Horizon-Labs commited on
Commit
01adcc6
·
verified ·
1 Parent(s): a3f01e8

Baselines: + gravitee distilbert, xlmr-large v2

Browse files
Files changed (1) hide show
  1. eval/baselines.json +192 -0
eval/baselines.json CHANGED
@@ -707,5 +707,197 @@
707
  "_civil_toxicity_auc": 0.8433979621843186,
708
  "_tdx_mean_auc": 0.6983786666666668,
709
  "_tdx_mean_f1": 0.2772211035293649
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
710
  }
711
  }
 
707
  "_civil_toxicity_auc": 0.8433979621843186,
708
  "_tdx_mean_auc": 0.6983786666666668,
709
  "_tdx_mean_f1": 0.2772211035293649
710
+ },
711
+ "gravitee-io/distilbert-multilingual-toxicity-classifier": {
712
+ "labels": [
713
+ "not-toxic",
714
+ "toxic"
715
+ ],
716
+ "mapped": [
717
+ "toxicity"
718
+ ],
719
+ "civil_test": {
720
+ "n": 20000,
721
+ "auc": {
722
+ "toxicity": 0.8089855626873078
723
+ },
724
+ "mean_auc": 0.8089855626873078,
725
+ "toxicity_auc": 0.8089855626873078
726
+ },
727
+ "tdx_am": {
728
+ "n": 1000,
729
+ "auc": 0.647036,
730
+ "f1_05": 0.684044233807267
731
+ },
732
+ "tdx_ar": {
733
+ "n": 1000,
734
+ "auc": 0.91521,
735
+ "f1_05": 0.8431372549019608
736
+ },
737
+ "tdx_de": {
738
+ "n": 1000,
739
+ "auc": 0.981946,
740
+ "f1_05": 0.9219562955254943
741
+ },
742
+ "tdx_en": {
743
+ "n": 1000,
744
+ "auc": 0.997914,
745
+ "f1_05": 0.9830508474576272
746
+ },
747
+ "tdx_es": {
748
+ "n": 1000,
749
+ "auc": 0.981574,
750
+ "f1_05": 0.9387755102040817
751
+ },
752
+ "tdx_fr": {
753
+ "n": 1000,
754
+ "auc": 0.99783,
755
+ "f1_05": 0.9810568295114656
756
+ },
757
+ "tdx_he": {
758
+ "n": 1000,
759
+ "auc": 0.913612,
760
+ "f1_05": 0.8340425531914893
761
+ },
762
+ "tdx_hi": {
763
+ "n": 1000,
764
+ "auc": 0.988272,
765
+ "f1_05": 0.955193482688391
766
+ },
767
+ "tdx_hin": {
768
+ "n": 1000,
769
+ "auc": 0.902658,
770
+ "f1_05": 0.8339622641509434
771
+ },
772
+ "tdx_it": {
773
+ "n": 1000,
774
+ "auc": 0.958374,
775
+ "f1_05": 0.8951841359773371
776
+ },
777
+ "tdx_ja": {
778
+ "n": 1000,
779
+ "auc": 0.975844,
780
+ "f1_05": 0.9194915254237288
781
+ },
782
+ "tdx_ru": {
783
+ "n": 1000,
784
+ "auc": 0.99792,
785
+ "f1_05": 0.9830845771144279
786
+ },
787
+ "tdx_tt": {
788
+ "n": 1000,
789
+ "auc": 0.991056,
790
+ "f1_05": 0.9592850049652433
791
+ },
792
+ "tdx_uk": {
793
+ "n": 1000,
794
+ "auc": 0.986376,
795
+ "f1_05": 0.9416498993963782
796
+ },
797
+ "tdx_zh": {
798
+ "n": 1000,
799
+ "auc": 0.87896,
800
+ "f1_05": 0.7629310344827587
801
+ },
802
+ "_civil_mean_auc": 0.8089855626873078,
803
+ "_civil_toxicity_auc": 0.8089855626873078,
804
+ "_tdx_mean_auc": 0.9409721333333334,
805
+ "_tdx_mean_f1": 0.8957896965865728
806
+ },
807
+ "textdetox/xlmr-large-toxicity-classifier-v2": {
808
+ "labels": [
809
+ "label_0",
810
+ "label_1"
811
+ ],
812
+ "mapped": [
813
+ "toxicity"
814
+ ],
815
+ "civil_test": {
816
+ "n": 20000,
817
+ "auc": {
818
+ "toxicity": 0.8947823432223221
819
+ },
820
+ "mean_auc": 0.8947823432223221,
821
+ "toxicity_auc": 0.8947823432223221
822
+ },
823
+ "tdx_am": {
824
+ "n": 1000,
825
+ "auc": 0.975296,
826
+ "f1_05": 0.9230769230769231
827
+ },
828
+ "tdx_ar": {
829
+ "n": 1000,
830
+ "auc": 0.98156,
831
+ "f1_05": 0.9408224674022067
832
+ },
833
+ "tdx_de": {
834
+ "n": 1000,
835
+ "auc": 0.994426,
836
+ "f1_05": 0.981018981018981
837
+ },
838
+ "tdx_en": {
839
+ "n": 1000,
840
+ "auc": 0.999892,
841
+ "f1_05": 0.9959919839679359
842
+ },
843
+ "tdx_es": {
844
+ "n": 1000,
845
+ "auc": 0.999264,
846
+ "f1_05": 0.9890329012961117
847
+ },
848
+ "tdx_fr": {
849
+ "n": 1000,
850
+ "auc": 0.99936,
851
+ "f1_05": 0.998003992015968
852
+ },
853
+ "tdx_he": {
854
+ "n": 1000,
855
+ "auc": 0.96858,
856
+ "f1_05": 0.9126016260162602
857
+ },
858
+ "tdx_hi": {
859
+ "n": 1000,
860
+ "auc": 0.993964,
861
+ "f1_05": 0.9779559118236473
862
+ },
863
+ "tdx_hin": {
864
+ "n": 1000,
865
+ "auc": 0.965486,
866
+ "f1_05": 0.9177403369672944
867
+ },
868
+ "tdx_it": {
869
+ "n": 1000,
870
+ "auc": 0.990784,
871
+ "f1_05": 0.9690927218344965
872
+ },
873
+ "tdx_ja": {
874
+ "n": 1000,
875
+ "auc": 0.984078,
876
+ "f1_05": 0.9562563580874873
877
+ },
878
+ "tdx_ru": {
879
+ "n": 1000,
880
+ "auc": 0.999964,
881
+ "f1_05": 0.996
882
+ },
883
+ "tdx_tt": {
884
+ "n": 1000,
885
+ "auc": 0.994964,
886
+ "f1_05": 0.9726443768996961
887
+ },
888
+ "tdx_uk": {
889
+ "n": 1000,
890
+ "auc": 0.996866,
891
+ "f1_05": 0.984984984984985
892
+ },
893
+ "tdx_zh": {
894
+ "n": 1000,
895
+ "auc": 0.958476,
896
+ "f1_05": 0.8995024875621891
897
+ },
898
+ "_civil_mean_auc": 0.8947823432223221,
899
+ "_civil_toxicity_auc": 0.8947823432223221,
900
+ "_tdx_mean_auc": 0.986864,
901
+ "_tdx_mean_f1": 0.9609817368636122
902
  }
903
  }