Skip to content

Instantly share code, notes, and snippets.

@mlabonne
Created February 4, 2024 19:47
Show Gist options
  • Save mlabonne/e8c99f1d522d40aebb99766cdbd272a9 to your computer and use it in GitHub Desktop.
Save mlabonne/e8c99f1d522d40aebb99766cdbd272a9 to your computer and use it in GitHub Desktop.
Model AGIEval GPT4All TruthfulQA Bigbench Average
Omnarch-7B 45.88 77.28 74.07 49.76 61.75

AGIEval

Task Version Metric Value Stderr
agieval_aqua_rat 0 acc 29.13 ± 2.86
acc_norm 28.35 ± 2.83
agieval_logiqa_en 0 acc 38.25 ± 1.91
acc_norm 38.86 ± 1.91
agieval_lsat_ar 0 acc 22.61 ± 2.76
acc_norm 23.04 ± 2.78
agieval_lsat_lr 0 acc 52.75 ± 2.21
acc_norm 53.73 ± 2.21
agieval_lsat_rc 0 acc 65.80 ± 2.90
acc_norm 64.68 ± 2.92
agieval_sat_en 0 acc 80.58 ± 2.76
acc_norm 81.07 ± 2.74
agieval_sat_en_without_passage 0 acc 45.15 ± 3.48
acc_norm 43.69 ± 3.46
agieval_sat_math 0 acc 34.55 ± 3.21
acc_norm 33.64 ± 3.19

Average: 45.88%

GPT4All

Task Version Metric Value Stderr
arc_challenge 0 acc 66.64 ± 1.38
acc_norm 67.66 ± 1.37
arc_easy 0 acc 87.58 ± 0.68
acc_norm 83.12 ± 0.77
boolq 1 acc 88.20 ± 0.56
hellaswag 0 acc 69.27 ± 0.46
acc_norm 86.89 ± 0.34
openbookqa 0 acc 38.40 ± 2.18
acc_norm 49.20 ± 2.24
piqa 0 acc 83.57 ± 0.86
acc_norm 85.47 ± 0.82
winogrande 0 acc 80.43 ± 1.12

Average: 77.28%

TruthfulQA

Task Version Metric Value Stderr
truthfulqa_mc 1 mc1 60.47 ± 1.71
mc2 74.07 ± 1.44

Average: 74.07%

Bigbench

Task Version Metric Value Stderr
bigbench_causal_judgement 0 multiple_choice_grade 56.84 ± 3.60
bigbench_date_understanding 0 multiple_choice_grade 65.31 ± 2.48
bigbench_disambiguation_qa 0 multiple_choice_grade 46.90 ± 3.11
bigbench_geometric_shapes 0 multiple_choice_grade 25.35 ± 2.30
exact_str_match 0.00 ± 0.00
bigbench_logical_deduction_five_objects 0 multiple_choice_grade 35.60 ± 2.14
bigbench_logical_deduction_seven_objects 0 multiple_choice_grade 25.29 ± 1.64
bigbench_logical_deduction_three_objects 0 multiple_choice_grade 62.00 ± 2.81
bigbench_movie_recommendation 0 multiple_choice_grade 53.00 ± 2.23
bigbench_navigate 0 multiple_choice_grade 54.70 ± 1.57
bigbench_reasoning_about_colored_objects 0 multiple_choice_grade 70.45 ± 1.02
bigbench_ruin_names 0 multiple_choice_grade 56.25 ± 2.35
bigbench_salient_translation_error_detection 0 multiple_choice_grade 37.88 ± 1.54
bigbench_snarks 0 multiple_choice_grade 70.17 ± 3.41
bigbench_sports_understanding 0 multiple_choice_grade 75.15 ± 1.38
bigbench_temporal_sequences 0 multiple_choice_grade 56.50 ± 1.57
bigbench_tracking_shuffled_objects_five_objects 0 multiple_choice_grade 22.00 ± 1.17
bigbench_tracking_shuffled_objects_seven_objects 0 multiple_choice_grade 20.23 ± 0.96
bigbench_tracking_shuffled_objects_three_objects 0 multiple_choice_grade 62.00 ± 2.81

Average: 49.76%

Average score: 61.75%

Elapsed time: 03:53:52

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment