Use following command to start training given Caffe's solver and configure file

./net_trainer.py <solver_file> <SNAPSHOT> <NUM_GPU> <SYNC_FREQ>

• solver_file is the file name in Caffe's solver format.
• SNAPSHOT is the index of the snapshot to start with (default: 0). If SNAPSHOT is not equal to 0, it means we continue training from the formal snapshot
• NUM_GPU is the number of gpu to use.
• SYNC_FREQ is the option to hide IO time consuming, set SYNC_FREQ > 1 usually have better speed but will consume more GPU RAM. Please set it properly to let Minerva run efficiently while won't exceed GPU RAM limitation. (default: 1)

Example:

./net_trainer.py /path/to/solver.txt 0 4 2

Use following command to perform testing on the given trained network. We could get top-1 or top-5 accuracy under single view or multiview

./net_tester.py <solver_file> <softmax_layer_name> <accuracy_layer_name> <SNAPSHOT> <GPU_IDX> <MULTIVIEW>

• solver_file is caffe solver configure file.
• softmax_layer_name indicate the layer to produce softmax distribution.
• accuracy_layer_name indicate the layer to produce accuracy. If you want to get top-5 accuracy, make sure to declare the accuracy_param top-k: 5 in the network configuration file.
• SNAPSHOT is the index of the snapshot to test with (default: 0).
• GPU_IDX is the id of the gpu on which you want the testing to be performed (default: 0).
• MULTIVIEW indicate whether to use multiview testing (default: 0).

Example:

./net_tester.py /path/to/solver.txt loss3/loss3 loss3/top-5 0 1 1

Use following command to extract the feature of a certain layer from the given trained network

./feature_extractor.py <solver_file> <layer_name> <feature_path> <SNAPSHOT> <GPU_IDX>

• solver_file is caffe solver configure file.
• layer_name is the name of the layer to extract feature
• The feature will be written to the feature_path in readable float format (not binary).
• SNAPSHOT is the index of the snapshot to test with (default: 0).
• GPU_IDX is the id of the gpu on which you want the testing to be performed (default: 0).

Example:

./feature_extractor.py /path/to/solver.txt fc6 /path/to/save/feature.txt 60 1

Use the following command to show what the nuerons interested in from a given model

./filter_visualizer.py <solver_file> <layer_name> <result_path> <SNAPSHOT> <GPU_IDX>

• solver_file is caffe solver configure file.
• the filters in layer_name will be visualized
• The visualization result will be saved in result_path, each filter/neuron will get a jpg visualization result.
• SNAPSHOT is the index of the snapshot to test with (default: 0).
• GPU_IDX is the id of the gpu on which you want the testing to be performed (default: 0).

Example:

./filter_visualizer.py /path/to/solver.txt conv4 /path/to/result/folder 60 0

Use the following command to show which part of the image activate the neurons

./heatmap_visualizer.py <solver_file> <layer_name> <result_path> <SNAPSHOT> <GPU_IDX>

• solver_file is caffe solver configure file.
• the filters in layer_name will be visualized
• The visualization result will be saved in result_path. For each testing image, we will generate a heatmap in jpg format.
• SNAPSHOT is the index of the snapshot to test with (default: 0).
• GPU_IDX is the id of the gpu on which you want the testing to be performed (default: 0).

Example:

./heatmap_visualizer.py /path/to/solver.txt conv4 /path/to/result/folder 60 0

For more documents on how to use NetTrainer and owl.net package yourself, please see here.