japanese-asr/whisper_transcriptions.reazon_speech_all.wer_10.0
收藏Hugging Face2024-09-14 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/japanese-asr/whisper_transcriptions.reazon_speech_all.wer_10.0
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个子集，每个子集包含音频数据和对应的转录文本。音频的采样率为16000Hz，转录文本包括原始转录、GPT-3.5生成的英文转录、Whisper生成的转录等。每个子集的训练集大小和样本数量不同，数据量从几百MB到几GB不等。
This dataset contains multiple subsets, each of which includes audio data and corresponding transcriptions. The audio has a sampling rate of 16000Hz, and the transcriptions include original transcriptions, GPT-3.5 generated English transcriptions, Whisper generated transcriptions, etc. The size and number of samples in the training set vary across subsets, with data sizes ranging from hundreds of MB to several GB.
提供机构：
japanese-asr
