Skip to content

v0.0.2 (Beta)

v0.0.21 (Beta) - 2025-02-18

💼 Other

  • (torch) Refactored torch device into cpu and cuda groups.

🧪 Testing

  • (fix) Fixed broken gym and buffer tests.

⚙️ Miscellaneous Tasks

  • (workflow) Updated workflows to use deps from docker containers.

v0.0.2 (Beta) - 2025-02-18

🚀 Features

  • (policy) Added CNN and MLP backbone policies.
  • (agent) Added agent base classes.
  • (lnn) Add Liquid Neural Network model.
  • (gym) Added Gymnasium and generic utility methods.
  • (buffer) Added Replay and Rollout buffer logic.
  • (ddpg) Added a Liquid variant of DDPG.
  • (cnn) Added basic CNN backbone.

🐛 Bug Fixes

  • (lnn) Fixed x dimension bug in forward method.
  • (lnn) Updated Cell implementation to fix training bug.
  • (lnn) Fixed LiquidNCP training bug.
  • (models) Fixed PyTorch hook and device transfer compatibility.

💼 Other

  • (poetry) Upgraded Poetry to v2.0.1 and deps to latest.
  • (poetry) Updated all packages to latest versions.

🚜 Refactor

  • (models) Migrated framework to Deep-RL methods.
  • (env) Overhauled GymEnvHandler to improve functionality.
  • (controller) Major overhaul of RLController for simplicity.
  • (old) Removed old ValueFunctions and EpsilonPolicy.
  • (lnn) Updated project code and docstrings for clarity.
  • (alpha) Removed v0.0.1 implementation for fresh start.
  • (lnn) Added device parameter for CUDA.
  • (buffer) Moved to_tensor as utility method and improved.
  • (ddpg) Updated DDPG for modularity and accuracy.
  • (lnn) Removed timespans logic to simplify.
  • (ddpg) Removed action_bounds to simplify.
  • (utils) Added and updated utility methods.
  • (alpha) Removed remaining old alpha logic. and files
  • (wiring) Moved Wiring -> models directory for clarity.
  • (gym) Utility method improvements.
  • (wiring) Updated wiring methods.

📚 Documentation

  • (lnn) Updated forward() docstrings for clarity.
  • (core) Added core documentation pages and content with assets.
  • (api) Added docs for the LNN model API reference.
  • (models) Improved model docstrings format and content.
  • (history) Added brief history and motivation of Velora.
  • (api) Added API reference documentation.

🧪 Testing

  • (missing) Added buffer and gym unit tests for coverage.

⚙️ Miscellaneous Tasks

  • (prep) Added scripts and packages for CI prep.
  • (readme) Updated README
  • (readme) Updated README