stable-diffusion-webui/modules/sd_models.py

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1035 lines
37 KiB
Python
Raw Normal View History

import collections
2024-06-15 23:04:31 -06:00
import importlib
2024-04-05 23:53:21 -06:00
import os
import sys
import threading
2024-06-15 23:04:31 -06:00
import enum
import torch
import re
2022-11-27 04:46:40 -07:00
import safetensors.torch
from omegaconf import OmegaConf, ListConfig
2022-12-08 17:14:35 -07:00
from urllib import request
import ldm.modules.midas as midas
from modules import paths, shared, modelloader, devices, script_callbacks, sd_vae, sd_disable_initialization, errors, hashes, sd_models_config, sd_unet, sd_models_xl, cache, extra_networks, processing, lowvram, sd_hijack, patches
from modules.timer import Timer
2024-02-26 21:43:27 -07:00
from modules.shared import opts
import tomesd
import numpy as np
model_dir = "Stable-diffusion"
model_path = os.path.abspath(os.path.join(paths.models_path, model_dir))
checkpoints_list = {}
2023-07-03 03:17:20 -06:00
checkpoint_aliases = {}
checkpoint_alisases = checkpoint_aliases # for compatibility with old name
checkpoints_loaded = collections.OrderedDict()
2023-01-13 23:56:59 -07:00
2024-06-15 23:04:31 -06:00
class ModelType(enum.Enum):
SD1 = 1
SD2 = 2
SDXL = 3
SSD = 4
SD3 = 5
def replace_key(d, key, new_key, value):
keys = list(d.keys())
d[new_key] = value
if key not in keys:
return d
index = keys.index(key)
keys[index] = new_key
new_d = {k: d[k] for k in keys}
d.clear()
d.update(new_d)
return d
2023-01-13 23:56:59 -07:00
class CheckpointInfo:
def __init__(self, filename):
self.filename = filename
abspath = os.path.abspath(filename)
2023-09-07 18:46:34 -06:00
abs_ckpt_dir = os.path.abspath(shared.cmd_opts.ckpt_dir) if shared.cmd_opts.ckpt_dir is not None else None
2023-01-13 23:56:59 -07:00
self.is_safetensors = os.path.splitext(filename)[1].lower() == ".safetensors"
2023-09-07 18:46:34 -06:00
if abs_ckpt_dir and abspath.startswith(abs_ckpt_dir):
name = abspath.replace(abs_ckpt_dir, '')
2023-01-13 23:56:59 -07:00
elif abspath.startswith(model_path):
name = abspath.replace(model_path, '')
else:
name = os.path.basename(filename)
if name.startswith("\\") or name.startswith("/"):
name = name[1:]
def read_metadata():
metadata = read_metadata_from_safetensors(filename)
self.modelspec_thumbnail = metadata.pop('modelspec.thumbnail', None)
return metadata
self.metadata = {}
if self.is_safetensors:
try:
self.metadata = cache.cached_data_for_file('safetensors-metadata', "checkpoint/" + name, filename, read_metadata)
except Exception as e:
errors.display(e, f"reading metadata for {filename}")
self.name = name
self.name_for_extra = os.path.splitext(os.path.basename(filename))[0]
2023-01-13 23:56:59 -07:00
self.model_name = os.path.splitext(name.replace("/", "_").replace("\\", "_"))[0]
self.hash = model_hash(filename)
self.sha256 = hashes.sha256_from_cache(self.filename, f"checkpoint/{name}")
self.shorthash = self.sha256[0:10] if self.sha256 else None
self.title = name if self.shorthash is None else f'{name} [{self.shorthash}]'
2023-07-30 04:48:27 -06:00
self.short_title = self.name_for_extra if self.shorthash is None else f'{self.name_for_extra} [{self.shorthash}]'
self.ids = [self.hash, self.model_name, self.title, name, self.name_for_extra, f'{name} [{self.hash}]']
if self.shorthash:
self.ids += [self.shorthash, self.sha256, f'{self.name} [{self.shorthash}]', f'{self.name_for_extra} [{self.shorthash}]']
2023-01-13 23:56:59 -07:00
def register(self):
checkpoints_list[self.title] = self
for id in self.ids:
2023-07-03 03:17:20 -06:00
checkpoint_aliases[id] = self
2023-01-13 23:56:59 -07:00
def calculate_shorthash(self):
self.sha256 = hashes.sha256(self.filename, f"checkpoint/{self.name}")
2023-02-04 01:38:56 -07:00
if self.sha256 is None:
return
shorthash = self.sha256[0:10]
if self.shorthash == self.sha256[0:10]:
return self.shorthash
self.shorthash = shorthash
2023-01-13 23:56:59 -07:00
if self.shorthash not in self.ids:
self.ids += [self.shorthash, self.sha256, f'{self.name} [{self.shorthash}]', f'{self.name_for_extra} [{self.shorthash}]']
2023-01-13 23:56:59 -07:00
old_title = self.title
self.title = f'{self.name} [{self.shorthash}]'
2023-07-30 04:48:27 -06:00
self.short_title = f'{self.name_for_extra} [{self.shorthash}]'
replace_key(checkpoints_list, old_title, self.title, self)
self.register()
2023-01-13 23:56:59 -07:00
return self.shorthash
try:
# this silences the annoying "Some weights of the model checkpoint were not used when initializing..." message at start.
2023-05-10 00:02:23 -06:00
from transformers import logging, CLIPModel # noqa: F401
logging.set_verbosity_error()
except Exception:
pass
def setup_model():
"""called once at startup to do various one-time tasks related to SD models"""
2023-05-29 01:18:15 -06:00
os.makedirs(model_path, exist_ok=True)
2022-12-08 17:14:35 -07:00
enable_midas_autodownload()
patch_given_betas()
2023-07-30 04:48:27 -06:00
def checkpoint_tiles(use_short=False):
return [x.short_title if use_short else x.title for x in checkpoints_list.values()]
def list_models():
checkpoints_list.clear()
2023-07-03 03:17:20 -06:00
checkpoint_aliases.clear()
cmd_ckpt = shared.cmd_opts.ckpt
2023-02-19 04:49:07 -07:00
if shared.cmd_opts.no_download_sd_model or cmd_ckpt != shared.sd_model_file or os.path.exists(cmd_ckpt):
2023-02-19 04:37:40 -07:00
model_url = None
expected_sha256 = None
2023-02-19 04:37:40 -07:00
else:
2024-04-05 23:53:21 -06:00
model_url = f"{shared.hf_endpoint}/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors"
expected_sha256 = '6ce0161689b3853acaa03779ec93eafe75a02f4ced659bee03f50797806fa2fa'
2023-02-19 04:37:40 -07:00
model_list = modelloader.load_models(model_path=model_path, model_url=model_url, command_path=shared.cmd_opts.ckpt_dir, ext_filter=[".ckpt", ".safetensors"], download_name="v1-5-pruned-emaonly.safetensors", ext_blacklist=[".vae.ckpt", ".vae.safetensors"], hash_prefix=expected_sha256)
2023-02-19 04:37:40 -07:00
if os.path.exists(cmd_ckpt):
2023-01-13 23:56:59 -07:00
checkpoint_info = CheckpointInfo(cmd_ckpt)
checkpoint_info.register()
shared.opts.data['sd_model_checkpoint'] = checkpoint_info.title
elif cmd_ckpt is not None and cmd_ckpt != shared.default_sd_model_file:
print(f"Checkpoint in --ckpt argument not found (Possible it was moved to {model_path}: {cmd_ckpt}", file=sys.stderr)
2023-01-13 23:56:59 -07:00
2023-07-30 04:48:27 -06:00
for filename in model_list:
2023-01-13 23:56:59 -07:00
checkpoint_info = CheckpointInfo(filename)
checkpoint_info.register()
2023-07-30 04:48:27 -06:00
re_strip_checksum = re.compile(r"\s*\[[^]]+]\s*$")
2023-01-13 23:56:59 -07:00
def get_closet_checkpoint_match(search_string):
if not search_string:
return None
2023-07-03 03:17:20 -06:00
checkpoint_info = checkpoint_aliases.get(search_string, None)
2023-01-13 23:56:59 -07:00
if checkpoint_info is not None:
2023-01-14 00:25:21 -07:00
return checkpoint_info
2023-01-13 23:56:59 -07:00
found = sorted([info for info in checkpoints_list.values() if search_string in info.title], key=lambda x: len(x.title))
if found:
return found[0]
2023-07-30 04:48:27 -06:00
search_string_without_checksum = re.sub(re_strip_checksum, '', search_string)
found = sorted([info for info in checkpoints_list.values() if search_string_without_checksum in info.title], key=lambda x: len(x.title))
if found:
return found[0]
2022-09-28 15:30:09 -06:00
return None
def model_hash(filename):
2023-01-13 23:56:59 -07:00
"""old hash that only looks at a small part of the file and is prone to collisions"""
try:
with open(filename, "rb") as file:
import hashlib
m = hashlib.sha256()
file.seek(0x100000)
m.update(file.read(0x10000))
return m.hexdigest()[0:8]
except FileNotFoundError:
return 'NOFILE'
def select_checkpoint():
"""Raises `FileNotFoundError` if no checkpoints are found."""
model_checkpoint = shared.opts.sd_model_checkpoint
2023-07-03 03:17:20 -06:00
checkpoint_info = checkpoint_aliases.get(model_checkpoint, None)
if checkpoint_info is not None:
return checkpoint_info
if len(checkpoints_list) == 0:
error_message = "No checkpoints found. When searching for checkpoints, looked at:"
if shared.cmd_opts.ckpt is not None:
error_message += f"\n - file {os.path.abspath(shared.cmd_opts.ckpt)}"
error_message += f"\n - directory {model_path}"
if shared.cmd_opts.ckpt_dir is not None:
error_message += f"\n - directory {os.path.abspath(shared.cmd_opts.ckpt_dir)}"
error_message += "Can't run without a checkpoint. Find and place a .ckpt or .safetensors file into any of those locations."
raise FileNotFoundError(error_message)
checkpoint_info = next(iter(checkpoints_list.values()))
if model_checkpoint is not None:
print(f"Checkpoint {model_checkpoint} not found; loading fallback {checkpoint_info.title}", file=sys.stderr)
return checkpoint_info
checkpoint_dict_replacements_sd1 = {
'cond_stage_model.transformer.embeddings.': 'cond_stage_model.transformer.text_model.embeddings.',
'cond_stage_model.transformer.encoder.': 'cond_stage_model.transformer.text_model.encoder.',
'cond_stage_model.transformer.final_layer_norm.': 'cond_stage_model.transformer.text_model.final_layer_norm.',
}
checkpoint_dict_replacements_sd2_turbo = { # Converts SD 2.1 Turbo from SGM to LDM format.
'conditioner.embedders.0.': 'cond_stage_model.',
}
def transform_checkpoint_dict_key(k, replacements):
for text, replacement in replacements.items():
if k.startswith(text):
k = replacement + k[len(text):]
return k
def get_state_dict_from_checkpoint(pl_sd):
pl_sd = pl_sd.pop("state_dict", pl_sd)
pl_sd.pop("state_dict", None)
is_sd2_turbo = 'conditioner.embedders.0.model.ln_final.weight' in pl_sd and pl_sd['conditioner.embedders.0.model.ln_final.weight'].size()[0] == 1024
sd = {}
for k, v in pl_sd.items():
if is_sd2_turbo:
new_key = transform_checkpoint_dict_key(k, checkpoint_dict_replacements_sd2_turbo)
else:
new_key = transform_checkpoint_dict_key(k, checkpoint_dict_replacements_sd1)
if new_key is not None:
sd[new_key] = v
2022-10-19 03:45:30 -06:00
pl_sd.clear()
pl_sd.update(sd)
return pl_sd
def read_metadata_from_safetensors(filename):
import json
with open(filename, mode="rb") as file:
metadata_len = file.read(8)
metadata_len = int.from_bytes(metadata_len, "little")
json_start = file.read(2)
assert metadata_len > 2 and json_start in (b'{"', b"{'"), f"{filename} is not a safetensors file"
res = {}
try:
json_data = json_start + file.read(metadata_len-2)
json_obj = json.loads(json_data)
for k, v in json_obj.get("__metadata__", {}).items():
res[k] = v
if isinstance(v, str) and v[0:1] == '{':
try:
res[k] = json.loads(v)
except Exception:
pass
2024-04-26 06:21:12 -06:00
except Exception:
errors.report(f"Error reading metadata from file: {filename}", exc_info=True)
2024-04-26 06:17:37 -06:00
return res
def read_state_dict(checkpoint_file, print_global_state=False, map_location=None):
_, extension = os.path.splitext(checkpoint_file)
if extension.lower() == ".safetensors":
device = map_location or shared.weight_load_location or devices.get_optimal_device_name()
if not shared.opts.disable_mmap_load_safetensors:
pl_sd = safetensors.torch.load_file(checkpoint_file, device=device)
else:
pl_sd = safetensors.torch.load(open(checkpoint_file, 'rb').read())
pl_sd = {k: v.to(device) for k, v in pl_sd.items()}
else:
pl_sd = torch.load(checkpoint_file, map_location=map_location or shared.weight_load_location)
if print_global_state and "global_step" in pl_sd:
print(f"Global Step: {pl_sd['global_step']}")
sd = get_state_dict_from_checkpoint(pl_sd)
return sd
def get_checkpoint_state_dict(checkpoint_info: CheckpointInfo, timer):
sd_model_hash = checkpoint_info.calculate_shorthash()
timer.record("calculate hash")
if checkpoint_info in checkpoints_loaded:
# use checkpoint cache
print(f"Loading weights [{sd_model_hash}] from cache")
2023-09-18 02:45:42 -06:00
# move to end as latest
checkpoints_loaded.move_to_end(checkpoint_info)
return checkpoints_loaded[checkpoint_info]
print(f"Loading weights [{sd_model_hash}] from {checkpoint_info.filename}")
res = read_state_dict(checkpoint_info.filename)
timer.record("load weights from disk")
return res
2023-08-06 08:01:07 -06:00
class SkipWritingToConfig:
"""This context manager prevents load_model_weights from writing checkpoint name to the config when it loads weight."""
skip = False
previous = None
def __enter__(self):
self.previous = SkipWritingToConfig.skip
SkipWritingToConfig.skip = True
return self
def __exit__(self, exc_type, exc_value, exc_traceback):
SkipWritingToConfig.skip = self.previous
2023-11-19 00:50:06 -07:00
def check_fp8(model):
if model is None:
return None
if devices.get_optimal_device_name() == "mps":
enable_fp8 = False
elif shared.opts.fp8_storage == "Enable":
enable_fp8 = True
elif getattr(model, "is_sdxl", False) and shared.opts.fp8_storage == "Enable for SDXL":
enable_fp8 = True
else:
enable_fp8 = False
return enable_fp8
2024-06-15 23:04:31 -06:00
def set_model_type(model, state_dict):
model.is_sd1 = False
model.is_sd2 = False
model.is_sdxl = False
model.is_ssd = False
2024-06-15 23:18:05 -06:00
model.is_sd3 = False
2024-06-15 23:04:31 -06:00
if "model.diffusion_model.x_embedder.proj.weight" in state_dict:
model.is_sd3 = True
model.model_type = ModelType.SD3
elif hasattr(model, 'conditioner'):
model.is_sdxl = True
if 'model.diffusion_model.middle_block.1.transformer_blocks.0.attn1.to_q.weight' not in state_dict.keys():
model.is_ssd = True
model.model_type = ModelType.SSD
else:
model.model_type = ModelType.SDXL
elif hasattr(model.cond_stage_model, 'model'):
model.is_sd2 = True
model.model_type = ModelType.SD2
else:
model.is_sd1 = True
model.model_type = ModelType.SD1
def set_model_fields(model):
if not hasattr(model, 'latent_channels'):
model.latent_channels = 4
def load_model_weights(model, checkpoint_info: CheckpointInfo, state_dict, timer):
2023-01-13 23:56:59 -07:00
sd_model_hash = checkpoint_info.calculate_shorthash()
timer.record("calculate hash")
2023-11-24 21:35:09 -07:00
if devices.fp8:
2023-11-19 00:50:06 -07:00
# prevent model to load state dict in fp8
model.half()
2023-08-06 08:01:07 -06:00
if not SkipWritingToConfig.skip:
shared.opts.data["sd_model_checkpoint"] = checkpoint_info.title
if state_dict is None:
state_dict = get_checkpoint_state_dict(checkpoint_info, timer)
2024-06-15 23:04:31 -06:00
set_model_type(model, state_dict)
set_model_fields(model)
if model.is_sdxl:
2023-07-11 12:16:43 -06:00
sd_models_xl.extend_sdxl(model)
2023-11-05 06:43:49 -07:00
if model.is_ssd:
sd_hijack.model_hijack.convert_sdxl_to_ssd(model)
2023-11-05 09:32:21 -07:00
if shared.opts.sd_checkpoint_cache > 0:
# cache newly loaded model
2023-10-13 23:01:04 -06:00
checkpoints_loaded[checkpoint_info] = state_dict.copy()
2023-10-07 01:36:01 -06:00
if hasattr(model, "before_load_weights"):
model.before_load_weights(state_dict)
2023-10-07 01:36:01 -06:00
model.load_state_dict(state_dict, strict=False)
timer.record("apply weights to model")
if hasattr(model, "after_load_weights"):
model.after_load_weights(state_dict)
del state_dict
2024-06-08 20:11:11 -06:00
# Set is_sdxl_inpaint flag.
2024-06-08 20:15:37 -06:00
# Checks Unet structure to detect inpaint model. The inpaint model's
# checkpoint state_dict does not contain the key
# 'diffusion_model.input_blocks.0.0.weight'.
2024-06-08 20:11:11 -06:00
diffusion_model_input = model.model.state_dict().get(
'diffusion_model.input_blocks.0.0.weight'
)
model.is_sdxl_inpaint = (
model.is_sdxl and
diffusion_model_input is not None and
diffusion_model_input.shape[1] == 9
)
if shared.cmd_opts.opt_channelslast:
model.to(memory_format=torch.channels_last)
timer.record("apply channels_last")
if shared.cmd_opts.no_half:
model.float()
2023-12-02 12:09:18 -07:00
model.alphas_cumprod_original = model.alphas_cumprod
2023-08-22 22:10:43 -06:00
devices.dtype_unet = torch.float32
2024-05-16 17:50:06 -06:00
assert shared.cmd_opts.precision != "half", "Cannot use --precision half with --no-half"
timer.record("apply float()")
else:
vae = model.first_stage_model
depth_model = getattr(model, 'depth_model', None)
# with --no-half-vae, remove VAE from model when doing half() to prevent its weights from being converted to float16
if shared.cmd_opts.no_half_vae:
model.first_stage_model = None
# with --upcast-sampling, don't convert the depth model weights to float16
if shared.cmd_opts.upcast_sampling and depth_model:
model.depth_model = None
2022-11-02 05:41:29 -06:00
alphas_cumprod = model.alphas_cumprod
model.alphas_cumprod = None
model.half()
model.alphas_cumprod = alphas_cumprod
model.alphas_cumprod_original = alphas_cumprod
model.first_stage_model = vae
if depth_model:
model.depth_model = depth_model
2022-11-02 05:41:29 -06:00
2023-08-22 22:10:43 -06:00
devices.dtype_unet = torch.float16
timer.record("apply half()")
apply_alpha_schedule_override(model)
for module in model.modules():
if hasattr(module, 'fp16_weight'):
del module.fp16_weight
if hasattr(module, 'fp16_bias'):
del module.fp16_bias
2023-11-19 00:50:06 -07:00
if check_fp8(model):
devices.fp8 = True
2023-11-19 00:50:06 -07:00
first_stage = model.first_stage_model
model.first_stage_model = None
for module in model.modules():
if isinstance(module, (torch.nn.Conv2d, torch.nn.Linear)):
if shared.opts.cache_fp16_weight:
module.fp16_weight = module.weight.data.clone().cpu().half()
if module.bias is not None:
module.fp16_bias = module.bias.data.clone().cpu().half()
2023-11-19 00:50:06 -07:00
module.to(torch.float8_e4m3fn)
model.first_stage_model = first_stage
2023-10-28 01:24:26 -06:00
timer.record("apply fp8")
else:
devices.fp8 = False
devices.unet_needs_upcast = shared.cmd_opts.upcast_sampling and devices.dtype == torch.float16 and devices.dtype_unet == torch.float16
model.first_stage_model.to(devices.dtype_vae)
timer.record("apply dtype to VAE")
2022-11-02 05:41:29 -06:00
# clean up cache if limit is reached
while len(checkpoints_loaded) > shared.opts.sd_checkpoint_cache:
checkpoints_loaded.popitem(last=False)
2022-10-31 03:27:27 -06:00
model.sd_model_hash = sd_model_hash
2023-01-13 23:56:59 -07:00
model.sd_model_checkpoint = checkpoint_info.filename
model.sd_checkpoint_info = checkpoint_info
shared.opts.data["sd_checkpoint_hash"] = checkpoint_info.sha256
2023-07-11 12:16:43 -06:00
if hasattr(model, 'logvar'):
model.logvar = model.logvar.to(devices.device) # fix for training
2023-01-01 14:38:09 -07:00
2022-11-12 21:11:14 -07:00
sd_vae.delete_base_vae()
sd_vae.clear_loaded_vae()
vae_file, vae_source = sd_vae.resolve_vae(checkpoint_info.filename).tuple()
sd_vae.load_vae(model, vae_file, vae_source)
timer.record("load VAE")
2022-12-08 17:14:35 -07:00
def enable_midas_autodownload():
"""
Gives the ldm.modules.midas.api.load_model function automatic downloading.
When the 512-depth-ema model, and other future models like it, is loaded,
it calls midas.api.load_model to load the associated midas depth model.
This function applies a wrapper to download the model to the correct
location automatically.
"""
midas_path = os.path.join(paths.models_path, 'midas')
2022-12-08 17:14:35 -07:00
# stable-diffusion-stability-ai hard-codes the midas model path to
# a location that differs from where other scripts using this model look.
# HACK: Overriding the path here.
for k, v in midas.api.ISL_PATHS.items():
file_name = os.path.basename(v)
midas.api.ISL_PATHS[k] = os.path.join(midas_path, file_name)
midas_urls = {
"dpt_large": "https://github.com/intel-isl/DPT/releases/download/1_0/dpt_large-midas-2f21e586.pt",
"dpt_hybrid": "https://github.com/intel-isl/DPT/releases/download/1_0/dpt_hybrid-midas-501f0c75.pt",
"midas_v21": "https://github.com/AlexeyAB/MiDaS/releases/download/midas_dpt/midas_v21-f6b98070.pt",
"midas_v21_small": "https://github.com/AlexeyAB/MiDaS/releases/download/midas_dpt/midas_v21_small-70d6b9c8.pt",
}
midas.api.load_model_inner = midas.api.load_model
def load_model_wrapper(model_type):
path = midas.api.ISL_PATHS[model_type]
if not os.path.exists(path):
if not os.path.exists(midas_path):
2024-04-05 23:53:21 -06:00
os.mkdir(midas_path)
2022-12-08 17:14:35 -07:00
print(f"Downloading midas model weights for {model_type} to {path}")
request.urlretrieve(midas_urls[model_type], path)
print(f"{model_type} downloaded")
return midas.api.load_model_inner(model_type)
midas.api.load_model = load_model_wrapper
def patch_given_betas():
import ldm.models.diffusion.ddpm
def patched_register_schedule(*args, **kwargs):
"""a modified version of register_schedule function that converts plain list from Omegaconf into numpy"""
if isinstance(args[1], ListConfig):
args = (args[0], np.array(args[1]), *args[2:])
original_register_schedule(*args, **kwargs)
original_register_schedule = patches.patch(__name__, ldm.models.diffusion.ddpm.DDPM, 'register_schedule', patched_register_schedule)
2024-06-15 23:04:31 -06:00
def repair_config(sd_config, state_dict=None):
if not hasattr(sd_config.model.params, "use_ema"):
sd_config.model.params.use_ema = False
2023-07-13 08:32:35 -06:00
if hasattr(sd_config.model.params, 'unet_config'):
if shared.cmd_opts.no_half:
sd_config.model.params.unet_config.params.use_fp16 = False
2024-05-17 11:34:04 -06:00
elif shared.cmd_opts.upcast_sampling or shared.cmd_opts.precision == "half":
2023-07-13 08:32:35 -06:00
sd_config.model.params.unet_config.params.use_fp16 = True
2024-06-15 23:04:31 -06:00
if hasattr(sd_config.model.params, 'first_stage_config'):
if getattr(sd_config.model.params.first_stage_config.params.ddconfig, "attn_type", None) == "vanilla-xformers" and not shared.xformers_available:
sd_config.model.params.first_stage_config.params.ddconfig.attn_type = "vanilla"
# For UnCLIP-L, override the hardcoded karlo directory
if hasattr(sd_config.model.params, "noise_aug_config") and hasattr(sd_config.model.params.noise_aug_config.params, "clip_stats_path"):
karlo_path = os.path.join(paths.models_path, 'karlo')
sd_config.model.params.noise_aug_config.params.clip_stats_path = sd_config.model.params.noise_aug_config.params.clip_stats_path.replace("checkpoints/karlo_models", karlo_path)
2024-05-16 14:39:02 -06:00
# Do not use checkpoint for inference.
# This helps prevent extra performance overhead on checking parameters.
2024-05-16 18:06:04 -06:00
# The perf overhead is about 100ms/it on 4090 for SDXL.
if hasattr(sd_config.model.params, "network_config"):
sd_config.model.params.network_config.params.use_checkpoint = False
if hasattr(sd_config.model.params, "unet_config"):
sd_config.model.params.unet_config.params.use_checkpoint = False
2024-05-16 14:39:02 -06:00
2024-03-01 20:54:11 -07:00
2024-06-15 23:04:31 -06:00
2024-03-01 20:54:11 -07:00
def rescale_zero_terminal_snr_abar(alphas_cumprod):
alphas_bar_sqrt = alphas_cumprod.sqrt()
# Store old values.
alphas_bar_sqrt_0 = alphas_bar_sqrt[0].clone()
alphas_bar_sqrt_T = alphas_bar_sqrt[-1].clone()
# Shift so the last timestep is zero.
alphas_bar_sqrt -= (alphas_bar_sqrt_T)
# Scale so the first timestep is back to the old value.
alphas_bar_sqrt *= alphas_bar_sqrt_0 / (alphas_bar_sqrt_0 - alphas_bar_sqrt_T)
# Convert alphas_bar_sqrt to betas
alphas_bar = alphas_bar_sqrt ** 2 # Revert sqrt
alphas_bar[-1] = 4.8973451890853435e-08
return alphas_bar
2024-02-26 21:43:27 -07:00
def apply_alpha_schedule_override(sd_model, p=None):
2024-03-01 20:54:11 -07:00
"""
Applies an override to the alpha schedule of the model according to settings.
- downcasts the alpha schedule to half precision
- rescales the alpha schedule to have zero terminal SNR
"""
if not hasattr(sd_model, 'alphas_cumprod') or not hasattr(sd_model, 'alphas_cumprod_original'):
return
sd_model.alphas_cumprod = sd_model.alphas_cumprod_original.to(shared.device)
if opts.use_downcasted_alpha_bar:
if p is not None:
p.extra_generation_params['Downcast alphas_cumprod'] = opts.use_downcasted_alpha_bar
sd_model.alphas_cumprod = sd_model.alphas_cumprod.half().to(shared.device)
if opts.sd_noise_schedule == "Zero Terminal SNR":
if p is not None:
p.extra_generation_params['Noise Schedule'] = opts.sd_noise_schedule
sd_model.alphas_cumprod = rescale_zero_terminal_snr_abar(sd_model.alphas_cumprod).to(shared.device)
sd1_clip_weight = 'cond_stage_model.transformer.text_model.embeddings.token_embedding.weight'
sd2_clip_weight = 'cond_stage_model.model.transformer.resblocks.0.attn.in_proj_weight'
2023-07-12 14:52:43 -06:00
sdxl_clip_weight = 'conditioner.embedders.1.model.ln_final.weight'
2023-07-14 00:16:01 -06:00
sdxl_refiner_clip_weight = 'conditioner.embedders.0.model.ln_final.weight'
class SdModelData:
def __init__(self):
self.sd_model = None
self.loaded_sd_models = []
self.was_loaded_at_least_once = False
self.lock = threading.Lock()
def get_sd_model(self):
if self.was_loaded_at_least_once:
return self.sd_model
if self.sd_model is None:
with self.lock:
if self.sd_model is not None or self.was_loaded_at_least_once:
return self.sd_model
try:
load_model()
except Exception as e:
errors.display(e, "loading stable diffusion model", full_traceback=True)
print("", file=sys.stderr)
print("Stable diffusion model failed to load", file=sys.stderr)
self.sd_model = None
return self.sd_model
def set_sd_model(self, v, already_loaded=False):
self.sd_model = v
if already_loaded:
sd_vae.base_vae = getattr(v, "base_vae", None)
sd_vae.loaded_vae_file = getattr(v, "loaded_vae_file", None)
sd_vae.checkpoint_info = v.sd_checkpoint_info
try:
self.loaded_sd_models.remove(v)
except ValueError:
pass
if v is not None:
self.loaded_sd_models.insert(0, v)
model_data = SdModelData()
2023-07-12 14:52:43 -06:00
def get_empty_cond(sd_model):
p = processing.StableDiffusionProcessingTxt2Img()
extra_networks.activate(p, {})
if hasattr(sd_model, 'get_learned_conditioning'):
2023-07-12 14:52:43 -06:00
d = sd_model.get_learned_conditioning([""])
else:
d = sd_model.cond_stage_model([""])
if isinstance(d, dict):
d = d['crossattn']
return d
2023-07-12 14:52:43 -06:00
def send_model_to_cpu(m):
if m is not None:
if m.lowvram:
lowvram.send_everything_to_cpu()
else:
m.to(devices.cpu)
devices.torch_gc()
2023-08-22 09:49:08 -06:00
def model_target_device(m):
if lowvram.is_needed(m):
return devices.cpu
else:
return devices.device
def send_model_to_device(m):
2023-08-22 09:49:08 -06:00
lowvram.apply(m)
if not m.lowvram:
m.to(shared.device)
def send_model_to_trash(m):
m.to(device="meta")
devices.torch_gc()
2024-06-15 23:04:31 -06:00
def instantiate_from_config(config, state_dict=None):
constructor = get_obj_from_str(config["target"])
params = {**config.get("params", {})}
if state_dict and "state_dict" in params and params["state_dict"] is None:
params["state_dict"] = state_dict
return constructor(**params)
def get_obj_from_str(string, reload=False):
module, cls = string.rsplit(".", 1)
if reload:
module_imp = importlib.import_module(module)
importlib.reload(module_imp)
return getattr(importlib.import_module(module, package=None), cls)
def load_model(checkpoint_info=None, already_loaded_state_dict=None):
from modules import sd_hijack
checkpoint_info = checkpoint_info or select_checkpoint()
timer = Timer()
if model_data.sd_model:
send_model_to_trash(model_data.sd_model)
model_data.sd_model = None
devices.torch_gc()
timer.record("unload existing model")
if already_loaded_state_dict is not None:
state_dict = already_loaded_state_dict
else:
state_dict = get_checkpoint_state_dict(checkpoint_info, timer)
checkpoint_config = sd_models_config.find_checkpoint_config(state_dict, checkpoint_info)
2023-07-14 00:19:08 -06:00
clip_is_included_into_sd = any(x for x in [sd1_clip_weight, sd2_clip_weight, sdxl_clip_weight, sdxl_refiner_clip_weight] if x in state_dict)
2022-11-26 11:28:44 -07:00
timer.record("find config")
sd_config = OmegaConf.load(checkpoint_config)
2024-06-15 23:04:31 -06:00
repair_config(sd_config, state_dict)
timer.record("load config")
print(f"Creating model from config: {checkpoint_config}")
sd_model = None
try:
with sd_disable_initialization.DisableInitialization(disable_clip=clip_is_included_into_sd or shared.cmd_opts.do_not_download_clip):
2023-07-24 13:08:08 -06:00
with sd_disable_initialization.InitializeOnMeta():
2024-06-15 23:04:31 -06:00
sd_model = instantiate_from_config(sd_config.model, state_dict)
2023-07-24 13:08:08 -06:00
except Exception as e:
errors.display(e, "creating model quickly", full_traceback=True)
if sd_model is None:
print('Failed to create model quickly; will retry using slow method.', file=sys.stderr)
2023-07-24 13:08:08 -06:00
with sd_disable_initialization.InitializeOnMeta():
2024-06-15 23:04:31 -06:00
sd_model = instantiate_from_config(sd_config.model, state_dict)
sd_model.used_config = checkpoint_config
timer.record("create model")
if shared.cmd_opts.no_half:
weight_dtype_conversion = None
else:
weight_dtype_conversion = {
'first_stage_model': None,
'alphas_cumprod': None,
'': torch.float16,
}
2023-08-22 09:49:08 -06:00
with sd_disable_initialization.LoadStateDictOnMeta(state_dict, device=model_target_device(sd_model), weight_dtype_conversion=weight_dtype_conversion):
2023-07-24 13:08:08 -06:00
load_model_weights(sd_model, checkpoint_info, state_dict, timer)
2024-06-15 23:04:31 -06:00
timer.record("load weights from state dict")
send_model_to_device(sd_model)
timer.record("move model to device")
sd_hijack.model_hijack.hijack(sd_model)
timer.record("hijack")
sd_model.eval()
model_data.set_sd_model(sd_model)
model_data.was_loaded_at_least_once = True
sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings(force_reload=True) # Reload embeddings after model load as they may or may not fit the model
timer.record("load textual inversion embeddings")
script_callbacks.model_loaded_callback(sd_model)
timer.record("scripts callbacks")
with devices.autocast(), torch.no_grad():
2023-07-12 14:52:43 -06:00
sd_model.cond_stage_model_empty_prompt = get_empty_cond(sd_model)
timer.record("calculate empty prompt")
print(f"Model loaded in {timer.summary()}.")
2022-12-31 09:27:02 -07:00
return sd_model
def reuse_model_from_already_loaded(sd_model, checkpoint_info, timer):
"""
Checks if the desired checkpoint from checkpoint_info is not already loaded in model_data.loaded_sd_models.
If it is loaded, returns that (moving it to GPU if necessary, and moving the currently loadded model to CPU if necessary).
If not, returns the model that can be used to load weights from checkpoint_info's file.
If no such model exists, returns None.
2024-03-03 23:37:23 -07:00
Additionally deletes loaded models that are over the limit set in settings (sd_checkpoints_limit).
"""
if sd_model is not None and sd_model.sd_checkpoint_info.filename == checkpoint_info.filename:
return sd_model
if shared.opts.sd_checkpoints_keep_in_cpu:
send_model_to_cpu(sd_model)
timer.record("send model to cpu")
already_loaded = None
for i in reversed(range(len(model_data.loaded_sd_models))):
loaded_model = model_data.loaded_sd_models[i]
if loaded_model.sd_checkpoint_info.filename == checkpoint_info.filename:
already_loaded = loaded_model
continue
if len(model_data.loaded_sd_models) > shared.opts.sd_checkpoints_limit > 0:
print(f"Unloading model {len(model_data.loaded_sd_models)} over the limit of {shared.opts.sd_checkpoints_limit}: {loaded_model.sd_checkpoint_info.title}")
del model_data.loaded_sd_models[i]
send_model_to_trash(loaded_model)
timer.record("send model to trash")
if already_loaded is not None:
send_model_to_device(already_loaded)
timer.record("send model to device")
model_data.set_sd_model(already_loaded, already_loaded=True)
2023-08-10 08:04:59 -06:00
if not SkipWritingToConfig.skip:
shared.opts.data["sd_model_checkpoint"] = already_loaded.sd_checkpoint_info.title
shared.opts.data["sd_checkpoint_hash"] = already_loaded.sd_checkpoint_info.sha256
print(f"Using already loaded model {already_loaded.sd_checkpoint_info.title}: done in {timer.summary()}")
sd_vae.reload_vae_weights(already_loaded)
return model_data.sd_model
elif shared.opts.sd_checkpoints_limit > 1 and len(model_data.loaded_sd_models) < shared.opts.sd_checkpoints_limit:
print(f"Loading model {checkpoint_info.title} ({len(model_data.loaded_sd_models) + 1} out of {shared.opts.sd_checkpoints_limit})")
model_data.sd_model = None
load_model(checkpoint_info)
return model_data.sd_model
elif len(model_data.loaded_sd_models) > 0:
sd_model = model_data.loaded_sd_models.pop()
model_data.sd_model = sd_model
sd_vae.base_vae = getattr(sd_model, "base_vae", None)
sd_vae.loaded_vae_file = getattr(sd_model, "loaded_vae_file", None)
sd_vae.checkpoint_info = sd_model.sd_checkpoint_info
print(f"Reusing loaded model {sd_model.sd_checkpoint_info.title} to load {checkpoint_info.title}")
return sd_model
else:
return None
2023-11-19 00:50:06 -07:00
def reload_model_weights(sd_model=None, info=None, forced_reload=False):
checkpoint_info = info or select_checkpoint()
timer = Timer()
if not sd_model:
sd_model = model_data.sd_model
if sd_model is None: # previous model load failed
current_checkpoint_info = None
else:
current_checkpoint_info = sd_model.sd_checkpoint_info
2023-11-19 00:50:06 -07:00
if check_fp8(sd_model) != devices.fp8:
# load from state dict again to prevent extra numerical errors
forced_reload = True
2023-12-06 00:16:10 -07:00
elif sd_model.sd_model_checkpoint == checkpoint_info.filename and not forced_reload:
return sd_model
2023-05-27 06:47:33 -06:00
sd_model = reuse_model_from_already_loaded(sd_model, checkpoint_info, timer)
2023-11-19 00:50:06 -07:00
if not forced_reload and sd_model is not None and sd_model.sd_checkpoint_info.filename == checkpoint_info.filename:
return sd_model
if sd_model is not None:
sd_unet.apply_unet("None")
send_model_to_cpu(sd_model)
sd_hijack.model_hijack.undo_hijack(sd_model)
state_dict = get_checkpoint_state_dict(checkpoint_info, timer)
checkpoint_config = sd_models_config.find_checkpoint_config(state_dict, checkpoint_info)
timer.record("find config")
if sd_model is None or checkpoint_config != sd_model.used_config:
if sd_model is not None:
send_model_to_trash(sd_model)
load_model(checkpoint_info, already_loaded_state_dict=state_dict)
return model_data.sd_model
try:
load_model_weights(sd_model, checkpoint_info, state_dict, timer)
2023-05-09 22:52:45 -06:00
except Exception:
print("Failed to load checkpoint, restoring previous")
load_model_weights(sd_model, current_checkpoint_info, None, timer)
raise
finally:
sd_hijack.model_hijack.hijack(sd_model)
timer.record("hijack")
2023-08-22 09:49:08 -06:00
if not sd_model.lowvram:
sd_model.to(devices.device)
timer.record("move model to device")
script_callbacks.model_loaded_callback(sd_model)
timer.record("script callbacks")
print(f"Weights loaded in {timer.summary()}.")
model_data.set_sd_model(sd_model)
sd_unet.apply_unet()
return sd_model
def unload_model_weights(sd_model=None, info=None):
2023-10-15 00:41:02 -06:00
send_model_to_cpu(sd_model or shared.sd_model)
return sd_model
def apply_token_merging(sd_model, token_merging_ratio):
"""
Applies speed and memory optimizations from tomesd.
"""
current_token_merging_ratio = getattr(sd_model, 'applied_token_merged_ratio', 0)
if current_token_merging_ratio == token_merging_ratio:
return
if current_token_merging_ratio > 0:
tomesd.remove_patch(sd_model)
if token_merging_ratio > 0:
tomesd.apply_patch(
sd_model,
ratio=token_merging_ratio,
use_rand=False, # can cause issues with some samplers
merge_attn=True,
merge_crossattn=False,
merge_mlp=False
)
sd_model.applied_token_merged_ratio = token_merging_ratio