Closed rojinsafavi closed 5 years ago
Why not just leave the code unchanged and represent your 1D signal as a 2D signal of height 1?
Thank you for submitting your issue @rojinsafavi.
The problem is that the layer stored input shape when the model is built, so when batch size was not defined (i.e. inLayer = Input(shape=(16889,))
), it stored None for one of the shape values. This results in errors when .call() happens and it tries to make a random tensor with a shape of (None, 1024).
The error is resolved in commit 745a892df89b1fc851b44daa0acb66d18304b232
Hi, Thanks for providing the code! I was wondering if you know how to apply the code to a 1D data instead an image? I have done some edits to the code, but I am getting the following error:
Here is my edits to the code: `class Darknet19Encoder(Architecture): ''' This encoder predicts distributions then randomly samples them. Regularization may be applied to the latent space output
and this is the error that I'm getting:
and here some printing results that may help?