This dataset is an enhanced version of the FLEURS (Few-shot Learning Evaluation of Universal Representations of Speech) test set, restructured with complete metadata for easier use in automatic speech recognition (ASR) and multilingual speech processing tasks.
Dataset Description
FLEURS is a multilingual speech benchmark dataset designed to evaluate universal speech representations. This particular version focuses on 25 European… See the full description on the dataset page: https://huggingface.co/datasets/rasgaard/fleurs_test.
