campa.tl.Estimator

class Estimator(exp)

Neural network estimator.

Handles training and evaluation of models.

Parameters

exp (Experiment) – Experiment with model config.

Attributes

test_dataset	tf.data.Dataset of test split.
train_dataset	Shuffled tf.data.Dataset of train split.
val_dataset	tf.data.Dataset of val split.

Methods

create_model()	Initialise neural network model.
evaluate_model([dataset])	Evaluate model using tf.data.Dataset.
predict_model(data[, batch_size])	Predict all elements in data.
train_model()	Train neural network model.