sn92-generalist-h4
SN92 (REAX) generalist entry v2 (eval v5.1). Fine-tune of decisionmodels/reax-1.1 @ 142b19355af00251f08db3d287307dd8b711c133: 1 LoRA run merged into the weights plus the fine-tuned readout head, trained on procedurally generated decision items with computed gold (arithmetic, logic, code output, chords, chess, rule application), Aegis-2.0-based safety labelling, self-distilled abstention, document-to-official-category items built from public labelled records (patent abstracts with CPC sections from ccdv/patent-classification; LEDGAR contract provisions, Supreme Court issue areas and CaseHOLD holdings from coastalcph/lex_glue; DBpedia ontology classes from fancyzhx/dbpedia_14; condition categories from TimSchopf/medical_abstracts; development-status and topic classifiers from the PyPI JSON API), each under its own terms, and the public SN92 generalist training set. Same architecture, tokenizer, loader and temperature as the base; see MODIFICATIONS.md and NOTICE. Licence: Apache-2.0 (upstream terms retained).
- Downloads last month
- 23