running for sequential dense model breaks it
d8sconz opened this issue · comments
I have a simple toy rnn lstm in keras. To run it through devol I set the convolution variables (max_conv_layers, max_filters) to 0 because I am not doing any convolutions. This obviously (in hindsight) triggered a division by zero error (or similar). Point being, running this just for a sequential dense model seems to break it. Scanning through the genome_handler.py file I see that convolution layers are required - they're absence is certainly not tested for.
The error I'm getting (still) is:
Traceback (most recent call last): File "gene.py", line 80, in <module> activations=["relu", "sigmoid", "linear"]) File "C:\Users\...\genome_handler.py", line 42, in __init__ [2**i for i in range(3, int(math.log(max_filters, 2)) + 1)], ValueError: math domain error
I've set the convolution variables max_conv_layers and max_filters to zero because there are none in my model. You've closed the issue so I'm wondering now if I've missed something obvious, or the program is only for testing cnn's.
Hey sorry about that. I closed it because I added the functionality to not include convolutional layers but tested it with max_conv_layers not equal to zero still. I've just fixed that - try pulling again.
Sorry about all the holes - this project was not thoroughly tested and all the attention was unexpected.