ruMTEB v1.1 results

#1
by BorisTM - opened
Task Score
GeoreviewClassification 21.8066
HeadlineClassification 21.8848
InappropriatenessClassification 50.1855
KinopoiskClassification 34.2933
MassiveIntentClassification 12.2657
MassiveScenarioClassification 17.2162
RuReviewsClassification 37.1680
RuSciBenchGRNTIClassification 4.9414
RuSciBenchOECDClassification 4.3750
GeoreviewClusteringP2P 19.9622
RuSciBenchGRNTIClusteringP2P 7.8454
RuSciBenchOECDClusteringP2P 8.3841
CEDRClassification 32.0138
SensitiveTopicsClassification 17.7783
TERRa 49.6673
MIRACLReranking 14.3370
RuBQReranking 26.0150
MIRACLRetrievalHardNegatives.v2 1.9620
RiaNewsRetrievalHardNegatives.v2 2.8130
RuBQRetrieval 0.4280
RUParaPhraserSTS 29.6974
STS22 −3.9589
RuSTSBenchmarkSTS 37.2306
Mean(Task) 19.4918
Mean(TaskType) 21.7442

Sign up or log in to comment