mirror of
https://github.com/kohya-ss/sd-scripts.git
synced 2026-04-08 22:35:09 +00:00
Replace print with logger if they are logs (#905)
* Add get_my_logger() * Use logger instead of print * Fix log level * Removed line-breaks for readability * Use setup_logging() * Add rich to requirements.txt * Make simple * Use logger instead of print --------- Co-authored-by: Kohya S <52813779+kohya-ss@users.noreply.github.com>
This commit is contained in:
Yuta Hayashibe
GitHub
167
sdxl_gen_img.py
167
sdxl_gen_img.py
@@ -55,6 +55,10 @@ from networks.lora import LoRANetwork
|
||||
from library.sdxl_original_unet import InferSdxlUNet2DConditionModel
|
||||
from library.original_unet import FlashAttentionFunction
|
||||
from networks.control_net_lllite import ControlNetLLLite
|
||||
from library.utils import setup_logging
|
||||
setup_logging()
|
||||
import logging
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
# scheduler:
|
||||
SCHEDULER_LINEAR_START = 0.00085
|
||||
@@ -76,12 +80,12 @@ CLIP_VISION_MODEL = "laion/CLIP-ViT-bigG-14-laion2B-39B-b160k"
|
||||
|
||||
def replace_unet_modules(unet: diffusers.models.unet_2d_condition.UNet2DConditionModel, mem_eff_attn, xformers, sdpa):
|
||||
if mem_eff_attn:
|
||||
print("Enable memory efficient attention for U-Net")
|
||||
logger.info("Enable memory efficient attention for U-Net")
|
||||
|
||||
# これはDiffusersのU-Netではなく自前のU-Netなので置き換えなくても良い
|
||||
unet.set_use_memory_efficient_attention(False, True)
|
||||
elif xformers:
|
||||
print("Enable xformers for U-Net")
|
||||
logger.info("Enable xformers for U-Net")
|
||||
try:
|
||||
import xformers.ops
|
||||
except ImportError:
|
||||
@@ -89,7 +93,7 @@ def replace_unet_modules(unet: diffusers.models.unet_2d_condition.UNet2DConditio
|
||||
|
||||
unet.set_use_memory_efficient_attention(True, False)
|
||||
elif sdpa:
|
||||
print("Enable SDPA for U-Net")
|
||||
logger.info("Enable SDPA for U-Net")
|
||||
unet.set_use_memory_efficient_attention(False, False)
|
||||
unet.set_use_sdpa(True)
|
||||
|
||||
@@ -106,7 +110,7 @@ def replace_vae_modules(vae: diffusers.models.AutoencoderKL, mem_eff_attn, xform
|
||||
|
||||
|
||||
def replace_vae_attn_to_memory_efficient():
|
||||
print("VAE Attention.forward has been replaced to FlashAttention (not xformers)")
|
||||
logger.info("VAE Attention.forward has been replaced to FlashAttention (not xformers)")
|
||||
flash_func = FlashAttentionFunction
|
||||
|
||||
def forward_flash_attn(self, hidden_states, **kwargs):
|
||||
@@ -162,7 +166,7 @@ def replace_vae_attn_to_memory_efficient():
|
||||
|
||||
|
||||
def replace_vae_attn_to_xformers():
|
||||
print("VAE: Attention.forward has been replaced to xformers")
|
||||
logger.info("VAE: Attention.forward has been replaced to xformers")
|
||||
import xformers.ops
|
||||
|
||||
def forward_xformers(self, hidden_states, **kwargs):
|
||||
@@ -218,7 +222,7 @@ def replace_vae_attn_to_xformers():
|
||||
|
||||
|
||||
def replace_vae_attn_to_sdpa():
|
||||
print("VAE: Attention.forward has been replaced to sdpa")
|
||||
logger.info("VAE: Attention.forward has been replaced to sdpa")
|
||||
|
||||
def forward_sdpa(self, hidden_states, **kwargs):
|
||||
residual = hidden_states
|
||||
@@ -352,7 +356,7 @@ class PipelineLike:
|
||||
token_replacements = self.token_replacements_list[tokenizer_index]
|
||||
|
||||
def replace_tokens(tokens):
|
||||
# print("replace_tokens", tokens, "=>", token_replacements)
|
||||
# logger.info("replace_tokens", tokens, "=>", token_replacements)
|
||||
if isinstance(tokens, torch.Tensor):
|
||||
tokens = tokens.tolist()
|
||||
|
||||
@@ -444,7 +448,7 @@ class PipelineLike:
|
||||
do_classifier_free_guidance = guidance_scale > 1.0
|
||||
|
||||
if not do_classifier_free_guidance and negative_scale is not None:
|
||||
print(f"negative_scale is ignored if guidance scalle <= 1.0")
|
||||
logger.info(f"negative_scale is ignored if guidance scalle <= 1.0")
|
||||
negative_scale = None
|
||||
|
||||
# get unconditional embeddings for classifier free guidance
|
||||
@@ -548,7 +552,7 @@ class PipelineLike:
|
||||
text_pool = text_pool[num_sub_prompts - 1 :: num_sub_prompts] # last subprompt
|
||||
|
||||
if init_image is not None and self.clip_vision_model is not None:
|
||||
print(f"encode by clip_vision_model and apply clip_vision_strength={self.clip_vision_strength}")
|
||||
logger.info(f"encode by clip_vision_model and apply clip_vision_strength={self.clip_vision_strength}")
|
||||
vision_input = self.clip_vision_processor(init_image, return_tensors="pt", device=self.device)
|
||||
pixel_values = vision_input["pixel_values"].to(self.device, dtype=text_embeddings.dtype)
|
||||
|
||||
@@ -715,7 +719,7 @@ class PipelineLike:
|
||||
if not enabled or ratio >= 1.0:
|
||||
continue
|
||||
if ratio < i / len(timesteps):
|
||||
print(f"ControlNet {j} is disabled (ratio={ratio} at {i} / {len(timesteps)})")
|
||||
logger.info(f"ControlNet {j} is disabled (ratio={ratio} at {i} / {len(timesteps)})")
|
||||
control_net.set_cond_image(None)
|
||||
each_control_net_enabled[j] = False
|
||||
|
||||
@@ -935,7 +939,7 @@ def get_prompts_with_weights(tokenizer: CLIPTokenizer, token_replacer, prompt: L
|
||||
if word.strip() == "BREAK":
|
||||
# pad until next multiple of tokenizer's max token length
|
||||
pad_len = tokenizer.model_max_length - (len(text_token) % tokenizer.model_max_length)
|
||||
print(f"BREAK pad_len: {pad_len}")
|
||||
logger.info(f"BREAK pad_len: {pad_len}")
|
||||
for i in range(pad_len):
|
||||
# v2のときEOSをつけるべきかどうかわからないぜ
|
||||
# if i == 0:
|
||||
@@ -965,7 +969,7 @@ def get_prompts_with_weights(tokenizer: CLIPTokenizer, token_replacer, prompt: L
|
||||
tokens.append(text_token)
|
||||
weights.append(text_weight)
|
||||
if truncated:
|
||||
print("warning: Prompt was truncated. Try to shorten the prompt or increase max_embeddings_multiples")
|
||||
logger.warning("warning: Prompt was truncated. Try to shorten the prompt or increase max_embeddings_multiples")
|
||||
return tokens, weights
|
||||
|
||||
|
||||
@@ -1238,7 +1242,7 @@ def handle_dynamic_prompt_variants(prompt, repeat_count):
|
||||
elif len(count_range) == 2:
|
||||
count_range = [int(count_range[0]), int(count_range[1])]
|
||||
else:
|
||||
print(f"invalid count range: {count_range}")
|
||||
logger.warning(f"invalid count range: {count_range}")
|
||||
count_range = [1, 1]
|
||||
if count_range[0] > count_range[1]:
|
||||
count_range = [count_range[1], count_range[0]]
|
||||
@@ -1308,7 +1312,7 @@ def handle_dynamic_prompt_variants(prompt, repeat_count):
|
||||
|
||||
|
||||
# def load_clip_l14_336(dtype):
|
||||
# print(f"loading CLIP: {CLIP_ID_L14_336}")
|
||||
# logger.info(f"loading CLIP: {CLIP_ID_L14_336}")
|
||||
# text_encoder = CLIPTextModel.from_pretrained(CLIP_ID_L14_336, torch_dtype=dtype)
|
||||
# return text_encoder
|
||||
|
||||
@@ -1378,7 +1382,7 @@ def main(args):
|
||||
replace_vae_modules(vae, mem_eff, args.xformers, args.sdpa)
|
||||
|
||||
# tokenizerを読み込む
|
||||
print("loading tokenizer")
|
||||
logger.info("loading tokenizer")
|
||||
tokenizer1, tokenizer2 = sdxl_train_util.load_tokenizers(args)
|
||||
|
||||
# schedulerを用意する
|
||||
@@ -1452,7 +1456,7 @@ def main(args):
|
||||
self.sampler_noises = noises
|
||||
|
||||
def randn(self, shape, device=None, dtype=None, layout=None, generator=None):
|
||||
# print("replacing", shape, len(self.sampler_noises), self.sampler_noise_index)
|
||||
# logger.info("replacing", shape, len(self.sampler_noises), self.sampler_noise_index)
|
||||
if self.sampler_noises is not None and self.sampler_noise_index < len(self.sampler_noises):
|
||||
noise = self.sampler_noises[self.sampler_noise_index]
|
||||
if shape != noise.shape:
|
||||
@@ -1461,7 +1465,7 @@ def main(args):
|
||||
noise = None
|
||||
|
||||
if noise == None:
|
||||
print(f"unexpected noise request: {self.sampler_noise_index}, {shape}")
|
||||
logger.warning(f"unexpected noise request: {self.sampler_noise_index}, {shape}")
|
||||
noise = torch.randn(shape, dtype=dtype, device=device, generator=generator)
|
||||
|
||||
self.sampler_noise_index += 1
|
||||
@@ -1493,7 +1497,7 @@ def main(args):
|
||||
# ↓以下は結局PipeでFalseに設定されるので意味がなかった
|
||||
# # clip_sample=Trueにする
|
||||
# if hasattr(scheduler.config, "clip_sample") and scheduler.config.clip_sample is False:
|
||||
# print("set clip_sample to True")
|
||||
# logger.info("set clip_sample to True")
|
||||
# scheduler.config.clip_sample = True
|
||||
|
||||
# deviceを決定する
|
||||
@@ -1522,7 +1526,7 @@ def main(args):
|
||||
|
||||
vae_dtype = dtype
|
||||
if args.no_half_vae:
|
||||
print("set vae_dtype to float32")
|
||||
logger.info("set vae_dtype to float32")
|
||||
vae_dtype = torch.float32
|
||||
vae.to(vae_dtype).to(device)
|
||||
vae.eval()
|
||||
@@ -1547,10 +1551,10 @@ def main(args):
|
||||
network_merge = args.network_merge_n_models
|
||||
else:
|
||||
network_merge = 0
|
||||
print(f"network_merge: {network_merge}")
|
||||
logger.info(f"network_merge: {network_merge}")
|
||||
|
||||
for i, network_module in enumerate(args.network_module):
|
||||
print("import network module:", network_module)
|
||||
logger.info(f"import network module: {network_module}")
|
||||
imported_module = importlib.import_module(network_module)
|
||||
|
||||
network_mul = 1.0 if args.network_mul is None or len(args.network_mul) <= i else args.network_mul[i]
|
||||
@@ -1568,7 +1572,7 @@ def main(args):
|
||||
raise ValueError("No weight. Weight is required.")
|
||||
|
||||
network_weight = args.network_weights[i]
|
||||
print("load network weights from:", network_weight)
|
||||
logger.info(f"load network weights from: {network_weight}")
|
||||
|
||||
if model_util.is_safetensors(network_weight) and args.network_show_meta:
|
||||
from safetensors.torch import safe_open
|
||||
@@ -1576,7 +1580,7 @@ def main(args):
|
||||
with safe_open(network_weight, framework="pt") as f:
|
||||
metadata = f.metadata()
|
||||
if metadata is not None:
|
||||
print(f"metadata for: {network_weight}: {metadata}")
|
||||
logger.info(f"metadata for: {network_weight}: {metadata}")
|
||||
|
||||
network, weights_sd = imported_module.create_network_from_weights(
|
||||
network_mul, network_weight, vae, [text_encoder1, text_encoder2], unet, for_inference=True, **net_kwargs
|
||||
@@ -1586,20 +1590,20 @@ def main(args):
|
||||
|
||||
mergeable = network.is_mergeable()
|
||||
if network_merge and not mergeable:
|
||||
print("network is not mergiable. ignore merge option.")
|
||||
logger.warning("network is not mergiable. ignore merge option.")
|
||||
|
||||
if not mergeable or i >= network_merge:
|
||||
# not merging
|
||||
network.apply_to([text_encoder1, text_encoder2], unet)
|
||||
info = network.load_state_dict(weights_sd, False) # network.load_weightsを使うようにするとよい
|
||||
print(f"weights are loaded: {info}")
|
||||
logger.info(f"weights are loaded: {info}")
|
||||
|
||||
if args.opt_channels_last:
|
||||
network.to(memory_format=torch.channels_last)
|
||||
network.to(dtype).to(device)
|
||||
|
||||
if network_pre_calc:
|
||||
print("backup original weights")
|
||||
logger.info("backup original weights")
|
||||
network.backup_weights()
|
||||
|
||||
networks.append(network)
|
||||
@@ -1613,7 +1617,7 @@ def main(args):
|
||||
# upscalerの指定があれば取得する
|
||||
upscaler = None
|
||||
if args.highres_fix_upscaler:
|
||||
print("import upscaler module:", args.highres_fix_upscaler)
|
||||
logger.info(f"import upscaler module: {args.highres_fix_upscaler}")
|
||||
imported_module = importlib.import_module(args.highres_fix_upscaler)
|
||||
|
||||
us_kwargs = {}
|
||||
@@ -1622,7 +1626,7 @@ def main(args):
|
||||
key, value = net_arg.split("=")
|
||||
us_kwargs[key] = value
|
||||
|
||||
print("create upscaler")
|
||||
logger.info("create upscaler")
|
||||
upscaler = imported_module.create_upscaler(**us_kwargs)
|
||||
upscaler.to(dtype).to(device)
|
||||
|
||||
@@ -1639,7 +1643,7 @@ def main(args):
|
||||
# control_nets.append(ControlNetInfo(ctrl_unet, ctrl_net, prep, weight, ratio))
|
||||
if args.control_net_lllite_models:
|
||||
for i, model_file in enumerate(args.control_net_lllite_models):
|
||||
print(f"loading ControlNet-LLLite: {model_file}")
|
||||
logger.info(f"loading ControlNet-LLLite: {model_file}")
|
||||
|
||||
from safetensors.torch import load_file
|
||||
|
||||
@@ -1670,7 +1674,7 @@ def main(args):
|
||||
control_nets.append((control_net, ratio))
|
||||
|
||||
if args.opt_channels_last:
|
||||
print(f"set optimizing: channels last")
|
||||
logger.info(f"set optimizing: channels last")
|
||||
text_encoder1.to(memory_format=torch.channels_last)
|
||||
text_encoder2.to(memory_format=torch.channels_last)
|
||||
vae.to(memory_format=torch.channels_last)
|
||||
@@ -1694,7 +1698,7 @@ def main(args):
|
||||
args.clip_skip,
|
||||
)
|
||||
pipe.set_control_nets(control_nets)
|
||||
print("pipeline is ready.")
|
||||
logger.info("pipeline is ready.")
|
||||
|
||||
if args.diffusers_xformers:
|
||||
pipe.enable_xformers_memory_efficient_attention()
|
||||
@@ -1736,7 +1740,7 @@ def main(args):
|
||||
|
||||
token_ids1 = tokenizer1.convert_tokens_to_ids(token_strings)
|
||||
token_ids2 = tokenizer2.convert_tokens_to_ids(token_strings)
|
||||
print(f"Textual Inversion embeddings `{token_string}` loaded. Tokens are added: {token_ids1} and {token_ids2}")
|
||||
logger.info(f"Textual Inversion embeddings `{token_string}` loaded. Tokens are added: {token_ids1} and {token_ids2}")
|
||||
assert (
|
||||
min(token_ids1) == token_ids1[0] and token_ids1[-1] == token_ids1[0] + len(token_ids1) - 1
|
||||
), f"token ids1 is not ordered"
|
||||
@@ -1766,7 +1770,7 @@ def main(args):
|
||||
|
||||
# promptを取得する
|
||||
if args.from_file is not None:
|
||||
print(f"reading prompts from {args.from_file}")
|
||||
logger.info(f"reading prompts from {args.from_file}")
|
||||
with open(args.from_file, "r", encoding="utf-8") as f:
|
||||
prompt_list = f.read().splitlines()
|
||||
prompt_list = [d for d in prompt_list if len(d.strip()) > 0]
|
||||
@@ -1795,7 +1799,7 @@ def main(args):
|
||||
for p in paths:
|
||||
image = Image.open(p)
|
||||
if image.mode != "RGB":
|
||||
print(f"convert image to RGB from {image.mode}: {p}")
|
||||
logger.info(f"convert image to RGB from {image.mode}: {p}")
|
||||
image = image.convert("RGB")
|
||||
images.append(image)
|
||||
|
||||
@@ -1811,14 +1815,14 @@ def main(args):
|
||||
return resized
|
||||
|
||||
if args.image_path is not None:
|
||||
print(f"load image for img2img: {args.image_path}")
|
||||
logger.info(f"load image for img2img: {args.image_path}")
|
||||
init_images = load_images(args.image_path)
|
||||
assert len(init_images) > 0, f"No image / 画像がありません: {args.image_path}"
|
||||
print(f"loaded {len(init_images)} images for img2img")
|
||||
logger.info(f"loaded {len(init_images)} images for img2img")
|
||||
|
||||
# CLIP Vision
|
||||
if args.clip_vision_strength is not None:
|
||||
print(f"load CLIP Vision model: {CLIP_VISION_MODEL}")
|
||||
logger.info(f"load CLIP Vision model: {CLIP_VISION_MODEL}")
|
||||
vision_model = CLIPVisionModelWithProjection.from_pretrained(CLIP_VISION_MODEL, projection_dim=1280)
|
||||
vision_model.to(device, dtype)
|
||||
processor = CLIPImageProcessor.from_pretrained(CLIP_VISION_MODEL)
|
||||
@@ -1826,22 +1830,22 @@ def main(args):
|
||||
pipe.clip_vision_model = vision_model
|
||||
pipe.clip_vision_processor = processor
|
||||
pipe.clip_vision_strength = args.clip_vision_strength
|
||||
print(f"CLIP Vision model loaded.")
|
||||
logger.info(f"CLIP Vision model loaded.")
|
||||
|
||||
else:
|
||||
init_images = None
|
||||
|
||||
if args.mask_path is not None:
|
||||
print(f"load mask for inpainting: {args.mask_path}")
|
||||
logger.info(f"load mask for inpainting: {args.mask_path}")
|
||||
mask_images = load_images(args.mask_path)
|
||||
assert len(mask_images) > 0, f"No mask image / マスク画像がありません: {args.image_path}"
|
||||
print(f"loaded {len(mask_images)} mask images for inpainting")
|
||||
logger.info(f"loaded {len(mask_images)} mask images for inpainting")
|
||||
else:
|
||||
mask_images = None
|
||||
|
||||
# promptがないとき、画像のPngInfoから取得する
|
||||
if init_images is not None and len(prompt_list) == 0 and not args.interactive:
|
||||
print("get prompts from images' metadata")
|
||||
logger.info("get prompts from images' metadata")
|
||||
for img in init_images:
|
||||
if "prompt" in img.text:
|
||||
prompt = img.text["prompt"]
|
||||
@@ -1870,17 +1874,17 @@ def main(args):
|
||||
h = int(h * args.highres_fix_scale + 0.5)
|
||||
|
||||
if init_images is not None:
|
||||
print(f"resize img2img source images to {w}*{h}")
|
||||
logger.info(f"resize img2img source images to {w}*{h}")
|
||||
init_images = resize_images(init_images, (w, h))
|
||||
if mask_images is not None:
|
||||
print(f"resize img2img mask images to {w}*{h}")
|
||||
logger.info(f"resize img2img mask images to {w}*{h}")
|
||||
mask_images = resize_images(mask_images, (w, h))
|
||||
|
||||
regional_network = False
|
||||
if networks and mask_images:
|
||||
# mask を領域情報として流用する、現在は一回のコマンド呼び出しで1枚だけ対応
|
||||
regional_network = True
|
||||
print("use mask as region")
|
||||
logger.info("use mask as region")
|
||||
|
||||
size = None
|
||||
for i, network in enumerate(networks):
|
||||
@@ -1905,14 +1909,14 @@ def main(args):
|
||||
|
||||
prev_image = None # for VGG16 guided
|
||||
if args.guide_image_path is not None:
|
||||
print(f"load image for ControlNet guidance: {args.guide_image_path}")
|
||||
logger.info(f"load image for ControlNet guidance: {args.guide_image_path}")
|
||||
guide_images = []
|
||||
for p in args.guide_image_path:
|
||||
guide_images.extend(load_images(p))
|
||||
|
||||
print(f"loaded {len(guide_images)} guide images for guidance")
|
||||
logger.info(f"loaded {len(guide_images)} guide images for guidance")
|
||||
if len(guide_images) == 0:
|
||||
print(f"No guide image, use previous generated image. / ガイド画像がありません。直前に生成した画像を使います: {args.image_path}")
|
||||
logger.warning(f"No guide image, use previous generated image. / ガイド画像がありません。直前に生成した画像を使います: {args.image_path}")
|
||||
guide_images = None
|
||||
else:
|
||||
guide_images = None
|
||||
@@ -1938,7 +1942,7 @@ def main(args):
|
||||
max_embeddings_multiples = 1 if args.max_embeddings_multiples is None else args.max_embeddings_multiples
|
||||
|
||||
for gen_iter in range(args.n_iter):
|
||||
print(f"iteration {gen_iter+1}/{args.n_iter}")
|
||||
logger.info(f"iteration {gen_iter+1}/{args.n_iter}")
|
||||
iter_seed = random.randint(0, 0x7FFFFFFF)
|
||||
|
||||
# バッチ処理の関数
|
||||
@@ -1950,7 +1954,7 @@ def main(args):
|
||||
# 1st stageのバッチを作成して呼び出す：サイズを小さくして呼び出す
|
||||
is_1st_latent = upscaler.support_latents() if upscaler else args.highres_fix_latents_upscaling
|
||||
|
||||
print("process 1st stage")
|
||||
logger.info("process 1st stage")
|
||||
batch_1st = []
|
||||
for _, base, ext in batch:
|
||||
|
||||
@@ -1995,7 +1999,7 @@ def main(args):
|
||||
images_1st = process_batch(batch_1st, True, True)
|
||||
|
||||
# 2nd stageのバッチを作成して以下処理する
|
||||
print("process 2nd stage")
|
||||
logger.info("process 2nd stage")
|
||||
width_2nd, height_2nd = batch[0].ext.width, batch[0].ext.height
|
||||
|
||||
if upscaler:
|
||||
@@ -2161,7 +2165,7 @@ def main(args):
|
||||
n.restore_weights()
|
||||
for n in networks:
|
||||
n.pre_calculation()
|
||||
print("pre-calculation... done")
|
||||
logger.info("pre-calculation... done")
|
||||
|
||||
images = pipe(
|
||||
prompts,
|
||||
@@ -2240,7 +2244,7 @@ def main(args):
|
||||
cv2.waitKey()
|
||||
cv2.destroyAllWindows()
|
||||
except ImportError:
|
||||
print("opencv-python is not installed, cannot preview / opencv-pythonがインストールされていないためプレビューできません")
|
||||
logger.error("opencv-python is not installed, cannot preview / opencv-pythonがインストールされていないためプレビューできません")
|
||||
|
||||
return images
|
||||
|
||||
@@ -2253,7 +2257,8 @@ def main(args):
|
||||
# interactive
|
||||
valid = False
|
||||
while not valid:
|
||||
print("\nType prompt:")
|
||||
logger.info("")
|
||||
logger.info("Type prompt:")
|
||||
try:
|
||||
raw_prompt = input()
|
||||
except EOFError:
|
||||
@@ -2302,74 +2307,74 @@ def main(args):
|
||||
|
||||
prompt_args = raw_prompt.strip().split(" --")
|
||||
prompt = prompt_args[0]
|
||||
print(f"prompt {prompt_index+1}/{len(prompt_list)}: {prompt}")
|
||||
logger.info(f"prompt {prompt_index+1}/{len(prompt_list)}: {prompt}")
|
||||
|
||||
for parg in prompt_args[1:]:
|
||||
try:
|
||||
m = re.match(r"w (\d+)", parg, re.IGNORECASE)
|
||||
if m:
|
||||
width = int(m.group(1))
|
||||
print(f"width: {width}")
|
||||
logger.info(f"width: {width}")
|
||||
continue
|
||||
|
||||
m = re.match(r"h (\d+)", parg, re.IGNORECASE)
|
||||
if m:
|
||||
height = int(m.group(1))
|
||||
print(f"height: {height}")
|
||||
logger.info(f"height: {height}")
|
||||
continue
|
||||
|
||||
m = re.match(r"ow (\d+)", parg, re.IGNORECASE)
|
||||
if m:
|
||||
original_width = int(m.group(1))
|
||||
print(f"original width: {original_width}")
|
||||
logger.info(f"original width: {original_width}")
|
||||
continue
|
||||
|
||||
m = re.match(r"oh (\d+)", parg, re.IGNORECASE)
|
||||
if m:
|
||||
original_height = int(m.group(1))
|
||||
print(f"original height: {original_height}")
|
||||
logger.info(f"original height: {original_height}")
|
||||
continue
|
||||
|
||||
m = re.match(r"nw (\d+)", parg, re.IGNORECASE)
|
||||
if m:
|
||||
original_width_negative = int(m.group(1))
|
||||
print(f"original width negative: {original_width_negative}")
|
||||
logger.info(f"original width negative: {original_width_negative}")
|
||||
continue
|
||||
|
||||
m = re.match(r"nh (\d+)", parg, re.IGNORECASE)
|
||||
if m:
|
||||
original_height_negative = int(m.group(1))
|
||||
print(f"original height negative: {original_height_negative}")
|
||||
logger.info(f"original height negative: {original_height_negative}")
|
||||
continue
|
||||
|
||||
m = re.match(r"ct (\d+)", parg, re.IGNORECASE)
|
||||
if m:
|
||||
crop_top = int(m.group(1))
|
||||
print(f"crop top: {crop_top}")
|
||||
logger.info(f"crop top: {crop_top}")
|
||||
continue
|
||||
|
||||
m = re.match(r"cl (\d+)", parg, re.IGNORECASE)
|
||||
if m:
|
||||
crop_left = int(m.group(1))
|
||||
print(f"crop left: {crop_left}")
|
||||
logger.info(f"crop left: {crop_left}")
|
||||
continue
|
||||
|
||||
m = re.match(r"s (\d+)", parg, re.IGNORECASE)
|
||||
if m: # steps
|
||||
steps = max(1, min(1000, int(m.group(1))))
|
||||
print(f"steps: {steps}")
|
||||
logger.info(f"steps: {steps}")
|
||||
continue
|
||||
|
||||
m = re.match(r"d ([\d,]+)", parg, re.IGNORECASE)
|
||||
if m: # seed
|
||||
seeds = [int(d) for d in m.group(1).split(",")]
|
||||
print(f"seeds: {seeds}")
|
||||
logger.info(f"seeds: {seeds}")
|
||||
continue
|
||||
|
||||
m = re.match(r"l ([\d\.]+)", parg, re.IGNORECASE)
|
||||
if m: # scale
|
||||
scale = float(m.group(1))
|
||||
print(f"scale: {scale}")
|
||||
logger.info(f"scale: {scale}")
|
||||
continue
|
||||
|
||||
m = re.match(r"nl ([\d\.]+|none|None)", parg, re.IGNORECASE)
|
||||
@@ -2378,25 +2383,25 @@ def main(args):
|
||||
negative_scale = None
|
||||
else:
|
||||
negative_scale = float(m.group(1))
|
||||
print(f"negative scale: {negative_scale}")
|
||||
logger.info(f"negative scale: {negative_scale}")
|
||||
continue
|
||||
|
||||
m = re.match(r"t ([\d\.]+)", parg, re.IGNORECASE)
|
||||
if m: # strength
|
||||
strength = float(m.group(1))
|
||||
print(f"strength: {strength}")
|
||||
logger.info(f"strength: {strength}")
|
||||
continue
|
||||
|
||||
m = re.match(r"n (.+)", parg, re.IGNORECASE)
|
||||
if m: # negative prompt
|
||||
negative_prompt = m.group(1)
|
||||
print(f"negative prompt: {negative_prompt}")
|
||||
logger.info(f"negative prompt: {negative_prompt}")
|
||||
continue
|
||||
|
||||
m = re.match(r"c (.+)", parg, re.IGNORECASE)
|
||||
if m: # clip prompt
|
||||
clip_prompt = m.group(1)
|
||||
print(f"clip prompt: {clip_prompt}")
|
||||
logger.info(f"clip prompt: {clip_prompt}")
|
||||
continue
|
||||
|
||||
m = re.match(r"am ([\d\.\-,]+)", parg, re.IGNORECASE)
|
||||
@@ -2404,47 +2409,47 @@ def main(args):
|
||||
network_muls = [float(v) for v in m.group(1).split(",")]
|
||||
while len(network_muls) < len(networks):
|
||||
network_muls.append(network_muls[-1])
|
||||
print(f"network mul: {network_muls}")
|
||||
logger.info(f"network mul: {network_muls}")
|
||||
continue
|
||||
|
||||
# Deep Shrink
|
||||
m = re.match(r"dsd1 ([\d\.]+)", parg, re.IGNORECASE)
|
||||
if m: # deep shrink depth 1
|
||||
ds_depth_1 = int(m.group(1))
|
||||
print(f"deep shrink depth 1: {ds_depth_1}")
|
||||
logger.info(f"deep shrink depth 1: {ds_depth_1}")
|
||||
continue
|
||||
|
||||
m = re.match(r"dst1 ([\d\.]+)", parg, re.IGNORECASE)
|
||||
if m: # deep shrink timesteps 1
|
||||
ds_timesteps_1 = int(m.group(1))
|
||||
ds_depth_1 = ds_depth_1 if ds_depth_1 is not None else -1 # -1 means override
|
||||
print(f"deep shrink timesteps 1: {ds_timesteps_1}")
|
||||
logger.info(f"deep shrink timesteps 1: {ds_timesteps_1}")
|
||||
continue
|
||||
|
||||
m = re.match(r"dsd2 ([\d\.]+)", parg, re.IGNORECASE)
|
||||
if m: # deep shrink depth 2
|
||||
ds_depth_2 = int(m.group(1))
|
||||
ds_depth_1 = ds_depth_1 if ds_depth_1 is not None else -1 # -1 means override
|
||||
print(f"deep shrink depth 2: {ds_depth_2}")
|
||||
logger.info(f"deep shrink depth 2: {ds_depth_2}")
|
||||
continue
|
||||
|
||||
m = re.match(r"dst2 ([\d\.]+)", parg, re.IGNORECASE)
|
||||
if m: # deep shrink timesteps 2
|
||||
ds_timesteps_2 = int(m.group(1))
|
||||
ds_depth_1 = ds_depth_1 if ds_depth_1 is not None else -1 # -1 means override
|
||||
print(f"deep shrink timesteps 2: {ds_timesteps_2}")
|
||||
logger.info(f"deep shrink timesteps 2: {ds_timesteps_2}")
|
||||
continue
|
||||
|
||||
m = re.match(r"dsr ([\d\.]+)", parg, re.IGNORECASE)
|
||||
if m: # deep shrink ratio
|
||||
ds_ratio = float(m.group(1))
|
||||
ds_depth_1 = ds_depth_1 if ds_depth_1 is not None else -1 # -1 means override
|
||||
print(f"deep shrink ratio: {ds_ratio}")
|
||||
logger.info(f"deep shrink ratio: {ds_ratio}")
|
||||
continue
|
||||
|
||||
except ValueError as ex:
|
||||
print(f"Exception in parsing / 解析エラー: {parg}")
|
||||
print(ex)
|
||||
logger.error(f"Exception in parsing / 解析エラー: {parg}")
|
||||
logger.error(f"{ex}")
|
||||
|
||||
# override Deep Shrink
|
||||
if ds_depth_1 is not None:
|
||||
@@ -2462,7 +2467,7 @@ def main(args):
|
||||
if len(predefined_seeds) > 0:
|
||||
seed = predefined_seeds.pop(0)
|
||||
else:
|
||||
print("predefined seeds are exhausted")
|
||||
logger.error("predefined seeds are exhausted")
|
||||
seed = None
|
||||
elif args.iter_same_seed:
|
||||
seeds = iter_seed
|
||||
@@ -2472,7 +2477,7 @@ def main(args):
|
||||
if seed is None:
|
||||
seed = random.randint(0, 0x7FFFFFFF)
|
||||
if args.interactive:
|
||||
print(f"seed: {seed}")
|
||||
logger.info(f"seed: {seed}")
|
||||
|
||||
# prepare init image, guide image and mask
|
||||
init_image = mask_image = guide_image = None
|
||||
@@ -2488,7 +2493,7 @@ def main(args):
|
||||
width = width - width % 32
|
||||
height = height - height % 32
|
||||
if width != init_image.size[0] or height != init_image.size[1]:
|
||||
print(
|
||||
logger.warning(
|
||||
f"img2img image size is not divisible by 32 so aspect ratio is changed / img2imgの画像サイズが32で割り切れないためリサイズされます。画像が歪みます"
|
||||
)
|
||||
|
||||
@@ -2548,7 +2553,7 @@ def main(args):
|
||||
process_batch(batch_data, highres_fix)
|
||||
batch_data.clear()
|
||||
|
||||
print("done!")
|
||||
logger.info("done!")
|
||||
|
||||
|
||||
def setup_parser() -> argparse.ArgumentParser:
|
||||
|
||||
Reference in New Issue
Block a user