OPUS-MT Dashboard

Language pair: [eng-fra] all languages
Models: [all models] [OPUS-MT] [external] [compare]
Benchmark: [all benchmarks] [average score] newstest2012
Evaluation metric: [bleu][spbleu][chrf] chrf++ [comet]
Highlighting colors: [model score] [model type] [model size]

Table of chrf++ score differences between the best OPUS-MT model and the best external modelfor language pairs with an existing external model (other cells are left blank):

	ces	deu	eng	fra	rus	spa
ces	ces	1.8	1.1	1.2	0.4	0.9
deu	3.4	deu	-2.1	1.1	0.2	1.4
eng	1.8	-3.6	eng	0.1	1.4	0.3
fra	2.7	2.6	0.5	fra	0.2	1.1
rus	2.7	1.8	-0.1	0.8	rus	0.9
spa	1.5	2.7	0.1	0.0	0.3	spa