Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
code
Size:
10K - 100K
ArXiv:
Tags:
code
License:
metadata
dataset_info:
features:
- name: commit
dtype: string
- name: old_file
dtype: string
- name: new_file
dtype: string
- name: old_contents
dtype: string
- name: new_contents
dtype: string
- name: subject
dtype: string
- name: message
dtype: string
- name: lang
dtype: string
- name: license
dtype: string
- name: repos
dtype: string
- name: ndiff
dtype: string
- name: instruction
dtype: string
- name: content
dtype: string
splits:
- name: train
num_bytes: 113752028
num_examples: 22602
download_size: 48124127
dataset_size: 113752028
configs:
- config_name: default
data_files:
- split: train
path: data/train-*