EMCqMRI.core.engine.build_model module

class EMCqMRI.core.engine.build_model.Configuration(mode)[source]

Bases: object

parse_command_line_configuration()[source]

parse_imported_cmd_configuration(config_file)[source]

set_dataset_config(config_parser)[source]

set_global_config()[source]

set_inference_config(config_parser)[source]

EMCqMRI.core.engine.build_model.check_override(config_object, override_modules)[source]

EMCqMRI.core.engine.build_model.configure_cuda(config_object)[source]

EMCqMRI.core.engine.build_model.load_checkpoint(config_object)[source]

EMCqMRI.core.engine.build_model.load_cost_function(config_object)[source]

EMCqMRI.core.engine.build_model.load_dataset_model(config_object)[source]

EMCqMRI.core.engine.build_model.load_inference_model(config_object)[source]

EMCqMRI.core.engine.build_model.load_likelihood_model(config_object)[source]

EMCqMRI.core.engine.build_model.load_optimizer(config_object)[source]

EMCqMRI.core.engine.build_model.load_signal_model(config_object)[source]

EMCqMRI.core.engine.build_model.make(mode='training', override_modules=None)[source]

Links all configurations to internal modules, which are dinamically imported.

Parameters

([type (config_object) – Configuration]): [Object containing all backend configuration settings]
config_object.args (Required) –
- useCUDA
- device
- runBenchmark
- likelihoodModel
- signalModel
- inferenceModel
- task
- optimizer
- learningRate
- lossFunction
- loadCheckpoint
- loadCheckpointPath

Returns

Configuration]: Updated config_object containing all imported Python modules

Return type

[type