Spotty drastically simplifies training of deep learning models on AWS and GCP:

it makes training on GPU instances as simple as training on your local machine
it automatically manages all necessary cloud resources including images, volumes, snapshots and SSH keys
it makes your model trainable in the cloud by everyone with a couple of commands
it uses tmux to easily detach remote processes from their terminals
it saves you up to 70% of the costs by using AWS Spot Instances and GCP Preemtible VMs

Documentation

Installation

Requirements:

Use pip to install or upgrade Spotty:

$ pip install -U spotty

Prepare a spotty.yaml file and put it to the root directory of your project:
- See the file specification here.
- Read this article for a real-world example.
Start an instance:
```
$ spotty start
```
It will run a Spot Instance, restore snapshots if any, synchronize the project with the running instance and start the Docker container with the environment.
Train a model or run notebooks.

To connect to the running container via SSH, use the following command:
```
$ spotty sh
```
It runs a tmux session, so you can always detach this session using Ctrl + b, then d combination of keys. To be attached to that session later, just use the spotty sh command again.

Also, you can run your custom scripts inside the Docker container using the spotty run <SCRIPT_NAME> command. Read more about custom scripts in the documentation: Configuration: "scripts" section.

Any feedback or contributions are welcome! Please check out the guidelines.