Config Module

class config.Config(cfgFile)[source]

Bases: object

Get and validate configuration parameters from a configuration file based on YAML. These parameters will be saved as class attributes.

Parameters:: cfgFile (str) – Configuration file path.

cpuUsed

Number of CPUs to be used.

Type:: int

trainPath

Directory path of train data.

Type:: str

testPath

Directory path of test data.

Type:: str

validationPath

Directory path of validation data.

Type:: str

modelsPath

Directory path of models that will be generated.

Type:: str

transformsHeight

Images transforms height.

Type:: int

transformsWidth

Images transforms width.

Type:: int

replications

Number of replications used at each trained model.

Type:: int

batchSize

Batch size.

Type:: int

modelNames

List of model names to be trained.

Type:: list

epochs

List of epochs to be trained.

Type:: list

learningRates

List of learning rates to be trained.

Type:: list

weightDecays

List of weight decays to be used at train.

Type:: list

tree

Configuration file tree (from YAML object).

Type:: dict

readConfigFile(configFile)[source]

Read the configuration file.

Parameters:: configFile (str) – Configuration file path.
Returns:: Configuration tree (from YAML object).
Return type:: (dict)