Purpose: to prevent overfitting
# early stoppping from keras.callbacks import EarlyStopping early_stopping = EarlyStopping(monitor='val_loss', patience=50, verbose=2) # Training history = model.fit(train_X, train_y, epochs=300, batch_size=20, validation_data=(test_X, test_y), verbose=2, shuffle=False, callbacks=[early_stopping])monitor: the amount to be monitored, val_loss, val_acc
patience: when early stop is activated (if it is found that the loss has not decreased compared to the previous epoch training), the training will be stopped after patience epochs
verbose: information display mode
mode: 'auto', One of 'min', 'max' , train in min mode, stop training if detections stop falling. In max mode, the training is stopped when the detection value no longer rises.