OPUS-MT Dashboard

Language pair: [eng-fra] all languages
Models: [all models] [OPUS-MT] [external] [compare]
Selected: Tatoeba-MT-models/zle-bat/opusTCv20210807_transformer-align_2022-03-17
Benchmark: all benchmarks [download]
Evaluation metric: bleu [spbleu][chrf][chrf++][comet]
Chart Type: [barchart] [heatmap]

ID	Language	Benchmark	Output	bleu
0	rus-lav	flores101-devtest	show	20.4
1	rus-lav	ntrex128	show	13.7
2	rus-lav	tatoeba-test-v2020-07-28	show	56.8
3	rus-lav	tatoeba-test-v2021-03-30	show	56.4
4	rus-lav	tatoeba-test-v2021-08-07	show	56.8
5	rus-lit	flores101-devtest	show	20.7
6	rus-lit	flores200-devtest	show	20.7
7	rus-lit	ntrex128	show	13.7
8	rus-lit	tatoeba-test-v2020-07-28	show	49.8
9	rus-lit	tatoeba-test-v2021-03-30	show	48.8
10	rus-lit	tatoeba-test-v2021-08-07	show	47.3
11	rus-ltg	flores200-devtest	show	3.5
12	ukr-lav	flores101-devtest	show	21.3
13	ukr-lav	ntrex128	show	13.8
14	ukr-lit	flores101-devtest	show	20.6
15	ukr-lit	flores200-devtest	show	20.6
16	ukr-lit	ntrex128	show	14.8
17	ukr-ltg	flores200-devtest	show	3.8
			average	27.972