OPUS-MT Dashboard

Language pair: [eng-fra] all languages
Models: [all models] [OPUS-MT] [external] [compare]
Selected: Tatoeba-MT-models/gmq-deu/opusTCv20210807_transformer-big_2022-07-22
Benchmark: all benchmarks [download]
Evaluation metric: bleu [spbleu][chrf][chrf++][comet]
Chart Type: [barchart] [heatmap]

ID	Language	Benchmark	Output	bleu
0	dan-deu	flores101-devtest	show	33.8
1	dan-deu	flores200-devtest	show	33.9
2	dan-deu	ntrex128	show	26.2
3	dan-deu	tatoeba-test-v2020-07-28	show	58.8
4	dan-deu	tatoeba-test-v2021-03-30	show	58.8
5	dan-deu	tatoeba-test-v2021-08-07	show	58.7
6	isl-deu	flores101-devtest	show	23.2
7	isl-deu	flores200-devtest	show	23.2
8	isl-deu	ntrex128	show	21.4
9	isl-deu	tatoeba-test-v2020-07-28	show	53
10	isl-deu	tatoeba-test-v2021-03-30	show	53.2
11	isl-deu	tatoeba-test-v2021-08-07	show	52.9
12	nno-deu	flores200-devtest	show	28
13	nno-deu	ntrex128	show	23.9
14	nob-deu	flores101-devtest	show	28
15	nob-deu	flores200-devtest	show	28
16	nob-deu	ntrex128	show	27.2
17	nob-deu	tatoeba-test-v2021-08-07	show	51.1
18	swe-deu	flores101-devtest	show	33.7
19	swe-deu	flores200-devtest	show	33.7
20	swe-deu	ntrex128	show	27.8
21	swe-deu	tatoeba-test-v2021-08-07	show	57.1
			average	37.982