# Dynamic Temperature Knowledge Distillation

Source: [Dynamic Temperature Knowledge Distillation](<https://www.sota2.com/research/paper/68fa72ee1e209131cea59562>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T02:37:39.153Z

Authors: Yukang Wei, Yu Bai

Published: 2024-04-19T08:40:52.000Z

## Abstract

Temperature plays a pivotal role in moderating label softness in the realm of knowledge distillation (KD). Traditional approaches often employ a static temperature throughout the KD process, which fails to address the nuanced complexities of samples with varying levels of difficulty and overlooks the distinct capabilities of different teacher-student pairings. This leads to a less-than-ideal transfer of knowledge. To improve the process of knowledge propagation, we proposed Dynamic Temperature Knowledge Distillation (DTKD) which introduces a dynamic, cooperative temperature control for both teacher and student models simultaneously within each training iterafion. In particular, we proposed "\\textbf{sharpness}" as a metric to quantify the smoothness of a model's output distribution. By minimizing the sharpness difference between the teacher and the student, we can derive sample-specific temperatures for them respectively. Extensive experiments on CIFAR-100 and ImageNet-2012 demonstrate that DTKD performs comparably to leading KD techniques, with added robustness in Target Class KD and None-target Class KD scenarios.The code is available at https://github.com/JinYu1998/DTKD.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2404.12711>)

[PDF](<https://arxiv.org/pdf/2404.12711.pdf>)
