RLkit is a reinforcement learning framework and collection of RL algorithms implemented in PyTorch. It includes implementations of major RL algorithms such as Soft Actor Critic (SAC), Twin Delayed Deep Deterministic Policy Gradient (TD3), Hindsight Experience Replay (HER), Deep Q-Network (DQN), Advantage Weighted Actor Critic (AWAC), and others. The repository provides example scripts and documentation for training agents using these algorithms.