feat(export): ara-diac-small-2.0 — the Muon optimizer rung

models/ara-diac-small/ara-diac-small-2.0.README.md

@@ -0,0 +1,20 @@
+# ara-diac-small-2.0
+
+Arabic diacritization (haraqat restoration). Client-tier ByT5-small
+student, identical corpus/labels/epochs to ara-diac-small-1.0 with one
+variable changed: **the Muon optimizer** (E3 factorial). That single
+change closes 2.96pp of the 5.68pp teacher-student gap:
+
+- 1.0 (AdamW): 8.26 full-set windowed DER-CE
+- **2.0 (Muon): 5.29** (teacher r6: 2.58; in-run reproduction 2.60)
+
+The strict teacher+0.5pp gate is still missed; the residual decomposes
+as ~0.70pp capacity + ~2.25pp domain coverage (E2/E3 factorial), and an
+r7-teacher re-distillation is in flight. Identical IMF v1 contract:
+dynamic fetch, sha256-verified, KV decode, margins JSON alongside.
+
+```python
+from interscript_ml import Model
+model = Model.load("ara-diac-small-2.0")
+model.translate("قوله فحكمها في الوفاة")
+```

models/ara-diac-small/ara-diac-small-2.0.metadata.yaml

@@ -0,0 +1,32 @@
+format: imf-v1
+id: ara-diac-small-2.0
+task: diacritization
+source_script: Arab
+target: Arab
+tokenizer: bytes
+opset: 14
+decoder: kv
+precision: fp32
+license: BSD-3-Clause
+trained_from: 'sequence-level KD from the r6 teacher (rababa_arabic_byt5/run-006-morph/best,
+  2.5793 windowed DER-CE full protocol): the same 29,322 greedy teacher labels as
+  ara-diac-small-1.0, with the Muon optimizer (E3 factorial, vanilla student arm).
+  Checkpoint rababa-checkpoints:/rababa_arabic_distill_small/run-005-muon/best. The
+  optimizer alone closes 2.96pp of the 5.68pp teacher-student gap measured for the
+  1.0 (AdamW) release: 8.259 -> 5.2945 full-set windowed DER-CE (teacher reproduces
+  2.5997 in-run). Still misses the strict teacher+0.5pp gate — the residual decomposes
+  as ~0.70pp capacity + ~2.25pp domain coverage per the E2/E3 factorial; miss disclosed.
+  An r7-teacher re-distillation is in flight and expected to supersede as a later rung.'
+metrics:
+- name: der_teacher_fullset
+  value: 2.5997
+  protocol: windowed DER-CE (1400-byte windows, word-boundary split, greedy,
+    haraqat-projected, Misraj evaluator); full 1,200-paragraph SadeedDiac-25;
+    in-run reproduction of the documented 2.5815/2.5793
+  source: interscript/interscript-ml docs/RESULTS.md#run-005-muon
+- name: der_student_fullset
+  value: 5.2945
+  protocol: same full-set harness; +2.69pp over the in-run teacher — the Muon
+    arm of the E3 2x2 factorial (vanilla student, optimizer as the only variable
+    vs the 8.259 AdamW release)
+  source: interscript/interscript-ml docs/RESULTS.md#run-005-muon

src/gpu/modal_export.py

@@ -114,6 +114,15 @@
         "test_data": "arabic-sadeed-imf/test.jsonl",
         "probe": "قوله",
     },
+    "ara-diac-small-2": {
+        "volume": "/volumes/rababa-checkpoints",
+        "checkpoint": "rababa_arabic_distill_small/run-005-muon/best",
+        "metadata": "models/ara-diac-small/ara-diac-small-2.0.metadata.yaml",
+        "readme": "models/ara-diac-small/ara-diac-small-2.0.README.md",
+        "test_volume": "/datasets/rababa",
+        "test_data": "arabic-sadeed-imf/test.jsonl",
+        "probe": "قوله",
+    },
     "urd-diac": {
         "volume": "/volumes/urdu-diacrit-checkpoints",
         "checkpoint": "urdu_diacrit/run-001/best",