Releases: keras-rl/keras-rl
Releases · keras-rl/keras-rl
v0.4.2
v0.4.1
- Add support for current version of Keras and Gym
v0.4.0
- Removes legacy code and now requires
Keras >= 2.0.7
(https://github.com/matthiasplappert/keras-rl/pull/157) - Adds two new policies (https://github.com/matthiasplappert/keras-rl/pull/156, https://github.com/matthiasplappert/keras-rl/pull/122)
- Fixes a few problems with replay memory sampling being incorrect for certain edge cases (https://github.com/matthiasplappert/keras-rl/pull/138)
- Fixes a problem with dropout in actor model for DDPG (https://github.com/matthiasplappert/keras-rl/pull/150)
v0.3.1
- Require Keras < 2.0.7 due to non-trivial and breaking API change in optimizer code
v0.3.0
- Full compatibility with the recently released Keras 2
- New
SarsaAgent
implementation - New
train_policy
onDQNAgent
andSarsaAgent
to allow for different policies at training and test time - Minor clean-ups
v0.2.2
- Fix version mix-up on release, otherwise identical to version
v0.2.1
v0.2.1
- Add Dueling DQN
- Add support for TensorFlow 1.0
- Fix problem with default policy if multiple DQN agents are used simultaneously
v0.2.0rc1
Various improvements for DQN, the replay memory, and CEM (#31) - Fix various problems with sequential memory - Fix various problems with CEMAgent - Remove duplicate forward pass from DQN agent - Remove performance bottleneck from TrainIntervalLogger - Add tests for core and memory classes - Add integration test for CEM