FASTRL
// [ASPLOS'26] Taming the Long-Tail: Efficient Reasoning RL Training with Adaptive Drafter
Fastrl
[ASPLOS'26] Taming the Long-Tail: Efficient Reasoning RL Training with Adaptive Drafter
25EmergingSlowing
What it does
[ASPLOS'26] Taming the Long-Tail: Efficient Reasoning RL Training with Adaptive Drafter
Getting Started
git
git clone https://github.com/mit-han-lab/fastrl