From 6c3928efc0abf6d50163c3292cb4ad4b75a2595b Mon Sep 17 00:00:00 2001 From: Ronald Tse Date: Fri, 30 Oct 2026 20:44:05 +0800 Subject: [PATCH] =?UTF-8?q?docs:=20WO68=20=E2=80=94=20B2=20EXPANSION=5FPAY?= =?UTF-8?q?S=20(40.25=20vs=2047.15=20CER);=20attempt=203=20=3D=20latn-conv?= =?UTF-8?q?ention=20run?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- TODO.sota/68-b7-khmer-lexicon.md | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/TODO.sota/68-b7-khmer-lexicon.md b/TODO.sota/68-b7-khmer-lexicon.md index db8af9f..ac3130c 100644 --- a/TODO.sota/68-b7-khmer-lexicon.md +++ b/TODO.sota/68-b7-khmer-lexicon.md @@ -145,3 +145,24 @@ student-vs-teacher on the same rows and surface. Gate: student CER < teacher CER on that split -> EXPANSION_PAYS. Note the translit test's anchor CER band is ~27.4 (run-081 reproduction); the teacher's own CER on the same rows is the number that matters. + + +**Stage B2 VERDICT: EXPANSION_PAYS** (2026-10-30, run-082d). Matched-surface +result on the translit romanization test split (895 rows, exact run-081 +protocol): student CER 40.25% / EM 52.4% vs the phonemic teacher +(khm-phonemic-1.0, same rows in-job) 47.15% / EM 44.6% — the 57K SBBIC +weak labels + 16K translit anchor beat the teacher by 6.9 CER points on +its own surface. Caveat carried forward: this compares against the +PHONEMIC teacher; the khm-latn lineage's own band is 27.4, and the +translit-surface score of a convention-blended model is not a lineage +benchmark. + +Attempt 3 (running): the convention-correct experiment — SBBIC labeled +with the SHIPPED khm-latn-1.0 artifact (run-082e-khm-latn-label, 24 +CPU processes on a100-large, rotating parts, sha256-verified Release +fetch; plain decoders are batch=1-traced but words are tiny) -> weak +labels in the translit convention -> train byt5-small (weak + translit +train anchor, exact run-081 split) -> gate: student CER < teacher CER +on the identical 895 test rows (the lineage comparison vs 27.4). +Verdict script: train_khm_expansion3.py (stage-b3 outputs under +bucket khmer-latn/).