload to the bucket.
• destination_path: The destination path in the bucket.

Note

If destination is not specified, the base name of the file is used.

• For example:

1. without destination:

datasource.upload(file_path='abc.txt')

this will be uploaded to ‘path1/abc.txt’ in the bucket

BUCKET:

path1/file1
path1/file2
path1/abc.txt

2. with destination:

datasource.upload(file_path='abc.txt', destination_path='myfolder/abc.txt')

this will be uploaded to ‘path1/myfolder/abc.txt’ in the bucket

BUCKET:

path1/file1
path1/file2
path1/myfolder/abc.txt

3. with duplicate destination:

datasource.upload(file_path='abc.txt', destination_path='path1/myfolder/abc.txt')

this will be uploaded to ‘path1/path1/myfolder/abc.txt’

BUCKET:

path1/file1
path1/file2
path1/path1//myfolder/abc.txt

Removing a file

Removes a file from the bucket.

datasource.remove_file(file_path)

• file_path: The path of the file you wish to remove from the bucket.

Note:

If the datasource has a path, the file_path is appended to the file_path.

Cloning a datasource

Clones the contents of the datasource to the current directory.

datasource.clone(page_size=100, max_workers=os.cpu_count(), skip_if_exists=False, force=False)

• page_size: The number of files to download per page.
• max_workers: The maximum number of parallel workers for downloading files.
• skip_if_exists: Whether to skip cloning if the directory already exists.
• force: Whether to force cloning by overwrite the existing directory.

Important Notes:

• If skip_if_exists is True and the directory exists, the method will return without cloning.
• If force is True and the directory exists, it will be removed before cloning.
• If the datasource has a path, files will be listed for the given path only.

Datasources Additions to Jobs

The Datasource will be adding to the job at Workspace or Experiment startup. When a Datasource is attached to a job, it will be cloned BEFORE the job starts. The datasource will be cloned to the /data directory

For Worcspaces:

p = cnvrg.projects.get('myproject')
p.clone()
# (create workspace must be done within a project context)
os.chdir(p.slug)
w = p.workspaces.create(
        title='My Workspace',
        templates=["small","medium"],
        overriding_datasources=[datasources_slugs_list],
        datasources=[datasources_slugs_list]) 
    )

The params are useful in the case the datasource may have already been clone (pvc etc)

• datasources: For these Datasources, if the Datasource has already been cloned it will be skipped.
• overriding_datasources: For these Datasources, if the Datasource has already been cloned, the folder will be deleted and the Datasource will be cloned again

For Experiments:

p = cnvrg.projects.get('myproject')
p.clone()
# (create experiment must be done within a project context)
os.chdir(p.slug)
e = p.experiments.create(
        title="my new exp",
        template_names=["small","medium"],
        command="python3 test.py",
        sync_before=False,
        sync_after=False,
        overriding_datasources=[datasources_slugs_list],
        datasources=[datasources_slugs_list]) 
    )

The params are useful in the case the datasource may have already been clone (pvc etc)

• datasources: For these Datasources, if the Datasource has already been cloned it will be skipped.
• overriding_datasources: For these Datasources, if the Datasource has already been cloned, the folder will be deleted and the Datasource will be cloned again

Validations:

• If the Datasource does not exist: throw an error and do not create job.
• If the current user is not assigned to one of the Datasources: it throws an error and does not create the job.

Datasources attached to the job check:

To retrieve and display the datasources associated with a specific job within the experiment and workspace contexts:

print(experiment.datasources)

print(workspace.datasources)

Dataset Operations

Create a Dataset

You can create a new Dataset in AI Studio:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.create(name="MyDataset",category='general')

Available Parameters:

Parameter	type	description	required	default
name	string	The name of the new Dataset	Yes
category	string	The type of dataset. options are: general, images, audio, video, text, tabular	No	"general"

Dataset ID

In some methods, you will need to use a dataset ID. The dataset ID is the unique name of the dataset, as seen in its URL.

For example, if you have a dataset that lives at: https://app.cnvrg.io/my_org/datasets/mydataset, the dataset ID is mydataset.

Get a Dataset

To fetch a Dataset from AI Studio you can use:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get('dataset-id')

Access Dataset attributes

You can access the Dataset attributes by using regular dot notation:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get('dataset-id')
ds.slug
ds.members
ds.last_commit

Available attributes:

Attribute	type	description
slug	string	The unique slug value of the Dataset
size	int	The size of the Dataset
title	int	The name of the Dataset
members	List	List of collaborators on this Dataset
category	string	The data structure category
description	string	Description of the Dataset
num_files	int	The number of files in the Dataset
last_commit	string	The last commit on this Dataset
current_commit	string	The current commit on this Dataset object

List all existing Datasets

You can list all the datasets in the current organization:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.list(sort="-created_at")

TIP

sort the list by: -key -> DESC | key -> ASC

Delete a Dataset

To delete a Dataset call the delete() function on it's instance:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get("dataset-id")
ds.delete()

Dataset File Operations

Clone the dataset to the current working directory

In order to clone dataset into your working directory, use following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get('dataset-id')
ds.clone()

Available Parameters:

Parameter	type	description	required	default
commit	string	Sha1 of the commit to clone	No	latest
override	boolean	re-clone even if the dataset already cloned	No	False
query_slug	string	slug of a query to clone	No	None
fullpath_filter	string	Filter on path of file by part of the path	No	None
use_cached	boolean	whether to use nfs cache link or not	No	False
threads	integer	number of threads that will be used in order to clone the project	No	40
check_disk_space	boolean	check there is enough space to clone dataset. clone won't execute if there isn't enough available space (leaving at least 5% of free space after clone)	No	True

TIP

The dataset will be cloned into a directory with its name, and the directory will be associated with the AI Studio dataset

Download dataset latest commit

In order to fetch dataset's data, use following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get('dataset-id')
ds.download()

WARNING

The Dataset must be cloned first

Upload files to a dataset

upload all file changes from local dataset directory

Upload all file changes from remote directory using following commad:

import os
from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get('dataset-id')
ds.clone()
os.chdir("dataset-id")
ds.upload()

Available Parameters:

Parameter	type	description	required	default
message	string	The commit message	No	""

WARNING

• This command must be used from a directory linked to a AI Studio Dataset
• Folder/file names must not contain special characters, which are - +?!@#$%&^*(){}[]

upload specific files from remote directory

ds.put_files(paths=['/files_dir/file1.txt', '/files_dir/file2.txt'])

Available Parameters:

Parameter	type	description	required	default
paths	List	The list of file paths that will be uploaded to the dataset	Yes
message	string	The commit message	No	""
override	boolean	Upload every file specified, even if it already exists in the latest commit	No	False
force	boolean	Create a new commit with only the files included in the current upload	No	False
git_diff	boolean	upload files from git diff output in addition to the given paths	No	Only upload files mentioned in the paths

WARNING

Folder/file names must not contain special characters, which are - +?!@#$%&^*(){}[]

NOTE

If a folder is given all the relevant files in that folder (that answers to the regex pattern) will be uploaded.

Remove files from a dataset

You can remove files from the dataset:

ds.remove_files(paths='*', 
          message='This will delete everything!')

Available Parameters:

Parameter	type	description	required	default
paths	List	The list of file/folder paths that will be deleted (regex and wildcard are allowed)	Yes
message	string	The commit message	No	""

NOTE

• When deleting files from a dataset paths parameter can be both a list of file paths or a string pattern like '*'
• In order to delete an enitre folder, specify its name followed by / for example: paths=["folder/"]

List Dataset content

You can list all of the files and folders that are in the dataset:

ds.list_files(query_raw='{"color": "yellow"}',
              sort='-id')
ds.list_folders(commit_sha1='xxxxxxxxx')

Available Parameters:

Parameter	type	description	required	default
commit_sha1	string	Sha1 string of the commit to list the files from	No	None
query	string	Query slug to list files from	No	None
query_raw	string	Raw query to list files according to Query language syntax	No	None
sort	string	Key to sort the list by (-key -> DESC / key -> ASC)	No	"-id"

Dataset Commits

Every Dataset in AI Studio may contain multiple data commits that you can interact with in the following manner:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get("dataset-id")  # Get the Dataset object

# Get A specific commit by its sha1 value
cm = ds.get_commit('xxxxxxxxx')
# OR list all available commits
commits = [cm for cm in ds.list_commits()]

# Last commits is available as attributes
last_commit = ds.last_commit

Checking Dataset Commit Indexing Status

To ascertain whether a dataset commit has completed its indexing process, execute the following code snippet:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get('dataset-id')
ds.get_commit('commit_sha1').is_indexed

Dataset Caching

Cache commit

In order to cache a dataset commit, use following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get("dataset-id")
dc= ds.get_commit("sha1")
dc.cache_commit('nfs-name')

NOTE

NFS connection needs to be configured in order to cache dataset commmits

Uncache Commit

In order to clear cached commit, use following command:

dc.clear_cached_commit('nfs-name')

Dataset Queries

You can create and use queries directly from AI Studio SDK to filter and use the Dataset exectly the way you want it by using Query language syntax

Create a new Query:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get("MyDataset")  # Get the Dataset object
ds.queries.create(name='OnlyPngFiles',
                  query='{"fullpath":"*.png"}',
            commit_sha1='xxxxxxxxx')

Available Parameters:

Parameter	type	description	required	default
name	string	The name of the query	Yes
query	string	The query string according to Query language syntax	Yes
commit_sha1	string	The sha1 value of the commit that this query will be based on	No	None

List all of the Dataset queries:

ds.queries.list(sort="-created_at")

TIP

sort the list by: -key -> DESC | key -> ASC

Get a specific query

q = ds.queries.get('slug')

Delete a query

q.delete()

Workspaces operations:

Workspace ID

In some methods, you will need to use the workspace ID(slug). The ID(slug) can be found in the last part of the workspace's URL .

For example, if you have a dataset that lives at: https://app.cnvrg.io/my_org/my_project/notebook_sessions/show/xaqn2mozyuivzwd7ajso, the ID is xaqn2mozyuivzwd7ajso.

Fetch the workspace object

Once you have the workspace'd id(slug) you can fetch it with the 'get' command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ws = myproj.workspaces.get("workspace-id")

NOTE

You can also reference the current running workspace from within its job scope:

from cnvrgv2 import Workspace
ws = Workspace()

Create a new workspace

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ws = myproj.workspaces.create(title="My Workspace", 
                              templates=["small","worker.medium"], 
                              notebook_type='jupyterlab')

If no parameters are provided, then the default values are used to further customize the created workspace you pass use the following parameters

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the workspace	No	None
templates	list	A list containing the names of the desired compute templates. when specifiying a compute template which is not on the default resource, add the name of the resouce followed by name of the template. for example: "intel.small"	No	project's default template
notebook_type	string	The notebook type (options are: "jupyterlab", "r_studio", "vscode")	No	'jupyterlab'
volume	string	The volume that will be attached to the workspace	No	None
datasets	list	A list of datasets to be connected and used in the workspace	No	None
image	object	Object of the image to be used for the workspace environment. for example: image=cnvrg.images.get(name="cnvrg", tag="v6.0")	No	project's default image
queue	string	Name of the queue to run on	No	default queue

Access workspace attributes

You can access the workspace attributes by using regular dot notation:

ws_slug = ws.slug
ws_title = ws.title
ws_datasets = ws.datasets
ws_notebook = ws.notebook_type

Sync workspace's file changes

To trigger sync in the running workspace:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ws = myproj.workspaces.get("workspace-slug")
ws.sync_remote()

Available Parameters:

Parameter	type	description	required	default
commit_msg	string	The commit message	No	None

Stop a running workspace

Stop a running workspace and sync it (the default is sync=False):

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ws = myproj.workspaces.get("workspace-slug")
ws.stop(sync=True)

Stop multiple workspaces at once:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
myproj.workspaces.stop(['workspace-slug1','workspace-slug2'],sync=True)

Start a stopped workspace

Start a stopped workspace:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ws = myproj.workspaces.get("workspace-slug")
ws.start()

List all of the workspaces

You can list all the workspaces in the current project, as well as sort them by a key in ASC or DESC order:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
wslist = myproj.workspaces.list(sort="-created_at")

TIP

sort the list by: -key for DESC or key for ASC

Delete workspaces

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ws = myproj.workspaces.get("workspace-slug")
ws.delete()

Delete multiple workspaces by listing their slugs:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
myproj.workspaces.delete(['workspace-slug1','workspace-slug2'])

Operate a Tensorboard

Start a Tensorboard session for an ongoing workspace

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ws = myproj.workspaces.get("workspace-slug")
ws.start_tensorboard()

Get the Tensorboard url:

ws.tensorboard_url

Stop the Tensorboard session:

ws.stop_tensorboard()

Experiment Operations

Experiment slug

In many commands, you will need to use an experiment slug. The experiment slug can be found in the URL for the experiment.

For example, if you have an experiment that lives at: https://app.cnvrg.io/my_org/projects/my_project/experiments/kxdjsuvfdcpqkjma5ppq, the experiment slug is kxdjsuvfdcpqkjma5ppq.

Fetch the experiment object

Once you have the experiment's slug you can fetch it with the 'get' command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.get("experiment-slug")

NOTE

You can also reference the current running experiment from within its job scope:

from cnvrgv2 import Experiment
e = Experiment()
print(e.slug) # Get the current running experiment's slug

Create a new experiment

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.create(title="my new exp", 
                              templates=["medium", "small"], 
                              command="python3 test.py")

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the experiment	No	None
templates	List	list of the compute templates to be used in the experiment (if the cluster will not be able to allocate the first template, then it will try the one after and so on..)	No	None
local	boolean	whether or not to run the experiment locally	No	False
local_arguments	dict	If local experiment and command is a function, local_arguments is a dictionary of the arguments to pass to the experiment's function	No	None
command	string	the starting command for the experiment (example: command='python3 train.py')	Yes
datasets	List	A list of dataset objects to use in the experiment	No	None
volume	Volume	Volume to be attached to this experiment	No	None
sync_before	boolean	Wheter or not to sync the environment before running the experiment	No	True
sync_after	boolean	Wheter or not to sync the environment after the experiment has finished	No	True
image	object	The image to run on (example: image=cnvrg.images.get(name="cnvrg", tag="v5.0")	No	project's default image
git_branch	string	The branch to pull files from for the experiment, in case project is git project	No	None
git_commit	string	The specific commit to pull files from for the experiment, in case project is git project	No	None
queue	string	Name of the queue to run on	No	default queue

Create a grid of experiments

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.create_grid(title="my new grid",  
                              command="python3 train.py", 
                              parameters=[{'key': 'learning_rate', 'type': 'discrete', 'values': [0.1, 0.01]}])

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the experiment	No	None
templates	List	list of the compute templates to be used in the experiment (if the cluster will not be able to allocate the first template, then it will try the one after and so on..)	No	None
parameters	list of dict	List of dictioneries of parameters to be passed to the grid. syntax is parameters=[{"key":"KEY", "type":"TYPE","value":"VALUE"},{..}]. example bellow	Yes
command	string	the starting command for the experiment (example: command='python3 train.py')	Yes
datasets	List	A list of dataset objects to use in the experiment	No	None
volume	Volume	Volume to be attached to this experiment	No	None
sync_before	boolean	Wheter or not to sync the environment before running the experiment	No	True
sync_after	boolean	Wheter or not to sync the environment after the experiment has finished	No	True
image	object	The image to run on (example: image=cnvrg.images.get(name="cnvrg", tag="v5.0")	No	project's default image
git_branch	string	The branch to pull files from for the experiment, in case project is git project	No	None
git_commit	string	The specific commit to pull files from for the experiment, in case project is git project	No	None
queue	string	Name of the queue to run on	No	default queue

Example of different types or parameters:

import yaml
params = """
    - key: "learning_rate"
      type: "discrete" # An array of numerical values
      values: [0.1, 0.01]

    - key: "kernel"
      type: "categorical" # An array of string values
      values: ["linear", "rbf"]

    - key: "epochs"
      type: "integer"
      min: 10 # inclusive
      max: 20 # not inclusive
      scale: "linear"
      steps: 2 # The number of linear steps to produce.
"""
e = myproj.experiments.create_grid(command="python3 train.py", parameters=yaml.safe_load(params))


### OR 
e = myproj.experiments.create_grid(command="python3 train.py", parameters=[{'key': 'learning_rate', 'type': 'discrete', 'values': [0.1, 0.01]}, {'key': 'kernel', 'type': 'categorical', 'values': ['linear', 'rbf']}, {'key': 'epochs', 'type': 'integer', 'min': 10, 'max': 20, 'scale': 'linear', 'steps': 2}])

Get an existing experiment

You can get the experiment object by using the experiment's slug:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.get("experiment-slug") 

list all experiment in the project

To get a list of all of the experiments in the project:

experiments = [e for e in myproj.experiments.list(sort="-created_at")]

TIP

sort the list by: -key for DESC or key for ASC

Delete experiment

You can delete a Experiment from a project by its slug value

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.get("experiment-slug")
e.delete()

# Do bulk delete on multiple experiments
myproj.experiments.delete(['experiment-slug1','experiment-slug2']) 

Available Parameters:

Parameter	type	description	required	default
delete_artifacts	boolean	Delete experiment's artifacts from storage	No	False

Stop a running experiment

Stop a running Experiment by passing its slug value (also All Experiment must be running)

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.get("experiment-slug")
e.stop()

# Do bulk stop on multiple experiments
myproj.experiments.stop(['experiment-slug1','experiment-slug2'])

Available Parameters:

Parameter	type	description	required	default
sync	boolean	sync experiment's file changes before stopping	No	True

Get experiment's system utilization

You can access the experiment's system resources usage data. For example, let's get the 5 last records for memomry utilization percentage:

utilization = e.get_utilization()
utilization.attributes['memory']['series'][0]['data'][-5:]  
[[1626601529000, 7.7], [1626601559000, 19.85], [1626601589000, 48.05], [1626601620000, 49.26]]

NOTE

The data syntax is [unix_timestamp, metric]

Track an experiment manually

You can initialize an empty Experiment in AI Studio:

cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.init("experiment-title") 

Now that the Experiment is initialized, its status is ONGOING and you can preform operations from within your code like with regular AI Studio Experiments in order to track

If you have initialized an Experiment object, you should conclude the experiment with the e.finish() command.

To conclude an experiment object:

exit_status = 0
e.finish(exit_status=exit_status)

NOTE

Options for exit_status: 0 for success, -1 for aborted, 1 and higher is error

Examples

To end an experiment with the status as Success:

e = myproj.experiments.init() #initialize experiment
# run code, track logs and metrics
e.finish(exit_status=0)

To end an experiment with the status as Error:

e = myproj.experiments.init() #initialize experiment
# run code, track logs and metrics
e.finish(exit_status=1)

To end an experiment with the status as Aborted:

e = myproj.experiments.init() #initialize experiment
# run code, track logs and metrics
e.finish(exit_status=-1)

Metadata operations on experiments

Add logs to a running experiment:

from datetime import datetime
e.log("my first log")
e.log(["my first log","my second log"])

Available Parameters:

Parameter	type	description	required	default
logs	array	Logs to be added to the experiment	Yes
timestamp	timestemp	timestemp to be added on	No	UTC Now
log_level	string	Log level. options are: "output", "error" ,"info", "warning"	No	"info"

Get the experiment's last 40 logs:

logs = e.logs()
for i in logs.attributes['logs']:
    print(i['message'])

Available Parameters:

Parameter	type	description	required	default
offset	integer	offset of logs chunk	No	-1 (latest)

Upload artifacts

You can add local files to the Experiments artifacts and create a new commit for it:

paths = ['output/model.h5']
e.log_artifacts(paths=paths)

Available Parameters:

Parameter	type	description	required	default
paths	list	list of paths of artifacts to save	Yes
git_diff	boolean	upload files from git diff output in addition to the given paths	No	False
work_dir	string	path to files. For example: when needed files are in dir ROOT/B. instead of sending paths B/., set working_dir="B", and send path as .	No	ROOT

Files will be uploaded to the experiment as output artifacts.

NOTE

Log only images available in the choosen path with log_images(file_paths=['*']) Images logged with this function will be displayed in the "Visuals" tab in the experiment's show page

Fetch a list of experiment's artifacts:

In order to see existing artifacts of your experiment, use following command:

cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.get("experiment-slug") 
a = e.list_artifacts()
for i in a:
    print(i.fullpath)

Download the experiment's artifacts

You can download the artifacts to your local working directory

e.pull_artifacts(wait_until_success=True, poll_interval=5)

Available Parameters:

Parameter	Type	Description	required	default
wait_until_success	boolean	Wait until current experiment is done before pulling artifacts	No	False
poll_interval	int	If wait_until_success is True, poll_interval represents the time between status poll loops in seconds	No	10

Merge experiment's commit into project files

In order to merge experiments artifacts into project's master branch, use following command:

e.merge_to_master()

Available Parameters:

Parameter	Type	Description	required	default
commit_sha1	string	Specific commit to be merges	No	latest commit created by the experiment

Download the experiment's artifacts

You can download the artifacts to your local working directory

e.pull_artifacts(wait_until_success=True, poll_interval=5)

Available Parameters:

Parameter	Type	Description	required	default
wait_until_success	boolean	Wait until current experiment is done before pulling artifacts	No	False
poll_interval	int	If wait_until_success is True, poll_interval represents the time between status poll loops in seconds	No	10

Merge experiment's commit into project files

In order to merge experiments artifacts into project's master branch, use following command:

e.merge_to_master()

Available Parameters:

Parameter	Type	Description	required	default
commit_sha1	string	Specific commit to be merges	No	latest commit created by the experiment

Create a Tag

To log a parameter to a running experiment, use:

e.log_param("key","value")

Parameters can be seen in the "Config and Metrics" section of the experiment's show page

Charts

You can create various charts by using the sdk, for example create a linechart showing the experiments loss:

from cnvrgv2 import Cnvrg, LineChart
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.get("exp-slug")

loss_vals = []
# experiment loop:
for epoch in range(8):
    loss_vals.append(loss_func())

# attach the chart to the experiment
loss_chart = LineChart('loss')
loss_chart.add_series(loss_vals, 's1')
e.log_metric(loss_chart)

You will immediately see the chart on the experiment's page:

WARNING

chart_name can't include "/"

Update existing chart:

Use following command to update existing chart with added values:

e.update_chart(key="loss",data=[1,2,3,4],series_name='s1')

You can create all different types of charts:

Heatmap:

In case of Heatmap, list of tuples that form a matrix. e.g, 2x2 matrix: [(0.5,1),(1,1)]

from cnvrgv2 import Heatmap
heatmap_chart = Heatmap('heatmap_example2',
                        x_ticks=['x', 'y'], y_ticks=['a', 'b'], min=0,max=10)
heatmap_chart.add_series([(0.5,1),(1,1)])
e.log_metric(heatmap_chart)

Typing information: x_ticks and y_ticks must be a List and matrix is a list of tuples in struct (x,y,z). color_stops is optional and is a List of Lists with size 2, where the nested first value is a float 0 <= X <= 1, and the second value is the hex value for the color to represent matrix values at that point of the scale. min and max are optional and should be numbers corresponding to the minimum and a maximum values for the key (scaling will be done automatically when these values are not submitted).

Each struct corresponds to a row in the matrix and to a label from the y_ticks list. The matrix is built from the bottom up, with the first struct and y_tick at the bottom edge. Each value inside the struct corresponds to each x_tick. Using steps and groups allow you to submit the same heatmap across different steps and visualize it in a single chart with a slider to easily switch between the charts. step should be an integer and group should be a string.

Steps and groups:

Using steps and groups allow you to submit heatmaps across different steps and visualize it in a single chart with a slider to easily move between the steps. step should be an integer and group should be a string. The group parameter requires the step parameter.

Bar chart:

• Single bar:

  from cnvrgv2 import BarChart
  bar_chart = BarChart('bar_example', x_ticks=['bar1', 'bar2'])
  bar_chart.add_series([1, 2],'s1')
  e.log_metric(bar_chart)
  

• Multiple bars:

  from cnvrgv2 import BarChart
  bar_chart = BarChart('bar_example1', x_ticks=['bar1', 'bar2'])
  bar_chart.add_series([1, 2, 3],'s2')
  bar_chart.add_series([3, 4],'s3')
  e.log_metric(bar_chart)
  

The x_ticks list will populate the labels for the bars, and the corresponding series values will dictate the value of the bar for that category. min and max are optional and are numbers that correspond the lower and upper bounds for the y values. Optionally, you can set each bar to be a specific color using the colors list of hex values, with each hex value corresponding to each x value.

Steps and groups:

Using steps and groups allow you to submit bar charts across different steps and visualize it in a single chart with a slider to easily move between the steps. step should be an integer and group should be a string. The group parameter requires the step parameter.

Operate a Tensorboard

Start a Tensorboard session for an ongoing experiment

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
e = myproj.experiments.get("experiment-slug")
e.start_tensorboard()

Get the Tensorboard url:

e.tensorboard_url

Stop the Tensorboard session:

e.stop_tensorboard()

Flow Operations

Flows can be created and run from any environment using the SDK. Creating flows requires using a flow configuration YAML file.

Create a Flow

You can use a flow YAML to create a flow inside a project. You can use either the absolute path to a YAML file or include the YAML content directly. Use the flows.create command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flow = proj.flows.create(yaml_path='PATH/TO/YAML')

Available Parameters:

Parameter	type	description	required	default
yaml_path	string	A path to the YAML configuration file	No	None
yaml_string	string	string of the yaml to create flow with	No	None

For example:

flow = """
---
flow: Flow 1
recurring: 
next_run_utc: 
tasks:
- input: python3 train.py
  params: []
  computes:
  - mini
  image: cnvrg:v5.0
  description: Task 1
  type: exec
  git_commit: 
  git_branch: 
  mount_folders: []
  icon: 
  output_dir: output
  confirmation: false
  standalone_mode: false
  notify_on_error: false
  notify_on_success: false
  emails: []
  objective: 
  objective_goal: 
  objective_function: min
  max_jobs: -1
  parallel_jobs: -1
  algorithm: GridSearch
  queue_slug: c9rlzkv5zazkdyxg7esg
  title: Task 1
  top: 50
  left: 50
  conditions: []
  commit: 908eea8c3aaa5059c55b4e3716a5bd96431c06b0
relations: []
"""

myproj.flows.create(yaml_string=flow)

Access Flow attributes

You can access the Flow's attributes by using regular dot notation:

Example:

>>> flow.title
'Training Task'

Flow Attributes:

Parameter	type	description
title	string	The name of the Flow
slug	string	The flow slug value
created_at	datetime	The time that the Flow was created
updated_at	datetime	The time that the Flow was last updated
cron_syntax	string	The schedule Cron expression string (If the Flow was scheduled)
webhook_url	string
trigger_dataset	string	A dataset that with every change will trigger this Flow

Flow slug

In some commands, you will need to use an Flow slug. The Flow slug can be found in the Flow page URL.

For example, if you have an Flow that lives at: https://app.cnvrg.io/my_org/projects/my_project/flows/iakzsmftgewhpxx9pqfo, the Flow slug is iakzsmftgewhpxx9pqfo.

Get a Flow

Get an existing Flow by passing its slug value or title

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flow = proj.flows.get("slug/title")

List Flows

You can list all existing flows:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flows = proj.flows.list()
for flow in flows:
    print(flow.title)

Run a Flow

To run The Flow's latest version:

flow.run()

Update Flow

You can update the existing Flow's title:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flow = proj.flows.get("slug/title")
flow.update(title="My Updated Flow")

Delete Flow

You can delete an existing Flow:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flow = proj.flows.get("slug/title")
flow.delete()

Or multiple Flows at once by listing all of the Flows slug values:

proj.flows.delete(["FLOW1_SLUG", "FLOW2_SLUG"])

Schedule a Flow

You can make the Flow run on schedule by using Cron expression syntax.

Set a new schedule:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flow = proj.flows.get("slug/title")
flow.set_schedule("* * * * *")  # Run every minute

Disable it with:

flow.clear_schedule()

Trigger webhook

You can create a webhook that will trigger the Flow run.

Toggle it on/off by setting the toggle parameter to True or False:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flow = proj.flows.get("slug/title")
flow.toggle_webhook(True)

Get the webhook url:

flow.webhook_url

NOTE

If you just toggled the webhook use flow.reload() before fetching the webhook_url

Toggle dataset update trigger

You can toggle the option to trigger on dataset update on/off, by setting the toggle parameter to True or False:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
ds = cnvrg.datasets.get("myds")
proj = cnvrg.projects.get("myproject")
flow = proj.flows.get("slug/title")

flow.toggle_dataset_update(True, cnvrg.datasets.get("dataset-slug"))

Flow versions

Every Flow have multiple versions and you can access them:

List all the flow versions:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flow = proj.flows.get("slug/title")
flow_versions = flow.flow_versions.list()
for fv in flow_versions:
    print(fv.title)

Get a specific flow version object by slug or title:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
flow = proj.flows.get("slug/title")
flow_version = flow.flow_versions.get("Version 1")

Get info of a flow version status:

info = flow_version.info()

Stop a running Flow version:

flow_version.stop()

#

Endpoint Operations

Create Endpoint

Use the following command to create a new endpoint:

from cnvrgv2 import Cnvrg
from cnvrgv2 import EndpointKind, EndpointEnvSetup
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
ep = proj.endpoints.create(title="myendpoint",
               templates=["small","medium"],
               kind=EndpointKind.WEB_SERVICE, 
               file_name="predict.py",
               stream=True,
               function_name="predict")

Available Parameters:

Parameter	type	description	required	default
title	string	Name of the Endpoint	Yes
kind	int	The kind of endpoint to deploy. options are: [WEB_SERVICE, STREAM, BATCH, TGI, GENERIC])	No	EndpointKind.WEB_SERVICE
templates	List	List of template names to be used	No	Default compute from project's settings
image	Image	Image object to create endpoint with	No	Default image from project's settings
file_name	string	The file containing the endpoint function	Yes
function_name	string	The name of the function the endpoint will route to	Yes
prep_file	string	The file containing the preprocess functions	No	None
prep_function	string	The name of the preprocess function	No	None
commit	string	Commit sha1 to use	No	latest
git_branch	string	Git branch to use	No	Default git branch from project's settings
git_commit	string	Git commit to use	No	latest
desired_percentage	integer	Traffic ratio for canary rollout	No	100 (full transition)
gunicorn_config	List	Gunicorn configurations in the following format: ["key=value", "key=value"]	No	None
flask_config	List	flask configurations in the following format: ["key=value", "key=value"]	No	None
input_file	boolean	Does endpoint accepts file	No	False
env_setup	string	The interpreter to use. options: [PYTHON2, PYTHON3, PYSPARK, RENDPOINT])	No	EndpointEnvSetup.PYTHON3
kafka_brokers	List	List of kafka brokers	No	None
kafka_input_topics	List	List of topics to register as input	No	None
queue	List	Name of the queue to run this job on	No	None
kafka_output_topics	List	List of topics to register as input	No	None
model_id	string	Name of the model id	No	None
stream	boolean	Does endpoint returns stream output	No	False
generic_command	string	Command for starting a generic endpoint	No	None

NOTE

Stream output is supported only at web service endpoints.

Endpoint slug

In many commands, you will need to use an endpoint slug. The endpoint slug can be found in the URL for the endpoint.

For example, if you have an endpoint that lives at: https://app.cnvrg.io/my_org/projects/my_project/endpoints/show/j46mbomoyyqj4xx5f53f, the endpoint slug is j46mbomoyyqj4xx5f53f.

Get Endpoint object

You can get Endpoints by passing their slug value:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
proj = cnvrg.projects.get("myproject")
ep = proj.endpoints.get('slug')

NOTE

You can also reference the current running endpoint from within its job scope:

from cnvrgv2 import Endpoint
ep = Endpoint()

List Endpoints

ep_list = proj.endpoints.list(sort='-created_at')  # Descending order

TIP

sort the list by: -key -> DESC | key -> ASC

Stop running Endpoints

Stop a running Endpoint using following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ep = myproj.endpoints.get("endpoint-slug")
ep.stop()

# Do bulk stop on multiple endpoints
myproj.endpoints.stop(['endpoint-slug1','endpoint-slug2'])

Start a stopped Endpoint

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ep = myproj.endpoints.get("endpoint-slug")
ep.start()

Delete Endpoints

You can delete a Endpoint from a project by its slug value

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ep = myproj.endpoints.get("endpoint-slug")
ep.delete()

# Do bulk delete on multiple endpoints
myproj.endpoints.delete(['endpoint-slug1','endpoint-slug2']) 

Endpoint Attributes

You can access the Endpoint attributes by using regular dot notation, for example:

>>> ep.api_key
'43iVTWTp55N7p62iSZYZLyuk'

Attribute	type	description
title	string	Name of the Endpoint
kind	int	The kind of endpoint (webservice, stream, batch)
updated_at	string	when was this Endpoint last updated
last_deployment	dict	details about the Endpoint's last deployment
deployments	List	list of dictionaries containing details about all of the Endpoint's deployments
deployments_count	int	The number of deployments that the Endpoint had
templates	List	List of compute templates that are assigned to the Endpoint
endpoint_url	string	The Endpoint's requests URL
url	string	The Endpoint's base URL
current_deployment	dict	The active deployment's data
compute_name	string	The name of the current compute template that is being used for the Endpoint to run
image_name	string	Name of the Endpoint's environment that is currently deployed
image_slug	string	The slug value of the Endpoint's deployed image
api_key	string	API key to access the Endpoint securely
created_at	string	The time that this endpoint was created
max_replica	int	Maximum number of pods to run this endpoint on
min_replica	int	Minimum number of pods0 to run this endpoint on
export_data	boolean	whether to export data or not
conditions	dict	conditions attached to this Endpoint and trigger a Flow/email every time one of them is met

Update The Endpoint's version

You can deploy a new version to the Endpoint and change some of its settings, for example:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ep = myproj.endpoints.get("endpoint-slug") 
ep.update_version(file_name="new_predict.py", commit="q7veenevzd83rewxgncx")

Available Parameters:

Parameter	type	description	required	default
title	string	Name of the Endpoint	Yes
kind	int	The kind of endpoint to deploy. options are: [WEB_SERVICE, STREAM, BATCH, TGI, GENERIC])	No	EndpointKind.WEB_SERVICE
templates	List	List of template names to be used	No	Default compute from project's settings
image	Image	Image object to create endpoint with	No	Default image from project's settings
file_name	string	The file containing the endpoint function	Yes
function_name	string	The name of the function the endpoint will route to	Yes
prep_file	string	The file containing the preprocess functions	No	None
prep_function	string	The name of the preprocess function	No	None
commit	string	Commit sha1 to use	No	latest
git_branch	string	Git branch to use	No	Default git branch from project's settings
git_commit	string	Git commit to use	No	latest
desired_percentage	integer	Traffic ratio for canary rollout	No	100 (full transition)
gunicorn_config	List	Gunicorn configurations in the following format: ["key=value", "key=value"]	No	None
flask_config	List	flask configurations in the following format: ["key=value", "key=value"]	No	None
input_file	boolean	Does endpoint accepts file	No	False
env_setup	string	The interpreter to use. options: [PYTHON2, PYTHON3, PYSPARK, RENDPOINT])	No	EndpointEnvSetup.PYTHON3
kafka_brokers	List	List of kafka brokers	No	None
kafka_input_topics	List	List of topics to register as input	No	None
queue	List	Name of the queue to run this job on	No	None
kafka_output_topics	List	List of topics to register as input	No	None
model_id	string	Name of the model id	No	None
stream	boolean	Does endpoint returns output stream	No	False
generic_endpoint	string	Command for starting a generic endpoint	'No' for all kinds other than 'generic'	False

Update the Endpoint's replica set

You can update the minimum and maximum number of pods to run the Endpoint on:

ep.update_replicas(min_replica=2, max_replica=5)

Rollback version

If you want to rollback the Endpoint version to a previous one you just need to pass the current version's slug value, for example:

ep.current_deployment['title']
3  # current version is 3
last_version_slug = ep.current_deployment["slug"] 
ep.rollback_version(version_slug=last_version_slug)
ep.reload()
ep.current_deployment['title']
2  # after the rollback the Endpoint's version is now 2

NOTE

To fetch the most updated attributes of the Endpoint, use ep.reload()

Get sample code

You can fetch the sample code to query the Endpoint (as shown in the Endpoint's main screen):

example:

sample_code = ep.get_sample_code()
sample_code['curl']
'curl -X POST \\\n    http://endpoint_title.cnvrg.io/api/v1/endpoints/q7veenevzd83rewx...'

Send prediction

In order to send a request to a running endpoint and get a prediction, use following command:

ep.predict(input_params)

Log metrics

To log mertics to an endpoint, use following command within the endpoint function:

### predict.py
from cnvrgv2 import Endpoint
def predict(data):
    ep = Endpoint()
    ep.log_metric("key",0.1)
    return(data)

Poll charts

You can fetch a dictionary with data about the Endpoint's latency performance, number of requests and user generated metrics from the Endpoint's charts:

>>> ep.poll_charts()

Add logs to a running endpoint

from datetime import datetime
ep.log("my first log")
ep.log(["my first log","my second log"])

Available Parameters:

Parameter	type	description	required	default
logs	array	Logs to be added to the experiment	Yes
timestamp	timestemp	timestemp to be added on	No	UTC Now
log_level	string	Log level. options are: "output", "error" ,"info", "warning"	No	"info"

Set feedback loop

You can grab all inbound data and feed it into a dataset for various uses, such as continuous learning for your models, for example:

from cnvrgv2 import Cnvrg
from cnvrgv2.modules.workflows.endpoint.endpoint import FeedbackLoopKind
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ep = myproj.endpoints.get("endpoint-slug") 
ds_slug = "dataset-name"
ep.configure_feedback_loop(dataset_slug=ds_slug,
                           scheduling_type=FeedbackLoopKind.IMMEDIATE)

Available Parameters:

Parameter	type	description	required	default
dataset_slug	string	slug of the receiving dataset	No	None
scheduling_type	integer	whether if the feedback loop is immediate (for every request) or recurring (for every time interval) (use FeedbackLoopKind.IMMEDIATE or FeedbackLoopKind.RECURRING)	No	FeedbackLoopKind.IMMEDIATE
cron_string	string	Cron syntax string if scheduling type is recurring	No	None

Disable the feedback loop:

ep.stop_feedback_loop()

NOTE

The data will be automatically uploaded to your dataset under predict/predictions.csv

Add Continual Learning rules

In order to set alerts to trigger email notification or flow execution, based on model behaviour in the endpoint, use following command:

rule = ep.add_rule(
    title="new-tule",
    description="some description",
    action={"type": "email","emails": ["EMAIL"]},
    metric="key",
    operation="lt",
    threshold="1",
    severity="info"
)

Available Parameters:

Parameter	type	description	required	default
title	string	title of the rule	Yes
description	string	rule description	No	None
metric	string	The cnvrg SDK metric used for the trigger. (Only tags with numeric values are supported)	Yes
threshold	float/integer	metric value to compare to trigger the alert	Yes
operation	string	The type of comparison used for comparing with your set value. options are: "gt" (greater than) or "lt" (less than)	Yes
action	json	Action to occur when the alert is triggered (a json of: type, webhook slug, flow_slug and emails)	Yes
severity	string	An indication of the importance of this alert (Info, Warning or Critical)	Yes
min_events	integer	Minimum number of events before triggering action	No	1
frequency	integer	How often to run condition (in minutes)	No	1

Disable the feedback loop:

ep.update_rule(rule.slug,is_active=False)

Delete feedback loop:

ep.delete_rule(rule.slug)

Control batch Endpoint

If the Endpoint is of batch type, then you can control it straight from the SDK:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
ep = myproj.endpoints.get("endpoint-slug")

# Check if it is running
ep.batch_is_running()

# Scale it up or down
ep.batch_scale_up()
ep.batch_scale_down()

#

Webapps Operations

Create a webapp

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
w = myproj.webapps.create(title="mywebapp",
                          templates=["small","medium"],
                          webapp_type="dash", 
                          file_name="app.py")

Available Parameters:

Parameter	type	description	required	default
webapp_type	string	The type of webapp to create ("shiny", "dash", "voila", "generic")	Yes
file_name	string	File name of the main app script	Yes (Except Generic type)
generic_command	string	The Generic command	No
title	string	Name of the webapp	No	None
templates	list	List of template names to be used	No	None
datasets	list	List of datasets to connect with the webapp	No	None
image	object	obgect of the image to be used (for example: image=cnvrg.images.get(name="cnvrg", tag="v6.0")	No	default image from project's settings
queue	string	Name of the queue to run this job on	No	default queue

Available attributes: You can access the WebApp attributes by using regular dot notation, for example:

>>> wb.webapp_type
'dash'

Attribute	type	description
webapp_type	string	The type of webapp ("shiny", "dash" or "voila")
template_ids	List	The size of the Dataset
title	int	The name of the Dataset
members	List	List of collaborators on this Dataset
category	string	The data structure category
description	string	Description of the Dataset
num_files	int	The number of files in the Dataset
last_commit	string	The last commit on this Dataset
current_commit	string	The current commit on this Dataset object

Webapp's slug

In many commands, you will need to use an webapp's slug. The webapp slug can be found in the URL of it.

For example, if you have an webapp that lives at: https://app.cnvrg.io/my_org/projects/my_project/webapps/kth1zj9xfh3gfvhfnh5k, the experiment slug iskth1zj9xfh3gfvhfnh5k`.

Get webapp object

Get a specific Webapp by its slug value

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
wb = myproj.webapps.get("webapp-slug")

NOTE

You can also reference the current running webapp from within its job scope:

from cnvrgv2 import Webapp
w = Webapp()

List all the webapps in the project

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
# sort them by decending order
wb = myproj.webapps.list(sort="-created_at")

TIP

sort the list by: -key -> DESC | key -> ASC

Delete webapp

You can delete a WebApp from a project by its slug value

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
wb = myproj.webapps.get("webapp-slug")
wb.delete()

# Do bulk delete on multiple webapps
myproj.webapps.delete(['webapp-slug1','webapp-slug2']) 

Stop running webapp

Stop a running WebApp by passing its slug value (sync=False by default, also All WebApps must be running)

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
wb = myproj.webapps.get("webapp-slug")
wb.stop(sync=False)

# Do bulk stop on multiple webapps
myproj.webapps.stop(['webapp-slug1','webapp-slug2'])

Share and Permissions

Share a running WebApp by using share and update_privacy commands

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
myproj = cnvrg.projects.get("myproject")
wb = myproj.webapps.get("webapp-slug")
wb.share(public_url_suffix="mysuffix")
wb.update_privacy(is_public=True) # to make the wb public

or in order to make the webapp only for project scope, the default is false:

wb.update_privacy(is_public=False)

Resource Operations

Connect your existing Kubernetes cluster

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
mycluster = cnvrg.clusters.create(resource_name="kubernetes_cluster",
                                  kube_config_yaml_path="kube_config.yaml",
                                  domain="https://app.cnvrg.io")

Available Parameters:

Parameter	type	description	required	default
resource_name	string	Name of the resource	Yes
kube_config_yaml_path	string	path to kubeconfig file	Yes
domain	string	domain url of existing kubernetes_cluster	Yes
provider_name	string	the provider name (options are: "aws","gke","eks","ics"...)	No	None
scheduler	string	supported schedulers to deploy cnvrg jobs	No	cnvrg_scheduler
namespace	string	the namespace to use inside the cluster	No	cnvrg
https_scheme	boolean	resource supports HTTP/S urls when accessing jobs from the browser	No	False
persistent_volumes	boolean	resource can dynamically create PVCs when running jobs	No	False
gaudi_enabled	boolean	the cluster support HPU devices	No	False

Build managed EKS cluster (Supported on MetaCloud)

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
mycluster = cnvrg.clusters.create(build_yaml_path="aws.yaml", provider_name="aws")

Available Parameters:

Parameter	type	description	required	default
network	string	if left blank cnvrg will automatically provision the network for your cluster	No	istio

Yaml example:

name: mycluster
version: '1.21'
roleARN: arn:aws:iam::123456789101:role/cnvrg_role
region: us-west-2
vpc: null
publicSubnets:
  - ''
privateSubnets:
  - ''
securityGroup: ''
nodeGroups:
  - availabilityZones:
      - us-west-2a
      - us-west-2b
      - us-west-2d
      - us-west-2c
    autoScaling: false
    instanceType: m5.metal
    desiredCapacity: 2
    minSize: 0
    maxSize: 2
    spotInstances: false
    volumeSize: 100
    privateNetwork: true
    securityGroups:
      - ''
    tags:
      - key: ''
        value: ''
    taints:
      - key: ''
        value: ''
    labels:
      - key: ''
        value: ''
    attachPolicies:
      - ''
    addonPolicies:
      - key: ''
        value: ''

Get an existing resource

You can get the resource object by using the resource's slug:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
mycluster = cnvrg.clusters.get(slug="cluster-slug")

You can also get all of the resources in the organization:

clusters = [c for c in cnvrg.clusters.list()]

Update an existing resource

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
mycluster = cnvrg.clusters.get(slug="cluster-slug")
mycluster.update(resource_name="new-name")

List of optional parameters:

Parameter	type	description	required
resource_name	string	Name of the resource	No
kube_config_yaml_path	string	path to kubeconfig file	No
domain	string	domain url of existing kubernetes_cluster	No
scheduler	string	supported schedulers to deploy cnvrg jobs	No
namespace	string	the namespace to use inside the cluster	No
https_scheme	boolean	resource supports HTTP/S urls when accessing jobs from the browser	No
persistent_volumes	boolean	resource can dynamically create PVCs when running jobs	No
gaudi_enabled	boolean	the cluster support HPU devices	No

Delete a resource

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
mycluster = cnvrg.clusters.delete(slug="cluster-slug")

or

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
mycluster = cnvrg.clusters.get(slug="cluster-slug")
mycluster.delete()

Templates Operations

Create a new template

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.create(title="Regular template",type="regular", 
                              cpu=2,memory=4,gpu=1)

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the template	Yes
template_type	string	The type of the template	No	regular
cpu	float	Number of CPUs	Yes
memory	float	Amount of Memory
gpu	integer	Number of GPUs	No	None
is_private	boolean	Does the template are private and not all users have access	No	False
users	array	the users that can use this template, if private (Administrators have permissions to all compute resources)	No	None
labels	string	Labels for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
taints	string	Taints for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.create(title="Mpi template",type="mpi", 
                              cpu=2,memory=4,gpu=1,worker_num_executors=2,
                              worker_cpu=2.0, worker_memory=2.0, worker_gpu=1,
                              worker_labels="gpu=v100,playground=dev")

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the template	Yes	None
type	string	The type of the template	Yes	regular
cpu	float	Number of CPUs	Yes	None
memory	float	Amount of Memory	Yes	None
gpu	float	Number of GPUs	No	None
shm	float	Amount of shared memory (/dev/shm) to mount to the pod	No	None
huge_pages	float	Amount of huge pages (Mi)	No	None
gaudi	float	The amount of HPU	No	None
is_private	boolean	Does the template are private and not all users have access	No	False
labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
taints	string	Taints or selector for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None
mig_device	string	The mig device to use	No	None
users	array	the users that can use this template (Administrators have permissions to all compute resources)	No	None
worker_num_executors	int	The amount of workers	Yes	None
worker_cpu	float	Number of worker CPUs	Yes	None
worker_memory	float	Amount of worker Memory	Yes	None
worker_huge_pages	string	Amount of worker huge pages (Mi)	No	None
worker_gpu	float	Number of worker GPUs	No	None
worker_mig_device	string	The mig device to use	No	None
worker_labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
worker_taints	string	Taints or selector for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.create(title="Pytorch template",type="pytorch",  
                              cpu=2,memory=4,gpu=1,worker_num_executors=2)

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the template	Yes	None
type	string	The type of the template	Yes	regular
cpu	float	Number of CPUs	Yes	None
memory	float	Amount of Memory	Yes	None
gpu	float	Number of GPUs	No	None
shm	float	Amount of shared memory (/dev/shm) to mount to the pod	No	None
huge_pages	float	Amount of huge pages (Mi)	No	None
gaudi	float	The amount of HPU	No	None
is_private	boolean	Does the template are private and not all users have access	No	False
labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
taints	string	Taints or selector for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None
mig_device	string	The mig device to use	No	None
users	array	the users that can use this template (Administrators have permissions to all compute resources)	No	None
worker_num_executors	int	The amount of workers	Yes	None
worker_huge_pages	string	Amount of worker huge pages (Mi)	No	None
worker_mig_device	string	The mig device to use	No	None
worker_labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
worker_taints	string	Taints or selector for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
spark_options="spark.executor.memory=2G\nspark.executor.cores=4\nspark.executor.instances=2"
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.create(type="spark_on_kubernetes",
                                    title="Spark template",
                                    cpu=2,memory=4,gpu=1,
                                    spark_config=spark_options)

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the template	Yes	None
type	string	The type of the template	Yes	regular
cpu	float	Number of CPUs	Yes	None
memory	float	Amount of Memory	Yes	None
gpu	float	Number of GPUs	No	None
shm	float	Amount of shared memory (/dev/shm) to mount to the pod	No	None
huge_pages	float	Amount of huge pages (Mi)	No	None
gaudi	float	The amount of HPU	No	None
is_private	boolean	Does the template are private and not all users have access	No	False
labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
taints	string	Taints or selector for the node pools the template should use. For example, value1:NoSchedule.To specify several, separate them with commas.	No	None
mig_device	string	The mig device to use	No	None
users	array	the users that can use this template (Administrators have permissions to all compute resources)	No	None
spark_config	string	spark configuration options, To specify several, separate them with \n	No	None

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.create(title="Ray template",type="ray", 
                              cpu=2,memory=4,gpu=1,worker_num_executors=2,
                              worker_cpu=2.0, worker_memory=2.0, worker_gpu=1,
                              worker_labels="gpu=v100,playground=dev")

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the template	Yes	None
type	string	The type of the template	Yes	regular
cpu	float	Number of CPUs	Yes	None
memory	float	Amount of Memory	Yes	None
gpu	float	Number of GPUs	No	None
shm	float	Amount of shared memory (/dev/shm) to mount to the pod	No	None
huge_pages	float	Amount of huge pages (Mi)	No	None
gaudi	float	The amount of HPU	No	None
is_private	boolean	Does the template are private and not all users have access	No	False
labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
taints	string	Taints or selector for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None
mig_device	string	The mig device to use	No	None
users	array	the users that can use this template (Administrators have permissions to all compute resources)	No	None
worker_num_executors	int	The amount of workers	Yes	None
worker_cpu	float	Number of worker CPUs	Yes	None
worker_memory	float	Amount of worker Memory	Yes	None
worker_huge_pages	string	Amount of worker huge pages (Mi)	No	None
worker_gpu	float	Number of worker GPUs	No	None
worker_mig_device	string	The mig device to use	No	None
worker_labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
worker_taints	string	Taints or selector for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.create(title="Ray modin template",type="ray_modin", 
                              cpu=2,memory=4,gpu=1,worker_num_executors=2,
                              worker_cpu=2.0, worker_memory=2.0, worker_gpu=1,
                              worker_labels="gpu=v100,playground=dev")

Available Parameters:

Parameter	type	description	required	default
title	string	The name of the template	Yes	None
type	string	The type of the template	Yes	regular
cpu	float	Number of CPUs	Yes	None
memory	float	Amount of Memory	Yes	None
gpu	float	Number of GPUs	No	None
shm	float	Amount of shared memory (/dev/shm) to mount to the pod	No	None
huge_pages	float	Amount of huge pages (Mi)	No	None
gaudi	float	The amount of HPU	No	None
is_private	boolean	Does the template are private and not all users have access	No	False
labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
taints	string	Taints or selector for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None
mig_device	string	The mig device to use	No	None
users	array	the users that can use this template (Administrators have permissions to all compute resources)	No	None
worker_num_executors	int	The amount of workers	Yes	None
worker_cpu	float	Number of worker CPUs	Yes	None
worker_memory	float	Amount of worker Memory	Yes	None
worker_huge_pages	string	Amount of worker huge pages (Mi)	No	None
worker_gpu	float	Number of worker GPUs	No	None
worker_mig_device	string	The mig device to use	No	None
worker_labels	string	Labels or selector for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.	No	None
worker_taints	string	Taints or selector for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.	No	None

Get an existing template

You can get the template object by using the template's slug:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.get("template_slug")

List all existing templates

List all templates that the current user is allowed to view

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
templates = cluster.templates.list()
for template in templates:
    print("Template Details: title: {} , slug: {} , cpu: {} , memory: {} "
          .format(template.title, template.slug, template.cpu, template.memory))

Update an existing template

You can update the existing template attributes:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.get("template_slug")
template.update(title="new title",cpu=3)

Parameters Available for Update:

Parameter	type	description
title	string	The name of the template
cpu	float	Number of CPUs
memory	float	Amount of Memory
gpu	integer	Number of GPUs
is_private	boolean	Does the template are private and not all users have access
users	array	the users that can use this template, if private (Administrators have permissions to all compute resources)
labels	string	Labels for the node pools the template should use. For example, gputype=v100.To specify several, separate them with commas.
taints	string	Taints for the node pools the template should use. For example, value1=NoSchedule.To specify several, separate them with commas.

WARNING

If you change the template's status from public to private, you must specify a list of authorized users. Failure to do so will result in an error message indicating that authorization is required. Please ensure that all specified users have the appropriate permissions to access the private template.

Delete an existing template

You can delete an existing template:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
cluster = cnvrg.clusters.get("cluster_slug")
template = cluster.templates.get("template_slug")
template.delete()

Registry commands

Create a registry

To create a registry in the cnvrg environment, use the following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
registry = cnvrg.registries.create(title="TITLE", url="URL")

Available Parameters:

Parameter	type	description	required	default
title	string	Registry title	Yes
url	string	Registry URL	Yes
type	string	Registry type. options are: cnvrg, dockerhub, gcr, acr, ecr, nvidia, other	No	'other'
username	string	Registry username	required for private registries	None
password	string	Registry password	required for private registries	None

Get a registry

To retrieve information about a registry, use the following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
registry = cnvrg.registries.get(slug="cnvrg")
print(registry.url) ## docker.io/cnvrg
print(registry.title) ## cnvrg

List all registries

To retrieve all the registries in the AI Studio environment, use the following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
registry = cnvrg.registries.list()

Update a registry

To update or modify a registry, use the following command with the required combination of options corresponding to the fields you want to modify:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
registry = cnvrg.registries.get(slug="my-registry")
registry.update(title="new-title")

Available Parameters:

Parameter	type	description	required
title	string	Registry title	No
url	string	Registry URL	No
type	string	Registry type. options are: cnvrg, dockerhub, gcr, acr, ecr, nvidia, other	No
username	string	Registry username	required for private registries
password	string	Registry password	required for private registries

Delete a registry

To delete a registry from the AI Studio environment, use the following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
registry = cnvrg.registries.get(slug="my-registry")
registry.delete()

Image commands

Create an image

To create an image, use the following command:

from cnvrgv2 import Cnvrg
cnvrg = Cnvrg()
image = cnvrg.images.create(name="cnvrg", tag="v6.0", registry="cnvrg")

Available Parameters:

Parameter	type	description	required	default
name	string	Image repository name	Yes
tag	string	The image tag	Yes
registry	string	The slug of the registry that the image will be added to	Yes
dockerfile	string	path to Dockerfile path to build a custom image	No	None
custom	boolean	hether the image is custom (needs to be built with dockerfile)	No	False
readme	string	path to README file	No	None

Get an image

To retrieve information about an image, use the following command:

image = cnvrg.images.get(name="cnvrg", tag="v6.0")
print(image.slug)
print(image.dockerfile) ## when reffering to an image built from dockerfile

List all images

To retrieve all the images in the AI Studio environment, use the following command:

images = cnvrg.images.list(sort='-id')

TIP

sort the list by: -key -> DESC | key -> ASC

Delete an image

To delete an image using its name and tag use the following command:

image = cnvrg.images.get(name="cnvrg", tag="v6.0")
image.delete()

TIP

You can also delete an image using its slug instead