Game AI & Simulations

Toolkit for training intelligent agents in games and simulations using deep reinforcement learning. Enables NPC behavior control, automated testing, and game design evaluation. Apache 2.0 licensed.

A training framework for Stable Baselines3 reinforcement learning agents with hyperparameter optimization, pre-trained agents, and extensive benchmark environments. MIT licensed.

Modular reinforcement learning library implemented in PyTorch, JAX, and NVIDIA Warp with support for Gymnasium, NVIDIA Isaac Lab, MuJoCo Playground, and other environments. MIT licensed.

Scalable and memory-optimized training of diffusion models from Hugging Face. Supports LoRA and full fine-tuning for video and image generation models. Apache 2.0 licensed.

Collection of environments and algorithms for research in general reinforcement learning and search/planning in games from Google DeepMind. Apache 2.0 licensed.