commit ab2552b83e73c626004f3f82e9a00a883bd478b5
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun Sep 6 01:54:34 2026 +0200

sci-ml/sentence-transformers: Keyword 6.0.1 for ~arm64

The Python package and its machine-learning dependency stack support aarch64.

commit d160f27e44a78921436f4c71bf939e520bf920d5
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun Sep 6 01:54:34 2026 +0200

sci-ml/sentence-transformers: Keyword 6.0.0 for ~arm64

The Python package and its machine-learning dependency stack support aarch64.

commit d8f0e345d65e4d20c56dfe128964a31e73666bcb
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun Sep 6 01:54:34 2026 +0200

sci-ml/sentence-transformers: Keyword 5.7.0 for ~arm64

The Python package and its machine-learning dependency stack support aarch64.

commit 3ad092f1fffbbb1b6cba2e9e276688b04edb30bd
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Tue Sep 1 11:20:41 2026 +0200

sci-ml/sentence-transformers: add 6.0.1

Upstream [project.dependencies] is byte-identical to 6.0.0 -- same nine
entries, same bounds, same requires-python. Pure patch release.

Also adds the <sci-ml/huggingface_hub-2 cap that upstream has declared
since 6.0.0 (huggingface-hub<2.0.0,>=1.3.0) but this ebuild only carried
the floor for. Inert today, since ::stuff tops out at
huggingface_hub-1.29.0, and it mirrors the <sci-ml/transformers-6 line
already next to it, so the two capped deps are now expressed the same
way.

commit 58026c91590baf6db13b27d2361c071610c8c538
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Tue Aug 18 23:02:15 2026 +0200

sci-ml/sentence-transformers: drop 5.6.1

commit 07365c3186822dc78343cd36d76fb511e78e8811
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Tue Aug 18 23:02:15 2026 +0200

sci-ml/sentence-transformers: add 6.0.0

commit 7035dd45acfa6d2ae5693d554e2735933afb28ad
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun Aug 16 15:41:56 2026 +0200

sci-ml/sentence-transformers: add tokenizers dependency

The 5.7.0 project metadata requires tokenizers 0.19 or newer. Propagate the selected
Python implementation to its Gentoo provider.

commit 963d4420f1530b1298280bf84369c20c46b36757
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Aug 6 17:31:53 2026 +0200

sci-ml/sentence-transformers: drop 5.6.0

commit 516ecf30e27ad843fe32684a1c281aad2a5d57c6
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Aug 6 17:31:34 2026 +0200

sci-ml/sentence-transformers: add 5.7.0

commit 7e7b3b4916e301dffc9400d8b409ddaff6fd5417
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Fri Jul 24 11:41:20 2026 +0200

sci-ml/sentence-transformers: drop 5.5.0, 5.5.1

commit 6882abfc880d715e0e8c2b2ac72c11eb0dbb861c
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Fri Jul 24 10:49:03 2026 +0200

sci-ml/sentence-transformers: add 5.6.1

commit fcd95705efbdea9edd686e0cd9578f2a0ffc7278
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Jun 18 07:38:48 2026 +0200

sci-ml/sentence-transformers: add 5.6.0

commit a87c604c4d1693e65cefaa0cd2d36cb9cc2563ed
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Wed May 20 18:23:08 2026 +0200

sci-ml/sentence-transformers: drop 5.4.1

commit e9a82d010852878642e5ea0ab50f268ceaa97050
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Wed May 20 18:23:08 2026 +0200

sci-ml/sentence-transformers: add 5.5.1

commit bd4329091f5735e4a50de3a4c58d2c4cd9de3da5
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sat May 16 10:06:16 2026 +0200

sci-ml/sentence-transformers: add 5.5.0

Routine github bump. Upstream pyproject deps unchanged (transformers/huggingface-
hub/torch/numpy/scikit-learn/scipy/ typing-extensions/tqdm).

commit 3d75f88f24ef3f8d38cde83c8910701e723bfea8
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sat May 9 11:39:23 2026 +0200

sci-ml/sentence-transformers: new package, add 5.4.1

Pure-Python embeddings / retrieval / reranking layer over sci-ml/transformers,
needed by sci-ml/amd-gaia's deferred ui extra (eager faiss + sentence_
transformers boot in gaia.ui.server) and broadly useful for RAG pipelines.

Core deps only — no IUSE for the upstream optional-dependencies sets
(image, audio, video, train, onnx, onnx-gpu, openvino, dev). Each of
those would need additional ebuilds we don't have (datasets[2.0+] for
train, optimum-onnx, optimum-intel) or already-tracked transformers
extras that aren't in our overlay yet. Fold them in on demand.

RESTRICT="test" — upstream tests download model checkpoints from
huggingface.co, which fails inside the sandboxed network namespace.