atmatechai/indspeech_news_tts_embedding

---
dataset_info:
features:
- name: input_ids
sequence: int32
- name: labels
sequence:
sequence: float32
- name: speaker_embeddings
sequence: float32
splits:
- name: train
num_bytes: 131023319.94035785
num_examples: 1810
- name: test
num_bytes: 14622492.059642147
num_examples: 202
download_size: 147030425
dataset_size: 145645812.0
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
---

The dataset includes three main features: input_ids (integer sequence), labels (nested float sequence), and speaker_embeddings (float sequence). The dataset is divided into a training set and a test set, containing 1810 and 202 samples respectively. The download size of the dataset is 147030425 bytes, and the actual size is 145645812 bytes. The dataset configuration is set to default, with the training and test data files stored in data/train-* and data/test-* paths respectively.

原始信息汇总

数据集概述

数据特征

• input_ids: 整数序列，数据类型为 int32。
• labels: 嵌套序列，内部数据类型为 float32。
• speaker_embeddings: 浮点数序列，数据类型为 float32。

数据分割

• train:
  • 样本数量: 1810
  • 数据大小: 131023319.94035785 字节
• test:
  • 样本数量: 202
  • 数据大小: 14622492.059642147 字节

数据集大小

• 下载大小: 147030425 字节
• 数据集大小: 145645812.0 字节

配置

• default:
  • train: 数据路径为 data/train-*
  • test: 数据路径为 data/test-*