Logo Lanfrica

Obiohagwu/tts-nigeria

Domaine:

natural language processing

Type de record:

model
Créateur:
Obi
Hôte:
a Nigerian accented tts model # tts-nigeria a Nigerian accented tts model ## Base Model The base model will be based on either Tortoise-tts or T5 from google ### Dataset Thank God for @benjaminogbonna. He provided a free audio dataset of accented nigerian speech. The Nigerian Accent Speech Data is a comprehensive dataset of about 8 hours of audio recordings featuring speakers from various regions of Nigeria, capturing the rich diversity of Nigerian accents. This dataset is specifically curated to address the gap in speech and language datasets for African accents, making it a valuable resource for researchers and developers working on Automatic Speech Recognition (ASR), Speech-to-text (STT), Text-to-Speech (TTS), Accent recognition, and Natural language processing (NLP) systems. ### Hardware The model was trained on an a100 sxm instance on runpod.