Datasetpublic

Zaneoiji/inneratlas-sefaria

INNER ATLAS — Sefaria quotation corpus 21,564 quotation records in 2 file(s), 7,670 of them marked default-visible. Each record carries its own rights block, verification status and attribution string; the per-record rights field is authoritative, not this page. Licence Public Domain, CC0, CC BY, CC BY-SA. Licence Records Public Domain 14,559 CC BY 2,843 CC0 2,648 CC BY-SA 1,514 Attribution What must be displayed with this text… See the full description on the dataset page: https://huggingface.co/datasets/Zaneoiji/inneratlas-sefaria.

sourceHugging Facecc0-1.0updated 6d agoView on Hugging Face

0likes136downloads

Dataset Card

INNER ATLAS — Sefaria quotation corpus

21,564 quotation records in 2 file(s), 7,670 of them marked default-visible. Each record carries its own rights block, verification status and attribution string; the per-record rights field is authoritative, not this page.

Licence

Public Domain, CC0, CC BY, CC BY-SA.

Licence	Records
Public Domain	14,559
CC BY	2,843
CC0	2,648
CC BY-SA	1,514

Attribution

What must be displayed with this text, as recorded by the ingestion that produced it:

Sefaria Jewish canon (licence-filtered)

—Source: Sefaria — <https://www.sefaria.org>
—Licence: mixed: CC0, CC BY, CC BY-SA, Public Domain
—Requires: Texts from Sefaria (sefaria.org). Each version is credited to its own edition and carries that version's licence; licences are not inherited between an original and its translation.
—Notes: Every version judged on its own recorded licence; nothing inherits from the base text. NonCommercial versions are excluded, which removes the William Davidson Talmud. Language detected from the text, since Sefaria's language tag is unreliable.

Files and integrity

File	Records	Bytes	sha256
`sefaria_en.jsonl.gz`	9,491	1,128,221	`3c9790394ef3ad67821401d9b28e8b2f7c538ecbda0293acb7cb8aeded2e8b70`
`sefaria_he.jsonl.gz`	12,073	1,829,144	`ce8802ab3240d6aa6776f82b6b1db74f2e813269b4d6ff43e5bb0af1a5c33c25`

The digests above are the authority. They are recorded in `corpus/corpus_files_v1.json` and verified on every download, so a file that arrives with records added or removed is refused rather than used.

Provenance

Produced by INNER ATLAS from public dumps and APIs, not by scraping. The ingesters, the rights model and the checks that enforce it are in that repository; DATA_LICENSE.md §3.6 records the licensing position and what it does and does not establish.

Not established here: whether the underlying quoted words are free of third-party copyright. This dataset clears the compilation licence of its provider and that provider's own rights determination. Nothing here is a finding about any individual quotation.