Modelpublic

Natashadonoh/yoruba-diacritic-restoration

sourceHugging Facecc-by-nc-4.0updated 2mo agoView on Hugging Face

0likes

Model Card

Yorùbá Diacritic Restoration

Fine-tuned Mixtral 8x7B Instruct model for restoring tone marks and underdots in undiacriticized Yorùbá text.

Model Details

—Base model: mistralai/Mixtral-8x7B-Instruct-v0.1
—Task: Diacritic restoration (text-to-text) — given undiacriticized Yorùbá text, restore correct tone marks and underdots
—Training data: yoruba-diacritic-restoration-dataset, derived from MENYO-20k, adapted via Adaption Labs
—Fine-tuning method: LoRA instruction tuning via Adaption Labs AutoScientist

Results

—Win rate: 69 (fine-tuned) vs. 32 (base model), evaluated on the target dataset
—Training converged with stable loss reduction and gradient norm over ~105 steps

Usage

Input: Yorùbá text with diacritics stripped (e.g. Nibi abajade ipade to waye ni ilu Abuja) Output: Correctly diacritized Yorùbá text (e.g. Níbi àbájádé ìpàdé tó wáyé ní ìlú Àbújá)

License

CC BY-NC 4.0, inherited from the MENYO-20k source corpus.

Author

Areo Mafadesere Natasha