example/vae/VAE_example.ipynb runs into type error

Discussion
1

A noob question.
Any pointers to the error described below?
My environment is python 3.6 and mxnet 1.3.0
I fire up jupyter notebook and open VAE_example.ipynb
when execution is at:

sample the latent variables z according to Normal(mu,var)

z = mu + np.multiply(mx.symbol.exp(0.5 * logvar),
mx.symbol.random_normal(loc=0, scale=1, shape=np.shape(logvar.get_internals()[“logvar_output”])))

It results into an error whose trace is reproduced below:

TypeError Traceback (most recent call last)
in ()
13 ## sample the latent variables z according to Normal(mu,var)
14 z = mu + np.multiply(mx.symbol.exp(0.5 * logvar),
—> 15 mx.symbol.random_normal(loc=0, scale=1, shape=np.shape(logvar.get_internals()[“logvar_output”])))

~/incubator-mxnet/python/mxnet/symbol/symbol.py in add(self, other)
103 return _internal._PlusScalar(self, scalar=other)
104 else:
–> 105 raise TypeError(‘type %s not supported’ % str(type(other)))
106
107 def bool(self):

TypeError: type <class ‘numpy.ndarray’> not supported

update===========================================================
print(type(mu))
print(type(np.multiply(mx.symbol.exp(0.5 * logvar),
mx.symbol.random_normal(loc=0, scale=1, shape=np.shape(logvar.get_internals()[“logvar_output”])))))

result

<class ‘mxnet.symbol.symbol.Symbol’>
<class ‘numpy.ndarray’>

2

Not all the example in master have been updated. Follow sym vae and gluon vae tutorials instead.

P.S. FYI @ThomasDelteil

1 Like