Huawei Cloud Developer Alliance

Sep 5, 2026 · Artificial Intelligence

THD Packing & MTP Speculative Decoding Tame 64K Code Agent RL on Ascend NPUs

AReaL-Ascend v1.0.5 enables stable 64K-context Code Agent RL for Qwen3.6-27B on Ascend NPUs, using THD Sequence Packing to eliminate padding waste and MTP speculative decoding to accelerate rollout, achieving reward improvement from 0.56 to 0.94 over 120 steps with train-rollout consistency near 1.0.