Pinned Loading

1. RL RL Public

   Forked from NVIDIA-NeMo/RL

   Scalable toolkit for efficient model reinforcement

   Python

2. SpecForge SpecForge Public

   Forked from sgl-project/SpecForge

   Train speculative decoding models effortlessly and port them smoothly to SGLang serving.

   Python

3. RL-MTP RL-MTP Public

   Official repository of Joint Training of Multi-Token Prediction in Reinforcement Learning via Optimal Coefficient Calibration

   Python 5

4. CSpD CSpD Public

   The official repo of continuous speculative decoding

   Python 36 1