[N] NNAISENSE releases EvoTorch (https://evotorch.ai): An open-source Evolutionary Algorithm Library with multi-CPU/multi-GPU support for massive evolutionary experiments!

Hi everyone!

We're excited to release EvoTorch: An open-source Evolutionary Algorithm Library with multi-CPU/multi-GPU support for massive evolutionary experiments!

We are researchers and engineers in industrial automation at NNAISENSE and this library is the latest version of the tool we've been using in our own work. It is built on top of PyTorch and Ray, and provides a collection of state-of-the-art evolutionary algorithms, out-of-the-box support for scaling experiments to arbitrary clusters of CPUs and GPUs, advanced tools for NeuroEvolution of any PyTorch module, and direct interfaces to modern logging libraries to track experiments and integrate with existing workflows.

Our goal is to make many more engineers be able to use EAs in their workflow, and help researchers do better research.

We are happy to have your feedback or questions here or on our Slack channel!

EvoTorch is pip installable: pip install evotorch

Read the docs: https://docs.evotorch.ai

The library is completely open source under an Apache 2.0 license: https://github.com/nnaisense/evotorch