OPUS-MT Dashboard

Language pair: [eng-fra] all languages
Models: [all models] [OPUS-MT] [external] [compare]
Selected: Tatoeba-MT-models/pqw-pqw/opus-2021-02-18
Benchmark: all benchmarks [download]
Evaluation metric: bleu [spbleu][chrf][chrf++][comet]
Chart Type: [barchart] [heatmap]

Selected Benchmark: [avg] [flores101-devtest] [flores200-devtest] [tatoeba-test-v2020-07-28] [tatoeba-test-v2021-03-30] [tatoeba-test-v2021-08-07]
Scores shown in percentage points

	ceb	cha	eng	ilo	ind	jav	pag	plt	sun	war	zsm
ceb	ceb		22.8	18.2	15.2	6.0	10.7	4.7	5.6	22.4
cha		cha	5.7
dtp			0.9
eng	16.0	3.3	eng	29.5	31.7	8.1	12.8	5.5	7.2	16.1
ilo	18.2		31.8	ilo	13.1	6.2	10.4	4.5	5.2	19.8
ind	12.0		32.2	11.7	ind	9.1	7.5	4.1	8.4	13.8	2.0
jav	8.1		10.2	7.4	13.1	jav	5.1	2.3	5.9	8.8
msa	12.7		31.7	11.8	22.1	7.5	7.9	4.4	7.0	14.7
pag	13.8		19.2	13.4	10.9	5.2	pag	3.6	4.7	14.5
plt	10.4		18.2	10.6	11.2	3.8	7.1	plt	4.3	11.7
sun	7.5		9.9	7.1	12.0	5.6	5.1	2.4	sun	8.3
war	22.3		20.6	19.1	16.7	6.5	11.4	4.2	5.8	war
zsm	12.8		25.2	11.8	22.1	7.6	7.9	4.4	7.0	14.7	zsm

color:
score:	0	10	20	30	40	50	60	70	80	90	100