Commit History
-
Sloving conflicts
88f49f7 -
Re training will finish this time for sure
8082844
-
Halfway through
de7e31f
-
Modify training code
bba6806 -
Remove logging statements
2473850 -
Print statements for logging
43b6c80 -
Add num_heads variables
Yash Nagraj92078be -
Make changes
Yash Nagrajf0ff580 -
Resolve run time errors
8d85e1f -
Get the buffers in order to open as image
Yash Nagraj199f027 -
Change any output errors
788a6e1
-
Move the pretrained vgg16 model to another directory
Yash Nagrajf6182a3 -
Add forward function to the Diffusion model UNet
Yash Nagrajdff68d4 -
Build the Unet from the config
Yash Nagraj6966d99 -
change variable names
Yash Nagraj2e8c272 -
Get all parameters from model_config
Yash Nagraj97679c9 -
Add discriminator for adverserial loss for VAE training
Yash Nagrajec937a4
-
Add lpips for VAE training
Yash Nagrajb311ae5 -
Clean up
Yash Nagraj2a1c7f2 -
Add AutoEncoder (VQVAE)
Yash Nagraj7377e9c
-
Add UpBlock Unet to concat downblock's output
Yash Nagraj70a401a -
Add all the layers in the UpBlock
Yash Nagraj6801d5a
-
Add forward function to the midblock
Yash Nagrajde5e356 -
Add residual convolution to the mid block
Yash Nagraj38d054a
-
Add MidBlocks and change cross_attn in down blocks
Yash Nagrajc2857b5 -
Add cross attn if needed for conditional latent diffusion
Yash Nagraj95b2cf2 -
Add forward function with attention
Yash Nagrajaee1300
-
Add cross attention and down sample layers
Yash Nagraj1c92417 -
Add attention to Down Blocks
Yash Nagrajcb6bd3a