# Distilling Efficient Language-Specific Models for Cross-Lingual Transfer

arXiv: https://arxiv.org/abs/2306.01709 · HTML: https://mlsanity.com/paper/2306.01709/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 39 of its 60 references that are in the map.
- First posted on arXiv 2 Jun 2023.

## Builds on (most cited first, 30 of 39)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [Parameter-Efficient Transfer Learning for NLP](https://mlsanity.com/paper/1902.00751/) (2019)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference](https://mlsanity.com/paper/1704.05426/) (2017)
- [The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks](https://mlsanity.com/paper/1803.03635/) (2018)
- [MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers](https://mlsanity.com/paper/2002.10957/) (2020)
- [DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing](https://mlsanity.com/paper/2111.09543/) (2021)
- [TinyBERT: Distilling BERT for Natural Language Understanding](https://mlsanity.com/paper/1909.10351/) (2019)
- [The State and Fate of Linguistic Diversity and Inclusion in the NLP World](https://mlsanity.com/paper/2004.09095/) (2020)
- [How multilingual is Multilingual BERT?](https://mlsanity.com/paper/1906.01502/) (2019)
- [On the Cross-lingual Transferability of Monolingual Representations](https://mlsanity.com/paper/1910.11856/) (2019)
- [Learning multiple visual domains with residual adapters](https://mlsanity.com/paper/1705.08045/) (2017)
- [MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices](https://mlsanity.com/paper/2004.02984/) (2020)
- [MAD-X: An Adapter-Based Framework for Multi-Task Cross-Lingual Transfer](https://mlsanity.com/paper/2005.00052/) (2020)
- [Deep Biaffine Attention for Neural Dependency Parsing](https://mlsanity.com/paper/1611.01734/) (2016)
- [XCOPA: A Multilingual Dataset for Causal Commonsense Reasoning](https://mlsanity.com/paper/2005.00333/) (2020)
- [Beto, Bentz, Becas: The Surprising Cross-Lingual Effectiveness of BERT](https://mlsanity.com/paper/1904.09077/) (2019)
- [MiniLMv2: Multi-Head Self-Attention Relation Distillation for Compressing Pretrained Transformers](https://mlsanity.com/paper/2012.15828/) (2020)
- [Training Neural Networks with Fixed Sparse Masks](https://mlsanity.com/paper/2111.09839/) (2021)
- [MasakhaNER: Named Entity Recognition for African Languages](https://mlsanity.com/paper/2103.11811/) (2021)
- [Adapting Pre-trained Language Models to African Languages via Multilingual Adaptive Fine-Tuning](https://mlsanity.com/paper/2204.06487/) (2022)
- [Lifting the Curse of Multilinguality by Pre-training Modular Transformers](https://mlsanity.com/paper/2205.06266/) (2022)
- [Composable Sparse Fine-Tuning for Cross-Lingual Transfer](https://mlsanity.com/paper/2110.07560/) (2021)
- [UNKs Everywhere: Adapting Multilingual Language Models to New Scripts](https://mlsanity.com/paper/2012.15562/) (2020)
- [Rethinking embedding coupling in pre-trained language models](https://mlsanity.com/paper/2010.12821/) (2020)
- [WECHSEL: Effective initialization of subword embeddings for cross-lingual transfer of monolingual language models](https://mlsanity.com/paper/2112.06598/) (2021)
- [Modular Deep Learning](https://mlsanity.com/paper/2302.11529/) (2023)

## Built on by (most cited first, 6 of 6)

- [Probing the Emergence of Cross-lingual Alignment during LLM Training](https://mlsanity.com/paper/2406.13229/) (2024)
- [The Privileged Students: On the Value of Initialization in Multilingual Knowledge Distillation](https://mlsanity.com/paper/2406.16524/) (2024)
- [Extracting Small Translation Specialists from LLMs by Aggressively Pruning Experts](https://mlsanity.com/paper/2605.28042/) (2026)
- [Extracting General-use Transformers for Low-resource Languages via Knowledge Distillation](https://mlsanity.com/paper/2501.12660/) (2025)
- [DeFTX: Denoised Sparse Fine-Tuning for Zero-Shot Cross-Lingual Transfer](https://mlsanity.com/paper/2505.15090/) (2025)
- [On Multilingual Encoder Language Model Compression for Low-Resource Languages](https://mlsanity.com/paper/2505.16956/) (2025)

## Abstract (arXiv)

Massively multilingual Transformers (MMTs), such as mBERT and XLM-R, are widely used for cross-lingual transfer learning. While these are pretrained to represent hundreds of languages, end users of NLP systems are often interested only in individual languages. For such purposes, the MMTs' language coverage makes them unnecessarily expensive to deploy in terms of model size, inference time, energy, and hardware cost. We thus propose to extract compressed, language-specific models from MMTs which retain the capacity of the original MMTs for cross-lingual transfer. This is achieved by distilling the MMT bilingually, i.e., using data from only the source and target language of interest. Specifically, we use a two-phase distillation approach, termed BiStil: (i) the first phase distils a general bilingual model from the MMT, while (ii) the second, task-specific phase sparsely fine-tunes the bilingual "student" model using a task-tuned variant of the original MMT as its "teacher". We evaluate this distillation technique in zero-shot cross-lingual transfer across a number of standard cross-lingual benchmarks. The key results indicate that the distilled models exhibit minimal degradation in target language performance relative to the base MMT despite being significantly smaller and faster. Furthermore, we find that they outperform multilingually distilled models such as DistilmBERT and MiniLMv2 while having a very modest training budget in comparison, even on a per-language basis. We also show that bilingual models distilled from MMTs greatly outperform bilingual models trained from scratch. Our code and models are available at https://github.com/AlanAnsell/bistil.