metadata
dataset_info:
features:
- name: train
dtype: string
splits:
- name: train
num_bytes: 364036
num_examples: 6451
download_size: 226707
dataset_size: 364036
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: other
pretty_name: Discord Messages
size_categories:
- 1K<n<10K
language:
- en
task_categories:
- text-generation
This dataset includes more than a thousand Discord messages, ready to be used for training. All messages are filtered with static rules to prevent spam messages, sharing password or accidental Doxxing.