Table 5.

Training loss, training accuracy, validation loss and validation accuracy derived from the epoch with the highest validation accuracy achieved through transfer learning across all models.

Model	Best Epoch	Epoch Time (s)	Train_loss	Train_acc	Val_loss	Val_acc
Tiny-SGD	97	59	2.638	0.489	2.381	0.539
Tiny-RMSprop	97	62	0.310	0.917	0.261	0.932
Tiny-Adam	44	58.5	0.133	0.968	0.104	0.976
Tiny-Lion	55	54.9	0.136	0.966	0.098	0.976
Small-SGD	38	62.2	5.791	0.008	5.793	0.014
Small-RMSprop	26	59.1	0.141	0.962	0.091	0.977
Small-Adam	67	58.5	0.111	0.972	0.090	0.978
Small-Lion	36	59	0.090	0.976	0.077	0.981
Medium-SGD	89	60.5	5.803	0.007	5.804	0.011
Medium-RMSprop	97	60.2	0.265	0.937	0.371	0.907
Medium-Adam	51	65.3	0.138	0.968	0.231	0.944
Medium-Lion	22	65.1	0.131	0.968	0.256	0.942

Train_loss—training loss; Train_acc—training accuracy; Val_loss—vValidation loss; Val_acc—validation accuracy.