japanese-asr/whisper_transcriptions.reazonspeech.all.wer_10.0
收藏Hugging Face2024-08-31 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/japanese-asr/whisper_transcriptions.reazonspeech.all.wer_10.0
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个分片（split_0到split_7），每个分片包含音频、文本、Whisper转录、输入长度和标签等特征。音频的采样率为16000Hz，文本为字符串类型，Whisper转录、输入长度和标签为序列类型。数据集主要用于训练，每个分片都有相应的训练数据文件路径。
The dataset contains multiple splits (split_0 to split_7), each of which includes features such as audio, text, Whisper transcript, input length, and labels. The audio has a sampling rate of 16000Hz, the text is of string type, and the Whisper transcript, input length, and labels are of sequence type. The dataset is primarily used for training, with each split having corresponding training data file paths.
提供机构：
japanese-asr
