OPUS-MT Dashboard

Language pair: [eng-fra] all languages
Models: [all models] [OPUS-MT] [external] [compare]
Selected: Tatoeba-MT-models/zls-deu/opusTCv20210807_transformer-big_2022-07-26
Benchmark: all benchmarks [download]
Evaluation metric: bleu [spbleu][chrf][chrf++][comet]
Chart Type: [barchart] [heatmap]

ID	Language	Benchmark	Output	bleu
0	bul-deu	flores101-devtest	show	28.4
1	bul-deu	flores200-devtest	show	28.3
2	bul-deu	ntrex128	show	23.4
3	bul-deu	tatoeba-test-v2020-07-28	show	54.4
4	bul-deu	tatoeba-test-v2021-03-30	show	54.3
5	bul-deu	tatoeba-test-v2021-08-07	show	54.5
6	hbs-deu	tatoeba-test-v2021-08-07	show	54.8
7	hrv-deu	flores101-devtest	show	27.4
8	hrv-deu	flores200-devtest	show	27.4
9	hrv-deu	ntrex128	show	25.5
10	hrv-deu	tatoeba-test-v2021-08-07	show	53.1
11	mkd-deu	flores101-devtest	show	29.3
12	mkd-deu	flores200-devtest	show	29.3
13	mkd-deu	ntrex128	show	23.9
14	slv-deu	flores101-devtest	show	26.7
15	slv-deu	flores200-devtest	show	26.7
16	slv-deu	ntrex128	show	23.9
17	slv-deu	tatoeba-test-v2020-07-28	show	23.1
18	slv-deu	tatoeba-test-v2021-03-30	show	22.9
19	slv-deu	tatoeba-test-v2021-08-07	show	21.1
20	srp_Cyrl-deu	flores101-devtest	show	30.7
21	srp_Cyrl-deu	flores200-devtest	show	30.7
22	srp_Cyrl-deu	ntrex128	show	20.4
23	srp_Latn-deu	ntrex128	show	23.1
24	srp_Latn-deu	tatoeba-test-v2021-08-07	show	56
			average	32.772