Error about newest quantized int8 `ssd-resnet50_v1_voc_int8` model

#1

I’m interesting in quantized Object Detection model, so I upgrade gluoncv to 0.4.0 and try to download the ssd_512_resnet50_v1_voc_int8 model. But there is something wrong when I get_model through model_zoo. It seems like missing some symbol file about the model.

>>> mxnet.__version__
'1.4.0'
>>> gluoncv.__version__
'0.4.0'
>>> from gluoncv import model_zoo, data, utils
>>> net = model_zoo.get_model('ssd_512_resnet50_v1_voc_int8', pretrained=True)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python3.6/site-packages/gluoncv/model_zoo/model_zoo.py", line 231, in get_model
    net = _models[name](**kwargs)
  File "/usr/local/lib/python3.6/site-packages/gluoncv/model_zoo/quantized/quantized.py", line 45, in func
    sym_net = SymbolBlock.imports(json_file, ['data'], None, ctx=ctx)
  File "/usr/local/lib/python3.6/site-packages/mxnet/gluon/block.py", line 1018, in imports
    sym = symbol.load(symbol_file)
  File "/usr/local/lib/python3.6/site-packages/mxnet/symbol/symbol.py", line 2620, in load
    check_call(_LIB.MXSymbolCreateFromFile(c_str(fname), ctypes.byref(handle)))
  File "/usr/local/lib/python3.6/site-packages/mxnet/base.py", line 252, in check_call
    raise MXNetError(py_str(_LIB.MXGetLastError()))
mxnet.base.MXNetError: [09:44:53] src/io/local_filesys.cc:199: Check failed: allow_null  LocalFileSystem::Open "/usr/local/lib/python3.6/site-packages/gluoncv/model_zoo/quantized/ssd_512_resnet50_v1_voc_int8-symbol.json": No such file or directory

Stack trace returned 8 entries:
[bt] (0) 0   libmxnet.so                         0x000000010ef3c560 mxnet::op::NDArrayOpProp::~NDArrayOpProp() + 3024
[bt] (1) 1   libmxnet.so                         0x000000010ef3c30f mxnet::op::NDArrayOpProp::~NDArrayOpProp() + 2431
[bt] (2) 2   libmxnet.so                         0x000000010ef3b589 mxnet::op::MKLDNNActForward::MKLDNNActForward(mxnet::op::ActivationParam const&, bool, mxnet::NDArray const&, mkldnn::memory const&) + 1625
[bt] (3) 3   libmxnet.so                         0x0000000110c947d0 mxnet::Storage::Get() + 234720
[bt] (4) 4   libmxnet.so                         0x0000000110c8da74 mxnet::Storage::Get() + 206724
[bt] (5) 5   libmxnet.so                         0x00000001105bed0d MXSymbolCreateFromFile + 93
[bt] (6) 6   _ctypes.cpython-36m-darwin.so       0x000000010e0723c7 ffi_call_unix64 + 79
[bt] (7) 7   ???                                 0x00007fff524174c0 0x0 + 140734573409472
#2

Hi @Aspirinkb,

Try:
pip install mxnet-cu90mkl --pre --upgrade (use the right flavour of mxnet for your setup)
pip install gluoncv --pre --upgrade

worked for me.

#3

My machine is MacBookPro 2015, no GPU. So I have to install mxnet-mkl