chrisdonahue / wavegan

WaveGAN: Learn to synthesize raw audio with generative adversarial networks
MIT License
1.33k stars 280 forks source link

Error on Generation Colab notebook #84

Closed Tylersuard closed 4 years ago

Tylersuard commented 4 years ago

KeyErrorTraceback (most recent call last)

in () 11 12 # Generate ---> 13 z = graph.get_tensor_by_name('z:0') 14 G_z = graph.get_tensor_by_name('G_z:0')[:, :, 0] 15 G_z_spec = graph.get_tensor_by_name('G_z_spec:0') 2 frames /usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/ops.pyc in _as_graph_element_locked(self, obj, allow_tensor, allow_operation) 3836 raise KeyError("The name %s refers to a Tensor which does not " 3837 "exist. The operation, %s, does not exist in the " -> 3838 "graph." % (repr(name), repr(op_name))) 3839 try: 3840 return op.outputs[out_n] KeyError: "The name 'z:0' refers to a Tensor which does not exist. The operation, 'z', does not exist in the graph."
Tylersuard commented 4 years ago

I'm not able to get this notebook up and running. The tensorflow version you recommend (1.12) is incompatible with Colab's version of CUDA (9.0).

Tylersuard commented 4 years ago

I'm really interested in getting this to work, and I spent 8 hours training a model, is there anything we can do to fix this problem?

xyz010 commented 4 years ago

Hello,

I think that the solution to this is to get the whole tensor by

G_z = graph.get_tensor_by_name('G_z:0')

instead of

G_z = graph.get_tensor_by_name('G_z:0')[:, :, 0]

Tylersuard commented 4 years ago

@xyz010 Thank you!