orcanet_contrib.parser_orcatrain

Use orga.train with a parser.

Usage:

parser_orcatrain.py [options] FOLDER LIST CONFIG MODEL parser_orcatrain.py (-h | –help)

Arguments:

FOLDER Path to the folder where everything gets saved to, e.g. the

summary.txt, the plots, the trained models, etc.

LIST A .toml file which contains the pathes of the training and

validation files. An example can be found in examples/list_file.toml

CONFIG A .toml file which sets up the training. An example can be

found in examples/config_file.toml. The possible parameters are listed in core.py in the class Configuration.

MODEL Path to a .toml file with infos about a model.

An example can be found in examples/explanation.toml.

Options:

-h –help Show this screen. –recompile Recompile the keras model, e.g. needed if the loss weights

are changed during the training.

Module Contents

Functions

orca_train(output_folder, list_file, config_file, ...)	Run orga.train with predefined ModelBuilder networks using a parser.
main()	Run the orca_train function with a parser.

orcanet_contrib.parser_orcatrain.orca_train(output_folder, list_file, config_file, model_file, recompile_model=False)

Run orga.train with predefined ModelBuilder networks using a parser.

Parameters

output_folderstr

Path to the folder where everything gets saved to, e.g. the summary log file, the plots, the trained models, etc.

list_filestr

Path to a list file which contains pathes to all the h5 files that should be used for training and validation.

config_filestr

Path to a .toml file which overwrites some of the default settings for training and validating a model.

model_filestr

Path to a file with parameters to build a model of a predefined architecture with OrcaNet.

recompile_modelbool

If the model should be recompiled or not. Necessary, if e.g. the loss_weights are changed during the training.

orcanet_contrib.parser_orcatrain.main()

Run the orca_train function with a parser.