Skip to contents

Do validation only once at end of training.

Usage

validation_after_training_cb(gen.val, validation_steps)

Arguments

gen.val

Validation generator

validation_steps

Number of validation steps.

Value

Keras callback, apply validation only after training.

Examples

if (FALSE) { # reticulate::py_module_available("tensorflow")
maxlen <- 20
model <- create_model_lstm_cnn(layer_lstm = 8, maxlen = maxlen)
gen <- get_generator(train_type = 'dummy_gen', model = model, batch_size = 4, maxlen = maxlen)
vat <- validation_after_training_cb(gen.val = gen, validation_steps = 10)
}