Top Free Jupyter Runtimes to get started with Machine Learning journey

Jupyter notebooks are a great tool for data analysis due to rich interface features it offers. But for training fairly good models require good amount CPU, GPU, RAM and storage. Here are some of the cloud GPU runtimes that you to use freely to spin up a Jupyter notebook instance.

1. Google Colab

Colab, or "Colaboratory" is a cloud based Jupyter notebook runtime provided by Google. You can start using Colab if you have a Google account.

Highlights

Colab offers Dual core 2.20 GHz processor with 12 GB RAM & 70 GB storage.
Access to GPU & TPU runtimes.(Limited use in free tier)
Ability to mount your Google Drive.
Save notebook files to google drive.

2. AWS Sagemaker

AWS Sagemaker is a cloud based Jupyter notebook runtime provided by Amazon. You have to request for an account by going to this link.

Highlights

Sagemaker offers dual core 2.50 GHz processor with 15 GB RAM & 15 GB persistent storage.
Access to GPU runtime.
CPU sessions last for 12 hours & GPU sessions run for 4 hours.
Once the session is done, you will have to restart the project runtime again, but all of your files will be saved on the persistent project storage.

3. Kaggle Kernel

Kaggle is a subsidiary of Google LLC, is an online community of data scientists and machine learning practitioners. Kaggle allows users to find and publish data sets, explore and build models in a web-based data-science environment, work with other data scientists and machine learning engineers, and enter competitions to solve data science challenges.

Highlights

Kaggle offers dual core 2.20 GHz processor with 16 GB RAM & 20 GB persistent storage.
Access to GPU runtime.(30 hours per week)
Access to TPU runtime.(20 hours per week)

4. Saturn cloud

Saturn cloud offers a Jupyter runtime that runs on top of AWS.

Highlights

CPU runtime with upto 64 GB RAM & 100 GB disk space.
GPU runtime with 16 GB RAM.
Only 30 hours of Jupyter runtime is available per month.
Offers 3 hours of DASK runtime is available.

5. Gradient

Gradient is a cloud based platform for building machine learning applications & they offer a basis free tier.

Highlights

-Offers 5GB of storage

6. Deepnote

Deepnote is a new kind of data science notebook. Jupyter compatible with real-time collaboration and runs in the cloud.

Highlights

-Deepnote offers an instance with a dual core 2.50 GHz processor with 11 GB of RAM and 5 GB of storage. -Provides 750 hours of compute monthly free.