Portuguese adaptation of ModernBERT, obtained through continued pretraining on a curated 12-billion-token Portuguese corpus
-
Tropic-AI/moBERTo
Fill-Mask • 0.1B • Updated • 55 -
Tropic-AI/moBERTo-orig-tokenizer
Fill-Mask • 0.1B • Updated • 23 -
Tropic-AI/moberto-pretraining-dataset-c4-compatible
Viewer • Updated • 11.4M • 318 -
moBERTo: A Modern Encoder for Portuguese via Continued Pretraining of ModernBERT
Paper • 2606.22722 • Published • 1