Table 2.

Tested configurations.

Configuration	Batch Size	Learning Rate	Layers	Accuracy
#1	16	$1.5\times {10}^{-5}$	24	0.948
#2 (Best Model)	16	${1.0\times 10}^{-5}$	24	0.948
#3	16	$0.5\times {10}^{-5}$	24	0.948
#4	16	$0.5\times {10}^{-6}$	24	0.945
#5	16	$1.0\times {10}^{-5}$	20	0.925
#6	16	$1.0\times {10}^{-5}$	16	0.902
#7	16	$1.0\times {10}^{-5}$	12	0.850
#8	16	$1.0\times {10}^{-5}$	10	0.817
#9	16	$1.0\times {10}^{-5}$	8	0.788
#10	16	$1.0\times {10}^{-5}$	4	0.707
#11	16	$1.0\times {10}^{-5}$	2	0.655