Simpan model dengan kerugian validasi terbaik keras
model = getModel()
model.summary()
earlyStopping = EarlyStopping(monitor='val_loss', patience=1000, verbose=0, mode='min')
model.fit(Xtr_more, Ytr_more, batch_size=32, epochs=50, verbose=0, callbacks=[earlyStopping], validation_split=0.25)
Real Raccoon