# Learning Light-Weight Translation Models from Deep Transformer

arXiv: https://arxiv.org/abs/2012.13866 · HTML: https://mlsanity.com/paper/2012.13866/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Dec 2020.

## Built on by (most cited first, 22 of 22)

- [Stacked Acoustic-and-Textual Encoding: Integrating the Pre-trained Models into Speech Translation Encoders](https://mlsanity.com/paper/2105.05752/) (2021)
- [The Low-Resource Double Bind: An Empirical Study of Pruning for Low-Resource Machine Translation](https://mlsanity.com/paper/2110.03036/) (2021)
- [Selective Knowledge Distillation for Neural Machine Translation](https://mlsanity.com/paper/2105.12967/) (2021)
- [ODE Transformer: An Ordinary Differential Equation-Inspired Model for Sequence Generation](https://mlsanity.com/paper/2203.09176/) (2022)
- [Metadata Archaeology: Unearthing Data Subsets by Leveraging Training Dynamics](https://mlsanity.com/paper/2209.10015/) (2022)
- [Learning Occlusion-Robust Vision Transformers for Real-Time UAV Tracking](https://mlsanity.com/paper/2504.09228/) (2025)
- [LightSeq2: Accelerated Training for Transformer-based Models on GPUs](https://mlsanity.com/paper/2110.05722/) (2021)
- [SMaLL-100: Introducing Shallow Multilingual Machine Translation Model for Low-Resource Languages](https://mlsanity.com/paper/2210.11621/) (2022)
- [Intriguing Properties of Compression on Multilingual Models](https://mlsanity.com/paper/2211.02738/) (2022)
- [Improved Knowledge Distillation for Pre-trained Language Models via Knowledge Selection](https://mlsanity.com/paper/2302.00444/) (2023)
- [What Do Compressed Multilingual Machine Translation Models Forget?](https://mlsanity.com/paper/2205.10828/) (2022)
- [Learning Multiscale Transformer Models for Sequence Generation](https://mlsanity.com/paper/2206.09337/) (2022)
- [An Effective Non-Autoregressive Model for Spoken Language Understanding](https://mlsanity.com/paper/2108.07005/) (2021)
- [The NiuTrans Machine Translation Systems for WMT21](https://mlsanity.com/paper/2109.10485/) (2021)
- [Predictor-Corrector Enhanced Transformers with Exponential Moving Average Coefficient Learning](https://mlsanity.com/paper/2411.03042/) (2024)
- [The NiuTrans System for the WMT21 Efficiency Task](https://mlsanity.com/paper/2109.08003/) (2021)
- [Capture Salient Historical Information: A Fast and Accurate Non-Autoregressive Model for Multi-turn Spoken Language Understanding](https://mlsanity.com/paper/2206.12209/) (2022)
- [The NiuTrans End-to-End Speech Translation System for IWSLT 2021 Offline Task](https://mlsanity.com/paper/2107.02444/) (2021)
- [EIT: Enhanced Interactive Transformer](https://mlsanity.com/paper/2212.10197/) (2022)
- [Optimizing Deep Transformers for Chinese-Thai Low-Resource Translation](https://mlsanity.com/paper/2212.12662/) (2022)
- [Multi-Path Transformer is Better: A Case Study on Neural Machine Translation](https://mlsanity.com/paper/2305.05948/) (2023)
- [Unraveling Key Factors of Knowledge Distillation](https://mlsanity.com/paper/2312.08585/) (2023)

## Abstract (arXiv)

Recently, deep models have shown tremendous improvements in neural machine translation (NMT). However, systems of this kind are computationally expensive and memory intensive. In this paper, we take a natural step towards learning strong but light-weight NMT systems. We proposed a novel group-permutation based knowledge distillation approach to compressing the deep Transformer model into a shallow model. The experimental results on several benchmarks validate the effectiveness of our method. Our compressed model is 8X shallower than the deep model, with almost no loss in BLEU. To further enhance the teacher model, we present a Skipping Sub-Layer method to randomly omit sub-layers to introduce perturbation into training, which achieves a BLEU score of 30.63 on English-German newstest2014. The code is publicly available at https://github.com/libeineu/GPKD.