metadata
language: multilingual
license: apache-2.0
tags:
- bert
- modernbert
- flexbert
- masked-language-modeling
Bert
This model is a ModernBERT/FlexBERT checkpoint uploaded during training.
Training Information
- Step: 2001
- Epoch: 0
- Samples Seen: 1070555
Metrics
Model Architecture
This model uses the FlexBERT architecture with modern improvements over traditional BERT.
Usage
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained("QuangDuy/Bert")
tokenizer = AutoTokenizer.from_pretrained("QuangDuy/Bert")
Citation
If you use this model, please cite the ModernBERT paper.