Skip to content

Instantly share code, notes, and snippets.

@mlabonne
Created March 6, 2024 12:42
Show Gist options
  • Save mlabonne/70dc278b76bdadffef9efcb7a096b0a3 to your computer and use it in GitHub Desktop.
Save mlabonne/70dc278b76bdadffef9efcb7a096b0a3 to your computer and use it in GitHub Desktop.
Model AGIEval GPT4All TruthfulQA Bigbench Average
phi-2-orange-v2 34.55 70.96 54.87 38.17 49.64

AGIEval

Task Version Metric Value Stderr
agieval_aqua_rat 0 acc 20.47 ± 2.54
acc_norm 18.90 ± 2.46
agieval_logiqa_en 0 acc 32.10 ± 1.83
acc_norm 30.57 ± 1.81
agieval_lsat_ar 0 acc 19.13 ± 2.60
acc_norm 17.83 ± 2.53
agieval_lsat_lr 0 acc 34.51 ± 2.11
acc_norm 35.29 ± 2.12
agieval_lsat_rc 0 acc 44.61 ± 3.04
acc_norm 44.61 ± 3.04
agieval_sat_en 0 acc 66.02 ± 3.31
acc_norm 64.56 ± 3.34
agieval_sat_en_without_passage 0 acc 38.35 ± 3.40
acc_norm 36.89 ± 3.37
agieval_sat_math 0 acc 30.00 ± 3.10
acc_norm 27.73 ± 3.02

Average: 34.55%

GPT4All

Task Version Metric Value Stderr
arc_challenge 0 acc 53.41 ± 1.46
acc_norm 55.72 ± 1.45
arc_easy 0 acc 80.89 ± 0.81
acc_norm 75.93 ± 0.88
boolq 1 acc 80.28 ± 0.70
hellaswag 0 acc 57.52 ± 0.49
acc_norm 75.95 ± 0.43
openbookqa 0 acc 41.60 ± 2.21
acc_norm 51.80 ± 2.24
piqa 0 acc 80.52 ± 0.92
acc_norm 80.03 ± 0.93
winogrande 0 acc 77.03 ± 1.18

Average: 70.96%

TruthfulQA

Task Version Metric Value Stderr
truthfulqa_mc 1 mc1 38.92 ± 1.71
mc2 54.87 ± 1.56

Average: 54.87%

Bigbench

Task Version Metric Value Stderr
bigbench_causal_judgement 0 multiple_choice_grade 53.16 ± 3.63
bigbench_date_understanding 0 multiple_choice_grade 54.20 ± 2.60
bigbench_disambiguation_qa 0 multiple_choice_grade 39.92 ± 3.05
bigbench_geometric_shapes 0 multiple_choice_grade 19.22 ± 2.08
exact_str_match 0.00 ± 0.00
bigbench_logical_deduction_five_objects 0 multiple_choice_grade 28.80 ± 2.03
bigbench_logical_deduction_seven_objects 0 multiple_choice_grade 19.00 ± 1.48
bigbench_logical_deduction_three_objects 0 multiple_choice_grade 50.67 ± 2.89
bigbench_movie_recommendation 0 multiple_choice_grade 38.00 ± 2.17
bigbench_navigate 0 multiple_choice_grade 47.50 ± 1.58
bigbench_reasoning_about_colored_objects 0 multiple_choice_grade 59.00 ± 1.10
bigbench_ruin_names 0 multiple_choice_grade 26.79 ± 2.09
bigbench_salient_translation_error_detection 0 multiple_choice_grade 33.37 ± 1.49
bigbench_snarks 0 multiple_choice_grade 72.93 ± 3.31
bigbench_sports_understanding 0 multiple_choice_grade 51.83 ± 1.59
bigbench_temporal_sequences 0 multiple_choice_grade 9.50 ± 0.93
bigbench_tracking_shuffled_objects_five_objects 0 multiple_choice_grade 18.72 ± 1.10
bigbench_tracking_shuffled_objects_seven_objects 0 multiple_choice_grade 13.71 ± 0.82
bigbench_tracking_shuffled_objects_three_objects 0 multiple_choice_grade 50.67 ± 2.89

Average: 38.17%

Average score: 49.64%

Elapsed time: 01:30:17

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