-
Hey Alex!
Awesome that you're using the StagingArea. However, the usage of it is incorrect, and you won't see any speedup.
You're correctly aggregating a list of the staging area ops (you call e…
-
I'm trying to build a variational autoencoder that encodes and decodes music sequence data (of shape (50, 84)). Simultaneously, I want to predict the next element inside the music sequence (vector of …
-
```
def train_single_model(train,unlabel,test,train_label,test_label):
# input shape: (nb_samples, 32)
encoder = containers.Sequential([Dense(16, input_dim=32), Dense(8)])
decoder = conta…
-
Nice work @gwaygenomics!
Is there any plan to include dimensionality reduction via PCA as was done [here](https://github.com/cognoma/machine-learning/blob/master/explore/pca-visualization/PCA_Visual…
-
I have two outputs and two objective functions in my network. I want the first objective function just trains a set of layers (e.g. the first 5 layers) while the second one trains the rest.
Is that p…
-
In the `autoencoders/autoencoder_models`directory, the generate() function of the `VariationalAutoencoder` is broken. The following error results from using it:
This is fixed by calling `numpy.…
-
Here is a modified version of the `variational_autoencoder_deconv.py` file. It does downsampling in the input and upsampling in the output it compiles and converges as expected but the estimated outpu…
-
It looks like Conv1d only accepts `FloatTensor`, and when it is fed `DoubleTensor` it errors out.
Here is a short example
```python
import torch
from torch.autograd import Variable
import tor…
-
Hi Eder,
I see that your VariationalDense layer implements the KL loss as a regularizer, how can I monitor it during training ?
Thank you for your help,
Sebastien
-
The way `mask` affects the cost function right now is not the best way. Note that the `mask` comes from the input and multiplies the final cost. This is right when we are doing sequence prediction, in…