I am trying to use torch.jit to build a serialized model.
But I got a bug called: excepted a single top-level function, like:
Traceback (most recent call last):
File “train_Kinetics.py”, line 100, in
model = SNN(batch_norm=True, num_action=opt.class_num, dropout=opt.dropout, detachout=opt.detachout, img_size=opt.img_size, syn_bn=opt.syn_bn, test_scheme=3)
File “/home/poli/anaconda2/envs/pytorch1.0/lib/python2.7/site-packages/torch/jit/init.py”, line 951, in init_then_register
_create_methods_from_stubs(self, methods)
File “/home/poli/anaconda2/envs/pytorch1.0/lib/python2.7/site-packages/torch/jit/init.py”, line 912, in _create_methods_from_stubs
self._create_methods(defs, rcbs, defaults)
File “/home/poli/anaconda2/envs/pytorch1.0/lib/python2.7/site-packages/torch/jit/annotations.py”, line 67, in get_num_params
raise RuntimeError(“expected a single top-level function”)
RuntimeError: expected a single top-level function