# UniMax: Fairer and more Effective Language Sampling for Large-Scale Multilingual Pretraining

arXiv: https://arxiv.org/abs/2304.09151 · HTML: https://mlsanity.com/paper/2304.09151/

## Key facts

- Cited by 151 papers in the map (data through 14 Oct 2026).
- Builds on 43 of its 47 references that are in the map.
- First posted on arXiv 18 Apr 2023.

## Builds on (most cited first, 30 of 43)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [Training Compute-Optimal Large Language Models](https://mlsanity.com/paper/2203.15556/) (2022)
- [Extracting Training Data from Large Language Models](https://mlsanity.com/paper/2012.07805/) (2020)
- [SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing](https://mlsanity.com/paper/1808.06226/) (2018)
- [Language Models as Knowledge Bases?](https://mlsanity.com/paper/1909.01066/) (2019)
- [mT5: A massively multilingual pre-trained text-to-text transformer](https://mlsanity.com/paper/2010.11934/) (2020)
- [Cross-lingual Language Model Pretraining](https://mlsanity.com/paper/1901.07291/) (2019)
- [XNLI: Evaluating Cross-lingual Sentence Representations](https://mlsanity.com/paper/1809.05053/) (2018)
- [Adafactor: Adaptive Learning Rates with Sublinear Memory Cost](https://mlsanity.com/paper/1804.04235/) (2018)
- [XTREME: A Massively Multilingual Multi-task Benchmark for Evaluating Cross-lingual Generalization](https://mlsanity.com/paper/2003.11080/) (2020)
- [How Much Knowledge Can You Pack Into the Parameters of a Language Model?](https://mlsanity.com/paper/2002.08910/) (2020)
- [Deduplicating Training Data Makes Language Models Better](https://mlsanity.com/paper/2107.06499/) (2021)
- [On the Cross-lingual Transferability of Monolingual Representations](https://mlsanity.com/paper/1910.11856/) (2019)
- [KILT: a Benchmark for Knowledge Intensive Language Tasks](https://mlsanity.com/paper/2009.02252/) (2020)
- [Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model](https://mlsanity.com/paper/2201.11990/) (2022)
- [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://mlsanity.com/paper/2105.13626/) (2021)
- [TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages](https://mlsanity.com/paper/2003.05002/) (2020)
- [MLQA: Evaluating Cross-lingual Extractive Question Answering](https://mlsanity.com/paper/1910.07475/) (2019)
- [PAWS-X: A Cross-lingual Adversarial Dataset for Paraphrase Identification](https://mlsanity.com/paper/1908.11828/) (2019)
- [Massively Multilingual Neural Machine Translation in the Wild: Findings and Challenges](https://mlsanity.com/paper/1907.05019/) (2019)
- [Quality at a Glance: An Audit of Web-Crawled Multilingual Datasets](https://mlsanity.com/paper/2103.12028/) (2021)
- [Scaling Up Models and Data with textttt5x and textttseqio](https://mlsanity.com/paper/2203.17189/) (2022)
- [Gradient Vaccine: Investigating and Improving Multi-task Optimization in Massively Multilingual Models](https://mlsanity.com/paper/2010.05874/) (2020)
- [Scaling Laws and Interpretability of Learning from Repeated Data](https://mlsanity.com/paper/2205.10487/) (2022)
- [Lifting the Curse of Multilinguality by Pre-training Modular Transformers](https://mlsanity.com/paper/2205.06266/) (2022)
- [Systematic Inequalities in Language Technology Performance across the World's Languages](https://mlsanity.com/paper/2110.06733/) (2021)

## Built on by (most cited first, 30 of 151)

- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [OLMo: Accelerating the Science of Language Models](https://mlsanity.com/paper/2402.00838/) (2024)
- [Aya Model: An Instruction Finetuned Open-Access Multilingual Language Model](https://mlsanity.com/paper/2402.07827/) (2024)
- [SkyReels-V2: Infinite-length Film Generative Model](https://mlsanity.com/paper/2504.13074/) (2025)
- [OpenMoE: An Early Effort on Open Mixture-of-Experts Language Models](https://mlsanity.com/paper/2402.01739/) (2024)
- [A Pretrainer's Guide to Training Data: Measuring the Effects of Data Age, Domain Coverage, Quality, & Toxicity](https://mlsanity.com/paper/2305.13169/) (2023)
- [MADLAD-400: A Multilingual And Document-Level Large Audited Dataset](https://mlsanity.com/paper/2309.04662/) (2023)
- [VSA: Faster Video Diffusion with Trainable Sparse Attention](https://mlsanity.com/paper/2505.13389/) (2025)
- [RELIC: Interactive Video World Model with Long-Horizon Memory](https://mlsanity.com/paper/2512.04040/) (2025)
- [LongCat-Video Technical Report](https://mlsanity.com/paper/2510.22200/) (2025)
- [Yume: An Interactive World Generation Model](https://mlsanity.com/paper/2507.17744/) (2025)
- [Uni3C: Unifying Precisely 3D-Enhanced Camera and Human Motion Controls for Video Generation](https://mlsanity.com/paper/2504.14899/) (2025)
- [Large Video Planner Enables Generalizable Robot Control](https://mlsanity.com/paper/2512.15840/) (2025)
- [WorldGym: World Model as An Environment for Policy Evaluation](https://mlsanity.com/paper/2506.00613/) (2025)
- [UniVerse-1: Unified Audio-Video Generation via Stitching of Experts](https://mlsanity.com/paper/2509.06155/) (2025)
- [ACE-Step: A Step Towards Music Generation Foundation Model](https://mlsanity.com/paper/2506.00045/) (2025)
- [Paloma: A Benchmark for Evaluating Language Model Fit](https://mlsanity.com/paper/2312.10523/) (2023)
- [Scaling Laws for Optimal Data Mixtures](https://mlsanity.com/paper/2507.09404/) (2025)
- [HAE-RAE Bench: Evaluation of Korean Knowledge in Language Models](https://mlsanity.com/paper/2309.02706/) (2023)
- [HuMo: Human-Centric Video Generation via Collaborative Multi-Modal Conditioning](https://mlsanity.com/paper/2509.08519/) (2025)
- [NeoVerse: Enhancing 4D World Model with in-the-wild Monocular Videos](https://mlsanity.com/paper/2601.00393/) (2026)
- [Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance](https://mlsanity.com/paper/2512.08765/) (2025)
- [MYTE: Morphology-Driven Byte Encoding for Better and Fairer Multilingual Language Modeling](https://mlsanity.com/paper/2403.10691/) (2024)
- [Breaking the Curse of Multilinguality with Cross-lingual Expert Language Models](https://mlsanity.com/paper/2401.10440/) (2024)
- [FantasyWorld: Geometry-Consistent World Modeling via Unified Video and 3D Prediction](https://mlsanity.com/paper/2509.21657/) (2025)
- [UniAVGen: Unified Audio and Video Generation with Asymmetric Cross-Modal Interactions](https://mlsanity.com/paper/2511.03334/) (2025)
- [PAN: A World Model for General, Actionable, and Long-Horizon World Simulation](https://mlsanity.com/paper/2511.09057/) (2025)
- [MoDoMoDo: Multi-Domain Data Mixtures for Multimodal LLM Reinforcement Learning](https://mlsanity.com/paper/2505.24871/) (2025)
- [VerseCrafter: Dynamic Realistic Video World Model with 4D Geometric Control](https://mlsanity.com/paper/2601.05138/) (2026)
- [ShotStream: Streaming Multi-Shot Video Generation for Interactive Storytelling](https://mlsanity.com/paper/2603.25746/) (2026)

## Abstract (arXiv)

Pretrained multilingual large language models have typically used heuristic temperature-based sampling to balance between different languages. However previous work has not systematically evaluated the efficacy of different pretraining language distributions across model scales. In this paper, we propose a new sampling method, UniMax, that delivers more uniform coverage of head languages while mitigating overfitting on tail languages by explicitly capping the number of repeats over each language's corpus. We perform an extensive series of ablations testing a range of sampling strategies on a suite of multilingual benchmarks, while varying model scale. We find that UniMax outperforms standard temperature-based sampling, and the benefits persist as scale increases. As part of our contribution, we release: (i) an improved and refreshed mC4 multilingual corpus consisting of 29 trillion characters across 107 languages, and (ii) a suite of pretrained umT5 model checkpoints trained with UniMax sampling.