ResNet

Spring 2023 - Deep Learning - Mini Project

Jash Rathod
Aneek Roy

Model	Model Parameters (in M)	Data Augmentation	Optimizer	Epochs	Batch Size	Weight Decay	Activation	Test Accuracy (in %)	Directory
ResNet-9	3.58	AutoAugment, Horizontal Flip, Random Crop	AdaDelta	200	512	0.0001	ReLU	93.84	auto_hf_rc_ep200_optAdadelta_bs512
ResNet-9	3.58	TrivialAugment, Horizontal Flip, Random Crop	Adam	20	512	0.0001	ReLU	0.1	tra_hf_rc_ep200_optAdam_bs512
ResNet-9	3.58	TrivialAugment, Horizontal Flip, Random Crop	AdaDelta	200	512	0.0001	ReLU	94.41	tra_hf_rc_ep200_optAdadelta_bs512
ResNet-9	3.58	TrivialAugment, Horizontal Flip, Random Crop	AdaDelta	200	1024	0.0001	ReLU	93.78	tra_hf_rc_ep200_optAdadelta_bs1024
ResNet-9	3.58	TrivialAugment, Horizontal Flip, Random Crop	AdaDelta	200	512	0	ReLU	94.26	tra_hf_rc_ep200_optAdadelta_bs512_wd0
Modified ResNet-9	4.61	TrivialAugment, Horizontal Flip, Random Crop	AdaDelta	200	512	0.0001	ReLU	94.20	linear_inplane36