darkproger/librispeech_asr

---
license: cc-by-4.0
---

This is a dataset is a fork of [librispeech_asr](https://huggingface.co/datasets/librispeech_asr) that defines each original split (like train-clean-100) as a split (named `train.clean.100`, with dots instead of hyphens). This allows you to download each part separately.

This fork also reports a `path` for each sample accurately.

原始信息汇总

数据集概述

数据集来源

• 本数据集是librispeech_asr的一个分支。

数据集结构

• 原始数据集中的每个分割（如train-clean-100）被重新定义为一个新的分割，命名方式为train.clean.100，使用点代替了原始的连字符。
• 每个样本都准确报告了一个path。

许可证

• 本数据集遵循CC-BY-4.0许可证。