I try to deploy an SSD trained network on Android device.
When I try to build the .so file I got this error from nnvm.
[15:43:41] src/nnvm/legacy_json_util.cc:209: Loading symbol saved by previous version v0.9.5. Attempting to upgrade...
[15:43:41] src/nnvm/legacy_json_util.cc:217: Symbol successfully upgraded!
Traceback (most recent call last):
File "tvm_minimal.py", line 105, in <module>
Fire(main)
File "/usr/local/lib/python3.6/dist-packages/fire/core.py", line 127, in Fire
component_trace = _Fire(component, args, context, name)
File "/usr/local/lib/python3.6/dist-packages/fire/core.py", line 366, in _Fire
component, remaining_args)
File "/usr/local/lib/python3.6/dist-packages/fire/core.py", line 542, in _CallCallable
result = fn(*varargs, **kwargs)
File "tvm_minimal.py", line 50, in main
nnvm_sym, nnvm_params = nnvm.frontend.from_mxnet(mx_sym, args, auxs)
File "/usr/local/lib/python3.6/dist-packages/nnvm-0.8.0-py3.6.egg/nnvm/frontend/mxnet.py", line 549, in from_mxnet
sym = _from_mxnet_impl(symbol, {})
File "/usr/local/lib/python3.6/dist-packages/nnvm-0.8.0-py3.6.egg/nnvm/frontend/mxnet.py", line 506, in _from_mxnet_impl
node = _convert_symbol(op_name, childs, attr)
File "/usr/local/lib/python3.6/dist-packages/nnvm-0.8.0-py3.6.egg/nnvm/frontend/mxnet.py", line 414, in _convert_symbol
_raise_not_supported('Operator: ' + op_name)
File "/usr/local/lib/python3.6/dist-packages/nnvm-0.8.0-py3.6.egg/nnvm/frontend/mxnet.py", line 24, in _raise_not_supported
raise NotImplementedError(err)
NotImplementedError: Operator: L2Normalization is not supported in nnvm.
The SSD support on nnvm seem to be done https://github.com/dmlc/tvm/pull/1214 and https://github.com/dmlc/tvm/pull/1157.
Anyone know how to fix this ?
>>> mxnet.__version__
'1.5.0'
>>> tvm.__version__
'0.5.dev'
>>> nnvm.__version__
'0.8.0'