moBERTo - a Tropic-AI Collection

Tropic-AI 's Collections

moBERTo

updated 3 days ago

Portuguese adaptation of ModernBERT, obtained through continued pretraining on a curated 12-billion-token Portuguese corpus

Tropic-AI/moBERTo

Fill-Mask • 0.1B • Updated 3 days ago • 19
Tropic-AI/moBERTo-orig-tokenizer

Fill-Mask • 0.1B • Updated 3 days ago • 14
Tropic-AI/moberto-pretraining-dataset-c4-compatible

Viewer • Updated 3 days ago • 11.4M • 227
moBERTo: A Modern Encoder for Portuguese via Continued Pretraining of ModernBERT

Paper • 2606.22722 • Published 6 days ago • 1