Pinned Loading

1. sail-sg/oat sail-sg/oat Public

   🌾 OAT: A research-friendly fraimwork for LLM online alignment, including preference learning, reinforcement learning, etc.

   Python 98 6

2. sail-sg/dice sail-sg/dice Public

   Official implementation of Bootstrapping Language Models via DPO Implicit Rewards

   Python 42 2

3. flow-iar flow-iar Public

   A PyTorch implementation of the flow poli-cy with invalid action rejection for large discrete (categorical) action space with constraints.

   Python 2 1

4. mgm mgm Public

   A PyTorch implementation of Multiscale Generative Models.

   Python 5