Async RL post-training gets an offline scheduler: ASP cuts iteration time by up to 76.8% against MindSpeed-RL
Once LLM reinforcement-learning post-training splits generation from training across machines, multi-instance rollout still wastes time on uneven output lengths and on length-sorted dispatch that makes the training queue fill and then drain. ASP runs one offline preprocessing pass before training: it predicts each prompt's output length, uses dynamic programming to build decoding minibatches of similar length, then reorders them into batch groups with balanced predicted load. On Ascend 910B3, complete ASP reduces mean iteration time by up to 76.8% relative to MindSpeed-RL and by up to 62.5% relative to verl.