SayantanJoker/All_Hindi_ASR_v1.1_Cleaned

该数据集包含了音频文件及其对应的转录文本，可用于训练语音识别模型。数据集分为训练集，共有41429个音频转录对，文件格式为音频文件的采样率为44100Hz，转录文本为字符串类型，同时记录了每个文件的名称。

The dataset includes audio files and their corresponding transcriptions, which can be used to train speech recognition models. The dataset is split into a training set, containing a total of 41,429 audio transcription pairs. The audio files have a sampling rate of 44,100 Hz, the transcriptions are of string type, and the file names are recorded for each file.