Files
Tucano2-qwen-1.5B-Think/evals.yaml

208 lines
10 KiB
YAML
Raw Permalink Normal View History

evaluations:
arc_challenge_poly_pt_acc: 0.37948717948717947
arc_challenge_poly_pt_acc_norm: 0.4282051282051282
arc_challenge_poly_pt_acc_norm_stderr: 0.014472341586181985
arc_challenge_poly_pt_acc_stderr: 0.014192754090886793
arc_challenge_poly_pt_alias: arc_challenge_poly_pt
assin2_rte_acc,all: 0.8705065359477124
assin2_rte_acc_stderr,all: 0.004798540986291823
assin2_rte_alias: assin2_rte
assin2_rte_f1_macro,all: 0.8691746625542952
assin2_rte_f1_macro_stderr,all: 0.004847278802663799
assin2_sts_alias: assin2_sts
assin2_sts_mse,all: 2.2781209150326798
assin2_sts_mse_stderr,all: N/A
assin2_sts_pearson,all: 0.47716809267162485
assin2_sts_pearson_stderr,all: 0.014407170423859306
assin_entailment_acc: 0.7325
assin_entailment_acc_stderr: 0.006999870502142285
assin_entailment_alias: assin_entailment
assin_paraphrase_acc: 0.73625
assin_paraphrase_acc_stderr: 0.006968401827607802
assin_paraphrase_alias: assin_paraphrase
belebele_por_Latn_acc: 0.6766666666666666
belebele_por_Latn_acc_norm: 0.6766666666666666
belebele_por_Latn_acc_norm_stderr: 0.015600294087844632
belebele_por_Latn_acc_stderr: 0.015600294087844632
belebele_por_Latn_alias: belebele_por_Latn
bluex_acc,all: 0.39221140472878996
bluex_acc,exam_id__UNICAMP_2018: 0.48148148148148145
bluex_acc,exam_id__UNICAMP_2019: 0.42
bluex_acc,exam_id__UNICAMP_2020: 0.41818181818181815
bluex_acc,exam_id__UNICAMP_2021_1: 0.4782608695652174
bluex_acc,exam_id__UNICAMP_2021_2: 0.29411764705882354
bluex_acc,exam_id__UNICAMP_2022: 0.5384615384615384
bluex_acc,exam_id__UNICAMP_2023: 0.4883720930232558
bluex_acc,exam_id__UNICAMP_2024: 0.4444444444444444
bluex_acc,exam_id__USP_2018: 0.24074074074074073
bluex_acc,exam_id__USP_2019: 0.35
bluex_acc,exam_id__USP_2020: 0.39285714285714285
bluex_acc,exam_id__USP_2021: 0.3076923076923077
bluex_acc,exam_id__USP_2022: 0.24489795918367346
bluex_acc,exam_id__USP_2023: 0.38636363636363635
bluex_acc,exam_id__USP_2024: 0.4634146341463415
bluex_acc_stderr,all: 0.010527552369043944
bluex_acc_stderr,exam_id__UNICAMP_2018: 0.03932520147192548
bluex_acc_stderr,exam_id__UNICAMP_2019: 0.040409592319981036
bluex_acc_stderr,exam_id__UNICAMP_2020: 0.038500996629012005
bluex_acc_stderr,exam_id__UNICAMP_2021_1: 0.04265317780472471
bluex_acc_stderr,exam_id__UNICAMP_2021_2: 0.0369309745845929
bluex_acc_stderr,exam_id__UNICAMP_2022: 0.04616267465694302
bluex_acc_stderr,exam_id__UNICAMP_2023: 0.04391771882582978
bluex_acc_stderr,exam_id__UNICAMP_2024: 0.04266417007528182
bluex_acc_stderr,exam_id__USP_2018: 0.03361325629183327
bluex_acc_stderr,exam_id__USP_2019: 0.04344929186946837
bluex_acc_stderr,exam_id__USP_2020: 0.037457169801895486
bluex_acc_stderr,exam_id__USP_2021: 0.03696695228727266
bluex_acc_stderr,exam_id__USP_2022: 0.03545185428690165
bluex_acc_stderr,exam_id__USP_2023: 0.042474562735645566
bluex_acc_stderr,exam_id__USP_2024: 0.04499354148452882
bluex_alias: bluex
calame_pt_acc: 0.11127167630057803
calame_pt_acc_stderr: 0.00690347530269077
calame_pt_alias: calame_pt
calame_pt_perplexity: 5965.407764074068
calame_pt_perplexity_stderr: 683.2371253201818
enem_challenge_acc,all: 0.3988803358992302
enem_challenge_acc,exam_id__2009: 0.46956521739130436
enem_challenge_acc,exam_id__2010: 0.5128205128205128
enem_challenge_acc,exam_id__2011: 0.46153846153846156
enem_challenge_acc,exam_id__2012: 0.4224137931034483
enem_challenge_acc,exam_id__2013: 0.3148148148148148
enem_challenge_acc,exam_id__2014: 0.41284403669724773
enem_challenge_acc,exam_id__2015: 0.40336134453781514
enem_challenge_acc,exam_id__2016: 0.36363636363636365
enem_challenge_acc,exam_id__2016_2: 0.4065040650406504
enem_challenge_acc,exam_id__2017: 0.39655172413793105
enem_challenge_acc,exam_id__2022: 0.3157894736842105
enem_challenge_acc,exam_id__2023: 0.32592592592592595
enem_challenge_acc_stderr,all: 0.00749197527112301
enem_challenge_acc_stderr,exam_id__2009: 0.02676517541031027
enem_challenge_acc_stderr,exam_id__2010: 0.02676561702634192
enem_challenge_acc_stderr,exam_id__2011: 0.026575632846671127
enem_challenge_acc_stderr,exam_id__2012: 0.02652807977680181
enem_challenge_acc_stderr,exam_id__2013: 0.02570106484445073
enem_challenge_acc_stderr,exam_id__2014: 0.02722456002129014
enem_challenge_acc_stderr,exam_id__2015: 0.025882945505526844
enem_challenge_acc_stderr,exam_id__2016: 0.025343746567122242
enem_challenge_acc_stderr,exam_id__2016_2: 0.025491304109387235
enem_challenge_acc_stderr,exam_id__2017: 0.0261642629559971
enem_challenge_acc_stderr,exam_id__2022: 0.02331831926729694
enem_challenge_acc_stderr,exam_id__2023: 0.023372974631758865
enem_challenge_alias: enem
faquad_nli_acc,all: 0.7892307692307692
faquad_nli_acc_stderr,all: 0.011279066985135663
faquad_nli_alias: faquad_nli
faquad_nli_f1_macro,all: 0.49256657036543183
faquad_nli_f1_macro_stderr,all: 0.01318624622535848
global_piqa_completions_por_latn_braz_acc: 0.75
global_piqa_completions_por_latn_braz_acc_bytes: 0.74
global_piqa_completions_por_latn_braz_acc_bytes_stderr: 0.0440844002276808
global_piqa_completions_por_latn_braz_acc_norm: 0.74
global_piqa_completions_por_latn_braz_acc_norm_stderr: 0.0440844002276808
global_piqa_completions_por_latn_braz_acc_stderr: 0.04351941398892446
global_piqa_completions_por_latn_braz_alias: global_piqa_completions_por_latn_braz
gsm8k_pt_alias: gsm8k_pt
gsm8k_pt_exact_match,flexible-extract: 0.228310502283105
gsm8k_pt_exact_match,strict-match: 0.0
gsm8k_pt_exact_match_stderr,flexible-extract: 0.011583822031121497
gsm8k_pt_exact_match_stderr,strict-match: 0.0
hatebr_offensive_acc,all: 0.7071428571428572
hatebr_offensive_acc_stderr,all: 0.008611736471385361
hatebr_offensive_alias: hatebr_offensive_binary
hatebr_offensive_f1_macro,all: 0.7019751844399732
hatebr_offensive_f1_macro_stderr,all: 0.008866946580220815
hellaswag_poly_pt_acc: 0.4334164048109221
hellaswag_poly_pt_acc_norm: 0.5494636471990465
hellaswag_poly_pt_acc_norm_stderr: 0.005179413791359488
hellaswag_poly_pt_acc_stderr: 0.005158588405358706
hellaswag_poly_pt_alias: hellaswag_poly_pt
humaneval_instruct_alias: humaneval_instruct
humaneval_instruct_pass@1,create_test: 0.0
humaneval_instruct_pass@1_stderr,create_test: 0.0
ifeval_pt_alias: ifeval_pt
ifeval_pt_inst_level_loose_acc: 0.4511627906976744
ifeval_pt_inst_level_loose_acc_stderr: N/A
ifeval_pt_inst_level_strict_acc: 0.3395348837209302
ifeval_pt_inst_level_strict_acc_stderr: N/A
ifeval_pt_prompt_level_loose_acc: 0.33666666666666667
ifeval_pt_prompt_level_loose_acc_stderr: 0.02732941756218688
ifeval_pt_prompt_level_strict_acc: 0.23333333333333334
ifeval_pt_prompt_level_strict_acc_stderr: 0.024459979523511425
lambada_poly_pt_acc: 0.267028915195032
lambada_poly_pt_acc_stderr: 0.00616360274242743
lambada_poly_pt_alias: lambada_poly_pt
lambada_poly_pt_perplexity: 214.12248648482947
lambada_poly_pt_perplexity_stderr: 13.012850955803207
mmlu_poly_pt_acc: 0.43297808465926146
mmlu_poly_pt_acc_stderr: 0.004292713120965234
mmlu_poly_pt_alias: mmlu_poly_pt
oab_exams_acc,all: 0.3425968109339408
oab_exams_acc,exam_id__2010-01: 0.3176470588235294
oab_exams_acc,exam_id__2010-02: 0.33
oab_exams_acc,exam_id__2011-03: 0.3333333333333333
oab_exams_acc,exam_id__2011-04: 0.3625
oab_exams_acc,exam_id__2011-05: 0.325
oab_exams_acc,exam_id__2012-06: 0.35
oab_exams_acc,exam_id__2012-06a: 0.325
oab_exams_acc,exam_id__2012-07: 0.3375
oab_exams_acc,exam_id__2012-08: 0.325
oab_exams_acc,exam_id__2012-09: 0.24675324675324675
oab_exams_acc,exam_id__2013-10: 0.325
oab_exams_acc,exam_id__2013-11: 0.2875
oab_exams_acc,exam_id__2013-12: 0.3375
oab_exams_acc,exam_id__2014-13: 0.3875
oab_exams_acc,exam_id__2014-14: 0.325
oab_exams_acc,exam_id__2014-15: 0.41025641025641024
oab_exams_acc,exam_id__2015-16: 0.3125
oab_exams_acc,exam_id__2015-17: 0.3974358974358974
oab_exams_acc,exam_id__2015-18: 0.375
oab_exams_acc,exam_id__2016-19: 0.41025641025641024
oab_exams_acc,exam_id__2016-20: 0.4375
oab_exams_acc,exam_id__2016-20a: 0.325
oab_exams_acc,exam_id__2016-21: 0.25
oab_exams_acc,exam_id__2017-22: 0.375
oab_exams_acc,exam_id__2017-23: 0.3125
oab_exams_acc,exam_id__2017-24: 0.4
oab_exams_acc,exam_id__2018-25: 0.3375
oab_exams_acc_stderr,all: 0.005854656489923565
oab_exams_acc_stderr,exam_id__2010-01: 0.02913844109530932
oab_exams_acc_stderr,exam_id__2010-02: 0.02720413223389449
oab_exams_acc_stderr,exam_id__2011-03: 0.02739274623256926
oab_exams_acc_stderr,exam_id__2011-04: 0.03103610079124539
oab_exams_acc_stderr,exam_id__2011-05: 0.030180181541041882
oab_exams_acc_stderr,exam_id__2012-06: 0.030663426761582978
oab_exams_acc_stderr,exam_id__2012-06a: 0.03019542051107743
oab_exams_acc_stderr,exam_id__2012-07: 0.030587368127940558
oab_exams_acc_stderr,exam_id__2012-08: 0.030150394921254077
oab_exams_acc_stderr,exam_id__2012-09: 0.028367169852259254
oab_exams_acc_stderr,exam_id__2013-10: 0.030312606625205758
oab_exams_acc_stderr,exam_id__2013-11: 0.0291233179987808
oab_exams_acc_stderr,exam_id__2013-12: 0.030508999749351583
oab_exams_acc_stderr,exam_id__2014-13: 0.03131790740981766
oab_exams_acc_stderr,exam_id__2014-14: 0.030290399093110758
oab_exams_acc_stderr,exam_id__2014-15: 0.032059169494021815
oab_exams_acc_stderr,exam_id__2015-16: 0.029833251327948757
oab_exams_acc_stderr,exam_id__2015-17: 0.03201117621919014
oab_exams_acc_stderr,exam_id__2015-18: 0.0311783808239491
oab_exams_acc_stderr,exam_id__2016-19: 0.032162738371024235
oab_exams_acc_stderr,exam_id__2016-20: 0.031934266796334064
oab_exams_acc_stderr,exam_id__2016-20a: 0.030215012978448332
oab_exams_acc_stderr,exam_id__2016-21: 0.027979645441821365
oab_exams_acc_stderr,exam_id__2017-22: 0.031096010380049006
oab_exams_acc_stderr,exam_id__2017-23: 0.02995046656597778
oab_exams_acc_stderr,exam_id__2017-24: 0.03158083933896483
oab_exams_acc_stderr,exam_id__2018-25: 0.030481185333792096
oab_exams_alias: oab_exams
portuguese_hate_speech_acc,all: 0.7097532314923619
portuguese_hate_speech_acc_stderr,all: 0.01097429657035187
portuguese_hate_speech_alias: portuguese_hate_speech_binary
portuguese_hate_speech_f1_macro,all: 0.6922452504317789
portuguese_hate_speech_f1_macro_stderr,all: 0.011709050058396507
tweetsentbr_acc,all: 0.6422885572139303
tweetsentbr_acc_stderr,all: 0.007563312856378464
tweetsentbr_alias: tweetsentbr
tweetsentbr_f1_macro,all: 0.6428335432993848
tweetsentbr_f1_macro_stderr,all: 0.007483756787178434
step: 3595