mirror of
https://github.com/AUTOMATIC1111/stable-diffusion-webui.git
synced 2025-02-01 11:13:00 +08:00
import Flux from https://github.com/black-forest-labs/flux/
License: Apache 2.0 original author: Tim Dockhorn @timudk
This commit is contained in:
Won-Kyu Park
parent
1b16c62608
commit
853551bd6e
30
modules/models/flux/math.py
Normal file
30
modules/models/flux/math.py
Normal file
@ -0,0 +1,30 @@
|
||||
import torch
|
||||
from einops import rearrange
|
||||
from torch import Tensor
|
||||
|
||||
|
||||
def attention(q: Tensor, k: Tensor, v: Tensor, pe: Tensor) -> Tensor:
|
||||
q, k = apply_rope(q, k, pe)
|
||||
|
||||
x = torch.nn.functional.scaled_dot_product_attention(q, k, v)
|
||||
x = rearrange(x, "B H L D -> B L (H D)")
|
||||
|
||||
return x
|
||||
|
||||
|
||||
def rope(pos: Tensor, dim: int, theta: int) -> Tensor:
|
||||
assert dim % 2 == 0
|
||||
scale = torch.arange(0, dim, 2, dtype=torch.float64, device=pos.device) / dim
|
||||
omega = 1.0 / (theta**scale)
|
||||
out = torch.einsum("...n,d->...nd", pos, omega)
|
||||
out = torch.stack([torch.cos(out), -torch.sin(out), torch.sin(out), torch.cos(out)], dim=-1)
|
||||
out = rearrange(out, "b n d (i j) -> b n d i j", i=2, j=2)
|
||||
return out.float()
|
||||
|
||||
|
||||
def apply_rope(xq: Tensor, xk: Tensor, freqs_cis: Tensor) -> tuple[Tensor, Tensor]:
|
||||
xq_ = xq.float().reshape(*xq.shape[:-1], -1, 1, 2)
|
||||
xk_ = xk.float().reshape(*xk.shape[:-1], -1, 1, 2)
|
||||
xq_out = freqs_cis[..., 0] * xq_[..., 0] + freqs_cis[..., 1] * xq_[..., 1]
|
||||
xk_out = freqs_cis[..., 0] * xk_[..., 0] + freqs_cis[..., 1] * xk_[..., 1]
|
||||
return xq_out.reshape(*xq.shape).type_as(xq), xk_out.reshape(*xk.shape).type_as(xk)
|
||||
112
modules/models/flux/model.py
Normal file
112
modules/models/flux/model.py
Normal file
@ -0,0 +1,112 @@
|
||||
from dataclasses import dataclass
|
||||
|
||||
import torch
|
||||
from torch import Tensor, nn
|
||||
|
||||
from flux.modules.layers import (DoubleStreamBlock, EmbedND, LastLayer,
|
||||
MLPEmbedder, SingleStreamBlock,
|
||||
timestep_embedding)
|
||||
|
||||
|
||||
@dataclass
|
||||
class FluxParams:
|
||||
in_channels: int
|
||||
vec_in_dim: int
|
||||
context_in_dim: int
|
||||
hidden_size: int
|
||||
mlp_ratio: float
|
||||
num_heads: int
|
||||
depth: int
|
||||
depth_single_blocks: int
|
||||
axes_dim: list[int]
|
||||
theta: int
|
||||
qkv_bias: bool
|
||||
guidance_embed: bool
|
||||
|
||||
|
||||
class Flux(nn.Module):
|
||||
"""
|
||||
Transformer model for flow matching on sequences.
|
||||
"""
|
||||
|
||||
def __init__(self, params: FluxParams):
|
||||
super().__init__()
|
||||
|
||||
self.params = params
|
||||
self.in_channels = params.in_channels
|
||||
self.out_channels = self.in_channels
|
||||
if params.hidden_size % params.num_heads != 0:
|
||||
raise ValueError(
|
||||
f"Hidden size {params.hidden_size} must be divisible by num_heads {params.num_heads}"
|
||||
)
|
||||
pe_dim = params.hidden_size // params.num_heads
|
||||
if sum(params.axes_dim) != pe_dim:
|
||||
raise ValueError(f"Got {params.axes_dim} but expected positional dim {pe_dim}")
|
||||
self.hidden_size = params.hidden_size
|
||||
self.num_heads = params.num_heads
|
||||
self.pe_embedder = EmbedND(dim=pe_dim, theta=params.theta, axes_dim=params.axes_dim)
|
||||
self.img_in = nn.Linear(self.in_channels, self.hidden_size, bias=True)
|
||||
self.time_in = MLPEmbedder(in_dim=256, hidden_dim=self.hidden_size)
|
||||
self.vector_in = MLPEmbedder(params.vec_in_dim, self.hidden_size)
|
||||
self.guidance_in = (
|
||||
MLPEmbedder(in_dim=256, hidden_dim=self.hidden_size) if params.guidance_embed else nn.Identity()
|
||||
)
|
||||
self.txt_in = nn.Linear(params.context_in_dim, self.hidden_size)
|
||||
|
||||
self.double_blocks = nn.ModuleList(
|
||||
[
|
||||
DoubleStreamBlock(
|
||||
self.hidden_size,
|
||||
self.num_heads,
|
||||
mlp_ratio=params.mlp_ratio,
|
||||
qkv_bias=params.qkv_bias,
|
||||
)
|
||||
for _ in range(params.depth)
|
||||
]
|
||||
)
|
||||
|
||||
self.single_blocks = nn.ModuleList(
|
||||
[
|
||||
SingleStreamBlock(self.hidden_size, self.num_heads, mlp_ratio=params.mlp_ratio)
|
||||
for _ in range(params.depth_single_blocks)
|
||||
]
|
||||
)
|
||||
|
||||
self.final_layer = LastLayer(self.hidden_size, 1, self.out_channels)
|
||||
|
||||
def forward(
|
||||
self,
|
||||
img: Tensor,
|
||||
img_ids: Tensor,
|
||||
txt: Tensor,
|
||||
txt_ids: Tensor,
|
||||
timesteps: Tensor,
|
||||
y: Tensor,
|
||||
guidance: Tensor | None = None,
|
||||
) -> Tensor:
|
||||
if img.ndim != 3 or txt.ndim != 3:
|
||||
raise ValueError("Input img and txt tensors must have 3 dimensions.")
|
||||
|
||||
# running on sequences img
|
||||
img = self.img_in(img)
|
||||
vec = self.time_in(timestep_embedding(timesteps, 256))
|
||||
if self.params.guidance_embed:
|
||||
if guidance is None:
|
||||
raise ValueError("Didn't get guidance strength for guidance distilled model.")
|
||||
vec = vec + self.guidance_in(timestep_embedding(guidance, 256))
|
||||
vec = vec + self.vector_in(y)
|
||||
txt = self.txt_in(txt)
|
||||
|
||||
ids = torch.cat((txt_ids, img_ids), dim=1)
|
||||
pe = self.pe_embedder(ids)
|
||||
|
||||
for block in self.double_blocks:
|
||||
img, txt = block(img=img, txt=txt, vec=vec, pe=pe)
|
||||
|
||||
img = torch.cat((txt, img), 1)
|
||||
for block in self.single_blocks:
|
||||
img = block(img, vec=vec, pe=pe)
|
||||
img = img[:, txt.shape[1] :, ...]
|
||||
|
||||
img = self.final_layer(img, vec) # (N, T, patch_size ** 2 * out_channels)
|
||||
return img
|
||||
253
modules/models/flux/modules/layers.py
Normal file
253
modules/models/flux/modules/layers.py
Normal file
@ -0,0 +1,253 @@
|
||||
import math
|
||||
from dataclasses import dataclass
|
||||
|
||||
import torch
|
||||
from einops import rearrange
|
||||
from torch import Tensor, nn
|
||||
|
||||
from flux.math import attention, rope
|
||||
|
||||
|
||||
class EmbedND(nn.Module):
|
||||
def __init__(self, dim: int, theta: int, axes_dim: list[int]):
|
||||
super().__init__()
|
||||
self.dim = dim
|
||||
self.theta = theta
|
||||
self.axes_dim = axes_dim
|
||||
|
||||
def forward(self, ids: Tensor) -> Tensor:
|
||||
n_axes = ids.shape[-1]
|
||||
emb = torch.cat(
|
||||
[rope(ids[..., i], self.axes_dim[i], self.theta) for i in range(n_axes)],
|
||||
dim=-3,
|
||||
)
|
||||
|
||||
return emb.unsqueeze(1)
|
||||
|
||||
|
||||
def timestep_embedding(t: Tensor, dim, max_period=10000, time_factor: float = 1000.0):
|
||||
"""
|
||||
Create sinusoidal timestep embeddings.
|
||||
:param t: a 1-D Tensor of N indices, one per batch element.
|
||||
These may be fractional.
|
||||
:param dim: the dimension of the output.
|
||||
:param max_period: controls the minimum frequency of the embeddings.
|
||||
:return: an (N, D) Tensor of positional embeddings.
|
||||
"""
|
||||
t = time_factor * t
|
||||
half = dim // 2
|
||||
freqs = torch.exp(-math.log(max_period) * torch.arange(start=0, end=half, dtype=torch.float32) / half).to(
|
||||
t.device
|
||||
)
|
||||
|
||||
args = t[:, None].float() * freqs[None]
|
||||
embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
|
||||
if dim % 2:
|
||||
embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
|
||||
if torch.is_floating_point(t):
|
||||
embedding = embedding.to(t)
|
||||
return embedding
|
||||
|
||||
|
||||
class MLPEmbedder(nn.Module):
|
||||
def __init__(self, in_dim: int, hidden_dim: int):
|
||||
super().__init__()
|
||||
self.in_layer = nn.Linear(in_dim, hidden_dim, bias=True)
|
||||
self.silu = nn.SiLU()
|
||||
self.out_layer = nn.Linear(hidden_dim, hidden_dim, bias=True)
|
||||
|
||||
def forward(self, x: Tensor) -> Tensor:
|
||||
return self.out_layer(self.silu(self.in_layer(x)))
|
||||
|
||||
|
||||
class RMSNorm(torch.nn.Module):
|
||||
def __init__(self, dim: int):
|
||||
super().__init__()
|
||||
self.scale = nn.Parameter(torch.ones(dim))
|
||||
|
||||
def forward(self, x: Tensor):
|
||||
x_dtype = x.dtype
|
||||
x = x.float()
|
||||
rrms = torch.rsqrt(torch.mean(x**2, dim=-1, keepdim=True) + 1e-6)
|
||||
return (x * rrms).to(dtype=x_dtype) * self.scale
|
||||
|
||||
|
||||
class QKNorm(torch.nn.Module):
|
||||
def __init__(self, dim: int):
|
||||
super().__init__()
|
||||
self.query_norm = RMSNorm(dim)
|
||||
self.key_norm = RMSNorm(dim)
|
||||
|
||||
def forward(self, q: Tensor, k: Tensor, v: Tensor) -> tuple[Tensor, Tensor]:
|
||||
q = self.query_norm(q)
|
||||
k = self.key_norm(k)
|
||||
return q.to(v), k.to(v)
|
||||
|
||||
|
||||
class SelfAttention(nn.Module):
|
||||
def __init__(self, dim: int, num_heads: int = 8, qkv_bias: bool = False):
|
||||
super().__init__()
|
||||
self.num_heads = num_heads
|
||||
head_dim = dim // num_heads
|
||||
|
||||
self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
|
||||
self.norm = QKNorm(head_dim)
|
||||
self.proj = nn.Linear(dim, dim)
|
||||
|
||||
def forward(self, x: Tensor, pe: Tensor) -> Tensor:
|
||||
qkv = self.qkv(x)
|
||||
q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
|
||||
q, k = self.norm(q, k, v)
|
||||
x = attention(q, k, v, pe=pe)
|
||||
x = self.proj(x)
|
||||
return x
|
||||
|
||||
|
||||
@dataclass
|
||||
class ModulationOut:
|
||||
shift: Tensor
|
||||
scale: Tensor
|
||||
gate: Tensor
|
||||
|
||||
|
||||
class Modulation(nn.Module):
|
||||
def __init__(self, dim: int, double: bool):
|
||||
super().__init__()
|
||||
self.is_double = double
|
||||
self.multiplier = 6 if double else 3
|
||||
self.lin = nn.Linear(dim, self.multiplier * dim, bias=True)
|
||||
|
||||
def forward(self, vec: Tensor) -> tuple[ModulationOut, ModulationOut | None]:
|
||||
out = self.lin(nn.functional.silu(vec))[:, None, :].chunk(self.multiplier, dim=-1)
|
||||
|
||||
return (
|
||||
ModulationOut(*out[:3]),
|
||||
ModulationOut(*out[3:]) if self.is_double else None,
|
||||
)
|
||||
|
||||
|
||||
class DoubleStreamBlock(nn.Module):
|
||||
def __init__(self, hidden_size: int, num_heads: int, mlp_ratio: float, qkv_bias: bool = False):
|
||||
super().__init__()
|
||||
|
||||
mlp_hidden_dim = int(hidden_size * mlp_ratio)
|
||||
self.num_heads = num_heads
|
||||
self.hidden_size = hidden_size
|
||||
self.img_mod = Modulation(hidden_size, double=True)
|
||||
self.img_norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
|
||||
self.img_attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias)
|
||||
|
||||
self.img_norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
|
||||
self.img_mlp = nn.Sequential(
|
||||
nn.Linear(hidden_size, mlp_hidden_dim, bias=True),
|
||||
nn.GELU(approximate="tanh"),
|
||||
nn.Linear(mlp_hidden_dim, hidden_size, bias=True),
|
||||
)
|
||||
|
||||
self.txt_mod = Modulation(hidden_size, double=True)
|
||||
self.txt_norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
|
||||
self.txt_attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias)
|
||||
|
||||
self.txt_norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
|
||||
self.txt_mlp = nn.Sequential(
|
||||
nn.Linear(hidden_size, mlp_hidden_dim, bias=True),
|
||||
nn.GELU(approximate="tanh"),
|
||||
nn.Linear(mlp_hidden_dim, hidden_size, bias=True),
|
||||
)
|
||||
|
||||
def forward(self, img: Tensor, txt: Tensor, vec: Tensor, pe: Tensor) -> tuple[Tensor, Tensor]:
|
||||
img_mod1, img_mod2 = self.img_mod(vec)
|
||||
txt_mod1, txt_mod2 = self.txt_mod(vec)
|
||||
|
||||
# prepare image for attention
|
||||
img_modulated = self.img_norm1(img)
|
||||
img_modulated = (1 + img_mod1.scale) * img_modulated + img_mod1.shift
|
||||
img_qkv = self.img_attn.qkv(img_modulated)
|
||||
img_q, img_k, img_v = rearrange(img_qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
|
||||
img_q, img_k = self.img_attn.norm(img_q, img_k, img_v)
|
||||
|
||||
# prepare txt for attention
|
||||
txt_modulated = self.txt_norm1(txt)
|
||||
txt_modulated = (1 + txt_mod1.scale) * txt_modulated + txt_mod1.shift
|
||||
txt_qkv = self.txt_attn.qkv(txt_modulated)
|
||||
txt_q, txt_k, txt_v = rearrange(txt_qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
|
||||
txt_q, txt_k = self.txt_attn.norm(txt_q, txt_k, txt_v)
|
||||
|
||||
# run actual attention
|
||||
q = torch.cat((txt_q, img_q), dim=2)
|
||||
k = torch.cat((txt_k, img_k), dim=2)
|
||||
v = torch.cat((txt_v, img_v), dim=2)
|
||||
|
||||
attn = attention(q, k, v, pe=pe)
|
||||
txt_attn, img_attn = attn[:, : txt.shape[1]], attn[:, txt.shape[1] :]
|
||||
|
||||
# calculate the img bloks
|
||||
img = img + img_mod1.gate * self.img_attn.proj(img_attn)
|
||||
img = img + img_mod2.gate * self.img_mlp((1 + img_mod2.scale) * self.img_norm2(img) + img_mod2.shift)
|
||||
|
||||
# calculate the txt bloks
|
||||
txt = txt + txt_mod1.gate * self.txt_attn.proj(txt_attn)
|
||||
txt = txt + txt_mod2.gate * self.txt_mlp((1 + txt_mod2.scale) * self.txt_norm2(txt) + txt_mod2.shift)
|
||||
return img, txt
|
||||
|
||||
|
||||
class SingleStreamBlock(nn.Module):
|
||||
"""
|
||||
A DiT block with parallel linear layers as described in
|
||||
https://arxiv.org/abs/2302.05442 and adapted modulation interface.
|
||||
"""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
hidden_size: int,
|
||||
num_heads: int,
|
||||
mlp_ratio: float = 4.0,
|
||||
qk_scale: float | None = None,
|
||||
):
|
||||
super().__init__()
|
||||
self.hidden_dim = hidden_size
|
||||
self.num_heads = num_heads
|
||||
head_dim = hidden_size // num_heads
|
||||
self.scale = qk_scale or head_dim**-0.5
|
||||
|
||||
self.mlp_hidden_dim = int(hidden_size * mlp_ratio)
|
||||
# qkv and mlp_in
|
||||
self.linear1 = nn.Linear(hidden_size, hidden_size * 3 + self.mlp_hidden_dim)
|
||||
# proj and mlp_out
|
||||
self.linear2 = nn.Linear(hidden_size + self.mlp_hidden_dim, hidden_size)
|
||||
|
||||
self.norm = QKNorm(head_dim)
|
||||
|
||||
self.hidden_size = hidden_size
|
||||
self.pre_norm = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
|
||||
|
||||
self.mlp_act = nn.GELU(approximate="tanh")
|
||||
self.modulation = Modulation(hidden_size, double=False)
|
||||
|
||||
def forward(self, x: Tensor, vec: Tensor, pe: Tensor) -> Tensor:
|
||||
mod, _ = self.modulation(vec)
|
||||
x_mod = (1 + mod.scale) * self.pre_norm(x) + mod.shift
|
||||
qkv, mlp = torch.split(self.linear1(x_mod), [3 * self.hidden_size, self.mlp_hidden_dim], dim=-1)
|
||||
|
||||
q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
|
||||
q, k = self.norm(q, k, v)
|
||||
|
||||
# compute attention
|
||||
attn = attention(q, k, v, pe=pe)
|
||||
# compute activation in mlp stream, cat again and run second linear layer
|
||||
output = self.linear2(torch.cat((attn, self.mlp_act(mlp)), 2))
|
||||
return x + mod.gate * output
|
||||
|
||||
|
||||
class LastLayer(nn.Module):
|
||||
def __init__(self, hidden_size: int, patch_size: int, out_channels: int):
|
||||
super().__init__()
|
||||
self.norm_final = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
|
||||
self.linear = nn.Linear(hidden_size, patch_size * patch_size * out_channels, bias=True)
|
||||
self.adaLN_modulation = nn.Sequential(nn.SiLU(), nn.Linear(hidden_size, 2 * hidden_size, bias=True))
|
||||
|
||||
def forward(self, x: Tensor, vec: Tensor) -> Tensor:
|
||||
shift, scale = self.adaLN_modulation(vec).chunk(2, dim=1)
|
||||
x = (1 + scale[:, None, :]) * self.norm_final(x) + shift[:, None, :]
|
||||
x = self.linear(x)
|
||||
return x
|
||||
201
modules/models/flux/util.py
Normal file
201
modules/models/flux/util.py
Normal file
@ -0,0 +1,201 @@
|
||||
import os
|
||||
from dataclasses import dataclass
|
||||
|
||||
import torch
|
||||
from einops import rearrange
|
||||
from huggingface_hub import hf_hub_download
|
||||
from imwatermark import WatermarkEncoder
|
||||
from safetensors.torch import load_file as load_sft
|
||||
|
||||
from flux.model import Flux, FluxParams
|
||||
from flux.modules.autoencoder import AutoEncoder, AutoEncoderParams
|
||||
from flux.modules.conditioner import HFEmbedder
|
||||
|
||||
|
||||
@dataclass
|
||||
class ModelSpec:
|
||||
params: FluxParams
|
||||
ae_params: AutoEncoderParams
|
||||
ckpt_path: str | None
|
||||
ae_path: str | None
|
||||
repo_id: str | None
|
||||
repo_flow: str | None
|
||||
repo_ae: str | None
|
||||
|
||||
|
||||
configs = {
|
||||
"flux-dev": ModelSpec(
|
||||
repo_id="black-forest-labs/FLUX.1-dev",
|
||||
repo_flow="flux1-dev.safetensors",
|
||||
repo_ae="ae.safetensors",
|
||||
ckpt_path=os.getenv("FLUX_DEV"),
|
||||
params=FluxParams(
|
||||
in_channels=64,
|
||||
vec_in_dim=768,
|
||||
context_in_dim=4096,
|
||||
hidden_size=3072,
|
||||
mlp_ratio=4.0,
|
||||
num_heads=24,
|
||||
depth=19,
|
||||
depth_single_blocks=38,
|
||||
axes_dim=[16, 56, 56],
|
||||
theta=10_000,
|
||||
qkv_bias=True,
|
||||
guidance_embed=True,
|
||||
),
|
||||
ae_path=os.getenv("AE"),
|
||||
ae_params=AutoEncoderParams(
|
||||
resolution=256,
|
||||
in_channels=3,
|
||||
ch=128,
|
||||
out_ch=3,
|
||||
ch_mult=[1, 2, 4, 4],
|
||||
num_res_blocks=2,
|
||||
z_channels=16,
|
||||
scale_factor=0.3611,
|
||||
shift_factor=0.1159,
|
||||
),
|
||||
),
|
||||
"flux-schnell": ModelSpec(
|
||||
repo_id="black-forest-labs/FLUX.1-schnell",
|
||||
repo_flow="flux1-schnell.safetensors",
|
||||
repo_ae="ae.safetensors",
|
||||
ckpt_path=os.getenv("FLUX_SCHNELL"),
|
||||
params=FluxParams(
|
||||
in_channels=64,
|
||||
vec_in_dim=768,
|
||||
context_in_dim=4096,
|
||||
hidden_size=3072,
|
||||
mlp_ratio=4.0,
|
||||
num_heads=24,
|
||||
depth=19,
|
||||
depth_single_blocks=38,
|
||||
axes_dim=[16, 56, 56],
|
||||
theta=10_000,
|
||||
qkv_bias=True,
|
||||
guidance_embed=False,
|
||||
),
|
||||
ae_path=os.getenv("AE"),
|
||||
ae_params=AutoEncoderParams(
|
||||
resolution=256,
|
||||
in_channels=3,
|
||||
ch=128,
|
||||
out_ch=3,
|
||||
ch_mult=[1, 2, 4, 4],
|
||||
num_res_blocks=2,
|
||||
z_channels=16,
|
||||
scale_factor=0.3611,
|
||||
shift_factor=0.1159,
|
||||
),
|
||||
),
|
||||
}
|
||||
|
||||
|
||||
def print_load_warning(missing: list[str], unexpected: list[str]) -> None:
|
||||
if len(missing) > 0 and len(unexpected) > 0:
|
||||
print(f"Got {len(missing)} missing keys:\n\t" + "\n\t".join(missing))
|
||||
print("\n" + "-" * 79 + "\n")
|
||||
print(f"Got {len(unexpected)} unexpected keys:\n\t" + "\n\t".join(unexpected))
|
||||
elif len(missing) > 0:
|
||||
print(f"Got {len(missing)} missing keys:\n\t" + "\n\t".join(missing))
|
||||
elif len(unexpected) > 0:
|
||||
print(f"Got {len(unexpected)} unexpected keys:\n\t" + "\n\t".join(unexpected))
|
||||
|
||||
|
||||
def load_flow_model(name: str, device: str | torch.device = "cuda", hf_download: bool = True):
|
||||
# Loading Flux
|
||||
print("Init model")
|
||||
ckpt_path = configs[name].ckpt_path
|
||||
if (
|
||||
ckpt_path is None
|
||||
and configs[name].repo_id is not None
|
||||
and configs[name].repo_flow is not None
|
||||
and hf_download
|
||||
):
|
||||
ckpt_path = hf_hub_download(configs[name].repo_id, configs[name].repo_flow)
|
||||
|
||||
with torch.device("meta" if ckpt_path is not None else device):
|
||||
model = Flux(configs[name].params).to(torch.bfloat16)
|
||||
|
||||
if ckpt_path is not None:
|
||||
print("Loading checkpoint")
|
||||
# load_sft doesn't support torch.device
|
||||
sd = load_sft(ckpt_path, device=str(device))
|
||||
missing, unexpected = model.load_state_dict(sd, strict=False, assign=True)
|
||||
print_load_warning(missing, unexpected)
|
||||
return model
|
||||
|
||||
|
||||
def load_t5(device: str | torch.device = "cuda", max_length: int = 512) -> HFEmbedder:
|
||||
# max length 64, 128, 256 and 512 should work (if your sequence is short enough)
|
||||
return HFEmbedder("google/t5-v1_1-xxl", max_length=max_length, torch_dtype=torch.bfloat16).to(device)
|
||||
|
||||
|
||||
def load_clip(device: str | torch.device = "cuda") -> HFEmbedder:
|
||||
return HFEmbedder("openai/clip-vit-large-patch14", max_length=77, torch_dtype=torch.bfloat16).to(device)
|
||||
|
||||
|
||||
def load_ae(name: str, device: str | torch.device = "cuda", hf_download: bool = True) -> AutoEncoder:
|
||||
ckpt_path = configs[name].ae_path
|
||||
if (
|
||||
ckpt_path is None
|
||||
and configs[name].repo_id is not None
|
||||
and configs[name].repo_ae is not None
|
||||
and hf_download
|
||||
):
|
||||
ckpt_path = hf_hub_download(configs[name].repo_id, configs[name].repo_ae)
|
||||
|
||||
# Loading the autoencoder
|
||||
print("Init AE")
|
||||
with torch.device("meta" if ckpt_path is not None else device):
|
||||
ae = AutoEncoder(configs[name].ae_params)
|
||||
|
||||
if ckpt_path is not None:
|
||||
sd = load_sft(ckpt_path, device=str(device))
|
||||
missing, unexpected = ae.load_state_dict(sd, strict=False, assign=True)
|
||||
print_load_warning(missing, unexpected)
|
||||
return ae
|
||||
|
||||
|
||||
class WatermarkEmbedder:
|
||||
def __init__(self, watermark):
|
||||
self.watermark = watermark
|
||||
self.num_bits = len(WATERMARK_BITS)
|
||||
self.encoder = WatermarkEncoder()
|
||||
self.encoder.set_watermark("bits", self.watermark)
|
||||
|
||||
def __call__(self, image: torch.Tensor) -> torch.Tensor:
|
||||
"""
|
||||
Adds a predefined watermark to the input image
|
||||
|
||||
Args:
|
||||
image: ([N,] B, RGB, H, W) in range [-1, 1]
|
||||
|
||||
Returns:
|
||||
same as input but watermarked
|
||||
"""
|
||||
image = 0.5 * image + 0.5
|
||||
squeeze = len(image.shape) == 4
|
||||
if squeeze:
|
||||
image = image[None, ...]
|
||||
n = image.shape[0]
|
||||
image_np = rearrange((255 * image).detach().cpu(), "n b c h w -> (n b) h w c").numpy()[:, :, :, ::-1]
|
||||
# torch (b, c, h, w) in [0, 1] -> numpy (b, h, w, c) [0, 255]
|
||||
# watermarking libary expects input as cv2 BGR format
|
||||
for k in range(image_np.shape[0]):
|
||||
image_np[k] = self.encoder.encode(image_np[k], "dwtDct")
|
||||
image = torch.from_numpy(rearrange(image_np[:, :, :, ::-1], "(n b) h w c -> n b c h w", n=n)).to(
|
||||
image.device
|
||||
)
|
||||
image = torch.clamp(image / 255, min=0.0, max=1.0)
|
||||
if squeeze:
|
||||
image = image[0]
|
||||
image = 2 * image - 1
|
||||
return image
|
||||
|
||||
|
||||
# A fixed 48-bit message that was chosen at random
|
||||
WATERMARK_MESSAGE = 0b001010101111111010000111100111001111010100101110
|
||||
# bin(x)[2:] gives bits of x as str, use int to convert them to 0/1
|
||||
WATERMARK_BITS = [int(bit) for bit in bin(WATERMARK_MESSAGE)[2:]]
|
||||
embed_watermark = WatermarkEmbedder(WATERMARK_BITS)
|
||||
Loading…
Reference in New Issue
Block a user