Speech and OCR - a TilQazyna Collection

TilQazyna 's Collections

First Releases (2024)

Kazakh Terminology

Kazakh Morphology and POS

Til Web, Crawls and Archive

Til Books — Sources and Scans

Til Instruct — Task Datasets

Til 256k Research Ladder

Kazakh GEC — All Models

Til — Multilingual Models

Til Core — Kazakh-only Models

Til Flagship Datasets

Speech and OCR

updated about 12 hours ago

Audio data, ASR models and OCR tools for speech and document pipelines. Аудиоданные, ASR-модели и OCR-инструменты для речи и документов.