Decision models that read text and images, in 19 languages. Typed questions in, calibrated probabilities out, in one pass.
AI & ML interests
None defined yet.
Recent Activity
Document → JSON extraction for English and African languages, with calibrated confidence per field. Text, photos and scans.
A collection of models for diacritic restoration. Explore cutting-edge tools for processing and enhancing text with diacritics.
A curated collection of language identification models, including specialized solutions for Nigerian and multilingual contexts.
A collection of token classification models designed for diacritic tagging and restoration.
showcases experimental and production-ready language models created using modern model merging techniques.
Decision models for African languages and English. Typed questions in, calibrated probabilities out, in one pass.
A collection of token classification models designed for diacritic tagging and restoration.
A compact multilingual question generator across 25 languages. Given a passage, it produces natural, search-style questions that the passage directly
A collection of models for diacritic restoration. Explore cutting-edge tools for processing and enhancing text with diacritics.
A collection of embedding models for sentence similarity, reranking and fill-mask tasks
contains pretrained tokenizers for multilingual language modeling. Designed with a focus on African and other underrepresented languages
Decision models that read text and images, in 19 languages. Typed questions in, calibrated probabilities out, in one pass.
Decision models for African languages and English. Typed questions in, calibrated probabilities out, in one pass.
Document → JSON extraction for English and African languages, with calibrated confidence per field. Text, photos and scans.
A collection of token classification models designed for diacritic tagging and restoration.
A collection of models for diacritic restoration. Explore cutting-edge tools for processing and enhancing text with diacritics.
A compact multilingual question generator across 25 languages. Given a passage, it produces natural, search-style questions that the passage directly
A curated collection of language identification models, including specialized solutions for Nigerian and multilingual contexts.
A collection of models for diacritic restoration. Explore cutting-edge tools for processing and enhancing text with diacritics.
A collection of token classification models designed for diacritic tagging and restoration.
A collection of embedding models for sentence similarity, reranking and fill-mask tasks
showcases experimental and production-ready language models created using modern model merging techniques.
contains pretrained tokenizers for multilingual language modeling. Designed with a focus on African and other underrepresented languages