”d4rl“ 的搜索结果

A collection of reference environments for offline reinforcement learning

High-quality single-file implementations of SOTA Offline and Offline-to-Online RL algorithms: AWAC, BC, CQL, DT, EDAC, IQL, SAC-N, TD3+BC, LB-SAC, SPOT, Cal-QL, ReBRAC

Datasets for data-driven deep reinforcement learning with Atari (wrapper for datasets released by Google)

Datasets for data-driven deep reinforcement learning with PyBullet environments

Challenges and Opportunities in Offline Reinforcement Learning from Visual Observations

A collection of robotics simulation environments for reinforcement learning

Scripts to recreate the D4RL datasets with Minari

Single-file SAC-N implementation on jax with flax and equinox. 10x faster than pytorch