BERT masked-language model
The captured configuration identifies BertForMaskedLM with model type bert and Transformers support.
Open Source Model Profile · pablocosta
bertabaporu-base-uncased
bertabaporu-base-uncased is a Brazilian Portuguese BERT fill-mask model from pablocosta. According to the model card, it targets the Twitter domain under mit licensing.
- Publisher
- pablocosta
- Task
- fill-mask
- Model type
- bert
- License
- mit
- Library
- transformers
- Publication status
- Accepted · not indexed
Model overview
bertabaporu-base-uncased is published by pablocosta as a fill-mask BERT model. The captured configuration identifies BertForMaskedLM with model type bert under mit. According to the model card, it is the BERT-Base variant of BERTabaporu, a Brazilian Portuguese Twitter-domain model.
Recorded capabilities
Brazilian Portuguese Twitter domain
According to the model card, the model addresses Brazilian Portuguese in the Twitter domain.
BERT-Base configuration
According to the model card, this variant is BERT-Base with 12 layers and 110M parameters.
Documented research use
According to the model card, loading code and a RANLP 2023 citation are provided for research reference.
Use cases in the source record
- Fill-mask prediction on Brazilian Portuguese Twitter-style text.
- Portuguese social-media language research using the card's documented paper, code, and citation.
Limitations and unknowns
- No Safetensors parameter count was extracted from this record; the only size figure is the publisher's 110M statement.
- No evaluation scores were extracted from this record.
- Provider state is historical snapshot data and should be refreshed before being presented as current.
- The publisher's claim of outperforming a general-purpose Portuguese model on Twitter tasks is unverified in this evidence and should not be treated as an Ethen measurement.
Source and provenance
Source: pablocosta/bertabaporu-base-uncased
Captured: Unknown. Processed: 2026-09-07T19:34:55.250396+00:00.
Paper: For more details, please refer to our paper: BERTabaporu: Assessing a Genre-Specific Language Model for Portuguese NLP Introduction BERTabaporu is a Brazilian Portuguese BERT model in the Twitter domain. The model has been built from a collection of 238 million tweets written by over 100 thousand unique Twitter users, and conveying over 2.9 billion tokens in total. Available models Model Arch. #Layers #Params pablocosta/bertabaporu-base-uncased BERT-Base 12 110M pablocosta/bertabaporu-large-uncased BERT-Large 24 335M Usage from transformers import AutoTokenizer # Or BertTokenizer from transformers import AutoModelForPreTraini…
F001F002F003F004F005F006F007F008F009F010F012F013F014