iNeil77/filtered-repos

---
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: repo_name
dtype: string
- name: repo_url
dtype: string
- name: snapshot_id
dtype: string
- name: revision_id
dtype: string
- name: directory_id
dtype: string
- name: branch_name
dtype: string
- name: visit_date
dtype: timestamp[ns]
- name: revision_date
dtype: timestamp[ns]
- name: committer_date
dtype: timestamp[ns]
- name: github_id
dtype: int64
- name: star_events_count
dtype: int64
- name: fork_events_count
dtype: int64
- name: gha_license_id
dtype: string
- name: gha_created_at
dtype: timestamp[ns]
- name: gha_updated_at
dtype: timestamp[ns]
- name: gha_pushed_at
dtype: timestamp[ns]
- name: gha_language
dtype: string
- name: files
list:
- name: blob_id
dtype: string
- name: path
dtype: string
- name: content_id
dtype: string
- name: language
dtype: string
- name: length_bytes
dtype: int64
- name: detected_licenses
list: string
- name: license_type
dtype: string
- name: src_encoding
dtype: string
- name: is_vendor
dtype: bool
- name: is_generated
dtype: bool
- name: alphanum_fraction
dtype: float32
- name: alpha_fraction
dtype: float32
- name: num_lines
dtype: int32
- name: avg_line_length
dtype: float32
- name: max_line_length
dtype: int32
- name: num_files
dtype: int64
- name: num_format_pass_prop
dtype: float64
- name: num_ext_pass_prop
dtype: float64
splits:
- name: train
num_bytes: 4008429812
num_examples: 402243
download_size: 2256630346
dataset_size: 4008429812
---