-
docs: separate Cadryl catalogues from the native PyTorch cascade
-
docs: English quick cards, variant guides and recorded comparisons
-
Add four verified v2.5 LiteRT cascade models, Cadryl contracts and bilingual 680-image benchmark
-
Refresh bilingual documentation, model evidence and continued-learning guides
unicorn who dev
Documentation and evidence presentation only. Preserve all weights, contracts, raw historical reports, source notices and repository visibility.
-
docs: add bilingual LiteRT evidence, figures and integration guides
unicorn who dev
