iamTangsang/

nepali_to_english_pipeline_evaluation

	Nepali-English Speech-to-Text Translation Evaluation Dataset




	
	Dataset Description

This dataset is designed for evaluating Nepali→English speech-to-text translation pipelines. It contains audio recordings of 300 Nepali sentences, spoken by three speakers, covering a range of sentence types (statements, questions, commands, complex sentences, and named entities/numbers). Each sentence is paired with:

Source text (Nepali) transcription Reference English translation Audio… See the full description on the dataset page: https://huggingface.co/datasets/iamTangsang/nepali_to_english_pipeline_evaluation.