epochs = mc_epochs,
validation_data = valid_gen,
steps_per_epoch = tsteps,
validation_steps = vsteps,
callbacks = [check_point, early_stop, csv_logger, lr_schedule.lr_scheduler()])
train_end_time = time.time()
m.sec_to_time_elapsed(train_end_time, train_start_time)
### Model Evaluation
###############################################################################
# Training Progress
m.plot_training_progress(csv_file_path = mc_csv_log_save_name,
train_metric = 'loss',
validation_metric = 'val_loss')
# Predict with Model on Test Set
saved_model = keras.models.load_model(mc_model_save_name)
pred_values = model.predict(test_x)
plot_i = 150
imm.plot_image_bounding_box(img_arr = test_x[plot_i],
xmin = [pred_values[plot_i][0]],
xmax = [pred_values[plot_i][1]],
steps_per_epoch=tsteps,
validation_steps=vsteps,
callbacks=[
check_point, early_stop, csv_logger,
lr_schedule.lr_scheduler()
],
class_weight=class_weight_dict)
train_end_time = time.time()
m.sec_to_time_elapsed(train_end_time, train_start_time)
### Plot Model Progress
###############################################################################
# Accuracy
m.plot_training_progress(csv_file_path=m.config_csv_save_name,
train_metric='categorical_accuracy',
validation_metric='val_categorical_accuracy')
# Entropy (Loss)
m.plot_training_progress(csv_file_path=m.config_csv_save_name,
train_metric='loss',
validation_metric='val_loss')
### Model Test Set Prediction
###############################################################################
# Predict with Model on Test Set
saved_model = keras.models.load_model(m.config_model_save_name)
pred_values = model.predict(test_x)
# Accuracy on Test Set
true_pos = [