metadata
dataset_info:
features:
- name: in_source_id
dtype: string
- name: issue
dtype: string
- name: before_files
list:
- name: content
dtype: string
- name: path
dtype: string
- name: after_files
list:
- name: content
dtype: string
- name: path
dtype: string
- name: pr_diff
dtype: string
splits:
- name: train
num_bytes: 7054760354.199002
num_examples: 67511
download_size: 2653477261
dataset_size: 7054760354.199002
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
# removed all repos of SWE-bench and RepoBench
repos = [
"astropy",
"django",
"flask",
"matplotlib",
"seaborn",
"requests",
"xarray",
"pylint",
"pytest",
"scikit-learn",
"sphinx",
"sympy",
]