Meta-Llama-3-12B-Instruct-Nous.md

Model	AGIEval	GPT4All	TruthfulQA	Bigbench	Average
Meta-Llama-3-12B-Instruct	41.7	67.71	52.75	40.58	50.69

AGIEval

Task	Version	Metric	Value		Stderr
agieval_aqua_rat	0	acc	28.35	±	2.83
		acc_norm	28.35	±	2.83
agieval_logiqa_en	0	acc	36.56	±	1.89
		acc_norm	38.10	±	1.90
agieval_lsat_ar	0	acc	21.30	±	2.71
		acc_norm	18.70	±	2.58
agieval_lsat_lr	0	acc	48.24	±	2.21
		acc_norm	47.06	±	2.21
agieval_lsat_rc	0	acc	59.85	±	2.99
		acc_norm	57.25	±	3.02
agieval_sat_en	0	acc	71.36	±	3.16
		acc_norm	68.45	±	3.25
agieval_sat_en_without_passage	0	acc	40.29	±	3.43
		acc_norm	40.29	±	3.43
agieval_sat_math	0	acc	40.00	±	3.31
		acc_norm	35.45	±	3.23

Average: 41.7%

GPT4All

Task	Version	Metric	Value		Stderr
arc_challenge	0	acc	49.57	±	1.46
		acc_norm	52.30	±	1.46
arc_easy	0	acc	79.46	±	0.83
		acc_norm	75.97	±	0.88
boolq	1	acc	78.81	±	0.71
hellaswag	0	acc	58.24	±	0.49
		acc_norm	76.18	±	0.43
openbookqa	0	acc	34.80	±	2.13
		acc_norm	43.60	±	2.22
piqa	0	acc	77.31	±	0.98
		acc_norm	76.71	±	0.99
winogrande	0	acc	70.40	±	1.28

Average: 67.71%

TruthfulQA

Task	Version	Metric	Value		Stderr
truthfulqa_mc	1	mc1	36.60	±	1.69
		mc2	52.75	±	1.57

Average: 52.75%

Bigbench

Task	Version	Metric	Value		Stderr
bigbench_causal_judgement	0	multiple_choice_grade	52.11	±	3.63
bigbench_date_understanding	0	multiple_choice_grade	61.25	±	2.54
bigbench_disambiguation_qa	0	multiple_choice_grade	36.05	±	3.00
bigbench_geometric_shapes	0	multiple_choice_grade	18.94	±	2.07
		exact_str_match	0.00	±	0.00
bigbench_logical_deduction_five_objects	0	multiple_choice_grade	31.80	±	2.08
bigbench_logical_deduction_seven_objects	0	multiple_choice_grade	21.00	±	1.54
bigbench_logical_deduction_three_objects	0	multiple_choice_grade	51.00	±	2.89
bigbench_movie_recommendation	0	multiple_choice_grade	30.40	±	2.06
bigbench_navigate	0	multiple_choice_grade	50.10	±	1.58
bigbench_reasoning_about_colored_objects	0	multiple_choice_grade	55.75	±	1.11
bigbench_ruin_names	0	multiple_choice_grade	47.54	±	2.36
bigbench_salient_translation_error_detection	0	multiple_choice_grade	25.35	±	1.38
bigbench_snarks	0	multiple_choice_grade	65.19	±	3.55
bigbench_sports_understanding	0	multiple_choice_grade	58.01	±	1.57
bigbench_temporal_sequences	0	multiple_choice_grade	33.80	±	1.50
bigbench_tracking_shuffled_objects_five_objects	0	multiple_choice_grade	23.84	±	1.21
bigbench_tracking_shuffled_objects_seven_objects	0	multiple_choice_grade	17.31	±	0.90
bigbench_tracking_shuffled_objects_three_objects	0	multiple_choice_grade	51.00	±	2.89

Average: 40.58%

Average score: 50.69%

Elapsed time: 03:15:26