Update mxnet version to fix CI error

[xinyu-intel]

No description provided.

[xinyu-intel]

@zhreshold Catch a bug after update to the latest nightly build. It seems a cudnn op introduce it:

import mxnet as mx
import gluoncv as gcv

ctx = mx.gpu(0)
x = mx.random.uniform(shape=(2, 3, 224, 224), ctx=ctx)
net = gcv.model_zoo.get_model('resnet18_v1b_0.89', pretrained=False)
net.initialize()
net.collect_params().reset_ctx(ctx)
net(x)
mx.nd.waitall()

[22:39:50] src/operator/nn/./cudnn/./cudnn_algoreg-inl.h:97: Running performance tests to find the best convolution algorithm, this can take a while... (set the environment variable MXNET_CUDNN_AUTOTUNE_DEFAULT to 0 to disable)
Traceback (most recent call last):
  File "test_model_zoo.py", line 415, in <module>
    test_imagenet_models()
  File "/home/chenxiny/gluon-cv/tests/unittests/common.py", line 43, in test_wrapper
    orig_test(*args, **kwargs)
  File "test_model_zoo.py", line 123, in test_imagenet_models
    _test_model_list(models, ctx, x)
  File "test_model_zoo.py", line 53, in _test_model_list
    mx.nd.waitall()
  File "/home/chenxiny/mxnet-gpu/python/mxnet/ndarray/ndarray.py", line 166, in waitall
    check_call(_LIB.MXNDArrayWaitAll())
  File "/home/chenxiny/mxnet-gpu/python/mxnet/base.py", line 253, in check_call
    raise MXNetError(py_str(_LIB.MXGetLastError()))
mxnet.base.MXNetError: [22:39:50] include/mxnet/././tensor_blob.h:290: Check failed: this->shape_.Size() == static_cast<size_t>(shape.Size()) (64 vs. 8) : TBlob.get_with_shape: new and old shape do not match total elements
Stack trace:
  [bt] (0) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(dmlc::LogMessageFatal::~LogMessageFatal()+0x2b) [0x7f46220c1c4b]
  [bt] (1) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(mshadow::Tensor<mshadow::gpu, 1, float> mxnet::TBlob::get_with_shape<mshadow::gpu, 1, float>(mshadow::Shape<1> const&, mshadow::Stream<mshadow::gpu>*) const+0x1e4) [0x7f4624f96194]
  [bt] (2) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(mxnet::op::CuDNNBatchNormOp<float>::Forward(mxnet::OpContext const&, std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob> > const&, std::vector<mxnet::OpReqType, std::allocator<mxnet::OpReqType> > const&, std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob> > const&, std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob> > const&)+0x6c5) [0x7f46264c8305]
  [bt] (3) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(void mxnet::op::BatchNormCompute<mshadow::gpu>(nnvm::NodeAttrs const&, mxnet::OpContext const&, std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob> > const&, std::vector<mxnet::OpReqType, std::allocator<mxnet::OpReqType> > const&, std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob> > const&)+0x90e) [0x7f46264bd96e]
  [bt] (4) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(mxnet::imperative::PushFCompute(std::function<void (nnvm::NodeAttrs const&, mxnet::OpContext const&, std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob> > const&, std::vector<mxnet::OpReqType, std::allocator<mxnet::OpReqType> > const&, std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob> > const&)> const&, nnvm::Op const*, nnvm::NodeAttrs const&, mxnet::Context const&, std::vector<mxnet::engine::Var*, std::allocator<mxnet::engine::Var*> > const&, std::vector<mxnet::engine::Var*, std::allocator<mxnet::engine::Var*> > const&, std::vector<mxnet::Resource, std::allocator<mxnet::Resource> > const&, std::vector<mxnet::NDArray*, std::allocator<mxnet::NDArray*> > const&, std::vector<mxnet::NDArray*, std::allocator<mxnet::NDArray*> > const&, std::vector<unsigned int, std::allocator<unsigned int> > const&, std::vector<mxnet::OpReqType, std::allocator<mxnet::OpReqType> > const&)::{lambda(mxnet::RunContext)#1}::operator()(mxnet::RunContext) const+0x372) [0x7f46245a5f22]
  [bt] (5) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(+0x3f06e64) [0x7f4624d27e64]
  [bt] (6) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(mxnet::engine::ThreadedEngine::ExecuteOprBlock(mxnet::RunContext, mxnet::engine::OprBlock*)+0x636) [0x7f4624d23576]
  [bt] (7) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(void mxnet::engine::ThreadedEnginePerDevice::GPUWorker<(dmlc::ConcurrentQueueType)0>(mxnet::Context, bool, mxnet::engine::ThreadedEnginePerDevice::ThreadWorkerBlock<(dmlc::ConcurrentQueueType)0>*, std::shared_ptr<dmlc::ManualEvent> const&)+0x150) [0x7f4624d35c70]
  [bt] (8) /home/chenxiny/mxnet-gpu/python/mxnet/../../lib/libmxnet.so(std::_Function_handler<void (std::shared_ptr<dmlc::ManualEvent>), mxnet::engine::ThreadedEnginePerDevice::PushToExecute(mxnet::engine::OprBlock*, bool)::{lambda()#4}::operator()() const::{lambda(std::shared_ptr<dmlc::ManualEvent>)#1}>::_M_invoke(std::_Any_data const&, std::shared_ptr<dmlc::ManualEvent>)+0x46) [0x7f4624d35ef6]

[zhreshold]

I'd say it's really a weird error

[xinyu-intel]

@zhreshold It looks like resnet18_v1b_0.89 is shown as resnet18_v1b_2.6x in gcv.model_zoo.pretrained_model_list(), so the CI tests this model without using pre-trained weight.

[zhreshold]

#871

[zhreshold]

@xinyu-intel I am trying to fix all these things together but get a new error: http://ci.mxnet.io/blue/organizations/jenkins/gluon-cv/detail/PR-870/11/pipeline

[xinyu-intel]

@zhreshold okay, i'll take a look at this bug.

[xinyu-intel]

@zhreshold It's weird that when I test this only one case, it returns no error. But when I test the whole file, it returns error...

[mli]

Job PR-863-10 is done.
Docs are uploaded to http://gluon-vision-staging.s3-website-us-west-2.amazonaws.com/PR-863/10/index.html
Code coverage of this PR: vs. Master:

[mli]

Job PR-863-11 is done.
Docs are uploaded to http://gluon-vision-staging.s3-website-us-west-2.amazonaws.com/PR-863/11/index.html
Code coverage of this PR: vs. Master:

[xinyu-intel]

@zhreshold ci pass:)

[zhreshold]

@xinyu-intel merged 😃

[xinyu-intel]

@zhreshold Thanks:)