News
EveryLLM

English first

Trained mostly on English. 206 models on EveryLLM so far.

English-first models were trained mostly on English text, because English dominates the web, books, code comments and the research papers that set the agenda. Most early language models, from BERT and GPT-2 to Chinchilla and the first Llama, fall in this group, and many small and research models still do. They often handle other languages, but with clearly lower quality.

Where the data came from

BERT (2018) learned from English Wikipedia and a large set of unpublished books. GPT-2 (2019) used WebText, about 40 GB of text from 8 million web pages linked from Reddit. EleutherAI's Pile (2020), an 825 GiB collection weighted towards English academic papers, books and code, trained GPT-J and Pythia. Common Crawl, a free archive of the web gathered since 2008, supplies the bulk of most training sets.

Why it still matters

English-first training shapes what models know and how they speak: they may default to American spelling, law and culture, and do worse on questions about other places. Many benchmarks exist only in English. Labs now add far more non-English data, but English remains the largest share for most general models, and research models built to study training, such as OLMo, often stay English to keep the experiment clean.

Famous english first models

Still to be written: OLMo (language model).

Further reading

XLNet WorldXLNet Permutation-based autoregressive transformer model created by Google AIResearch · Open weights · 2019 GPT-J WorldGPT-J A 6-billion-parameter open English text generator from EleutherAI, released in 2021.General · Open weights · 2021 BERT WorldBERT An encoder-only transformer that reads text in both directions at once.Research · Open weights · 2018 GPT-1 WorldGPT-1 OpenAI's first GPT model, a 117M-parameter decoder-only transformer from 2018.Research · Open weights · 2018 GPT-2 🇺🇸 United StatesGPT-2 OpenAI's 2019 text model that predicts the next word in a sequenceGeneral · Open weights · 2019 Talkie 🇺🇸 United StatesTalkie A 13B small language model trained only on pre-1931 public domain English textSmall · Open weights · 2026 Pythia 🇺🇸 United StatesPythia A suite of 16 models for interpretability research.Research · Open weights · 2023 INTELLECT-1 WorldINTELLECT-1 A 10 billion parameter model trained by a global community.General · Open weights · 2024 OpenELM 🇺🇸 United StatesOpenELM Apple's small open language models with a full training and inference releaseSmall · Open weights · 2024 Flan-T5 🇺🇸 United StatesFlan-T5 Instruction-finetuned T5 checkpoints released by Google for zero-shot and few-shot workGeneral · Open weights · 2022 DialoGPT 🇺🇸 United StatesDialoGPT A GPT-style model trained on Reddit threads to write chat repliesGeneral · Open weights · 2019 DeBERTa 🇺🇸 United StatesDeBERTa A BERT-style encoder with disentangled attention and an enhanced mask decoder.Encoder · Open weights · 2020 IndicBART 🇮🇳 IndiaIndicBART Sequence-to-sequence model for Indic languagesRegional · Open weights · 2021 BioGPT 🇺🇸 United StatesBioGPT A generative Transformer trained on biomedical writing for text and mining tasksDomain · Open weights · 2022 Ganga-1B 🇮🇳 IndiaGanga-1B Hindi-focused autoregressive language model from IIT GandhinagarRegional · Open weights Cerebras-GPT 🇺🇸 United StatesCerebras-GPT Open compute-optimal language models from Cerebras SystemsGeneral · Open weights · 2023 GPT-NeoX-20B 🇺🇸 United StatesGPT-NeoX-20B Open 20 billion parameter autoregressive language model by EleutherAIGeneral · Open weights · 2022 Phi-2 🇺🇸 United StatesPhi-2 A 2.7 billion parameter Microsoft research model for QA, chat and codeResearch · Open weights Tülu 3 🇺🇸 United StatesTülu 3 Open post-trained model family built on Llama 3.1General · Open weights · 2024 ALBERT 🇺🇸 United StatesALBERT A slimmed-down BERT encoder that shares weights across its layersEncoder · Open weights · 2019 RoBERTa 🇺🇸 United StatesRoBERTa A robustly optimised BERT pretraining study with released models and codeEncoder · Open weights · 2019 T0 🇫🇷 FranceT0 Encoder-decoder model for zero-shot task generalisation via promptsGeneral · Open weights · 2021

The weekly digest

Every Monday: the model of the day, new pages on EveryLLM, and the conferences and anniversaries of the week ahead.