veziriii/Creativeguy-persian-asr

这是一个波斯语自动语音识别数据集，专为文本转语音和自动语音识别任务设计。数据集包含音频文件，已重采样至16000 Hz，并配有波斯语转录文本。数据分为训练集和测试集：训练集有13165个样本，总时长100小时37分钟38秒，体积2.712 GB；测试集有5643个样本，总时长43小时8分钟16秒，体积1.163 GB。总计18808个样本，总时长143小时45分钟55秒，总体积3.874 GB。

This is a Persian automatic speech recognition dataset designed for text-to-speech and automatic speech recognition tasks. The dataset includes audio files that have been resampled to 16000 Hz and come with Persian transcripts. It is split into training and test sets: the training set contains 13,165 samples with a total duration of 100 hours, 37 minutes, and 38 seconds, and a volume of 2.712 GB; the test set contains 5,643 samples with a total duration of 43 hours, 8 minutes, and 16 seconds, and a volume of 1.163 GB. In total, there are 18,808 samples with a total duration of 143 hours, 45 minutes, and 55 seconds, and an overall volume of 3.874 GB.