NEWS.md

ReinforcementLearning.jl Release Notes

ReinforcementLearning.jl@v0.10.0

ReinforcementLearningExperiments.jl

v0.1.4

• Support device_rng in SAC #606

v0.1.3

• Test experiments on GPU by default #549

v0.1.2

• Added an experiment for DQN training on discrete PendulumEnv (#537)

ReinforcementLearningEnvironments.jl

v0.6.12

• Bugfix bug with is_discrete_space #566

v0.6.11

• Bugfix of CartPoleEnv with keyword arguments

v0.6.10

• Bugfix of CartPoleEnv with Float32

v0.6.9

• Added a continuous option for CartPoleEnv #543.

v0.6.8

• Support action_space(::TicTacToeEnv, player).

v0.6.7

• Fixed bugs in plotting MountainCarEnv (#537)
• Implemented plotting for PendulumEnv (#537)

v0.6.6

• Bugfix with ZeroTo #534

ReinforcementLearningCore.jl

v0.8.11

• When sending a CircularArrayBuffer to GPU devices, convert CircularArrayBuffer into CuArray instead of the adapted CircularArrayBuffer of CuArray. #606

v0.8.10

• Update dependency of CircularArrayBuffers to v0.1.9. #602
• Add CovGaussianNetwork. #597

v0.8.8

• Fix warning about vararg.data in Julia@v1.7 #560

v0.8.7

• Make GaussianNetwork differentiable. #549

v0.8.6

• Fixed a bug [1] with the DoOnExit hook (#537)
• Added some convenience hooks for rendering rollout episodes (#537)

v0.8.5

• Fixed the method overwritten warning of device from CUDA.jl.

ReinforcementLearningZoo.jl

v0.5.11

• Fix multi-dimension action space in TD3. #624

v0.5.10

• Support device_rng in SAC #606

v0.5.7

• Fix warning about vararg.data in Julia@v1.7 #560

v0.5.6

• Make BC GPU compatible #553

v0.5.5

• Make most algorithms GPU compatible #549

v0.5.4

• Support length method for VectorWSARTTrajectory.

v0.5.3

• Revert part of the unexpected change of PPO in the last PR.

v0.5.2

• Fixed the bug with MaskedPPOTrajectory reported here

v0.5.0

• Update the complete SAC implementation and modify some details based on the original paper. #365
• Add some extra keyword parameters for BehaviorCloningPolicy to use it online. #390

ReinforcementLearningDatasets.jl

v0.1.0

• Add functionality for fetching d4rl datasets as an iterable DataSet. Credits: https://arxiv.org/abs/2004.07219
• This supports d4rl and d4rl-pybullet and Google Research DQN atari datasets.
• Uses DataDeps for data dependency management.
• This package also supports RL Unplugged Datasets.
• Support for google-research/deep_ope added.

ReinforcementLearning.jl@v0.9.0

ReinforcementLearningBase.jl

v0.9.6

• Implement Base.:(==) for Space. #428

v0.9.5

• Add default Base.:(==) and Base.hash method for AbstractEnv. #348

ReinforcementLearningCore.jl

v0.8.3

• Add extra two optional keyword arguments (min_σ and max_σ) in GaussianNetwork to clip the output of logσ. #428

v0.8.2

• Add GaussianNetwork and DuelingNetwork into ReinforcementLearningCore.jl as general components. #370
• Export WeightedSoftmaxExplorer. #382

v0.8.1

• Minor bug & typo fixes

v0.8.0

• Removed ResizeImage preprocessor to reduce the dependency of ImageTransformations.
• Show unicode plot at the end of an experiment in the TotalRewardPerEpisode hook.

ReinforcementLearningZoo.jl

v0.4.1

• Make keyword argument n_actions in TabularPolicy optional. #300

v0.4.0

• Moved all the experiments into a new package ReinforcementLearningExperiments.jl. The related dependencies are also removed (BSON.jl, StableRNGs.jl, TensorBoardLogger.jl).

ReinforcementLearningEnvironments.jl

v0.6.4-dev

• Add GraphShortestPathEnv. #445

v0.6.3

• Add StockTradingEnv from the paper Deep Reinforcement Learning for Automated Stock Trading: An Ensemble Strategy. This environment is a good testbed for multi-continuous action space algorithms. #428

v0.6.2

• Add SequentialEnv environment wrapper to turn a simultaneous environment into a sequential one.

v0.6.1

• Drop GR in RLEnvs and lazily load ploting functions.#309, #310

v0.6.0

• Set AcrobotEnv into lazy loading to reduce the dependency of OrdinaryDiffEq.

ReinforcementLearningExperiments.jl

v0.1.0