paper-with-me

Papers

Student-friendly Knowledge Distillation

2023-05-18 · Mengyang Yuan, Bo Lang, Fengnan Quan

In knowledge distillation, the knowledge from the teacher model is often too complex for the student model to thoroughly process. However, good teachers in real life always simplify complex material before teaching it to students. Inspired by this fact, we propose student-friendly knowledge distillation (SKD) to simplify teacher output into new knowledge representations, which makes the learning of the student model easier and more effective. SKD contains a softening processing and a learning simplifier. First, the softening processing uses the temperature hyperparameter to soften the output logits of the teacher model, which simplifies the output to some extent and makes it easier for the learning simplifier to process. The learning simplifier utilizes the attention mechanism to further simplify the knowledge of the teacher model and is jointly trained with the student model using the distillation loss, which means that the process of simplification is correlated with the training objective of the student model and ensures that the simplified new teacher knowledge representation is more suitable for the specific student model. Furthermore, since SKD does not change the form of the distillation loss, it can be easily combined with other distillation methods that are based on the logits or features of intermediate layers to enhance its effectiveness. Therefore, SKD has wide applicability. The experimental results on the CIFAR-100 and ImageNet datasets show that our method achieves state-of-the-art performance while maintaining high training efficiency.

📄 PDF Abstract BibTeX arXiv:2305.10893

Code (0)

등록된 구현이 없습니다.

Tasks

Knowledge Distillation

Methods 이 논문이 사용한 방법론

Knowledge Distillation A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions.…

Similar Papers 제목 키워드 기반

Learning Student-Friendly Teacher Networks for Knowledge Distillation

2021-02-12 · NeurIPS 2021 12 · Dae Young Park, Moon-Hyun Cha, Changwook Jeong, Dae Sin Kim 외

We propose a novel knowledge distillation approach to facilitate the transfer of dark knowledge from a teacher to a student. Contrary to most of the existing methods that rely on effective training of student models give…

Knowledge DistillationTransfer Learning

Parameter-Efficient and Student-Friendly Knowledge Distillation

2022-05-28 · Jun Rao, Xv Meng, Liang Ding, Shuhan Qi 외

Knowledge distillation (KD) has been extensively employed to transfer the knowledge from a large teacher model to the smaller students, where the parameters of the teacher are fixed (or partially) during training. Recent…

Knowledge DistillationTransfer Learning

PromptKD: Distilling Student-Friendly Knowledge for Generative Language Models via Prompt Tuning

2024-02-20 · Gyeongman Kim, Doohyuk Jang, Eunho Yang

Recent advancements in large language models (LLMs) have raised concerns about inference costs, increasing the need for research into model compression. While knowledge distillation (KD) is a prominent method for this, r…

Instruction FollowingKnowledge DistillationModel Compression

Pixel-Wise Contrastive Distillation

2022-11-01 · ICCV 2023 1 · Junqiang Huang, Zichao Guo

We present a simple but effective pixel-level self-supervised distillation framework friendly to dense prediction tasks. Our method, called Pixel-Wise Contrastive Distillation (PCD), distills knowledge by attracting the …

Prediction

One Student Knows All Experts Know: From Sparse to Dense

2022-01-26 · Fuzhao Xue, Xiaoxin He, Xiaozhe Ren, Yuxuan Lou 외

Human education system trains one student by multiple experts. Mixture-of-experts (MoE) is a powerful sparse architecture including multiple experts. However, sparse MoE model is easy to overfit, hard to deploy, and not …

AllKnowledge DistillationMixture-of-Experts