Reinforced-lib 1.1.4

• Add experimental masked MAB agent.

Reinforced-lib 1.1.3

• Add epsilon decay in e-greedy MAB.

Reinforced-lib 1.1.2

• Update dependencies.
• Fix error with action space size of one.

Reinforced-lib 1.1.1

Improvements:

• Update documentation.
• Add reference to the SoftwareX paper.

Fix:

• Normal Thompson sampling allows the lam parameter to be zero.
• Bernoulli Thompson sampling is stationary by default.
• Update the default value of the decay parameter the in ra-sim example.

Reinforced-lib 1.1.0

Major API changes:

• Migrate from haiku (deprecated) to flax as the base naural network library.
• Update agent names to match literature:
  • QLearning (deep Q-learning) -> DQN,
  • DQN (deep double Q-learning) -> DDQN.
• Move particle filter from agents to utils.
• New behavior of loggers - all declared loggers get values from all sources.

New functionalities:

• Add Weights & Biases logger.

Other important changes:

• Fix updates with empty replay buffer.
• Fix logging of arrays to TensorBoard.
• Minor improvements in documentation.
• Rewrite Gymnasium integration example in documentation.
• Improve the CCOD example to better reflect the original implementation.

Reinforced-lib 1.0.4

Improvements:

• Update documentation.
• Enable the use of 64-bit JAX.

Reinforced-lib 1.0.3

New functionalities:

• Add the normal-gamma Thompson sampling agent.
• Add the log-normal Thompson sampling agent.

Reinforced-lib 1.0.2

Fix:

• Make it easier to import the BasicMab extension.

Reinforced-lib 1.0.1

Important changes:

• Move to pyproject.toml configuration file.
• Add basic extension for MABs.
• Update dependencies.
• Fix bug modifying user values passed to library functions.
• Fix agents behavior with multiple optimal actions - now agents draw one of the optimal actions instead of selecting the first one.

Reinforced-lib 1.0.0

Major API changes:

• Added a support for deep reinforcement learning agents.
• Relaxation of the requirements for the implementation of custom agents.
• Major changes in the logging module (e.g., custom logging, synchronization).
• Removed ability of the sample method to change state.
• Introduced an inference only mode.

New functionalities:

• Added new deep learning agents: deep Q-learning, deep expected SARSA, DQN, DDPG.
• Added the Exp3 algorithm.
• Added the Gymnasium extension.
• Added the TensorBoard logger.
• Added an easy export to TensorFlow Lite.
• Added an automatic checkpointing.

Other important changes:

• Upgraded the library to Python 3.9.
• Updated and polished the documentation.
• Added several new examples.
• Moved Wi-Fi specific classes to examples.
• Fixed known bugs.