How piskvorky/gensim-data works
Gensim-data is a durable catalog and distribution store for unstructured-text corpora and pretrained models. The documented workflow is that a user requests information or a named resource through Gensim’s downloader; it uses repository metadata to discover resources and retrieves immutable release attachments, which the downloader can return as a local path or open as a corpus/model. README documents datasets, models, release notes/examples, and per-dataset licensing. Runtime downloader implementation is outside this repository, so the interaction edges reflect the README rather than inspected downloader source.
Main entry points: list.json Resource metadata
Components 6
Resource catalog
- Resource metadata
list.json - Usage and catalog guide
README.md
Release storage
- GitHub release attachments
Other components
- Gensim user
- Gensim downloader
- Local corpus or model
Connections 5
- Gensim user → Gensim downloader: requests resource
- Gensim downloader → Resource metadata: reads catalog
- Gensim downloader → GitHub release attachments: downloads attachment
- GitHub release attachments → Gensim downloader: serves asset
- Gensim downloader → Local corpus or model: returns or opens