NEWAnnouncing TPUv2 support for Gradient°. Learn more →

Effortless infrastructure for Machine Learning and Data Science



Watch the intro video

What is Gradient° ?

Gradient° is a full suite of tools for exploring data, training neural networks, and building production-grade machine learning pipelines. Gradient° includes 1-click jupyter notebooks, a powerful job runner, and a python module to run any code with full power of GPU cloud infrastructure.

01

Launch a GPU-enabled Jupyter Notebook with 1-Click

These interactive coding environments let you combine code with equations, visualizations, rich text, and other media.

They make it easy to explore data and coding concepts, collaborate with other people on projects, and easily get a coding environment provisioned in seconds.

Hello, TensorFlowRun
Fast.AI Pytorch 0.3Run
Datascience toolbox Run

02

Train and deploy directly from your command line.

With the Gradient° job runner you work on your local machine and submit "jobs" to the cloud to be processed. It only takes a few lines of code to start running on Gradient. Note, if you don't already have a free Paperspace account you can easily create one here.

A Gradient° job consists of:

  • a collection of files (code, resources, etc)
  • a docker container (with code dependencies and packages pre-installed)
  • a command to execute (i.e. python main.py or nvidia-smi)


First you need to install the Paperspace Command Line Interface (CLI).

Next, login through the CLI and download the sample project which runs style transfer on a video.

# Login to the CLI
$ paperspace login

# Then clone the test repository to your local machine
$ git clone https://github.com/Paperspace/fast-style-transfer && cd fast-style-transfer

Now, run the paperspace jobs command to submit this code to Gradient° . Check the documentation to learn more about this command.

# Run the repo as a GPU job
$ paperspace jobs create --container Test-Container --machineType GPU+ --command "/paperspace/run.sh"

Powerful toolkit for training ML models.

Bring the cloud to you with our powerful CLI and native Python integrations. Integrate any existing project with just a few lines of code.

Accelerate your ML workflow.

Gradient° is a powerful cloud IDE for modern Machine Learning, AI, and data science. Built on best practices and a commitment to performance, Gradient° helps you develop better and faster.

Continuous integration (CI/CD) for AI

Move your ML pipeline in to the future with modern, reproducible, and deterministic processes. Everything is tagged with a UUID, and native support for containers lets you control every part of the pipeline. Remove unnecessary complexity to accelerate your development.

Visualize training metrics

Easily visualize your training metrics. You have complete flexibility to easily graph any metric. Simply log out a
{"chart": "<identifier>", "axis": "<label>"}
in any job to preserve metrics.

03

One Line of code to run any python script on a GPU in the cloud.

If you don't already have a free Paperspace account, go ahead and create one first. Once you have done that, you will need to download the Paperspace python module. It is compatible with Python 2 and 3.


# Use pip
$ pip install paperspace

# Or if you use Anaconda
$ conda install paperspace -c paperspace

Import the package to your main python file

# within your python file at the top
import paperspace

Run your code with the paperspace.run() command in your python file.

# add this line to the end of your main python file
paperspace.run()

Pricing

Gradient° runs on the Paperspace cloud which means you get the best prices for GPUs in the industry. Easily scale up or down as your needs change.

Individuals
Teams

Individual

For hobbyists, explorers, and adventurous learners

Pay-as-you-go
 
VM access (desktop, terminal, SSH)
Standard GPU access
10 GPU Jobs Limit (1 concurrent)
10 Notebooks Limit (1 concurrent)

G1

Developer

For ML/AI developers, researchers, and professional users

$8/mo
+ machine utilization charges
VM access (desktop, terminal, SSH)
Priority GPU access
Persistent Drive
50 GPU Jobs Limit (5 concurrent)
25 Notebooks Limit (5 concurrent)
API

 

Enterprise

Production-grade tools for ML/AI companies and teams.

contact us
 
VM access (desktop, terminal, SSH)
Priority GPU access
Persistent Drive
Unlimited GPU Jobs
Unlimited Notebooks
API
Priority support
Team features / ACL

04

Enterprise deep learning at scale

A single platform for ML development, collaboration, deployment, and monitoring.

Gradient is build on GRAI, our AI orchestration fabric. Learn more about how we provide production-grade ML tooling for companies of all sizes. Contact sales today.


Built by Paperspace on the world's most advanced GPU cloud.