From the makers of

PyTorch Lightning

Seamlessly train hundreds of Machine Learning models on the cloud from your laptop.

Focus on machine learning, not infrastructure.
Join our Waitlist
How it Works
Easy install
Install the Grid CLI and log in with your GitHub account.
Grid Train
Train on the cloud seamlessly from your laptop with one command.
Grid View
Monitor and manage your training infrastructure with our Web App or CLI.
Analyze with TensorBoard
Analyze your experiments with Tensorboard (or whatever logger you love).
Grid saves your Artifacts
Artifacts like weights and logs are automatically backed up for you - find them whenever you need!
Notebooks on your Infrastructure
Analyze and debug using GPU-powered notebooks or VSCode.
Focus on machine learning,
let Grid handle the engineering.

Built for Lightning Users

Grid and Lightning (or whichever framework you use) seamlessly automate all of your engineering.

100% Managed Infrastructure

Train instantly on CPUs, GPUs, or TPUs in the cloud - Grid handles all of it on your behalf.

Effective Resource Allocation

Grid makes sure your experiments are distributed as efficiently as possible across cloud resources.

Built for Lightning Users

Grid and Lightning (or whichever framework you use) seamlessly automate all of your engineering.

Automated Artifact Backup

Experiment artifacts, like weights and logs, are automatically saved to your cloud - access them anytime during or after your run.

100% Managed Infrastructure

Train instantly on CPUs, GPUs, or TPUs in the cloud - Grid handles all of it on your behalf.

No Code Changes

To use Grid, simply git clone your project, change python to grid train, then hit enter.

Automated Artifact Backup

Experiment artifacts, like weights and logs, are automatically saved to your cloud - access them anytime during or after your run.

No Code Changes

To use Grid, simply git clone your project, change python to grid train, then hit enter.

You own your data

Your data and artifacts stay on your infrastructure, Grid simply orchestrates.

Effective Resource Allocation

Grid makes sure your experiments are distributed as efficiently as possible across cloud resources.

You own your data

Your data and artifacts stay on your
infrastructure, Grid simply orchestrates.

By Lightning, for Lightning.

The makers of PyTorch Lightning brought the philosophy of power, simplicity, and usability to model training. Other frameworks can leverage some of the magic too.

We offer 3 different plans

Individual

Subline

Team

Subline

Enterprise

Subline

Base features

Unlimited projects
Artifacts storage
Ulimited history
Journals
10 TB
10 TB
10 TB

Compute

AWS
Concurrent GPUs
SLURM
Team Management
Unlimited Teams
On Prem Deployments
Pay-as-you-go
Max 32
Free
_
_
_
Use your own credentials
Max 1,024
Management Dashboard
Dashboard
_
_
Use your own credentials
Unlimited
Management Dashboard
Dashboard

Support

Slack Support
Private Slack Channel
1 Day response
Private Slack Channel
Less than 30 mins response

Base features

Unlimited projects
Artifacts storage
Ulimited history
Journals
10 TB

Compute

AWS
Concurrent GPUs
SLURM
Team Management
Unlimited Teams
On Prem Deployments
Pay-as-you-go
Max 32
Free
_
_
_

Support

Slack Support

Base features

Unlimited projects
Artifacts storage
Ulimited history
Journals
10 TB

Compute

AWS
Concurrent GPUs
SLURM
Team Management
Unlimited Teams
On Prem Deployments
Use your own credentials
Max 1,024
Management Dashboard
Dashboard
_
_

Support

Private Slack Channel
1 Day response

Base features

Unlimited projects
Artifacts storage
Ulimited history
Journals
10 TB

Compute

AWS
Concurrent GPUs
SLURM
Team Management
Unlimited Teams
On Prem Deployments
Use your own credentials
Unlimited
Management Dashboard
Dashboard

Support

Private Slack Channel
Less than 30 mins response
Join our waitlist for early access to Grid
Join our waitlist
Copyright Grid AI