MCPcopy Create free account
hub / github.com/PeizeSun/SparseR-CNN / save_graph

Method save_graph

detectron2/export/api.py:243–263  ·  view source on GitHub ↗

Save the graph as SVG format. Args: output_file (str): a SVG file inputs: optional inputs given to the model. If given, the inputs will be used to run the graph to record shape of every tensor. The shape information will be

(self, output_file, inputs=None)

Source from the content-addressed store, hash-verified

241 f.write(self._init_net.SerializeToString())
242
243 def save_graph(self, output_file, inputs=None):
244 """
245 Save the graph as SVG format.
246
247 Args:
248 output_file (str): a SVG file
249 inputs: optional inputs given to the model.
250 If given, the inputs will be used to run the graph to record
251 shape of every tensor. The shape information will be
252 saved together with the graph.
253 """
254 from .caffe2_export import run_and_save_graph
255
256 if inputs is None:
257 save_graph(self._predict_net, output_file, op_only=False)
258 else:
259 size_divisibility = get_pb_arg_vali(self._predict_net, "size_divisibility", 0)
260 device = get_pb_arg_vals(self._predict_net, "device", b"cpu").decode("ascii")
261 inputs = convert_batched_inputs_to_c2_format(inputs, size_divisibility, device)
262 inputs = [x.cpu().numpy() for x in inputs]
263 run_and_save_graph(self._predict_net, self._init_net, inputs, output_file)
264
265 @staticmethod
266 def load_protobuf(dir):

Callers 2

_test_modelMethod · 0.80

Calls 5

save_graphFunction · 0.85
get_pb_arg_valiFunction · 0.85
get_pb_arg_valsFunction · 0.85
run_and_save_graphFunction · 0.85

Tested by 1

_test_modelMethod · 0.64