File size: 9,625 Bytes
d39ef9e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 | # Copyright (c) Black Forest Labs.
# Copyright (c) 2026 The GeoSET Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
# Adapted from black-forest-labs/flux2 (Apache-2.0). Modified by the GeoSET authors.
"""Frozen FLUX.2 autoencoder with the packed (2x2 space-to-depth), BatchNorm-normalised 128-channel latent."""
from __future__ import annotations
import torch
from diffusers.configuration_utils import ConfigMixin, register_to_config
from diffusers.models.modeling_utils import ModelMixin
from einops import rearrange
from torch import Tensor, nn
def swish(x: Tensor) -> Tensor:
return x * torch.sigmoid(x)
class AttnBlock(nn.Module):
def __init__(self, in_channels: int):
super().__init__()
self.norm = nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)
self.q = nn.Conv2d(in_channels, in_channels, kernel_size=1)
self.k = nn.Conv2d(in_channels, in_channels, kernel_size=1)
self.v = nn.Conv2d(in_channels, in_channels, kernel_size=1)
self.proj_out = nn.Conv2d(in_channels, in_channels, kernel_size=1)
def forward(self, x: Tensor) -> Tensor:
h_ = self.norm(x)
q, k, v = self.q(h_), self.k(h_), self.v(h_)
b, c, h, w = q.shape
q, k, v = (rearrange(t, "b c h w -> b 1 (h w) c").contiguous() for t in (q, k, v))
h_ = nn.functional.scaled_dot_product_attention(q, k, v)
return x + self.proj_out(rearrange(h_, "b 1 (h w) c -> b c h w", h=h, w=w, c=c, b=b))
class ResnetBlock(nn.Module):
def __init__(self, in_channels: int, out_channels: int):
super().__init__()
self.in_channels = in_channels
self.out_channels = out_channels
self.norm1 = nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)
self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1, padding=1)
self.norm2 = nn.GroupNorm(num_groups=32, num_channels=out_channels, eps=1e-6, affine=True)
self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=1, padding=1)
if in_channels != out_channels:
self.nin_shortcut = nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, padding=0)
def forward(self, x: Tensor) -> Tensor:
h = self.conv1(swish(self.norm1(x)))
h = self.conv2(swish(self.norm2(h)))
if self.in_channels != self.out_channels:
x = self.nin_shortcut(x)
return x + h
class Downsample(nn.Module):
def __init__(self, in_channels: int):
super().__init__()
self.conv = nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=2, padding=0)
def forward(self, x: Tensor) -> Tensor:
return self.conv(nn.functional.pad(x, (0, 1, 0, 1), mode="constant", value=0))
class Upsample(nn.Module):
def __init__(self, in_channels: int):
super().__init__()
self.conv = nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=1, padding=1)
def forward(self, x: Tensor) -> Tensor:
return self.conv(nn.functional.interpolate(x, scale_factor=2.0, mode="nearest"))
class Encoder(nn.Module):
"""Image -> latent moments (mean and log-variance, 2 * z_channels at 1/8 resolution)."""
def __init__(self, in_channels: int, ch: int, ch_mult: list[int], num_res_blocks: int, z_channels: int):
super().__init__()
self.quant_conv = nn.Conv2d(2 * z_channels, 2 * z_channels, 1)
self.num_resolutions = len(ch_mult)
self.num_res_blocks = num_res_blocks
self.conv_in = nn.Conv2d(in_channels, ch, kernel_size=3, stride=1, padding=1)
in_ch_mult = (1,) + tuple(ch_mult)
self.down = nn.ModuleList()
for i_level in range(self.num_resolutions):
block = nn.ModuleList()
block_in = ch * in_ch_mult[i_level]
block_out = ch * ch_mult[i_level]
for _ in range(num_res_blocks):
block.append(ResnetBlock(block_in, block_out))
block_in = block_out
down = nn.Module()
down.block = block
if i_level != self.num_resolutions - 1:
down.downsample = Downsample(block_in)
self.down.append(down)
self.mid = nn.Module()
self.mid.block_1 = ResnetBlock(block_in, block_in)
self.mid.attn_1 = AttnBlock(block_in)
self.mid.block_2 = ResnetBlock(block_in, block_in)
self.norm_out = nn.GroupNorm(num_groups=32, num_channels=block_in, eps=1e-6, affine=True)
self.conv_out = nn.Conv2d(block_in, 2 * z_channels, kernel_size=3, stride=1, padding=1)
def forward(self, x: Tensor) -> Tensor:
h = self.conv_in(x)
for i_level in range(self.num_resolutions):
for i_block in range(self.num_res_blocks):
h = self.down[i_level].block[i_block](h)
if i_level != self.num_resolutions - 1:
h = self.down[i_level].downsample(h)
h = self.mid.block_2(self.mid.attn_1(self.mid.block_1(h)))
return self.quant_conv(self.conv_out(swish(self.norm_out(h))))
class Decoder(nn.Module):
"""Latent (z_channels at 1/8 resolution) -> image."""
def __init__(self, ch: int, out_ch: int, ch_mult: list[int], num_res_blocks: int, z_channels: int):
super().__init__()
self.post_quant_conv = nn.Conv2d(z_channels, z_channels, 1)
self.num_resolutions = len(ch_mult)
self.num_res_blocks = num_res_blocks
block_in = ch * ch_mult[-1]
self.conv_in = nn.Conv2d(z_channels, block_in, kernel_size=3, stride=1, padding=1)
self.mid = nn.Module()
self.mid.block_1 = ResnetBlock(block_in, block_in)
self.mid.attn_1 = AttnBlock(block_in)
self.mid.block_2 = ResnetBlock(block_in, block_in)
self.up = nn.ModuleList()
for i_level in reversed(range(self.num_resolutions)):
block = nn.ModuleList()
block_out = ch * ch_mult[i_level]
for _ in range(num_res_blocks + 1):
block.append(ResnetBlock(block_in, block_out))
block_in = block_out
up = nn.Module()
up.block = block
if i_level != 0:
up.upsample = Upsample(block_in)
self.up.insert(0, up)
self.norm_out = nn.GroupNorm(num_groups=32, num_channels=block_in, eps=1e-6, affine=True)
self.conv_out = nn.Conv2d(block_in, out_ch, kernel_size=3, stride=1, padding=1)
def trunk(self, z: Tensor) -> Tensor:
"""Every layer except ``conv_out``: returns the ``ch``-channel features at full resolution."""
h = self.conv_in(self.post_quant_conv(z))
h = self.mid.block_2(self.mid.attn_1(self.mid.block_1(h)))
h = h.to(next(self.up.parameters()).dtype)
for i_level in reversed(range(self.num_resolutions)):
for i_block in range(self.num_res_blocks + 1):
h = self.up[i_level].block[i_block](h)
if i_level != 0:
h = self.up[i_level].upsample(h)
return swish(self.norm_out(h))
def forward(self, z: Tensor) -> Tensor:
return self.conv_out(self.trunk(z))
class AutoencoderFlux2(ModelMixin, ConfigMixin):
"""FLUX.2 autoencoder: ``encode`` maps ``[B, 3, H, W]`` in [-1, 1] to ``[B, 128, H/16, W/16]``, ``decode`` inverts it.
The latent is the posterior mean, packed 2x2 space-to-depth and normalised by a BatchNorm with fixed running
statistics. The module is frozen and always in eval mode.
"""
@register_to_config
def __init__(
self,
in_channels: int = 3,
out_ch: int = 3,
ch: int = 128,
ch_mult: tuple[int, ...] = (1, 2, 4, 4),
num_res_blocks: int = 2,
z_channels: int = 32,
):
super().__init__()
self.encoder = Encoder(in_channels, ch, list(ch_mult), num_res_blocks, z_channels)
self.decoder = Decoder(ch, out_ch, list(ch_mult), num_res_blocks, z_channels)
self.bn = nn.BatchNorm2d(4 * z_channels, eps=1e-4, momentum=0.1, affine=False, track_running_stats=True)
self.requires_grad_(False)
self.eval()
@staticmethod
def pack(z: Tensor) -> Tensor:
return rearrange(z, "... c (i pi) (j pj) -> ... (c pi pj) i j", pi=2, pj=2)
@staticmethod
def unpack(z: Tensor) -> Tensor:
return rearrange(z, "... (c pi pj) i j -> ... c (i pi) (j pj)", pi=2, pj=2)
def normalize(self, z: Tensor) -> Tensor:
self.bn.eval()
return self.bn(z)
def inv_normalize(self, z: Tensor) -> Tensor:
self.bn.eval()
s = torch.sqrt(self.bn.running_var.view(1, -1, 1, 1) + self.bn.eps)
m = self.bn.running_mean.view(1, -1, 1, 1)
return z * s + m
def encode(self, x: Tensor) -> Tensor:
moments = self.encoder(x)
return self.normalize(self.pack(torch.chunk(moments, 2, dim=1)[0]))
def decode(self, z: Tensor) -> Tensor:
return self.decoder(self.unpack(self.inv_normalize(z)))
def train(self, mode: bool = True) -> "AutoencoderFlux2":
return super().train(False)
|