diff --git a/models.yaml b/models.yaml index 2006f5b..d3d71de 100644 --- a/models.yaml +++ b/models.yaml @@ -230,3 +230,29 @@ models: - {name: der_student_fullset, value: 8.259, source: interscript/interscript-ml docs/RESULTS.md#ara-diac-small-1.0} parity: {samples: 800, cer_delta: 0.0} license: BSD-3-Clause + ara-diac-small-2.0: + task: diacritization + scripts: [Arab, Arab] + precision: fp32 + filename: ara-diac-small-2.0-fp32.zip + url: https://github.com/interscript/interscript-ml/releases/download/ara-diac-small-2.0/ara-diac-small-2.0-fp32.zip + sha256: d9aa95d0e8fd0d7da80fd2f7dabedf0ac65f55ba984de565d252f5e14906d98c + size: 1418516306 + metrics: + - {name: der_teacher_fullset, value: 2.289, source: interscript/interscript-ml docs/RESULTS.md#run-006-r7-muon} + - {name: der_student_fullset, value: 4.8218, source: interscript/interscript-ml docs/RESULTS.md#run-006-r7-muon} + parity: {samples: 600, cer_delta: 0.1187} + license: BSD-3-Clause + ara-diac-small-2.0-int8: + task: diacritization + scripts: [Arab, Arab] + precision: int8 + filename: ara-diac-small-2.0-int8.zip + url: https://github.com/interscript/interscript-ml/releases/download/ara-diac-small-2.0/ara-diac-small-2.0-int8.zip + sha256: f4d07acdaf9c18cded97e15e0c5636e6cee90de2fa32c5a373a44c0cc912bf50 + size: 264043016 + metrics: + - {name: der_teacher_fullset, value: 2.289, source: interscript/interscript-ml docs/RESULTS.md#run-006-r7-muon} + - {name: der_student_fullset, value: 4.8218, source: interscript/interscript-ml docs/RESULTS.md#run-006-r7-muon} + parity: {samples: 600, cer_delta: 0.1393} + license: BSD-3-Clause diff --git a/models/ara-diac-small/ara-diac-small-2.0.metadata.yaml b/models/ara-diac-small/ara-diac-small-2.0.metadata.yaml index 79af49c..64642c4 100644 --- a/models/ara-diac-small/ara-diac-small-2.0.metadata.yaml +++ b/models/ara-diac-small/ara-diac-small-2.0.metadata.yaml @@ -11,23 +11,29 @@ license: BSD-3-Clause trained_from: 'sequence-level KD from the r7 canonical teacher (rababa_arabic_byt5/run-007-news/best, 2.2864 windowed DER-CE full protocol): fresh greedy r7 labels on the same r5-units corpus/limits as ara-diac-small-1.0, Muon optimizer (E3-adopted), vanilla ByT5-small - (E4, pre-registered gate <= 6.26). Checkpoint - rababa-checkpoints:/rababa_arabic_distill_small/run-006-r7-muon/best. The two - measured wins compound: 8.259 -> 4.8218 full-set windowed DER-CE (teacher reproduces - 2.289 in-run vs documented 2.2864) — a 42% error reduction on the 1.0 release at the - same architecture and artifact size. Still misses the strict teacher+0.5pp gate - (+2.53pp; miss disclosed); the E2/E3 factorial attributes the residual to domain + (E4, pre-registered gate <= 6.26). Checkpoint rababa-checkpoints:/rababa_arabic_distill_small/run-006-r7-muon/best. + The two measured wins compound: 8.259 -> 4.8218 full-set windowed DER-CE (teacher + reproduces 2.289 in-run vs documented 2.2864) — a 42% error reduction on the 1.0 + release at the same architecture and artifact size. Still misses the strict teacher+0.5pp + gate (+2.53pp; miss disclosed); the E2/E3 factorial attributes the residual to domain coverage.' metrics: - name: der_teacher_fullset value: 2.289 - 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 2.2864 (r7 canonical teacher) + 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 2.2864 (r7 canonical teacher) source: interscript/interscript-ml docs/RESULTS.md#run-006-r7-muon - name: der_student_fullset value: 4.8218 - protocol: same full-set harness; E4 (r7 teacher labels + Muon, vanilla - ByT5-small) vs the 8.259 AdamW/r6-labels 1.0 release — a 42% error - reduction at identical architecture and artifact size + protocol: same full-set harness; E4 (r7 teacher labels + Muon, vanilla ByT5-small) + vs the 8.259 AdamW/r6-labels 1.0 release — a 42% error reduction at identical + architecture and artifact size source: interscript/interscript-ml docs/RESULTS.md#run-006-r7-muon +parity: + samples: 600 + cer_delta: 0.1187 +sha256: + decoder-kv.onnx: 61ef645345997eb5b9746ac77b39363228e5d0873093e6ebaca9a5903d3ac846 + decoder.onnx: 56d81cc644e2c27b56f5efd07f3e6bac0244b74e1383782b13e2d615fbcd43c6 + encoder.onnx: 1eab5f6d418bd718e0628e0cee3124143dd7f277e5b14669a8df76744d12c5ce