Skip to content

Fastgan patch #108

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
victor-navarro wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Fastgan patch #108

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
5386a6a
added messy Dockerfile
victor-navarro May 3, 2024
7161b21
fastgan patch for big images
victor-navarro Jun 4, 2024
d758dce
removing Dockerfile
victor-navarro Jun 4, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
83 changes: 60 additions & 23 deletions pg_modules/networks_fastgan.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,15 @@
# modified by Axel Sauer for "Projected GANs Converge Faster"
#
import torch.nn as nn
from pg_modules.blocks import (InitLayer, UpBlockBig, UpBlockBigCond, UpBlockSmall, UpBlockSmallCond, SEBlock, conv2d)
from pg_modules.blocks import (
InitLayer,
UpBlockBig,
UpBlockBigCond,
UpBlockSmall,
UpBlockSmallCond,
SEBlock,
conv2d,
)


def normalize_second_moment(x, dim=1, eps=1e-8):
Expand All @@ -25,25 +33,35 @@ def __init__(self, ngf=128, z_dim=256, nc=3, img_resolution=256, lite=False):
self.z_dim = z_dim

# channel multiplier
nfc_multi = {2: 16, 4:16, 8:8, 16:4, 32:2, 64:2, 128:1, 256:0.5,
512:0.25, 1024:0.125}
nfc_multi = {
2: 16,
4: 16,
8: 8,
16: 4,
32: 2,
64: 2,
128: 1,
256: 0.5,
512: 0.25,
1024: 0.125,
}
nfc = {}
for k, v in nfc_multi.items():
nfc[k] = int(v*ngf)
nfc[k] = int(v * ngf)

# layers
self.init = InitLayer(z_dim, channel=nfc[2], sz=4)

UpBlock = UpBlockSmall if lite else UpBlockBig

self.feat_8 = UpBlock(nfc[4], nfc[8])
self.feat_16 = UpBlock(nfc[8], nfc[16])
self.feat_32 = UpBlock(nfc[16], nfc[32])
self.feat_64 = UpBlock(nfc[32], nfc[64])
self.feat_8 = UpBlock(nfc[4], nfc[8])
self.feat_16 = UpBlock(nfc[8], nfc[16])
self.feat_32 = UpBlock(nfc[16], nfc[32])
self.feat_64 = UpBlock(nfc[32], nfc[64])
self.feat_128 = UpBlock(nfc[64], nfc[128])
self.feat_256 = UpBlock(nfc[128], nfc[256])

self.se_64 = SEBlock(nfc[4], nfc[64])
self.se_64 = SEBlock(nfc[4], nfc[64])
self.se_128 = SEBlock(nfc[8], nfc[128])
self.se_256 = SEBlock(nfc[16], nfc[256])

Expand All @@ -64,7 +82,7 @@ def forward(self, input, c, **kwargs):
feat_16 = self.feat_16(feat_8)
feat_32 = self.feat_32(feat_16)
feat_64 = self.se_64(feat_4, self.feat_64(feat_32))
feat_128 = self.se_128(feat_8, self.feat_128(feat_64))
feat_128 = self.se_128(feat_8, self.feat_128(feat_64))

if self.img_resolution >= 128:
feat_last = feat_128
Expand All @@ -82,26 +100,39 @@ def forward(self, input, c, **kwargs):


class FastganSynthesisCond(nn.Module):
def __init__(self, ngf=64, z_dim=256, nc=3, img_resolution=256, num_classes=1000, lite=False):
def __init__(
self, ngf=64, z_dim=256, nc=3, img_resolution=256, num_classes=1000, lite=False
):
super().__init__()

self.z_dim = z_dim
nfc_multi = {2: 16, 4:16, 8:8, 16:4, 32:2, 64:2, 128:1, 256:0.5,
512:0.25, 1024:0.125, 2048:0.125}
nfc_multi = {
2: 16,
4: 16,
8: 8,
16: 4,
32: 2,
64: 2,
128: 1,
256: 0.5,
512: 0.25,
1024: 0.125,
2048: 0.125,
}
nfc = {}
for k, v in nfc_multi.items():
nfc[k] = int(v*ngf)
nfc[k] = int(v * ngf)

self.img_resolution = img_resolution

self.init = InitLayer(z_dim, channel=nfc[2], sz=4)

UpBlock = UpBlockSmallCond if lite else UpBlockBigCond

self.feat_8 = UpBlock(nfc[4], nfc[8], z_dim)
self.feat_16 = UpBlock(nfc[8], nfc[16], z_dim)
self.feat_32 = UpBlock(nfc[16], nfc[32], z_dim)
self.feat_64 = UpBlock(nfc[32], nfc[64], z_dim)
self.feat_8 = UpBlock(nfc[4], nfc[8], z_dim)
self.feat_16 = UpBlock(nfc[8], nfc[16], z_dim)
self.feat_32 = UpBlock(nfc[16], nfc[32], z_dim)
self.feat_64 = UpBlock(nfc[32], nfc[64], z_dim)
self.feat_128 = UpBlock(nfc[64], nfc[128], z_dim)
self.feat_256 = UpBlock(nfc[128], nfc[256], z_dim)

Expand All @@ -112,10 +143,10 @@ def __init__(self, ngf=64, z_dim=256, nc=3, img_resolution=256, num_classes=1000
self.to_big = conv2d(nfc[img_resolution], nc, 3, 1, 1, bias=True)

if img_resolution > 256:
self.feat_512 = UpBlock(nfc[256], nfc[512])
self.feat_512 = UpBlock(nfc[256], nfc[512], z_dim)
self.se_512 = SEBlock(nfc[32], nfc[512])
if img_resolution > 512:
self.feat_1024 = UpBlock(nfc[512], nfc[1024])
self.feat_1024 = UpBlock(nfc[512], nfc[1024], z_dim)

self.embed = nn.Embedding(num_classes, z_dim)

Expand All @@ -130,7 +161,7 @@ def forward(self, input, c, update_emas=False):
feat_16 = self.feat_16(feat_8, c)
feat_32 = self.feat_32(feat_16, c)
feat_64 = self.se_64(feat_4, self.feat_64(feat_32, c))
feat_128 = self.se_128(feat_8, self.feat_128(feat_64, c))
feat_128 = self.se_128(feat_8, self.feat_128(feat_64, c))

if self.img_resolution >= 128:
feat_last = feat_128
Expand Down Expand Up @@ -158,7 +189,7 @@ def __init__(
ngf=128,
cond=0,
mapping_kwargs={},
synthesis_kwargs={}
synthesis_kwargs={},
):
super().__init__()
self.z_dim = z_dim
Expand All @@ -170,7 +201,13 @@ def __init__(
# Mapping and Synthesis Networks
self.mapping = DummyMapping() # to fit the StyleGAN API
Synthesis = FastganSynthesisCond if cond else FastganSynthesis
self.synthesis = Synthesis(ngf=ngf, z_dim=z_dim, nc=img_channels, img_resolution=img_resolution, **synthesis_kwargs)
self.synthesis = Synthesis(
ngf=ngf,
z_dim=z_dim,
nc=img_channels,
img_resolution=img_resolution,
**synthesis_kwargs
)

def forward(self, z, c, **kwargs):
w = self.mapping(z, c)
Expand Down