doof-ferb/fpt_fosd

---
license: cc-by-4.0
task_categories:
- automatic-speech-recognition
- text-to-speech
language:
- vi
pretty_name: FPT Open Speech Dataset (FOSD)
size_categories:
- 10K<n<100K
dataset_info:
features:
- name: audio
dtype: audio
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 684961355.008
num_examples: 25917
download_size: 819140462
dataset_size: 684961355.008
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

# unofficial mirror of FPT Open Speech Dataset (FOSD)

released publicly in 2018 by FPT Corporation

100h, 25.9k samples

official link (dead): https://fpt.ai/fpt-open-speech-data/

mirror: https://data.mendeley.com/datasets/k9sxg2twv4/4

DOI: `10.17632/k9sxg2twv4.4`

pre-process:
- remove non-sense strings: `-N` `\r\n`
- remove 4 files because missing transcription:
- `Set001_V0.1_008210.mp3`
- `Set001_V0.1_010753.mp3`
- `Set001_V0.1_011477.mp3`
- `Set001_V0.1_011841.mp3`

need to do: check misspelling

usage with HuggingFace:
```python
# pip install -q "datasets[audio]"
from datasets import load_dataset
from torch.utils.data import DataLoader

dataset = load_dataset("doof-ferb/fpt_fosd", split="train", streaming=True)
dataset.set_format(type="torch", columns=["audio", "transcription"])
dataloader = DataLoader(dataset, batch_size=4)
```

原始信息汇总

FPT Open Speech Dataset (FOSD)

基本信息

• 许可证: cc-by-4.0
• 任务类别:
  • 自动语音识别
  • 文本转语音
• 语言: 越南语
• 数据集名称: FPT Open Speech Dataset (FOSD)
• 数据量: 10K<n<100K

数据集详情

• 特征:
  • 音频
  • 转录文本
• 分割:
  • 训练集: 25917个样本, 684961355.008字节
• 下载大小: 819140462字节
• 数据集大小: 684961355.008字节

配置

• 默认配置:
  • 数据文件路径: data/train-*