ValueError: steps_per_epoch=None is only valid for a generator based on the keras.utils.Sequence

https://github.com/keras-team/keras/tree/master/examples

运行如上代码cifar10_resnet.py文件出现下列错误

ValueError: steps_per_epoch=None is only valid for a generator based on the keras.utils.Sequence class. Please specify steps_per_epoch or use the keras.utils.Sequence class.

只需要把

  model.fit_generator(datagen.flow(x_train, y_train, batch_size=batch_size),
                        validation_data=(x_test, y_test),
                        epochs=epochs, verbose=1, workers=4,
                        callbacks=callbacks)

换成

model.fit_generator(datagen.flow(x_train, y_train, batch_size=batch_size),
                        steps_per_epoch=x_train.shape[0] // batch_size,
                        validation_data=(x_test, y_test),
                        epochs=epochs, verbose=1, workers=4,
                        callbacks=callbacks)