84.4 English (US) english_mgsm
english math
exact_match: 84.4sample_len: 250.0
exact_match: 84.4sample_len: 250.0
82.4 English (US) english_gsm8k
english math
exact_match: 82.4sample_len: 1319.0
exact_match: 82.4sample_len: 1319.0
79.6 Portuguese (Portugal) portuguese_hatebr
portuguese classification
f1_macro: 79.6sample_len: 1400.0
f1_macro: 79.6sample_len: 1400.0
77.7 English (US) english_belebele
english mcq
f1_macro: 77.7sample_len: 900.0
f1_macro: 77.7sample_len: 900.0
77.0 Arabic (Saudi Arabia) arabic_sib200
arabic classification
f1_macro: 77.0sample_len: 204.0
f1_macro: 77.0sample_len: 204.0
76.0 French (France) french_sib200
french classification
f1_macro: 76.0sample_len: 204.0
f1_macro: 76.0sample_len: 204.0
72.2 Ukrainian (Ukraine) ukrainian_sib200
ukrainian classification
f1_macro: 72.2sample_len: 204.0
f1_macro: 72.2sample_len: 204.0
70.5 Urdu (Pakistan) urdu_uquad
urdu qa
llm_judge_score: 70.5sample_len: 139.0
llm_judge_score: 70.5sample_len: 139.0
70.4 French (France) french_belebele
french mcq
f1_macro: 70.4sample_len: 900.0
f1_macro: 70.4sample_len: 900.0
70.2 Spanish (Spain) spanish_belebele
spanish mcq
f1_macro: 70.2sample_len: 900.0
f1_macro: 70.2sample_len: 900.0
69.6 English (US) ifeval
ifeval
inst_level_loose_acc: 76.3inst_level_strict_acc: 72.3prompt_level_loose_acc: 66.7prompt_level_strict_acc: 63.0sample_len: 541.0
inst_level_loose_acc: 76.3inst_level_strict_acc: 72.3prompt_level_loose_acc: 66.7prompt_level_strict_acc: 63.0sample_len: 541.0
66.8 French (France) french_mgsm
french math
exact_match: 66.8sample_len: 250.0
exact_match: 66.8sample_len: 250.0
61.8 Ukrainian (Ukraine) ukrainian_belebele
ukrainian mcq
f1_macro: 61.8sample_len: 900.0
f1_macro: 61.8sample_len: 900.0
58.9 French (France) french_fquad
french qa
exact_match: 47.0f1: 70.7sample_len: 400.0
exact_match: 47.0f1: 70.7sample_len: 400.0
56.1 Urdu (Pakistan) urdu_factcheckbench
urdu claim
f1_macro: 56.1sample_len: 387.0
f1_macro: 56.1sample_len: 387.0
55.1 Spanish (Spain) spanish_xquad_es
spanish xquad es
exact_match: 46.1f1: 64.1sample_len: 1190.0
exact_match: 46.1f1: 64.1sample_len: 1190.0
55.0 Urdu (Pakistan) urdu_fake_news
urdu classification
f1_macro: 55.0sample_len: 300.0
f1_macro: 55.0sample_len: 300.0
53.9 Portuguese (Portugal) portuguese_hate_speech
portuguese classification
f1_macro: 53.9sample_len: 851.0
f1_macro: 53.9sample_len: 851.0
51.3 Arabic (Saudi Arabia) arabic_belebele
arabic mcq
f1_macro: 51.3sample_len: 900.0
f1_macro: 51.3sample_len: 900.0
51.2 Arabic (Saudi Arabia) arabic_tydiqa
arabic qa
exact_match: 39.1f1: 63.4sample_len: 921.0
exact_match: 39.1f1: 63.4sample_len: 921.0
51.2 Portuguese (Portugal) portuguese_tweetsentbr
portuguese classification
f1_macro: 51.2sample_len: 2010.0
f1_macro: 51.2sample_len: 2010.0
49.0 Spanish (Spain) spanish_global_mmlu
spanish mcq
f1_macro: 49.0sample_len: 400.0
f1_macro: 49.0sample_len: 400.0
48.9 Albanian (Albania) albanian_sib200
albanian classification
f1_macro: 48.9sample_len: 204.0
f1_macro: 48.9sample_len: 204.0
48.1 Arabic (Saudi Arabia) arabic_aratrust
arabic mcq
f1_macro: 48.1sample_len: 522.0
f1_macro: 48.1sample_len: 522.0
48.0 Urdu (Pakistan) urdu_bingcheck
urdu claim
f1_macro: 48.0sample_len: 102.0
f1_macro: 48.0sample_len: 102.0
47.0 Urdu (Pakistan) urdu_facttool_qa
urdu claim
f1_macro: 47.0sample_len: 160.0
f1_macro: 47.0sample_len: 160.0
44.8 Swahili (Tanzania) swahili_sib200
swahili classification
f1_macro: 44.8sample_len: 204.0
f1_macro: 44.8sample_len: 204.0
44.7 Ukrainian (Ukraine) ukrainian_squad
ukrainian qa
exact_match: 33.6f1: 55.9sample_len: 3812.0
exact_match: 33.6f1: 55.9sample_len: 3812.0
44.3 French (France) french_mmmlu
french mcq
f1_macro: 44.3sample_len: 14042.0
f1_macro: 44.3sample_len: 14042.0
44.2 English (US) english_mmlu_pro
english mmlu pro
exact_match: 44.2sample_len: 2100.0
exact_match: 44.2sample_len: 2100.0
39.2 Ukrainian (Ukraine) ukrainian_global_mmlu
ukrainian mcq
f1_macro: 39.2sample_len: 2850.0
f1_macro: 39.2sample_len: 2850.0
37.5 Albanian (Albania) albanian_belebele
albanian mcq
f1_macro: 37.5sample_len: 900.0
f1_macro: 37.5sample_len: 900.0
37.2 Yoruba (Nigeria) yoruba_sib200
yoruba classification
f1_macro: 37.2sample_len: 204.0
f1_macro: 37.2sample_len: 204.0
34.1 Ukrainian (Ukraine) ukrainian_polywrite
ukrainian open generation
open_quality_score: 34.1sample_len: 154.0
open_quality_score: 34.1sample_len: 154.0
34.1 Arabic (Saudi Arabia) arabic_mmlu
arabic mcq
f1_macro: 34.1sample_len: 14316.0
f1_macro: 34.1sample_len: 14316.0
32.6 Albanian (Albania) albanian_global_mmlu
albanian mcq
f1_macro: 32.6sample_len: 400.0
f1_macro: 32.6sample_len: 400.0
32.2 Hausa (Nigeria) hausa_sib200
hausa classification
f1_macro: 32.2sample_len: 204.0
f1_macro: 32.2sample_len: 204.0
28.9 Yoruba (Nigeria) yoruba_naijasenti
yoruba sentiment
f1_macro: 28.9sample_len: 4515.0
f1_macro: 28.9sample_len: 4515.0
28.9 Yoruba (Nigeria) yoruba_afrimmlu
yoruba mcq
f1_macro: 28.9sample_len: 500.0
f1_macro: 28.9sample_len: 500.0
27.3 Igbo (Nigeria) igbo_sib200
igbo classification
f1_macro: 27.3sample_len: 204.0
f1_macro: 27.3sample_len: 204.0
26.6 Igbo (Nigeria) igbo_afrimmlu
igbo mcq
f1_macro: 26.6sample_len: 500.0
f1_macro: 26.6sample_len: 500.0
26.2 Igbo (Nigeria) igbo_belebele
igbo mcq
f1_macro: 26.2sample_len: 900.0
f1_macro: 26.2sample_len: 900.0
26.0 Albanian (Albania) albanian_aya
albanian open generation
llm_judge_score: 26.0sample_len: 200.0
llm_judge_score: 26.0sample_len: 200.0
24.7 Hausa (Nigeria) hausa_afrimmlu
hausa mcq
f1_macro: 24.7sample_len: 500.0
f1_macro: 24.7sample_len: 500.0
24.7 Yoruba (Nigeria) yoruba_belebele
yoruba mcq
f1_macro: 24.7sample_len: 900.0
f1_macro: 24.7sample_len: 900.0
24.2 Hausa (Nigeria) hausa_belebele
hausa mcq
f1_macro: 24.2sample_len: 900.0
f1_macro: 24.2sample_len: 900.0
23.1 Portuguese (Portugal) portuguese_oab_exams
portuguese mcq
exact_match: 23.1sample_len: 2210.0
exact_match: 23.1sample_len: 2210.0
22.7 Igbo (Nigeria) igbo_naijasenti
igbo sentiment
f1_macro: 22.7sample_len: 3682.0
f1_macro: 22.7sample_len: 3682.0
22.4 Urdu (Pakistan) urdu_emotion_class
urdu classification
f1_macro: 22.4sample_len: 200.0
f1_macro: 22.4sample_len: 200.0
21.9 Ukrainian (Ukraine) ukrainian_zno
ukrainian mcq
f1_macro: 21.9sample_len: 751.0
f1_macro: 21.9sample_len: 751.0
21.0 Hausa (Nigeria) hausa_naijasenti
hausa sentiment
f1_macro: 21.0sample_len: 5303.0
f1_macro: 21.0sample_len: 5303.0
20.3 Portuguese (Portugal) portuguese_enem
portuguese mcq
exact_match: 20.3sample_len: 1432.0
exact_match: 20.3sample_len: 1432.0
19.2 Portuguese (Portugal) portuguese_bluex
portuguese mcq
exact_match: 19.2sample_len: 724.0
exact_match: 19.2sample_len: 724.0
17.6 Hausa (Nigeria) hausa_afrixnli
hausa nli
f1_macro: 17.6sample_len: 600.0
f1_macro: 17.6sample_len: 600.0
16.7 Igbo (Nigeria) igbo_afrixnli
igbo nli
f1_macro: 16.7sample_len: 600.0
f1_macro: 16.7sample_len: 600.0
16.7 Yoruba (Nigeria) yoruba_afrixnli
yoruba nli
f1_macro: 16.7sample_len: 600.0
f1_macro: 16.7sample_len: 600.0
16.5 Swahili (Tanzania) swahili_afrixnli
swahili nli
f1_macro: 16.5sample_len: 600.0
f1_macro: 16.5sample_len: 600.0
9.6 Swahili (Tanzania) swahili_afrimgsm
swahili afrimgsm
exact_match: 9.6sample_len: 250.0
exact_match: 9.6sample_len: 250.0
9.5 Albanian (Albania) albanian_polywrite
albanian open generation
open_quality_score: 9.5sample_len: 155.0
open_quality_score: 9.5sample_len: 155.0
6.0 Yoruba (Nigeria) yoruba_afrimgsm
yoruba afrimgsm
exact_match: 6.0sample_len: 250.0
exact_match: 6.0sample_len: 250.0
4.0 Hausa (Nigeria) hausa_afrimgsm
hausa afrimgsm
exact_match: 4.0sample_len: 250.0
exact_match: 4.0sample_len: 250.0
3.7 Urdu (Pakistan) urdu_freshqa
urdu qa
llm_judge_score: 3.7sample_len: 323.0
llm_judge_score: 3.7sample_len: 323.0
2.4 Igbo (Nigeria) igbo_afrimgsm
igbo afrimgsm
exact_match: 2.4sample_len: 250.0
exact_match: 2.4sample_len: 250.0
1.5 Igbo (Nigeria) igbo_afriqa
igbo qa
exact_match: 0.5f1: 2.5sample_len: 409.0
exact_match: 0.5f1: 2.5sample_len: 409.0
1.3 Yoruba (Nigeria) yoruba_afriqa
yoruba qa
exact_match: 0.3f1: 2.2sample_len: 332.0
exact_match: 0.3f1: 2.2sample_len: 332.0
0.5 Urdu (Pakistan) urdu_simpleqa
urdu qa
llm_judge_score: 0.5sample_len: 200.0
llm_judge_score: 0.5sample_len: 200.0
0.4 Hausa (Nigeria) hausa_afriqa
hausa qa
exact_match: 0.0f1: 0.9sample_len: 300.0
exact_match: 0.0f1: 0.9sample_len: 300.0