theodorr/mls_eng_10k_encodec

---
dataset_info:
features:
- name: audio_duration
dtype: float64
- name: audio_token
sequence:
sequence:
sequence:
sequence: int64
- name: text
dtype: string
splits:
- name: train
num_bytes: 87042057050
num_examples: 2420047
download_size: 13484958846
dataset_size: 87042057050
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

This dataset is primarily designed for processing audio and text data, featuring audio duration, audio tokens, and text. The audio tokens are a complex multi-layer sequence structure containing integer data. The dataset includes only a training set with 2420047 samples, totaling 87042057050 bytes. The download size of the dataset is 13484958846 bytes. The training data files under the default configuration are stored in the data/train-* path.

原始信息汇总

数据集概述

数据集特征

• audio_duration: 数据类型为 float64。
• audio_token: 数据类型为 int64，具有多层序列结构。
• text: 数据类型为 string。

数据集划分

• 训练集 (train):
  • 数据量: 2420047 个样本
  • 存储大小: 87042057050 字节

数据集大小

• 下载大小: 13484958846 字节
• 总数据集大小: 87042057050 字节