Sadia Khair Rodela

When Does Knowledge Distillation Hurt? Reliability-Aware Distillation for Low-Resource Language Summarization

Dipto Sumit, Ankan Kumar Roy Srizon, Sadia Khair Rodela, Atia Haque Asha, Mourchona Afrin, Niloy Farhan, et al.

Knowledge distillation (KD) is a standard approach for compressing sequence-to-sequence models, but its per-sample effects are rarely examined. On the BanSum Bangla summarization benchmark, we find that standard KD improves ROUGE-L by only +0.0003 over a cross-entropy baseline, a…Read more