BEER 2.0 is a trained machine translation evaluation metric with high correlation with human judgment both on sentence and corpus level. For the papers that describe BEER in more detail look at the this References section.

 Download from GitHub

Implementation of domain adaptation algorithm based on the algorithm described in "Latent Domain Translation Models in Mix-of-Domains Haystack"

BEER is trained a machine translation evaluation metric which puts special attention on the sentence level evaluation. For details look at the paper BEER: BEtter Evaluation as Ranking (bibtex). Its features include:

• Full evaluation metric (published in BEER: BEtter Evaluation as Ranking and Fitting Sentence Level Translation Evaluation with Many Dense Features)
• Reordering evaluation metric (published in Evaluating Word Order Recursively over Permutation-Forests)
• Support for statistical testing trough multeval
• Permutation parser and visualizer

 Download from GitHub