AI Potluck
Model components / Dataset Processing Tools

CCNet

Meta

The CCNet pipeline: downloads and cleans Common Crawl into high-quality monolingual corpora via language identification, KenLM perplexity filtering, and deduplication. Built CC-100.

Meta facebookresearch/cc_net; MIT, ARCHIVED read-only 2023-10-31 (no longer maintained). Foundational Common Crawl pipeline that built CC-100 and seeded the perplexity-filtering approach reused across later multilingual corpora.

Openness

5 high confidence
5.0
license
MIT(OSI)
source
public
core-gated
ungated

Fully OSI-licensed (MIT), full source public.

Adoption

2 medium confidence
2.0

~1,000 GitHub stars; research repo, no PyPI package. Archived since 2023 but a foundational, widely-reused pipeline.

Capability

3 medium confidence
3.0

Historically foundational (CC-100), but a single-method pipeline and frozen since 2023.

Unchanged since 2026-07-30 (last edited, not re-checked)