Spaces:
Running
Running
Expand org README: datasets-first reframe + clearer contribution paths
#3
by davanstrien HF Staff - opened
The current README has good bones but the substance is hidden inside collapsible
blocks, and the framing has drifted relative to where the org actually sits today (46+ datasets vs 5 mostly-2023 models). Opening this as a PR rather than a direct commit since BigLAM is a community org with many contributors — easier to gather feedback this way.
Changes:
- Kept the frontmatter unchanged.
- Replaced the corporate "We aim to" mission-statement structure with a more descriptive intro that names what BigLAM actually solves (LAM data lives in IIIF / METS / MODS / catalogue formats that need ML-prep work).
- Unfolded the sections — the content was good, hiding it served no one.
- "What's here" reframes the org as datasets-first (since that's what the numbers say). Models and spaces noted honestly as BigScience-era experiments.
- Cross-link to small-models-for-glam as the focused-models sibling org — clarifies the division of labour.
- Contributing section now recommends institutional orgs as the preferred home (with NLS as the worked example), positions BigLAM as the fallback when an institution doesn't have a Hub presence yet, and offers to transfer existing BigLAM datasets to institutional orgs that want to take ownership.
- Honest "light-touch maintenance" footer to set expectations.
Happy to take this in, partially merge, or pass on it — flag any pushback in a comment.
davanstrien changed pull request status to merged