knoveleng/ms
收藏Hugging Face2025-12-11 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/knoveleng/ms
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
features:
- name: transcription
dtype: string
- name: gender
dtype: string
- name: country
dtype: string
- name: utterance_pitch_mean
dtype: float64
- name: utterance_pitch_std
dtype: float64
- name: snr
dtype: float64
- name: c50
dtype: float64
- name: speech_duration
dtype: float64
- name: stoi
dtype: float64
- name: si-sdr
dtype: float64
- name: pesq
dtype: float64
- name: pitch
dtype: string
- name: speaking_rate
dtype: string
- name: noise
dtype: string
- name: reverberation
dtype: string
- name: speech_monotony
dtype: string
- name: sdr_noise
dtype: string
- name: audio_filename
dtype: string
- name: prompt
dtype: string
- name: audio
dtype: audio
splits:
- name: train
num_bytes: 5844482059.5
num_examples: 18000
- name: validation
num_bytes: 324693447.75
num_examples: 1000
- name: test
num_bytes: 324693447.75
num_examples: 1000
download_size: 6274493720
dataset_size: 6493868955.0
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*
---
This dataset consists of 18,000 training samples, 1,000 validation samples, and 1,000 test samples. Each sample includes audio transcription text, speaker gender, speaker's country, multiple acoustic features of the audio (e.g., average pitch, pitch standard deviation, signal-to-noise ratio, speech duration, etc.), audio quality metrics (e.g., STOI, SI-SDR, PESQ, etc.), audio filename, and prompt text. The dataset also includes the audio files themselves.
提供机构：
knoveleng
