LazarusNLP/indobert-lite-base-p1-indonli-multilingual-nli-distil-mdeberta

Text Classification • Updated Jan 31

Note `indobenchmark/indobert-lite-base-p1` fine-tuned on IndoNLI and Multilingual NLI with knowledge distillation from `MoritzLaurer/mDeBERTa-v3-base-xnli-multilingual-nli-2mil7`. Improvement of the model below with the addition of machine-translated NLI training data by 4% F1 score).

LazarusNLP/indobert-lite-base-p1-indonli-distil-mdeberta

Text Classification • Updated Jan 31 • 1

Note `indobenchmark/indobert-lite-base-p1` fine-tuned on IndoNLI with knowledge distillation from `MoritzLaurer/mDeBERTa-v3-base-xnli-multilingual-nli-2mil7`. Improves original supervised baseline's performance on test-expert subset of IndoNLI.

Note Indonesian subsets of `MoritzLaurer/multilingual-NLI-26lang-2mil7 ` dataset. Machine-translated natural language inference dataset.

w11wo/indonesian-roberta-base-indonli

Text Classification • Updated May 2, 2023 • 12 • 2

Note `flax-community/indonesian-roberta-base` fine-tuned on IndoNLI. Baseline supervised model, to be compared with distilled models. **Update**: smaller, distilled models already outperforms this model!