Llama3-8B-DPO-function-calling-Nous.md

Model	AGIEval	GPT4All	TruthfulQA	Bigbench	Average
Llama3-8B-DPO-function-calling	37.43	71.05	48.52	40.73	49.43

AGIEval

Task	Version	Metric	Value		Stderr
agieval_aqua_rat	0	acc	27.17	±	2.80
		acc_norm	24.41	±	2.70
agieval_logiqa_en	0	acc	35.18	±	1.87
		acc_norm	36.10	±	1.88
agieval_lsat_ar	0	acc	18.26	±	2.55
		acc_norm	22.17	±	2.75
agieval_lsat_lr	0	acc	40.78	±	2.18
		acc_norm	38.04	±	2.15
agieval_lsat_rc	0	acc	49.44	±	3.05
		acc_norm	46.84	±	3.05
agieval_sat_en	0	acc	67.48	±	3.27
		acc_norm	61.65	±	3.40
agieval_sat_en_without_passage	0	acc	43.69	±	3.46
		acc_norm	38.83	±	3.40
agieval_sat_math	0	acc	35.45	±	3.23
		acc_norm	31.36	±	3.14

Average: 37.43%

GPT4All

Task	Version	Metric	Value		Stderr
arc_challenge	0	acc	54.44	±	1.46
		acc_norm	57.08	±	1.45
arc_easy	0	acc	82.41	±	0.78
		acc_norm	79.88	±	0.82
boolq	1	acc	82.17	±	0.67
hellaswag	0	acc	59.43	±	0.49
		acc_norm	78.25	±	0.41
openbookqa	0	acc	36.00	±	2.15
		acc_norm	45.80	±	2.23
piqa	0	acc	80.09	±	0.93
		acc_norm	80.47	±	0.92
winogrande	0	acc	73.72	±	1.24

Average: 71.05%

TruthfulQA

Task	Version	Metric	Value		Stderr
truthfulqa_mc	1	mc1	31.95	±	1.63
		mc2	48.52	±	1.49

Average: 48.52%

Bigbench

Task	Version	Metric	Value		Stderr
bigbench_causal_judgement	0	multiple_choice_grade	60.53	±	3.56
bigbench_date_understanding	0	multiple_choice_grade	69.65	±	2.40
bigbench_disambiguation_qa	0	multiple_choice_grade	30.62	±	2.88
bigbench_geometric_shapes	0	multiple_choice_grade	20.06	±	2.12
		exact_str_match	0.00	±	0.00
bigbench_logical_deduction_five_objects	0	multiple_choice_grade	30.00	±	2.05
bigbench_logical_deduction_seven_objects	0	multiple_choice_grade	22.00	±	1.57
bigbench_logical_deduction_three_objects	0	multiple_choice_grade	49.33	±	2.89
bigbench_movie_recommendation	0	multiple_choice_grade	31.20	±	2.07
bigbench_navigate	0	multiple_choice_grade	50.00	±	1.58
bigbench_reasoning_about_colored_objects	0	multiple_choice_grade	65.70	±	1.06
bigbench_ruin_names	0	multiple_choice_grade	37.72	±	2.29
bigbench_salient_translation_error_detection	0	multiple_choice_grade	25.65	±	1.38
bigbench_snarks	0	multiple_choice_grade	58.56	±	3.67
bigbench_sports_understanding	0	multiple_choice_grade	55.58	±	1.58
bigbench_temporal_sequences	0	multiple_choice_grade	37.80	±	1.53
bigbench_tracking_shuffled_objects_five_objects	0	multiple_choice_grade	22.72	±	1.19
bigbench_tracking_shuffled_objects_seven_objects	0	multiple_choice_grade	16.74	±	0.89
bigbench_tracking_shuffled_objects_three_objects	0	multiple_choice_grade	49.33	±	2.89

Average: 40.73%

Average score: 49.43%

Elapsed time: 02:44:12