ResNet50 & ResNet101
https://storage.googleapis.com/download.tensorflow.org/example_images/flower_photos.tgz
The key difference of the full preactivation 'v2' variant compared to the 'v1' variant in [1] is the use of batch normalization before every weight layer.
- resnet(V1) ** ResNet V1 models use vgg pre-processing and input image size of 224*224 ** https://github.com/tensorflow/models/blob/master/research/slim/preprocessing/vgg_preprocessing.py
- resnet(V2) according to the https://github.com/alexchungio/models/tree/master/research/slim ** ResNet V2 models use Inception pre-processing and input image size of 299*299 ** https://github.com/tensorflow/models/blob/master/research/slim/preprocessing/inception_preprocessing.py
https://github.com/alexchungio/models/tree/master/research/slim
- batch size: 32
- learning rate: 0.01
- decay rate: 0.1
- num epoch percent decay: 20
- weight decay: 0.0001
- epoch: 60
| loss | accuracy | |
|---|---|---|
| train | 1.2061034440994263 | 96875 |
| val | 1.203824520111084 | 0.9619565010070801 |
- batch size: 32
- learning rate: 0.01
- decay rate: 0.1
- num epoch percent decay: 20
- weight decay: 0.0001
- epoch: 60
| loss | accuracy | |
|---|---|---|
| train | 1.2062734365463257 | 0.96875 |
| val | 1.2127000093460083 | 0.95652174949646 |