This dataset repo is the canonical public ASR result archive for the May 2026 Adja thesis release. I
Somali ASR Corpus is a cleaned Somali speech dataset designed for Automatic Speech Recognition (ASR)