-
(data) further clean
-
(data) clean too short files, strip yaml front matter
-
Require readme to be found during cleaning.
-
(dataset) update gooddocs.parquet
-
(debug) wrong source file changed.
-
(optim) optim rebuild to run only over readme and add workers
-
(data) better filtering
-
(optim) faster rebuild
-
(format) format
-
feat: add README detection and download to GitHub repo scraping
-
(data) Exclude repo by name substring
-
Filtering data further
-
refactor: reorganize repository filtering into separate clean/ module with metadata and content filters
-
(rename) data_util to data_stats
-
feat: add DuckDB persistence and optimize GraphQL batching for repo scraping
-
feat: config to yaml
-
GraphQL for top1000 repos scraping
-
Persistence for scraping
-
GraphQL for fetching reliability
-
Merge branch 'main' of https://huggingface.co/datasets/MRiabov/gooddocs-v0
-
(data) top 1000 repos with links
-
(format) format
-
update text and filter good docs
-
Remove embedding via TEI
-
Move awesome repos to a separate repository
-
Update README.md
-
Add fetching of READMEs to final repos
-
TEI embedding generation for filtering.
-
Add missing description fetching from GitHub API
-
Remove the .txt for storing data
-
Stars and source for filtering
-
Extract awesome repos to parquet
-
gh scraping improved and split to utils.
-
(devops update) windsurf rules
-
(data) scrape from awesome repos.
-
(debug) don't clone forks
-
(data WIP) filtered out 100k duplicates and unquality data
-
gitignore the github_links.txt
-
Update the readme and remove the md-failed file for HF.
-
Regenerate the dataset to include language and READMEs
-
(update) CLI to YAML config update
-
(debug) deduplicate final parquet generation
-
(refactor) try-except blocks reduced
-
Update README
-
(devops) requirements.txt
-
(devops) windsurf rules for faster coding
-
Delete .txt to avoid hf conflicts
-
Move utils to folder
-
Update README.md
-
Update dataset card
