The
democratization of
Deep Learning and Artificial Intelligence
Don't you wish you could just build a huge ML model with unlimited
computer power, without any coding or expensive hardware?
Well, now you can!
With EzSpark, you can build and train AI models of any size through a DISTRIBUTED network of
users with or without coding. And it's
free!
So what are you waiting for? Request beta access now.
You don't need to be a programmer or an engineer to create AI. Use easily our configuration setup interface, you'll create your own AI in minutes.
You've been waiting for this. It's finally here! We are proud to present: API access to train your own ML model. You don't need to wait any longer, you can get started right now, even without a GPU or Google colab, all it takes is an internet connection.
Use our social to make other people join your training! They don't need any software, login or setup, just an internet connection and a browser!
EzSpark developed distributed algorithms to train in the fastest possible way your AI. Trainers can use either the browser through a complex webassembly front-end, or python packages to train your models
Yes. You can request the beta access and you will be put on a waiting list. All the features currently available are free
Since we are in beta, we'd like making stress tests through an appropriate workload we've pre-defined. We expect from the testers to be fully active. For this reason we'll privilege programmers, game-developers and researchers since these are the people most used to exploit ML models for their projects.
We've defined 2 type of users: Hosts and Trainers. The hosts are the beta testers, they can create a configuration file for their training and save it. Then, they use the API of EzSpark and the appropriate package we've created to train their models. The trainers can be anyone. They can train other's people models by web without installing anything or without any login, or with the appropriate python package.
For the hosting of the training you will use Pyezspark, a python package. For the inference of the models you can use Pyllab, a python package, or llab a C/C++ package, or jsllab a javascript package. For the training you can use Pyezspark or you can train directly from the browser
In beta we are focusing on reinforcement learning with genetic algorithms. Then, after the beta, will be released an update with a private and powerfull distributed deep learning algorithm we've created and succesfully tested.
Look at our roadmap to be kept updated!