ThinkingMachinesDataScience/Ratchada-Fang-Thon-Whisper
Automatic Speech Recognition
•
Updated
The dataset includes recordings from earnings calls of publicly traded companies in Thailand. Each audio file is accompanied by a transcription and metadata such as company name, reporting period, and other relevant details.
Total Duration:
File Count:
The dataset consists of recordings from various financial presentations categorized into different subfolders based on event and company.
Features:
Dataset Size: 2.11 GB
from datasets import load_dataset
dataset = load_dataset("ThinkingMachinesDataScience/Ratchada-STT")
train = dataset['train']
test = dataset['test']
print(train[0])
This dataset is distributed under the MIT
License.
If you use this dataset in your research or work, please cite it as:
@dataset{thinkingmachinesdatascience/ratchada-stt,
title = {Ratchada-STT Dataset},
publisher = {Hugging Face},
year = {2024},
author = {Thinking Machines Data Science},
license = {MIT},
}