Datasets:
metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 2265821001.125
num_examples: 6215
- name: test
num_bytes: 237323546
num_examples: 637
download_size: 2166856666
dataset_size: 2503144547.125
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
license: cc-by-sa-4.0
task_categories:
- automatic-speech-recognition
language:
- my
pretty_name: >-
Myanmar Speech Dataset for training and fine-tuning the ASR (Automatic Speech
Recognition) model.
Please visit to the GitHub repository for other Myanmar Langauge datasets.
Myanmar Speech Dataset for ASR
This dataset is a comprehensive collection of Myanmar language speech data specifically curated for Automatic Speech Recognition (ASR) task. It combines following datasets:
By merging these complementary resources, this dataset provides a more robust foundation for training and evaluating ASR models for the Myanmar language, offering greater diversity in speakers, speaking styles, and acoustic conditions.