In the paper there is a mention of the colorizing model. What are the details?
Specifically I'm interested in:
- training hyperparameters, especially learning rates, lambda_1 and learning rate decay
- what were the sizes of kernels in generator and discriminator
In the paper there is a mention of the colorizing model. What are the details?
Specifically I'm interested in: