phi-2-sft-dpo-gpt4_en-ep1-Nous.md

Model	AGIEval	GPT4All	TruthfulQA	Bigbench	Average
phi-2-sft-dpo-gpt4_en-ep1	30.61	71.13	48.74	35.23	46.43

AGIEval

Task	Version	Metric	Value		Stderr
agieval_aqua_rat	0	acc	19.29	±	2.48
		acc_norm	21.26	±	2.57
agieval_logiqa_en	0	acc	27.65	±	1.75
		acc_norm	32.10	±	1.83
agieval_lsat_ar	0	acc	17.39	±	2.50
		acc_norm	20.87	±	2.69
agieval_lsat_lr	0	acc	31.96	±	2.07
		acc_norm	28.82	±	2.01
agieval_lsat_rc	0	acc	32.34	±	2.86
		acc_norm	32.71	±	2.87
agieval_sat_en	0	acc	52.91	±	3.49
		acc_norm	43.20	±	3.46
agieval_sat_en_without_passage	0	acc	38.35	±	3.40
		acc_norm	35.92	±	3.35
agieval_sat_math	0	acc	33.64	±	3.19
		acc_norm	30.00	±	3.10

Average: 30.61%

GPT4All

Task	Version	Metric	Value		Stderr
arc_challenge	0	acc	54.86	±	1.45
		acc_norm	55.97	±	1.45
arc_easy	0	acc	80.09	±	0.82
		acc_norm	76.89	±	0.86
boolq	1	acc	82.23	±	0.67
hellaswag	0	acc	57.26	±	0.49
		acc_norm	75.63	±	0.43
openbookqa	0	acc	40.20	±	2.19
		acc_norm	52.60	±	2.24
piqa	0	acc	78.35	±	0.96
		acc_norm	78.35	±	0.96
winogrande	0	acc	76.24	±	1.20

Average: 71.13%

TruthfulQA

Task	Version	Metric	Value		Stderr
truthfulqa_mc	1	mc1	34.15	±	1.66
		mc2	48.74	±	1.55

Average: 48.74%

Bigbench

Task	Version	Metric	Value		Stderr
bigbench_causal_judgement	0	multiple_choice_grade	56.84	±	3.60
bigbench_date_understanding	0	multiple_choice_grade	58.27	±	2.57
bigbench_disambiguation_qa	0	multiple_choice_grade	37.98	±	3.03
bigbench_geometric_shapes	0	multiple_choice_grade	10.03	±	1.59
		exact_str_match	0.00	±	0.00
bigbench_logical_deduction_five_objects	0	multiple_choice_grade	24.20	±	1.92
bigbench_logical_deduction_seven_objects	0	multiple_choice_grade	17.86	±	1.45
bigbench_logical_deduction_three_objects	0	multiple_choice_grade	43.00	±	2.86
bigbench_movie_recommendation	0	multiple_choice_grade	36.60	±	2.16
bigbench_navigate	0	multiple_choice_grade	44.70	±	1.57
bigbench_reasoning_about_colored_objects	0	multiple_choice_grade	54.45	±	1.11
bigbench_ruin_names	0	multiple_choice_grade	25.67	±	2.07
bigbench_salient_translation_error_detection	0	multiple_choice_grade	24.95	±	1.37
bigbench_snarks	0	multiple_choice_grade	55.80	±	3.70
bigbench_sports_understanding	0	multiple_choice_grade	51.52	±	1.59
bigbench_temporal_sequences	0	multiple_choice_grade	17.50	±	1.20
bigbench_tracking_shuffled_objects_five_objects	0	multiple_choice_grade	18.88	±	1.11
bigbench_tracking_shuffled_objects_seven_objects	0	multiple_choice_grade	12.91	±	0.80
bigbench_tracking_shuffled_objects_three_objects	0	multiple_choice_grade	43.00	±	2.86

Average: 35.23%

Average score: 46.43% Elapsed time: 02:39:33