我正在尝试将我训练过的模型（.h5）转换为coreml，但它失败了 .
coremltools（0.8）h5py（2.7.1）Keras（2.1.3）tensorflow（1.5.0）tensorflow-tensorboard（1.5.1）
错误信息：
/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site-packages/keras/engine/topology.py:1269:UserWarning：更新对Keras 2 API的InputLayer调用：InputLayer（dtype = “float32”，batch_input_shape = [None，416 ...，sparse = False，name =“input_1”）返回cls（** config）/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/ site-packages / keras / engine / topology.py:1269:UserWarning：更新你对Keras 2 API的Conv2D调用：Conv2D（kernel_initializer =“glorot_uniform”，kernel_constraint = None，activity_regularizer = None，trainable = True，padding =“same “，strides = [1,1]，filters = 32，use_bias = False，name =”convolution2d_1“，bias_regularizer = None，bias_constraint = None，data_format =”channels_last“，kernel_regularizer = {u'l2'：0 ... . ，activation =“linear”，kernel_size =（3,3））return cls（** config）Traceback（最近一次调用last）：文件“python.py”，第2行，coreml_model = coremltools.converters.keras . convert（'/ home / depam / Downloads / Archive / keyboard.h5'）文件“/ home / depam / Desktop / mlvirtualenv / pythonenv / local / lib / python2.7 / site-packages / coremltools / converters / keras / _keras_converter.py“，第745行，转换为custom_conversion_functions = custom_conversion_functions）文件”/ home / depam / Desktop / mlvirtualenv / pythonenv / local /lib/python2.7/site-packages/coremltools/converters/keras/_keras_converter.py“，第543行，在convertToSpec custom_objects = custom_objects中）文件”/ home / depam / Desktop / mlvirtualenv / pythonenv / local / lib / python2 . 7 / site-packages / coremltools / converters / keras / _keras2_converter.py“，第182行，在_convert model = _keras.models.load_model（model，custom_objects = custom_objects）文件”/ home / depam / Desktop / mlvirtualenv / pythonenv / local /lib/python2.7/site-packages/keras/models.py“，第243行，在load_model model = model_from_config（model_config，custom_objects = custom_objects）文件”/ home / depam / Desktop / mlvirtualenv / pythonenv / local / lib / python2.7 / site-packages / keras / models.py“，第317行，在model_from_config中返回layer_module.deserialize（config，custom_objects = custom_ob （）</ home / depam / Desktop home / depam / Desktop / mlvirtualenv / pythonenv / local / lib / python2.7 / site-packages / keras / utils / generic_utils.py“，第143行，在deserialize_keras_object列表中（custom_objects.items（））））文件”/ home /depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site-packages/keras/engine/topology.py“，第2507行，in_config process_layer（layer_data）文件”/ home / depam / Desktop / mlvirtualenv / pythonenv / local / lib / python2.7 / site-packages / keras / engine / topology.py“，第2493行，在process_layer中custom_objects = custom_objects）文件”/ home / depam / Desktop / mlvirtualenv / pythonenv / local / lib / python2 .7 / site-packages / keras / layers / init.py“，第55行，反序列化printable_module_name ='layer'）文件”/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site- packages / keras / utils / generic_utils.py“，第145行，在deserialize_keras_object中返回n cls.from_config（config ['config']）文件“/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site-packages/keras/engine/topology.py”，第1269行， from_config返回cls（** config）文件“/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site-packages/keras/legacy/interfaces.py”，第91行，在包装返回函数中（ * args，** kwargs）文件“/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site-packages/keras/layers/convolutional.py”，第462行，在init ** kwargs中）文件“/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site-packages/keras/layers/convolutional.py”，第116行，在init self.kernel_regularizer = regularizers.get（kernel_regularizer）文件中“/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site-packages/keras/regularizers.py”，第80行，在获取返回反序列化（标识符）文件“/ home / depam / Desktop / mlvirtualenv / pythonenv / local / lib / python2.7 / site-packages / keras / regularizers.py“，第73行，反序列化printable_module_name = 'regularrizer'）文件“/home/depam/Desktop/mlvirtualenv/pythonenv/local/lib/python2.7/site-packages/keras/utils/generic_utils.py”，第126行，在deserialize_keras_object中引发ValueError（'不正确的配置格式：'str（config））ValueError：不正确的配置格式：{u'l2'：0.0005000000237487257，u'name'：u'L1L2Regularizer'，u'l1'：0.0}