@ Commit message Author Date Author
f6d5f07 configuring saving method and inference method for model Browse code 2023-07-11
5225b78 corrected wrong ... binary mask. It was inverted Browse code 2023-07-05
c3d88f8 corrected the inference output Browse code 2023-07-05
9773fd6 Tried to make logger with dash, however it did not work as I have planned, so instead, visdom is used. Also, now the dataloader shuffles the dataset by default Browse code 2023-07-04
2fbce64 Its running! Browse code 2023-07-03
a4559ba finished training code for generator, now the rest is discriminator. Browse code 2023-06-30
fa53112 Added comment Browse code 2023-06-30
6f62307 something is weird about this gamma correction Browse code 2023-06-29
6d3bc38 something is weird about this gamma correction Browse code 2023-06-29
84f45a0 Still working on loss functions Browse code 2023-06-29
e5885c1 integrated loss functions into neural network class Browse code 2023-06-28
a3a1a70 Added comment Browse code 2023-06-27
744883d Loss function for Autoencoder, also we may have to use other pretrained network for feature extractor for the loss function or even not using feature extractor. I ... honestly do not think using neural network for VGG is strictly nesccesary, and may have to be replaced with other image preprocessing like MSCN which was implimented before. Browse code 2023-06-27
dd1c901 Loss function for Autoencoder Browse code 2023-06-27
8264532 Loss function for Autoencoder Browse code 2023-06-27
3f5673b Discriminator Loss Browse code 2023-06-27
415bdf4 Discriminator Loss Browse code 2023-06-27
722005d wrote draft of logger and writing train.py Browse code 2023-06-26
2bbc716 changing return type into dict to not make additional cost for chaning return types of functions just in case. Browse code 2023-06-23
2c09074 discriminator test code Browse code 2023-06-23
3690a9e updating loss function Browse code 2023-06-23
163a5d0 theorizing train code for GAN Browse code 2023-06-22
d344876 Relu -> leakyRelu Browse code 2023-06-21
bfdeaf1 discrminator is separated Browse code 2023-06-21
cecfc8d Creation of Attentive RNN GAN, change of naming scheme for the shake of conciseness. Browse code 2023-06-21