Github user mjpost commented on the issue:
https://github.com/apache/incubator-joshua/pull/25
Okay, this checks out, producing virtually the same scores in two settings:
(1) re-decoding a test set with an existing model (built with the
non-state-minimizing LM), but swapping in the StateMinimizingLanguageModel, and
(2) te-tuning with the StateMinimizingLanguageModel (allowing it to learn its
own weight), with all other items the same.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
