D³-MOPD: Dynamic Scheduling for Multi-Teacher Distillation

Zechen Sun · hf · 2026-08-27

D³-MOPD dynamically adjusts domain sampling ratios during multi-teacher distillation by monitoring per-domain reverse-KL trajectories. It improves convergence efficiency and closes most of the student-to-teacher performance gap.

Original post →

More from Research

Research channel →