Forem: NoteDance

Multiprocessed Experience Replay Pool for Reinforcement Learning

NoteDance — Mon, 02 Jun 2025 12:22:49 +0000

The Pool class is designed for efficient, parallelized data collection from multiple environments, particularly useful in reinforcement learning settings. It leverages Python's multiprocessing module to manage shared memory and execute environment interactions concurrently.

https://github.com/NoteDance/Pool

TensorFlow implementation for optimizers

NoteDance — Thu, 08 May 2025 13:42:45 +0000

The NoteDance/optimizers repository is a comprehensive collection of optimization algorithms for TensorFlow/Keras. It extends the standard TensorFlow optimizer offerings with a wide range of state-of-the-art optimization techniques, enabling users to select the most appropriate optimizer for their specific machine learning tasks.

https://github.com/NoteDance/optimizers

A machine learning library

NoteDance — Thu, 08 May 2025 13:40:38 +0000

Note is a deep learning framework built on top of TensorFlow that provides a comprehensive set of tools for constructing and training neural networks.

https://github.com/NoteDance/Note