Releases: for-ai/rl
Releases · for-ai/rl
v1.0.1
Alpha Release
This release includes a fully generic, reinforcement learning codebase built with TensorFlow 1.0. We have provided a number of algorithms, models, and benchmarked environments. This release will be useful to both early-stage ML learners looking to see how RL algorithms are implemented and experienced researchers aiming to build of top of existing work. To this end, we have tried to make the codebase generic by separating models, algos, hyperameters, and other utilities.
Later releases of this codebase will include additional algorithms and models. The next major release will involve porting all existing TensorFlow code to TensorFlow 2.0.