metadata
license: cc-by-4.0
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: code
dtype: string
- name: repo_path
dtype: string
splits:
- name: train
num_bytes: 619559312188
num_examples: 77760861
download_size: 35038291124
dataset_size: 619559312188
This is starcoderdata, but with leading boilerplate text/license text removed, and with short sequences filtered out. It also removes the extra tags at the beginning of some of the files, like <reponame>.