GrowMTP: Can RL Grow Its Own Draft Head?
Reinforcement learning (RL) post-training drives the frontier capabilities of large language models, with its wall-clock dominated by autoregressive rollout generation. Speculative decoding is an established remedy for this bottleneck, but existing draft heads must be pretrained or warmed up before RL, introducing subs...
Ming-Hua He, Ling-Zhe Zhang, Yuan Liu et al.
· 0 citations