AttributeError：模块“keras.optimizers”没有属性“rmsprop”

Question

我收到 AttributeError 错误：模块 'keras.optimizers' 没有以下代码的属性 'rmsprop'。我正在将 tensorflow 2 与 python 3.8 一起使用。

# training
batch_size = 64
epochs=125

from keras import regularizers, optimizers
from keras.callbacks import ModelCheckpoint   

checkpointer = ModelCheckpoint(filepath='model.125epochs.hdf5', verbose=1, save_best_only=True)

optimizer = keras.optimizers.rmsprop(lr=0.0003,decay=1e-6)
model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy'])
history = model.fit_generator(datagen.flow(x_train, y_train, batch_size=batch_size), callbacks=[checkpointer],
                steps_per_epoch=x_train.shape[0] // batch_size, epochs=epochs,verbose=2,
                validation_data=(x_test,y_test))

AttributeError: module 'keras.optimizers' has no attribute 'rmsprop'

Answer 1

因为 rms 应该大写。

optimizer = keras.optimizers.RMSprop(lr=0.0003, decay=1e-6)