pytorch-pretrained-cnns

Reproduce training

python train.py --data_dir /data/datasets/ --classifier <Classifier Name> --dataset <Dataset Name> --batch_size 256 --gpu_id <GPU ID> --num_workers 8

Model	Epoch	Train Loss	Train Accuracy	Validation Loss	Validation Accuracy
lowres_alexnet	95	0.563	80.717	0.604	79.297
lowres_densenet121	96	0.047	99.716	0.247	93.780
lowres_densenet161	97	0.042	99.830	0.232	94.311
lowres_densenet169	98	0.046	99.742	0.235	94.171
lowres_googlenet	97	0.064	99.651	0.242	92.919
lowres_inception_v3	96	0.062	99.501	0.254	93.550
lowres_mobilenet_v2	96	0.074	98.892	0.237	93.760
lowres_resnet101	96	0.041	99.329	0.242	93.399
lowres_resnet14	97	0.040	99.567	0.254	92.588
lowres_resnet152	99	0.032	99.629	0.249	93.490
lowres_resnet18	97	0.033	99.685	0.250	92.929
lowres_resnet18_noresidual	99	0.059	99.371	0.282	92.738
lowres_resnet34	99	0.027	99.714	0.253	93.399
lowres_resnet50	97	0.039	99.473	0.227	93.780
lowres_resnet9	93	0.016	99.996	0.174	94.792
lowres_vgg11	95	0.024	99.860	0.254	92.258
lowres_vgg11_bn	95	0.024	99.860	0.254	92.258
lowres_vgg13	98	0.021	99.942	0.198	94.111
lowres_vgg13_bn	98	0.021	99.942	0.198	94.111
lowres_vgg16	99	0.021	99.912	0.228	93.930
lowres_vgg16_bn	99	0.021	99.912	0.228	93.930
lowres_vgg19	97	0.022	99.878	0.242	93.800
lowres_vgg19_bn	97	0.022	99.878	0.242	93.800

Model	Epoch	Train Loss	Train Accuracy	Validation Loss	Validation Accuracy
lowres_alexnet	4	4.605	0.919	4.605	1.122
lowres_densenet121	94	0.192	98.678	1.082	75.040
lowres_densenet161	98	0.171	99.373	1.044	76.412
lowres_densenet169	97	0.177	99.171	1.063	75.341
lowres_googlenet	97	0.331	98.109	1.077	73.417
lowres_inception_v3	97	0.276	98.395	1.055	75.040
lowres_mobilenet_v2	93	0.342	94.940	1.009	75.200
lowres_resnet101	95	0.133	98.596	1.070	74.740
lowres_resnet14	98	0.380	93.321	1.110	70.673
lowres_resnet152	97	0.127	98.846	1.059	74.720
lowres_resnet18	98	0.163	98.257	1.103	72.536
lowres_resnet18_noresidual	98	0.309	96.020	1.196	71.084
lowres_resnet34	97	0.103	99.165	1.161	72.546
lowres_resnet50	96	0.131	98.834	1.062	74.159
lowres_resnet9	91	0.075	99.806	1.000	75.591
lowres_vgg11	98	0.095	99.303	1.307	69.621
lowres_vgg11_bn	98	0.095	99.303	1.307	69.621
lowres_vgg13	94	0.088	99.393	1.158	73.017
lowres_vgg13_bn	94	0.088	99.393	1.158	73.017
lowres_vgg16	96	0.110	98.702	1.267	72.907
lowres_vgg16_bn	96	0.110	98.702	1.267	72.907
lowres_vgg19	96	0.136	97.917	1.349	71.945
lowres_vgg19_bn	96	0.136	97.917	1.349	71.945

Model	Epoch	Train Loss	Train Accuracy	Validation Loss	Validation Accuracy
lowres_alexnet	86	0.053	98.444	0.045	98.668
lowres_densenet121	92	0.035	99.980	0.044	99.579
lowres_densenet161	93	0.035	99.983	0.043	99.609
lowres_densenet169	96	0.036	99.977	0.042	99.649
lowres_googlenet	90	0.046	99.873	0.045	99.579
lowres_inception_v3	98	0.046	99.873	0.041	99.679
lowres_mobilenet_v2	96	0.035	99.983	0.042	99.659
lowres_resnet101	87	0.020	99.978	0.032	99.539
lowres_resnet14	80	0.021	99.985	0.030	99.669
lowres_resnet152	94	0.019	99.990	0.030	99.589
lowres_resnet18	83	0.020	99.993	0.030	99.639
lowres_resnet18_noresidual	84	0.034	99.988	0.041	99.619
lowres_resnet34	79	0.019	99.977	0.029	99.609
lowres_resnet50	86	0.020	99.982	0.032	99.559
lowres_resnet9	83	0.006	99.998	0.016	99.679
lowres_vgg11	75	0.017	99.995	0.027	99.639
lowres_vgg11_bn	75	0.017	99.995	0.027	99.639
lowres_vgg13	90	0.017	99.998	0.026	99.649
lowres_vgg13_bn	90	0.017	99.998	0.026	99.649
lowres_vgg16	90	0.017	99.992	0.026	99.639
lowres_vgg16_bn	90	0.017	99.992	0.026	99.639
lowres_vgg19	85	0.017	99.988	0.027	99.649
lowres_vgg19_bn	85	0.017	99.988	0.027	99.649

Model	Epoch	Train Loss	Train Accuracy	Validation Loss	Validation Accuracy
lowres_alexnet	97	0.047	98.775	0.191	94.872
lowres_densenet121	88	0.037	99.982	0.092	98.668
lowres_densenet161	87	0.038	99.972	0.084	98.688
lowres_densenet169	89	0.037	99.987	0.096	98.518
lowres_googlenet	97	0.044	99.970	0.112	97.947
lowres_inception_v3	97	0.044	99.970	0.090	98.658
lowres_mobilenet_v2	89	0.036	99.988	0.095	98.468
lowres_resnet101	90	0.020	99.993	0.097	98.147
lowres_resnet14	73	0.021	99.975	0.070	98.748
lowres_resnet152	85	0.020	99.988	0.090	98.197
lowres_resnet18	73	0.020	99.997	0.077	98.628
lowres_resnet18_noresidual	80	0.035	99.993	0.090	98.608
lowres_resnet34	83	0.018	99.998	0.081	98.538
lowres_resnet50	80	0.020	99.982	0.097	98.067
lowres_resnet9	54	0.008	99.997	0.069	98.528
lowres_vgg11	62	0.016	99.998	0.078	98.427
lowres_vgg11_bn	62	0.016	99.998	0.078	98.427
lowres_vgg13	54	0.016	99.980	0.069	98.698
lowres_vgg13_bn	54	0.016	99.980	0.069	98.698
lowres_vgg16	79	0.015	99.998	0.070	98.698
lowres_vgg16_bn	79	0.015	99.998	0.070	98.698
lowres_vgg19	99	0.015	99.998	0.078	98.518
lowres_vgg19_bn	99	0.015	99.998	0.078	98.518

Model	Epoch	Train Loss	Train Accuracy	Validation Loss	Validation Accuracy
lowres_alexnet	97	0.296	89.248	0.308	89.042
lowres_densenet121	92	0.037	99.947	0.266	93.950
lowres_densenet161	93	0.038	99.937	0.264	94.171
lowres_densenet169	97	0.038	99.933	0.258	94.291
lowres_googlenet	96	0.044	99.973	0.240	93.439
lowres_inception_v3	97	0.050	99.861	0.244	94.441
lowres_mobilenet_v2	98	0.040	99.913	0.252	93.860
lowres_resnet101	94	0.019	99.985	0.281	93.740
lowres_resnet14	89	0.021	99.997	0.228	94.040
lowres_resnet152	91	0.020	99.970	0.286	93.770
lowres_resnet18	86	0.020	99.998	0.228	93.970
lowres_resnet18_noresidual	97	0.034	99.977	0.260	93.730
lowres_resnet34	96	0.018	99.993	0.261	93.910
lowres_resnet50	89	0.019	99.985	0.261	93.810
lowres_resnet9	63	0.009	99.998	0.203	94.071
lowres_vgg11	91	0.017	100.000	0.229	93.600
lowres_vgg11_bn	91	0.017	100.000	0.229	93.600
lowres_vgg13	80	0.017	99.998	0.211	94.111
lowres_vgg13_bn	80	0.017	99.998	0.211	94.111
lowres_vgg16	86	0.018	99.995	0.226	94.030
lowres_vgg16_bn	86	0.018	99.995	0.226	94.030
lowres_vgg19	95	0.018	99.987	0.244	93.960
lowres_vgg19_bn	95	0.018	99.987	0.244	93.960