python - 转换为 CoreML 时 Keras 'InputLayer object has no attribute ' inbound_nodes'

Question

尝试将我的 Keras 模型转换为 CoreML 模型时，出现错误“InputLayer 对象没有属性‘inbound_nodes’”。

这是我的代码:

     loaded_model = load_model("diffinception.h5")
     coreml_model = coremltools.converters.keras.convert(loaded_model, 
         input_names="imageSculp", output_names="category")
     coreml_model.save("transfertestinception.mlmodel")

“diffinception.h5”是从 Keras 导入的 Inception 模型，带有我为迁移学习训练的附加层。

这是我生成该模型的代码:

    model = applications.InceptionV3(weights = "imagenet", include_top=False,         
    input_shape = (img_width, img_height, 3), pooling = max)

    # Freeze layers
    for layer in model.layers:
        layer.trainable = False

    #Adding custom Layers
    x = model.output
    x = Flatten()(x)
    x = Dense(1024, activation="relu")(x)
    x = Dropout(0.5)(x)
    x = Dense(1024, activation="relu")(x)
    predictions = Dense(2, activation="softmax")(x)

    # creating the final model
    model_final = Model(inputs = model.input, outputs = predictions)
    # compile the model
    model_final.compile(loss = "categorical_crossentropy", optimizer =                                         
        optimizers.SGD(lr=0.001, momentum=0.9), metrics=["accuracy"])

我更新了 Keras 的版本。使用 Python 2.7

Answer 1

我更新了我机器上的 _topology2.py 代码以匹配以下版本(2018 年 1 月 17 日更新):

https://github.com/apple/coremltools/blob/master/coremltools/converters/keras/_topology2.py

这解决了问题。