AReaL
A reinforcement learning training framework designed for the post-training stage of large language models. It supports complex sampling, reward computation, and distributed training coordination, and has been incorporated into the official PyTorch ecosystem.
Timeline (last 90 days)
The PyTorch ecosystem working group announced the addition of 10 new projects to the PyTorch Ecosystem Landscape: Perforated, AReaL, TorchJD, RLinf, Miles, SMG, FiftyOne, TokenSpeed, VisualTorch, and TorchSurv.
AReaL and RLinf both focus on the reinforcement learning domain and are RL training frameworks for large models.
All Facts (2)
AReaL and RLinf both focus on the reinforcement learning domain and are RL training frameworks for large models.
50%UnverifiedThe PyTorch ecosystem working group announced the addition of 10 new projects to the PyTorch Ecosystem Landscape: Perforated, AReaL, TorchJD, RLinf, Miles, SMG, FiftyOne, TokenSpeed, VisualTorch, and TorchSurv.
50%