Model zoo
Models you can scry with
Ids resolve through the models.yaml index. Teachers are the strongest trained models; students are the compact distillations that actually ship in the crystals.
Zoo
| id / family | reveals | measured quality |
|---|---|---|
ara-diac-1.0teacher | Arabic consonants → harakāt | 2.58% windowed DER-CE on SadeedDiac-25; 1.53% on word-final iʿrāb endings |
ara-diac-small-1.0student | Arabic → harakāt (compact, 1.3 GiB) | 3.66% on the same harness — the +2.34pp cost of the ByT5-small client rung |
heb-diac-1.1student | Hebrew consonants → nikud | 16.44% DER greedy / 16.43% beam-4 — the current best Hebrew rung |
heb-diac-1.0student | Hebrew → nikud | 17.46% DER beam-4 (DictaBERT baseline 35.6%) |
heb-diac-small-1.0student | Hebrew → nikud (compact) | 30.37% DER vs teacher 24.79% — +5.6pp for the 2× shrink |
fas-g2p-1.0student | Persian → Latin reading | 1.6% CER; 77.34% SentenceBench homograph accuracy |
tha-g2p-base-1.0student | Thai → IPA | teacher at 4.43% PER (public baseline 6.37%); student gated at +4.76pp |
tha-g2p-small-1.0student | Thai → IPA (int8, 246 MiB) | 2.85% PER greedy on the runtime protocol; parity 0.08pp |
urd-diac-1.0student | Urdu → vocalized text | 3.74% CER |
urd-g2p-1.0student | Urdu → phonemes (IPA) | 14.77% CER — first learned Urdu baselines we know of |
khm-latn-1.0student | Khmer text → Latin reading | 59.66% exact match, 27.42 CER (forward model) |
11 of 11 entries
Where the numbers come from
Every metric above was produced by the benchmark’s own evaluator on its full test set — Misraj’s evaluator for SadeedDiac-25, SentenceBench for Persian homographs, the fixed 1,219-sentence Kaikki set for Thai — under published, zero-skip protocols. Methodology and run IDs live with each project’s results file; nothing here is self-graded on a private split.
Conformance methodology
Teachers are trained in secryst-train (secryst’s own monorepo); distilled students that pass the export gate are published into the interscript-ml zoo. Before any model enters the index, the gate writes its parity record into the artifact: the student is diffed against its teacher on held-out samples, and a model whose behavior differs from its published identity does not ship. At the runtime level, the three crystals are diffed against each other on shared golden sets (generated by the Python crystal) in CI — the release gate for gem 1.0 and every subsequent crystal release.