hugpadhugpad

xiaomi-research / MiLMMT-46-4B-Pretrain

Vision4.3B223Updated Feb 13, 2026
Hugging Face

Downloads

22

Likes

3

Last updated

Feb 13, 2026

License

gemma

MiLMMT-46-4B-Pretrain is a language model developed through continual pretraining of Gemma3-4B using a mix of 143 billion tokens from both monolingual and parallel data across 46 different languages. Please find more details in our paper: Scaling Model and Data for Multilingua…

Tasks & library

  • image text to text
  • transformers

Tags

  • transformers
  • gemma3
  • image-text-to-text

Technical specifications

Parameters4.3B
Librarytransformers
LanguagesNot available
Base modelgoogle/gemma-3-4b-pt
Licensegemma