OPUS-MT Dashboard

blue = OPUS-MT / Tatoeba-MT models, grey = external models, purple = user-contributed
render chart with [gd] [plotly]

ID	Benchmark (chrf++)	Output	OPUS-MT	chrf++	external	chrf++	Diff
0	flores101-devtest	compare	fra-sla/opus...2022-09-15	0.4938	facebook/nll...illed-1.3B	0.47409	0.020
1	flores200-devtest	compare	deu+eng+fra+...2024-05-30	0.51175	facebook/nll...illed-1.3B	0.47409	0.038
2	multi30k_test_2016_flickr	compare	deu+eng+fra+...2024-05-30	0.53788	facebook/nll...illed-1.3B	0.48645	0.051
3	multi30k_test_2018_flickr	compare	deu+eng+fra+...2024-05-30	0.50387	facebook/nll...illed-1.3B	0.45825	0.046
4	newssyscomb2009	compare	deu+eng+fra+...2024-05-30	0.49434	facebook/nllb-200-3.3B	0.4625	0.032
5	newstest2008	compare	deu+eng+fra+...2024-05-30	0.47745	facebook/nllb-200-1.3B	0.45011	0.027
6	newstest2009	compare	deu+eng+fra+...2024-05-30	0.48031	facebook/nllb-200-3.3B	0.45178	0.029
7	newstest2010	compare	deu+eng+fra+...2024-05-30	0.48458	facebook/nll...illed-1.3B	0.45612	0.028
8	newstest2011	compare	deu+eng+fra+...2024-05-30	0.48576	facebook/nllb-200-3.3B	0.45845	0.027
9	newstest2012	compare	deu+eng+fra+...2024-05-30	0.46958	facebook/nll...illed-1.3B	0.44295	0.027
10	newstest2013	compare	deu+eng+fra+...2024-05-30	0.48302	facebook/nll...illed-1.3B	0.45303	0.030
11	ntrex128	compare	deu+eng+fra+...2024-05-30	0.4739	facebook/nll...illed-1.3B	0.45229	0.022
12	tatoeba-test-v2020-07-28	compare	fra-ces/opus-2021-02-18	0.66612	facebook/nll...illed-1.3B	0.64791	0.018
13	tatoeba-test-v2021-03-30	compare	fra-ces/opus-2021-02-18	0.66752	facebook/nll...illed-1.3B	0.64775	0.020
14	tatoeba-test-v2021-08-07	compare	fra-ces/opus-2021-02-18	0.66733	facebook/nll...illed-1.3B	0.64784	0.019
		average		0.526		0.498	0.029