release: tha-g2p-small-1.0 (int8, parity cer_delta 0.0803pp on 1219 samples)

Author: Ronald Tse

models.yaml

@@ -93,3 +93,16 @@ models:
       - {name: per_student, value: 9.19, source: interscript/ml-models docs/RESULTS.md#tha-g2p-base-1.0}
     parity: {samples: 1219, cer_delta: 0.0}
     license: BSD-3-Clause
+  tha-g2p-small-1.0:
+    task: g2p
+    scripts: [Thai, IPA]
+    precision: int8
+    filename: tha-g2p-small-1.0-int8.zip
+    url: https://github.com/interscript/interscript-ml/releases/download/tha-g2p-small-1.0/tha-g2p-small-1.0-int8.zip
+    sha256: f2cfa93a42d98aabf08fdc82dfc6bd1574e21993f36558aa581fbe29e0588a94
+    size: 257468517
+    metrics:
+      - {name: per_teacher, value: 4.43, source: interscript/interscript-ml docs/RESULTS.md#tha-g2p-small-1.0}
+      - {name: per_student, value: 12.06, source: interscript/interscript-ml docs/RESULTS.md#tha-g2p-small-1.0}
+    parity: {samples: 1219, cer_delta: 0.0803}
+    license: BSD-3-Clause

models/tha-g2p-small/tha-g2p-small-1.0.metadata.yaml

@@ -6,25 +6,33 @@ target: IPA
 tokenizer: bytes
 opset: 14
 decoder: kv
-precision: fp32
+precision: int8
 license: BSD-3-Clause
-trained_from: >-
-  distilled from the recovered Thai umt5 teacher (transformers 5.14.1
-  re-finetune of B-K/umt5-thai-g2p-v2-0.5k on the 60K Kaikki+epitran
-  corpus; the secryst-saved umt5 artifacts are unusable — 5.15 dropped
-  the untied lm_head) via sequence-level KD; ByT5-small init
-  google/byt5-small; checkpoint
-  secryst-checkpoints:/secryst_thai_g2p_distill_small/run-002/best
+trained_from: 'sequence-level KD from the B-K/umt5-thai-g2p-v2-0.5k teacher (4.43%
+  PER on this harness; the secryst-saved umt5 artifacts are unusable — transformers
+  5.15 dropped the untied lm_head); 48,757 usable beam-4 teacher-generated labels
+  (51,321 raw, deduplicated, degenerate outputs filtered); ByT5-small init google/byt5-small;
+  checkpoint secryst-checkpoints:/secryst_thai_g2p_distill_small/run-003/best. Client
+  tier: capacity-limited at +7.63pp over the teacher — the +5pp gate is met by tha-g2p-base-1.0;
+  this is the smallest rung that does not collapse (docs/RESULTS.md frontier table,
+  ~300MB int8).'
 metrics:
-  - name: per_teacher
-    value: 2.32
-    protocol: >-
-      greedy decode; 1,219 Kaikki Thai test sentences; umt5 teacher,
-      sentencepiece tokenizer; secryst RESULTS.md protocol
-    source: secryst/docs/RESULTS.md#thai-g2p
-  - name: per_student
-    value: PLACEHOLDER_STUDENT_PER
-    protocol: >-
-      greedy decode; same 1,219 test sentences, same harness as the
-      teacher (interscript/ml-models src/gpu/modal_distill.py::evaluate_per)
-    source: interscript/ml-models release tha-g2p-small-1.0
+- name: per_teacher
+  value: 4.43
+  protocol: beam-4, corpus-level PER (total_ed/total_gold over chars of joined-piece
+    decode); 1,219 Kaikki Thai test sentences; B-K/ umt5-thai-g2p-v2-0.5k teacher;
+    exact match 95.57%
+  source: interscript/interscript-ml docs/RESULTS.md#tha-g2p-small-1.0
+- name: per_student
+  value: 12.06
+  protocol: beam-4, corpus-level PER, same harness as the teacher (src/gpu/ modal_distill.py::evaluate_per,
+    run-003 checkpoint re-measured 2026-08-22); exact match 87.94%; +7.63pp client-tier
+    shrink cost
+  source: interscript/interscript-ml docs/RESULTS.md#tha-g2p-small-1.0
+parity:
+  samples: 1219
+  cer_delta: 0.0803
+sha256:
+  decoder-kv.onnx: b8d67ecbec9f4a29beafd3e6d431422f084c703a35467d836e81bdb8ceff4cba
+  decoder.onnx: 81b3580d1d26f78e2aa1786218f50ebe970b23309386fa06bb73b0f27333c0a8
+  encoder.onnx: 53416ec4c14ede50d2a5b3cd4b0de73c2b7c62e744a7d48d60e12a81ee89244d