8.1.1.1.1.5. blueoil.cmd.export

8.1.1.1.1.5.1. Module Contents

8.1.1.1.1.5.1.1. Functions

_pre_process(raw_image, pre_processor, data_format)

_save_all_operation_outputs(image_path, output_dir, image, raw_image, all_outputs, image_size)

_minimal_operations(sess)

Get inference operations.

_export(config, restore_path, image_path)

run(experiment_id, restore_path=None, image_size=(None, None), image=DEFAULT_INFERENCE_TEST_DATA_IMAGE, config_file=None)

main(experiment_id, restore_path, image_size, image, config_file)

Exporting a trained model to proto buffer files and meta config yaml.

blueoil.cmd.export.DEFAULT_INFERENCE_TEST_DATA_IMAGE
blueoil.cmd.export._pre_process(raw_image, pre_processor, data_format)
blueoil.cmd.export._save_all_operation_outputs(image_path, output_dir, image, raw_image, all_outputs, image_size)
blueoil.cmd.export._minimal_operations(sess)

Get inference operations.

blueoil.cmd.export._export(config, restore_path, image_path)
blueoil.cmd.export.run(experiment_id, restore_path=None, image_size=(None, None), image=DEFAULT_INFERENCE_TEST_DATA_IMAGE, config_file=None)
blueoil.cmd.export.main(experiment_id, restore_path, image_size, image, config_file)

Exporting a trained model to proto buffer files and meta config yaml.

In the case with image option, create each layer output value npy files into export/{restore_path}/{image_size}/inference_test_data/**.npy as expected value for inference test and debug.