metadata
dataset_info:
features:
- name: sentence
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 216779209.875
num_examples: 2545
- name: test
num_bytes: 60513801
num_examples: 604
download_size: 275550950
dataset_size: 277293010.875
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
language:
- ms
license: cc-by-nc-nd-4.0
size_categories:
- 1K<n<10K
Malay Conversational Speech Corpus
This dataset is contain 2.4 hours, processing train test split with test_size=0.2 from hugging face dataset (malaysia-ai/malay-conversational-speech-corpus) for further research purpose.
Everyone are welcome to download the dataset and see the conversion from iterable dataset to dataset in notebook !!
Mirror for Malay Conversational Speech Corpus