ylacombe/parler-tts-large-v1-a_speaker_similarity

该数据集包含多个特征，如原始文本、说话者ID、路径、章节ID、ID、信噪比（SNR）、C50、语音持续时间、语速、音素、STOI、SI-SDR、PESQ、性别、音高标准差、音高均值、音高、噪声、混响、语音单调性、噪声SDR、语音质量PESQ、口音、文本描述、音频、文本、生成音频和相似度等。数据集包含一个训练集，大小为626426671字节，包含340个样本。

The dataset includes various features related to speech and text, such as original text, speaker ID, audio path, chapter ID, SNR, C50, speech duration, speaking rate, phonemes, STOI, SI-SDR, PESQ, gender, pitch statistics, noise, reverberation, speech monotony, noise SDR, speech quality PESQ, accent, text description, audio, generated audio, and similarity. The dataset is divided into a training set with 340 samples, with a download size of 376855193 bytes and a dataset size of 626426671.0 bytes.

原始信息汇总

数据集概述

数据集信息

特征

• text_original: 字符串
• speaker_id: 字符串
• path: 字符串
• chapter_id: 字符串
• id: 字符串
• snr: 浮点数 (float32)
• c50: 浮点数 (float32)
• speech_duration: 浮点数 (float32)
• speaking_rate: 字符串
• phonemes: 字符串
• stoi: 浮点数 (float32)
• si-sdr: 浮点数 (float32)
• pesq: 浮点数 (float32)
• gender: 字符串
• utterance_pitch_std: 浮点数 (float32)
• utterance_pitch_mean: 浮点数 (float32)
• pitch: 字符串
• noise: 字符串
• reverberation: 字符串
• speech_monotony: 字符串
• sdr_noise: 字符串
• pesq_speech_quality: 字符串
• accent: 字符串
• text_description: 字符串
• audio: 音频 (采样率: 16000)
• text: 字符串
• generated_audio: 音频 (采样率: 16000)
• similarity: 浮点数 (float64)

数据分割

• train: 340个样本, 626426671.0字节

数据集大小

• 下载大小: 376855193字节
• 数据集大小: 626426671.0字节

配置

• config_name: default
  • data_files:
    • split: train
    • path: data/train-*