diff --git a/conformance/tha-g2p-lex-vectors.json b/conformance/tha-g2p-lex-vectors.json new file mode 100644 index 00000000..ad1fb3c7 --- /dev/null +++ b/conformance/tha-g2p-lex-vectors.json @@ -0,0 +1,58 @@ +{ + "model": "tha-g2p-lex-1.0", + "index_version": 12, + "decode": { + "mode": "single-row-greedy" + }, + "source": "run-080-thai-lex-v6a oepoch-3 (clean-split PER 19.29)", + "vectors": [ + { + "src": "ก", + "out": "k ɔː ˧" + }, + { + "src": "ก กา", + "out": "k a ˨˩ . k aː ˧" + }, + { + "src": "ก ข ไม่กระดิกหู", + "out": "k o k̚ ˨˩ . kʰ a ˨˩ . k a ˧ . k uː ˧" + }, + { + "src": "ก หัน", + "out": "k a ˨˩ . h a n ˩˩˦" + }, + { + "src": "กก", + "out": "k o k̚ ˨˩" + }, + { + "src": "กกช้าง", + "out": "k o k̚ ˨˩ . t͡ɕʰ aː ŋ ˦˥" + }, + { + "src": "กกธูป", + "out": "k o k̚ ˨˩ . kʰ uː p̚ ˥˩" + }, + { + "src": "กกหู", + "out": "k o k̚ ˨˩ . h uː ˩˩˦" + }, + { + "src": "กกุธภัณฑ์", + "out": "k a ˨˩ . k u t̚ ˨˩ . pʰ a n ˧" + }, + { + "src": "กกเสา", + "out": "k o k̚ ˨˩ . s a w ˩˩˦" + }, + { + "src": "กง", + "out": "k o ŋ ˧" + }, + { + "src": "กงกอน", + "out": "k o ŋ ˧ . k ɔː n ˧" + } + ] +} \ No newline at end of file diff --git a/models.yaml b/models.yaml index bc2ccbb9..d1a7d18d 100644 --- a/models.yaml +++ b/models.yaml @@ -1,5 +1,27 @@ version: 1 models: + tha-g2p-lex-1.0: + task: g2p + precision: fp32 + url: https://github.com/interscript/interscript-models/releases/download/UNPUBLISHED/tha-g2p-lex-1.0.zip + sha256: 534bc68960517da36c38d9ef79f1178722751e36e72a3313c0e21b6f66e95198 + size: 1111939721 + metrics: + - name: per + value: 19.29 + protocol: phoneme-token PER, clean v1 test split (3,210 rows), single-row greedy + source: interscript/interscript-train run-080-thai-lex-v6a eval (run-080-thai-lex-v6a/eval-verdict.json lineage; WO67) + parity: + samples: 0 + note: >- + PENDING — intermediate comparison exact (encoder 0.000000, step-1 logits + 0.0000, greedy ids identical); the legacy harness contradicts the + proven-exact artifact (CER 1.02 at BS=1) and is under investigation — + gate NOT claimed. + license: BSD-3-Clause + trained_from: >- + ByT5-small; corpora/tha-phonemic-v5.yaml (92,669 length-filtered + pairs: orst+thai_words+volubilis+wiki-agree); 4 epochs BS4; WO67 khm-latn-1.0: task: translit scripts: