CODEWITHSUNDEEP
pythonpython-3.xmxnet
Foivos
Foivos

Reputation: 545

Get HybridBlock layer shape on runtime

I am trying to build a custom pooling layer (both for ndarray and Symbol) and I need to know the input shape at runtime. According to the documentation, HybridBlock has the function "infer_shape", but I can't make it work. Any pointers into what I am doing wrong?

mxnet version

1.0.0 , build from conda, python3.

Minimum reproducible example

For example: 

import mxnet as mx
import mxnet.ndarray as nd
from mxnet.gluon import HybridBlock

class runtime_shape(HybridBlock):


    def __init__(self,  **kwards):
        HybridBlock.__init__(self,**kwards)


    def hybrid_forward(self,F,_input):

        print (self.infer_shape(_input))

        return _input

xx = nd.random_uniform(shape=[5,5,16,16])

mynet = runtime_shape()
mynet.hybrid_forward(nd,xx)

Error Message:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-41-3f539a940958> in <module>()
----> 1 mynet.hybrid_forward(nd,xx)

<ipython-input-38-afc9785b716d> in hybrid_forward(self, F, _input)
     17     def hybrid_forward(self,F,_input):
     18 
---> 19         print (self.infer_shape(_input))
     20 
     21         return _input

 /home/dia021/anaconda2/lib/python2.7/site-packages/mxnet/gluon/block.pyc in infer_shape(self, *args)
    460     def infer_shape(self, *args):
    461         """Infers shape of Parameters from inputs."""
--> 462         self._infer_attrs('infer_shape', 'shape', *args)
    463 
    464     def infer_type(self, *args):

/home/dia021/anaconda2/lib/python2.7/site-packages/mxnet/gluon/block.pyc in _infer_attrs(self, infer_fn, attr, *args)
    448     def _infer_attrs(self, infer_fn, attr, *args):
    449         """Generic infer attributes."""
--> 450         inputs, out = self._get_graph(*args)
    451         args, _ = _flatten(args)
    452         arg_attrs, _, aux_attrs = getattr(out, infer_fn)(

/home/dia021/anaconda2/lib/python2.7/site-packages/mxnet/gluon/block.pyc in _get_graph(self, *args)
    369             params = {i: j.var() for i, j in self._reg_params.items()}
    370             with self.name_scope():
--> 371                 out = self.hybrid_forward(symbol, *grouped_inputs, **params)  # pylint: disable=no-value-for-parameter
    372             out, self._out_format = _flatten(out)
    373 

/home/dia021/anaconda2/lib/python2.7/site-packages/mxnet/gluon/block.pyc in __exit__(self, ptype, value, trace)
     78         if self._block._empty_prefix:
     79             return
---> 80         self._name_scope.__exit__(ptype, value, trace)
     81         self._name_scope = None
     82         _BlockScope._current = self._old_scope

AttributeError: 'NoneType' object has no attribute '__exit__'

Upvotes: 2

Views: 562

Answers (1)

Thomas
Thomas

Reputation: 701

The idea of the HybridBlock is to make it easy to debug in the imperative world, where you can simply put a breakpoint, or a print statement, and see what data is flowing through your network. When you are confident the network is doing what you want, you can call .hybridize() and enjoy the speed improvements.

Whilst developing your network and using the imperative mode, you can simply print: print('shape',_input.shape)

and remove this line when using the hybridized version of your network, as this will only work for NDArrays.

If this does not answer your question, can you precise what is the goal you are trying to achieve by getting the shape of your input data?

Upvotes: 2

Related Questions