diff --git a/src/interscript/ml/registry.py b/src/interscript/ml/registry.py index 5f9185c..a7c0fb8 100644 --- a/src/interscript/ml/registry.py +++ b/src/interscript/ml/registry.py @@ -21,7 +21,7 @@ import yaml DEFAULT_INDEX_URL = ( - "https://github.com/interscript/interscript-models/releases/download/index-v8/models-index.yaml" + "https://github.com/interscript/interscript-models/releases/download/index-v9/models-index.yaml" ) # primary env vars; the SECRYST_* names are deprecated aliases (kept one transition period) ENV_INDEX = "INTERSCRIPT_ML_INDEX" diff --git a/src/interscript/ml/thai_hybrid.py b/src/interscript/ml/thai_hybrid.py index bdd0a6d..377329b 100644 --- a/src/interscript/ml/thai_hybrid.py +++ b/src/interscript/ml/thai_hybrid.py @@ -3,6 +3,9 @@ The FastThaiG2P response: dictionary hits decode at dict speed, out-of-vocabulary runs go to the neural model as whole spans. Maximal- munch longest-match segmentation; spaces/punct pass through untouched. + +Canonical lexicon artifact (CC BY-SA 3.0, attribution Wiktionary/Kaikki): +https://github.com/interscript/interscript-models/releases/download/tha-lexicon-kaikki-1.0/tha-lexicon-kaikki.json """ from __future__ import annotations