ModelTrainer Class Reference

Public Member Functions
	__init__ (self, config)
	train (self)

Public Attributes
	config = config
	model = self._load_model()
	train_generator = self._create_data_generator(config['train_data_dir'])
	val_generator = self._create_data_generator(config['val_data_dir'], shuffle=False)

Protected Member Functions
	_load_model (self)
	_create_data_generator (self, data_directory, shuffle=True)

Detailed Description

A class to handle the training of a model using a custom data generator.

Attributes:
    model (tf.keras.Model): The model to be trained.
    train_generator (CustomDataGenerator): Data generator for training data.
    val_generator (CustomDataGenerator): Data generator for validation data.

Constructor & Destructor Documentation

__init__()

__init__	(		self,
			config )

Initialize the ModelTrainer.

Args:
    config (dict): Configuration dictionary with training parameters.

Member Function Documentation

_create_data_generator()

_create_data_generator ( self, data_directory, shuffle = True )

protected

Create an instance of CustomDataGenerator.

Args:
    data_directory (str or list): Directory containing the dataset.
    shuffle (bool): Whether to shuffle data. Defaults to True.

Returns:
    CustomDataGenerator: Initialized data generator.

_load_model()

_load_model ( self )

protected

Load the model using ModelLoader.

Returns:
    tf.keras.Model: Compiled model.

train()

train

(

self

)

Train the model using the data generators.

Member Data Documentation

config

config = config

model

model = self._load_model()

train_generator

train_generator = self._create_data_generator(config['train_data_dir'])

val_generator

val_generator = self._create_data_generator(config['val_data_dir'], shuffle=False)

---

The documentation for this class was generated from the following file:

• model_trainer.py