COMET (Crosslingual Optimized Metric for Evaluation of Translation) is a new neural framework for training multilingual machine translation (MT) evaluation models. COMET predicts human judgments of MT quality. This “ready to use” trained COMET model is available as open-source to benefit the wider MT R&D community.
https://resources.unbabel.com/comet-unbabel-research
press: https://unbabel.com/news/Unbabel-Releases-New-Framework-and-Metric-for-Automated-Machine-Translation-Evaluation/
