File size: 8,530 Bytes
d0ffe9c |
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 |
import logging
from functools import wraps
from pathlib import Path
from typing import Optional, TypeVar
from diffusers import StableDiffusionPipeline, StableDiffusionXLPipeline
from huggingface_hub import hf_hub_download
from torch import nn
from animatediff import HF_HUB_CACHE, HF_MODULE_REPO, get_dir
from animatediff.settings import CKPT_EXTENSIONS
from animatediff.utils.huggingface import get_hf_pipeline, get_hf_pipeline_sdxl
from animatediff.utils.util import path_from_cwd
logger = logging.getLogger(__name__)
data_dir = get_dir("data")
checkpoint_dir = data_dir.joinpath("models/sd")
pipeline_dir = data_dir.joinpath("models/huggingface")
# for the nop_train() monkeypatch
T = TypeVar("T", bound=nn.Module)
def nop_train(self: T, mode: bool = True) -> T:
"""No-op for monkeypatching train() call to prevent unfreezing module"""
return self
def get_base_model(model_name_or_path: str, local_dir: Path, force: bool = False, is_sdxl:bool=False) -> Path:
model_name_or_path = Path(model_name_or_path)
model_save_dir = local_dir.joinpath(str(model_name_or_path).split("/")[-1]).resolve()
model_is_repo_id = False if model_name_or_path.joinpath("model_index.json").exists() else True
# if we have a HF repo ID, download it
if model_is_repo_id:
logger.debug("Base model is a HuggingFace repo ID")
if model_save_dir.joinpath("model_index.json").exists():
logger.debug(f"Base model already downloaded to: {path_from_cwd(model_save_dir)}")
else:
logger.info(f"Downloading base model from {model_name_or_path}...")
if is_sdxl:
_ = get_hf_pipeline_sdxl(model_name_or_path, model_save_dir, save=True, force_download=force)
else:
_ = get_hf_pipeline(model_name_or_path, model_save_dir, save=True, force_download=force)
model_name_or_path = model_save_dir
return Path(model_name_or_path)
def fix_checkpoint_if_needed(checkpoint: Path, debug:bool):
def dump(loaded):
for a in loaded:
logger.info(f"{a} {loaded[a].shape}")
if debug:
from safetensors.torch import load_file, save_file
loaded = load_file(checkpoint, "cpu")
dump(loaded)
return
try:
pipeline = StableDiffusionPipeline.from_single_file(
pretrained_model_link_or_path=str(checkpoint.absolute()),
local_files_only=False,
load_safety_checker=False,
)
logger.info("This file works fine.")
return
except:
from safetensors.torch import load_file, save_file
loaded = load_file(checkpoint, "cpu")
convert_table_bias={
"first_stage_model.decoder.mid.attn_1.to_k.bias":"first_stage_model.decoder.mid.attn_1.k.bias",
"first_stage_model.decoder.mid.attn_1.to_out.0.bias":"first_stage_model.decoder.mid.attn_1.proj_out.bias",
"first_stage_model.decoder.mid.attn_1.to_q.bias":"first_stage_model.decoder.mid.attn_1.q.bias",
"first_stage_model.decoder.mid.attn_1.to_v.bias":"first_stage_model.decoder.mid.attn_1.v.bias",
"first_stage_model.encoder.mid.attn_1.to_k.bias":"first_stage_model.encoder.mid.attn_1.k.bias",
"first_stage_model.encoder.mid.attn_1.to_out.0.bias":"first_stage_model.encoder.mid.attn_1.proj_out.bias",
"first_stage_model.encoder.mid.attn_1.to_q.bias":"first_stage_model.encoder.mid.attn_1.q.bias",
"first_stage_model.encoder.mid.attn_1.to_v.bias":"first_stage_model.encoder.mid.attn_1.v.bias",
}
convert_table_weight={
"first_stage_model.decoder.mid.attn_1.to_k.weight":"first_stage_model.decoder.mid.attn_1.k.weight",
"first_stage_model.decoder.mid.attn_1.to_out.0.weight":"first_stage_model.decoder.mid.attn_1.proj_out.weight",
"first_stage_model.decoder.mid.attn_1.to_q.weight":"first_stage_model.decoder.mid.attn_1.q.weight",
"first_stage_model.decoder.mid.attn_1.to_v.weight":"first_stage_model.decoder.mid.attn_1.v.weight",
"first_stage_model.encoder.mid.attn_1.to_k.weight":"first_stage_model.encoder.mid.attn_1.k.weight",
"first_stage_model.encoder.mid.attn_1.to_out.0.weight":"first_stage_model.encoder.mid.attn_1.proj_out.weight",
"first_stage_model.encoder.mid.attn_1.to_q.weight":"first_stage_model.encoder.mid.attn_1.q.weight",
"first_stage_model.encoder.mid.attn_1.to_v.weight":"first_stage_model.encoder.mid.attn_1.v.weight",
}
for a in list(loaded.keys()):
if a in convert_table_bias:
new_key = convert_table_bias[a]
loaded[new_key] = loaded.pop(a)
elif a in convert_table_weight:
new_key = convert_table_weight[a]
item = loaded.pop(a)
if len(item.shape) == 2:
item = item.unsqueeze(dim=-1).unsqueeze(dim=-1)
loaded[new_key] = item
new_path = str(checkpoint.parent / checkpoint.stem) + "_fixed"+checkpoint.suffix
logger.info(f"Saving file to {new_path}")
save_file(loaded, Path(new_path))
def checkpoint_to_pipeline(
checkpoint: Path,
target_dir: Optional[Path] = None,
save: bool = True,
) -> StableDiffusionPipeline:
logger.debug(f"Converting checkpoint {path_from_cwd(checkpoint)}")
if target_dir is None:
target_dir = pipeline_dir.joinpath(checkpoint.stem)
pipeline = StableDiffusionPipeline.from_single_file(
pretrained_model_link_or_path=str(checkpoint.absolute()),
local_files_only=False,
load_safety_checker=False,
)
if save:
target_dir.mkdir(parents=True, exist_ok=True)
logger.info(f"Saving pipeline to {path_from_cwd(target_dir)}")
pipeline.save_pretrained(target_dir, safe_serialization=True)
return pipeline, target_dir
def checkpoint_to_pipeline_sdxl(
checkpoint: Path,
target_dir: Optional[Path] = None,
save: bool = True,
) -> StableDiffusionXLPipeline:
logger.debug(f"Converting checkpoint {path_from_cwd(checkpoint)}")
if target_dir is None:
target_dir = pipeline_dir.joinpath(checkpoint.stem)
pipeline = StableDiffusionXLPipeline.from_single_file(
pretrained_model_link_or_path=str(checkpoint.absolute()),
local_files_only=False,
load_safety_checker=False,
)
if save:
target_dir.mkdir(parents=True, exist_ok=True)
logger.info(f"Saving pipeline to {path_from_cwd(target_dir)}")
pipeline.save_pretrained(target_dir, safe_serialization=True)
return pipeline, target_dir
def get_checkpoint_weights(checkpoint: Path):
temp_pipeline: StableDiffusionPipeline
temp_pipeline, _ = checkpoint_to_pipeline(checkpoint, save=False)
unet_state_dict = temp_pipeline.unet.state_dict()
tenc_state_dict = temp_pipeline.text_encoder.state_dict()
vae_state_dict = temp_pipeline.vae.state_dict()
return unet_state_dict, tenc_state_dict, vae_state_dict
def get_checkpoint_weights_sdxl(checkpoint: Path):
temp_pipeline: StableDiffusionXLPipeline
temp_pipeline, _ = checkpoint_to_pipeline_sdxl(checkpoint, save=False)
unet_state_dict = temp_pipeline.unet.state_dict()
tenc_state_dict = temp_pipeline.text_encoder.state_dict()
tenc2_state_dict = temp_pipeline.text_encoder_2.state_dict()
vae_state_dict = temp_pipeline.vae.state_dict()
return unet_state_dict, tenc_state_dict, tenc2_state_dict, vae_state_dict
def ensure_motion_modules(
repo_id: str = HF_MODULE_REPO,
fp16: bool = False,
force: bool = False,
):
"""Retrieve the motion modules from HuggingFace Hub."""
module_files = ["mm_sd_v14.safetensors", "mm_sd_v15.safetensors"]
module_dir = get_dir("data/models/motion-module")
for file in module_files:
target_path = module_dir.joinpath(file)
if fp16:
target_path = target_path.with_suffix(".fp16.safetensors")
if target_path.exists() and force is not True:
logger.debug(f"File {path_from_cwd(target_path)} already exists, skipping download")
else:
result = hf_hub_download(
repo_id=repo_id,
filename=target_path.name,
cache_dir=HF_HUB_CACHE,
local_dir=module_dir,
local_dir_use_symlinks=False,
resume_download=True,
)
logger.debug(f"Downloaded {path_from_cwd(result)}")
|