mdeberta-id-20k

mDeBERTa-ID-20k

An aggressively vocabulary-pruned version of microsoft/mdeberta-v3-base with a 20k-token Indonesian vocabulary, designed for downstream Indonesian NLP tasks and resource-constrained applications.

This model was developed using VocabPrune, a deterministic, language-aware, frequency-based vocabulary pruning method designed to reduce vocabulary-related model overhead.

The model is a base checkpoint and should be fine-tuned for a specific downstream task.

Model Details

This checkpoint uses a more aggressive vocabulary reduction than the 30k models in the VocabPrune collection.

Resources

• Paper: IEEE Access
• Paper DOI: 10.1109/ACCESS.2026.3679735
• Source Code: GitHub – VocabPrune
• Reproducibility Archive: Zenodo – VocabPrune v1.0.1
• Model Collection: Pruned mDeBERTa

For the methodology, experimental setup, and detailed evaluation results, please refer to the published paper.

Citation

If you use this model or the VocabPrune methodology in your research, please cite:

@article{fuadi2026efficient,
  author  = {Fuadi, Mukhlish and Wibawa, Adhi Dharma and Sumpeno, Surya},
  title   = {Efficient Transformer Models via Language-Aware
             Frequency-Based Vocabulary Pruning},
  journal = {IEEE Access},
  volume  = {14},
  pages   = {50993--51006},
  year    = {2026},
  doi     = {10.1109/ACCESS.2026.3679735}
}