Datasets:
Follow
DICTA: The Israel Center for Text Analysis 185
Oct 15 16 days ago 23:30 Release v1.1.5: fix passive-participle lemmas (default config) (#9) - Release v1.1.5: fix passive-participle lemmas (default config) (6a81d6280b9f837dd3be2a32dafc774abe7abc1b) 20:33 lookup-1.1.1: id-based dedup + aramaic ranking (#8 ) - lookup-1.1.1: id-based dedup + aramaic ranking (4f854427fa5ec3f0f17fcac6ca8a9ac90375795d) Oct 14 17 days ago 15:37 Release v1.1.4: active-participle lemmas + irregular-root fixes (default config) (#7 ) - Release v1.1.4: active-participle lemmas + irregular-root fixes (default config) (462d84334fb0ef965c6b8f4a6fa95518d2318dd5) 15:08 lookup-1.1.0: add low_memory + disk modes to helper (#6 ) - lookup-1.1.0: add low_memory + disk modes to helper (c79a1993d93b6ed7af79743826f5071a4256732b)
- fix(disk): rebuild index on content change, not just row count (230676de2e24627a40eb9112dcf0525ddb605740)
- docs: note disk index auto-rebuilds on new release (603416618b7643a9527d499a2710e21fbfea274f)
- fix(disk): address review — dual-config fingerprint, atomic rebuild, ordering, cleanup (184c51a2a7238a7c68e02fef3f87ee894e3405ec)
- fix(disk): close probe in finally, atomic build cleanup, restore _combines (b85f8916d856f7e4535ba66d39541f2dd2862bac)
- fix(disk): cap build memory, document close() and build cost (fcb87ac49b71cc0db65dd64ee6bd58c8d65eb4a7) Oct 12 19 days ago 17:56 Release v1.1.2: lex_maleh fixes (default config) (#5 ) - Release v1.1.2: lex_maleh fixes (default config) (6115147def22580c8f97a133b55a1d11383aa738) Sep 28 20:32 default v1.1.0: recover ~9,300 wrongly-dropped wordforms (#4 ) - default v1.1.0: recover ~9,300 wrongly-dropped wordforms (6380e59d8a5fa74b10ce8c980121b6b8eac254ce)
- Card: default -> v1.1.0 (4,066,087 rows) (7a91ec80f85c1ad4f91a037335b83cde89f4eeb8)
- Changelog: v1.1.0 + real release dates (f4a7c6e962d969f3d9d5ef1469179574699e604f)
- Remove superseded wordlist-v1.0.0.parquet (default now v1.1.0) (b2f1e0867a7d3d1ad6af3bda4324bd6e03e6d013) Sep 14 18:09 Add prefix-aware lookup helper + docs (#3 ) - Add prefix-aware lookup helper + docs (f15809bf36498d85b0c85a816a73a812cf7b3230)
- Add wordlist_query.py (612d3a0bcf4893205f5773f498f80a99420647e5)
- Card: prefix-aware lookup section (3a329c48dc331543f08a44fd91feda9678f680ce)
- Changelog: helper entry (ee677facf102f95dada43abd6482dcc9505f52c1)
- wordlist_query.py: describe rules directly (remove implementation-provenance comments) (bbbf7652e8a5e17c55b12036f1eb16d5e083760d)
- Drop stale spelling-tiebreak comment (feature was reverted) (1bb7a2bf48c32f8fbaac7b62cfe03ed4148bfcb2)
- Soften build-cost estimates (hardware-dependent ballparks) (172ae5e30b5429cdaa1a8b4a4cd36e4bfbf22bab)
- Soften build-cost estimates in card (4d7efee2cc7f74a265da061f63a3ed7a7267ed1f)
- Cost note wording (d1e0f1296fc4e1623122c9da4d159da916f68be5)
- Cost note wording (c6cff50520f350e5d8dc723dc82cc626840b0cb1) Aug 26 19:14 Add attested wordforms table (#2 ) - Add attested wordforms table (11d6858f6a1439c0bc88ec3a5f25e3edf7eebe01)
- Card: document attested config (2e7df5f787efb563f8d660e387b03da6074fa482)
- Add CHANGELOG.md (c5aed478aa3e6d30ab1971a816153aea814de0de) Aug 6 12:10 Update README.md 09:41 Upload wordlist-v1.0.0.parquet with huggingface_hub 09:37 Upload README.md with huggingface_hub 09:37 initial commit 
