Skip to main content

SDK to develop using Lightning AI Studios

Project description

Lightning SDK

Software Development Kit (SDK) for Lightning AI

Installation

The package can be installed using pip install lightning-sdk

Usage

To use the SDK, you need to export the environment variables LIGHTNING_USER_ID and LIGHTNING_API_KEY with the values found in your account settings -> Keys -> Programmatic Login.

If you want to use it from within a Studio, these variables are already available for you.

Example

from lightning_sdk import Machine, Studio

# or s =  Studio("my-studio", "my-teamspace", org="my-org", create_ok=True)
# or (inside a Studio) s = Studio()  # will infer name, teamspace and owner of the current studio automatically.
#    can also just pass some arguments: s = Studio("my-new-studio", create_ok=True)
s = Studio("my-studio", "my-teamspace", user="my-username", create_ok=True)

print("starting Studio...")
s.start()

# prints Machine.CPU-4
print(s.machine)

# or start directly on this machine with s.start(Machine.A10G)
print("switching Studio machine...")
s.switch_machine(Machine.A10G)

# prints Machine.A10G
print(s.machine)

# prints Status.Running
print(s.status)

print(s.run("nvidia-smi"))

# install plugins for jobs and multi-machine training
s.install_plugin("jobs")
s.install_plugin("multi-machine-training")

# run the resulting plugins to start 1 job and 1 multi-machine training
s.installed_plugins["jobs"].run("python my_dummy_file", name="my_first_job", machine=Machine.A10G)
s.installed_plugins["multi-machine-training"].run("python my_dummy_file", name="my_first_mmt", machine=Machine.T4, num_instances=42)

print("Stopping Studio")
s.stop()

# duplicates Studio, this will also duplicate the environment and all files in the Studio
duplicate = s.duplicate()

# delete original Studio, duplicated Studio is it's own entity and not related to original anymore
s.delete()

# stop and delete duplicated Studio
duplicate.stop()
duplicate.delete()

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distribution

lightning_sdk-0.1.29-py3-none-any.whl (1.5 MB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page