RSS / README.md
VladS159's picture
Resampled Romanian speech dataset from 48kHz to 16kHz with preprocessing
7ee4cf1 verified
---
dataset_info:
features:
- name: sentence
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 429209311.32
num_examples: 3180
- name: test
num_bytes: 61342282.0
num_examples: 536
download_size: 489592289
dataset_size: 490551593.32
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
---