Thomcles/YodaLingua-Swedish

YodaLingua-Swedish YodaLingua is a high-quality speech dataset designed for training text-to-speech (TTS) systems, ASR models, and any application requiring clean, well-aligned audio–text pairs.This release contains the Swedish portion of the multilingual YodaLingua collection. 🧾 Dataset Overview Property Value Total clips 43,048 audio–transcription pairs Total duration 112 hours Speakers 1,946 distinct speakers Audio format MP3 • mono • 24 kHz •… See the full description on the dataset page: https://huggingface.co/datasets/Thomcles/YodaLingua-Swedish.