Kotib/

eval_lang_id_multilingual_99

	Multilingual Language Identification (LID) Evaluation Dataset




	
	Developed by Kotibai & Rubai Team (@islomov)

A balanced multilingual speech dataset for training and evaluating language identification models.

This dataset combines audio samples from three major multilingual speech corpora:

Common Voice 17 (Mozilla) - Crowdsourced read speech with transcriptions FLEURS (Google) - Professional recordings in 102 languages VoxLingua107 -… See the full description on the dataset page: https://huggingface.co/datasets/Kotib/eval_lang_id_multilingual_99.