def convert_local_response_normalization(onnx_graph, node):
kwargs = {}
kwargs['size'] = oc.try_get_attribute(node.attribute_args.keywords['n'])
kwargs['bias'] = float(
oc.try_get_attribute(node.attribute_args.keywords['k']))
kwargs['alpha'] = float(
oc.try_get_attribute(node.attribute_args.keywords['alpha']) *
kwargs['size'])
kwargs['beta'] = float(
oc.try_get_attribute(node.attribute_args.keywords['beta']))
onnx_graph.add_node(
"LRN",
[node.inputs[0]],
node.outputs,
str(node.lineprop),
**kwargs,
)
def convert_unpooling_2d(onnx_graph, node: 'nodes.NodeCall'):
ksize = oc.try_get_attribute(node.attribute_args.keywords['ksize'])
stride = oc.try_get_attribute(node.attribute_args.keywords['stride'])
pad = oc.try_get_attribute(node.attribute_args.keywords['pad'])
outsize = oc.try_get_attribute(node.attribute_args.keywords['outsize'])
cover_all = oc.try_get_attribute(node.attribute_args.keywords['cover_all'])
assert (stride is None) # TODO(hamaji): Not supported yet.
assert (pad == 0) # TODO(hamaji): Not supported yet.
assert (outsize is None) # TODO(hamaji): Not supported yet.
assert (cover_all is False) # TODO(hamaji): Not supported yet.
scales = np.array([1, 1] + list(_pair(ksize)), dtype=np.float32)
scales_ = oc.ONNXValue(onnx_graph,
scales, [node, '/Scale'],
is_constant=True)
onnx_graph.add_node("Upsample", [node.inputs[0], scales_],
[node.outputs[0]],
name=str(node.lineprop))
def convert_concat(onnx_graph, node):
xs = oc.ONNXValue(onnx_graph, node.args.keywords['xs'])
axis = oc.try_get_attribute(node.attribute_args.keywords['axis'])
onnx_graph.add_node(
"ChainerSequenceConcat",
[xs.create_sequence()],
node.outputs,
str(node.lineprop),
axis=axis,
)
def convert_dropout(onnx_graph, node):
x = oc.ONNXValue(onnx_graph, node.args.keywords['x'])
ratio = oc.try_get_attribute(node.attribute_args.keywords['ratio'])
onnx_graph.add_node(
"Dropout",
[x],
node.outputs,
str(node.lineprop),
ratio=ratio,
)
def convert_roi_average_pooling_2d(onnx_graph, node):
x = oc.ONNXValue(onnx_graph, node.args.keywords['x'])
rois = oc.ONNXValue(onnx_graph, node.args.keywords['rois'])
roi_indices = oc.ONNXValue(onnx_graph, node.args.keywords['roi_indices'])
outsize = oc.ONNXValue(onnx_graph, node.args.keywords['outsize'])
spatial_scale = oc.ONNXValue(onnx_graph,
node.args.keywords['spatial_scale'])
def _pair(x):
if isinstance(x, collections.Iterable):
return x
return (x, x)
onnx_graph.add_node(
"ChainerROIAveragePool2D", [
x.create_tensor(node.lineprop),
rois.create_tensor(node.lineprop),
roi_indices.create_tensor(node.lineprop)
],
node.outputs,
str(node.lineprop),
output_shape=_pair(oc.try_get_attribute(outsize.value)),
spatial_scale=oc.try_get_attribute(spatial_scale.value))
return
def convert_max_pooling_2d(onnx_graph, node):
def _pair(x):
if isinstance(x, collections.Iterable):
return x
return (x, x)
ksize = oc.try_get_attribute(node.attribute_args.keywords['ksize'])
stride = oc.try_get_attribute(node.attribute_args.keywords['stride'])
pad = oc.try_get_attribute(node.attribute_args.keywords['pad'])
cover_all = oc.try_get_attribute(node.attribute_args.keywords['cover_all'])
return_indices = oc.try_get_attribute(
node.attribute_args.keywords['return_indices'])
assert not return_indices # TODO(hamaji): Not implemented yet.
kwargs = {}
kwargs['kernel_shape'] = _pair(ksize)
if stride is not None:
kwargs['strides'] = _pair(stride)
else:
kwargs['strides'] = _pair(ksize)
if pad is not None:
kwargs['pads'] = _pair(pad) * 2
else:
kwargs['pads'] = _pair(0)
onnx_graph.add_node(
"MaxPool",
[node.inputs[0]],
[node.outputs[0]],
name=str(node.lineprop),
chainer_cover_all=cover_all,
**kwargs,
)
def convert_softmax_cross_entropy(onnx_graph, node):
normalize = oc.try_get_attribute(node.attribute_args.keywords['normalize'])
cache_score = oc.try_get_attribute(
node.attribute_args.keywords['cache_score'])
class_weight = oc.try_get_attribute(
node.attribute_args.keywords['class_weight'])
ignore_label = oc.try_get_attribute(
node.attribute_args.keywords['ignore_label'])
reduce = oc.try_get_attribute(node.attribute_args.keywords['reduce'])
enable_double_backprop = oc.try_get_attribute(
node.attribute_args.keywords['enable_double_backprop'])
assert normalize # TODO(hamaji): Not supported yet.
assert cache_score # TODO(hamaji): Not supported yet.
assert class_weight is None # TODO(hamaji): Not supported yet.
assert ignore_label == -1 # TODO(hamaji): Not supported yet.
assert reduce == 'mean' # TODO(hamaji): Not supported yet.
assert not enable_double_backprop # TODO(hamaji): Not supported yet.
onnx_graph.add_node("ChainerSoftmaxCrossEntropy", node.inputs[0:2],
node.outputs, str(node.lineprop))
def convert_softmax(onnx_graph, node):
onnx_graph.add_node("Softmax", [node.inputs[0]], [node.outputs[0]],
str(node.lineprop),
axis=oc.try_get_attribute(node.inputs[1]),
chainer_is_onnx_semantics=False)
def convert_onnx_chainer_linear(onnx_graph: 'ONNXGraph',
node: 'nodes.NodeCall'):
chainer_inst = node.func.owner.inst # type: chainer.links.Linear
onnx_name = oc.node2onnx_parameter[node].onnx_name
x = oc.ONNXValue(onnx_graph, node.args.get_value('x'))
axes = oc.try_get_attribute(node.args.get_value('n_batch_axes'), node)
o = oc.ONNXValue(onnx_graph, node.outputs[0])
if chainer_inst.W.data is None:
print("W is unknown. Please infer this model.")
w = oc.ONNXValue(onnx_graph, chainer_inst.W)
if axes != 1:
inputs = [x, w]
if chainer_inst.b is not None:
b = oc.ONNXValue(onnx_graph, chainer_inst.b)
inputs.append(b)
onnx_graph.add_node('ChainerLinear',
inputs, [o],
str(node.lineprop),
n_batch_axes=axes)
return
(x_shape, ) = onnx_graph.add_node('Shape', [x], [None], str(node.lineprop))
(batch_size_1, ) = onnx_graph.add_node('Gather', [
x_shape,
oc.ONNXValue(onnx_graph, np.array(0, dtype=np.int64),
[onnx_name, '/Zero'])
], [None], str(node.lineprop))
(batch_size_2, ) = onnx_graph.add_node('Unsqueeze', [batch_size_1], [None],
str(node.lineprop),
axes=[0])
(mat_shape, ) = onnx_graph.add_node('Concat', [
batch_size_2,
oc.ONNXValue(onnx_graph, np.array([-1], dtype=np.int64),
[onnx_name, '/Minus1'])
], [None],
str(node.lineprop),
axis=0)
(x_reshape, ) = onnx_graph.add_node('Reshape', [x, mat_shape], [None],
str(node.lineprop))
if chainer_inst.b is not None:
b = oc.ONNXValue(onnx_graph, chainer_inst.b)
onnx_graph.add_node('Gemm', [x_reshape, w, b], [o],
str(node.lineprop),
transA=0,
transB=1)
else:
temp = oc.ONNXValue(onnx_graph, np.float32, [onnx_name, '/Temp'])
onnx_graph.add_node('Transpose', [w], [temp],
str(node.lineprop),
perm=[1, 0])
onnx_graph.add_node('MatMul', [x_reshape, temp], [o],
str(node.lineprop))