Skip to content

Instantly share code, notes, and snippets.

@mlabonne
Created January 5, 2024 00:36
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save mlabonne/14687f1eb3425b166db511f31f8e66f6 to your computer and use it in GitHub Desktop.
Save mlabonne/14687f1eb3425b166db511f31f8e66f6 to your computer and use it in GitHub Desktop.
Model AGIEval GPT4All TruthfulQA Bigbench Average
NeuralHermes-2.5-Mistral-7B 43.67 73.24 55.37 41.76 53.51

AGIEval

Task Version Metric Value Stderr
agieval_aqua_rat 0 acc 22.44 ± 2.62
acc_norm 22.44 ± 2.62
agieval_logiqa_en 0 acc 40.71 ± 1.93
acc_norm 41.32 ± 1.93
agieval_lsat_ar 0 acc 24.78 ± 2.85
acc_norm 23.48 ± 2.80
agieval_lsat_lr 0 acc 49.41 ± 2.22
acc_norm 50.00 ± 2.22
agieval_lsat_rc 0 acc 60.22 ± 2.99
acc_norm 59.11 ± 3.00
agieval_sat_en 0 acc 74.27 ± 3.05
acc_norm 73.79 ± 3.07
agieval_sat_en_without_passage 0 acc 45.63 ± 3.48
acc_norm 44.66 ± 3.47
agieval_sat_math 0 acc 35.00 ± 3.22
acc_norm 34.55 ± 3.21

Average: 43.67%

GPT4All

Task Version Metric Value Stderr
arc_challenge 0 acc 57.34 ± 1.45
acc_norm 60.58 ± 1.43
arc_easy 0 acc 83.75 ± 0.76
acc_norm 81.31 ± 0.80
boolq 1 acc 86.42 ± 0.60
hellaswag 0 acc 63.76 ± 0.48
acc_norm 82.22 ± 0.38
openbookqa 0 acc 34.20 ± 2.12
acc_norm 44.80 ± 2.23
piqa 0 acc 81.66 ± 0.90
acc_norm 82.75 ± 0.88
winogrande 0 acc 74.59 ± 1.22

Average: 73.24%

TruthfulQA

Task Version Metric Value Stderr
truthfulqa_mc 1 mc1 37.58 ± 1.70
mc2 55.37 ± 1.54

Average: 55.37%

Bigbench

Task Version Metric Value Stderr
bigbench_causal_judgement 0 multiple_choice_grade 53.68 ± 3.63
bigbench_date_understanding 0 multiple_choice_grade 67.48 ± 2.44
bigbench_disambiguation_qa 0 multiple_choice_grade 36.05 ± 3.00
bigbench_geometric_shapes 0 multiple_choice_grade 25.07 ± 2.29
exact_str_match 4.46 ± 1.09
bigbench_logical_deduction_five_objects 0 multiple_choice_grade 28.80 ± 2.03
bigbench_logical_deduction_seven_objects 0 multiple_choice_grade 20.14 ± 1.52
bigbench_logical_deduction_three_objects 0 multiple_choice_grade 46.33 ± 2.88
bigbench_movie_recommendation 0 multiple_choice_grade 36.80 ± 2.16
bigbench_navigate 0 multiple_choice_grade 49.90 ± 1.58
bigbench_reasoning_about_colored_objects 0 multiple_choice_grade 66.00 ± 1.06
bigbench_ruin_names 0 multiple_choice_grade 44.64 ± 2.35
bigbench_salient_translation_error_detection 0 multiple_choice_grade 25.15 ± 1.37
bigbench_snarks 0 multiple_choice_grade 70.17 ± 3.41
bigbench_sports_understanding 0 multiple_choice_grade 66.63 ± 1.50
bigbench_temporal_sequences 0 multiple_choice_grade 30.30 ± 1.45
bigbench_tracking_shuffled_objects_five_objects 0 multiple_choice_grade 21.04 ± 1.15
bigbench_tracking_shuffled_objects_seven_objects 0 multiple_choice_grade 17.09 ± 0.90
bigbench_tracking_shuffled_objects_three_objects 0 multiple_choice_grade 46.33 ± 2.88

Average: 41.76%

Average score: 53.51%

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