chainer.exporters.caffe.export

chainer.exporters.caffe.export(model, args, directory=None, export_params=True, graph_name='Graph')

(Experimental) Export a computational graph as Caffe format.

Parameters

• model (Chain) – The model object you want to export in Caffe format. It should have __call__() method because the second argument args is directly given to the model by the () accessor.

• args (list of ~chainer.Variable) – The arguments which are given to the model directly.

• directory (str) – The directory used for saving the resulting Caffe model. If None, nothing is saved to the disk.

• export_params (bool) – If True, this function exports all the parameters included in the given model at the same time. If False, the exported Caffe model doesn’t include any parameter values.

• graph_name (str) – A string to be used for the name field of the graph in the exported Caffe model.

Note

Currently, this function supports networks that created by following layer functions.

• linear()

• convolution_2d()

• deconvolution_2d()

• max_pooling_2d()

• average_pooling_2d()

• batch_normalization()

• local_response_normalization()

• relu()

• leaky_relu()

• concat()

• softmax()

• reshape()

• add()

This function can export at least following networks.

• GoogLeNet

• ResNet

• VGG

And, this function use testing (evaluation) mode.

Example

>>> from chainer.exporters import caffe
>>>
>>> class Model(chainer.Chain):
...    def __init__(self):
...        super(Model, self).__init__()
...        with self.init_scope():
...            self.l1 = L.Convolution2D(None, 1, 1, 1, 0)
...            self.b2 = L.BatchNormalization(1)
...            self.l3 = L.Linear(None, 1)
...
...    def __call__(self, x):
...        h = F.relu(self.l1(x))
...        h = self.b2(h)
...        return self.l3(h)
...
>>> x = chainer.Variable(np.zeros((1, 10, 10, 10), np.float32))
>>> caffe.export(Model(), [x], None, True, 'test')