AI & ML interests
AI for the developing world
Recent Activity
A series of Large Language Models (LLMs), specifically trained for 32+ of Uganda's languages
ASR models and dataset for Kinyarwanda to explore how much data is needed to train accurate low-resource speech models.
-
evie-8/kinyarwanda-speech-hackathon
Viewer • Updated • 641k • 18 -
akera/whisper-large-v3-kin-1h-v2
Automatic Speech Recognition • 2B • Updated • 36 -
akera/whisper-large-v3-kin-50h-v2
Automatic Speech Recognition • 2B • Updated • 46 -
akera/whisper-large-v3-kin-100h-v2
Automatic Speech Recognition • 2B • Updated • 44
A series of Large Language Models (LLMs), specifically trained for 32+ of Uganda's languages
Adaptation of whisper-large-v3 for the languages widely spoken in Uganda: Luganda, Acholi, Lugbara, Ateso, etc + English (Ugandan accent)
ASR models and dataset for Kinyarwanda to explore how much data is needed to train accurate low-resource speech models.
-
evie-8/kinyarwanda-speech-hackathon
Viewer • Updated • 641k • 18 -
akera/whisper-large-v3-kin-1h-v2
Automatic Speech Recognition • 2B • Updated • 36 -
akera/whisper-large-v3-kin-50h-v2
Automatic Speech Recognition • 2B • Updated • 46 -
akera/whisper-large-v3-kin-100h-v2
Automatic Speech Recognition • 2B • Updated • 44