import spaces
import os
from stablepy import (
Model_Diffusers,
SCHEDULE_TYPE_OPTIONS,
SCHEDULE_PREDICTION_TYPE_OPTIONS,
check_scheduler_compatibility,
TASK_AND_PREPROCESSORS,
FACE_RESTORATION_MODELS,
scheduler_names,
)
from constants import (
DIRECTORY_UPSCALERS,
TASK_STABLEPY,
TASK_MODEL_LIST,
UPSCALER_DICT_GUI,
UPSCALER_KEYS,
PROMPT_W_OPTIONS,
WARNING_MSG_VAE,
SDXL_TASK,
MODEL_TYPE_TASK,
POST_PROCESSING_SAMPLER,
DIFFUSERS_CONTROLNET_MODEL,
IP_MODELS,
MODE_IP_OPTIONS,
)
from stablepy.diffusers_vanilla.style_prompt_config import STYLE_NAMES
import torch
import re
import time
from PIL import ImageFile
from utils import (
get_model_list,
extract_parameters,
get_model_type,
extract_exif_data,
create_mask_now,
download_diffuser_repo,
get_used_storage_gb,
delete_model,
progress_step_bar,
html_template_message,
escape_html,
)
from image_processor import preprocessor_tab
from datetime import datetime
import gradio as gr
import logging
import diffusers
import warnings
from stablepy import logger
from diffusers import FluxPipeline
# import urllib.parse
import subprocess
subprocess.run("rm -rf /data-nvme/zerogpu-offload/*", env={}, shell=True)
ImageFile.LOAD_TRUNCATED_IMAGES = True
torch.backends.cuda.matmul.allow_tf32 = True
# os.environ["PYTORCH_NO_CUDA_MEMORY_CACHING"] = "1"
print(os.getenv("SPACES_ZERO_GPU"))
## BEGIN MOD
from modutils import (list_uniq, download_private_repo, get_model_id_list, get_tupled_embed_list,
get_lora_model_list, get_all_lora_tupled_list, update_loras, apply_lora_prompt, set_prompt_loras,
get_my_lora, upload_file_lora, move_file_lora, search_civitai_lora, select_civitai_lora,
update_civitai_selection, get_civitai_tag, CIVITAI_SORT, CIVITAI_PERIOD, CIVITAI_BASEMODEL,
set_textual_inversion_prompt, get_model_pipeline, change_interface_mode, get_t2i_model_info, download_link_model,
get_tupled_model_list, save_gallery_images, save_gallery_history, set_optimization, set_sampler_settings,
set_quick_presets, process_style_prompt, optimization_list, save_images, download_things, valid_model_name,
preset_styles, preset_quality, preset_sampler_setting, translate_to_en, EXAMPLES_GUI, RESOURCES)
from env import (HF_TOKEN, CIVITAI_API_KEY, HF_LORA_ESSENTIAL_PRIVATE_REPO, HF_VAE_PRIVATE_REPO,
HF_SDXL_EMBEDS_NEGATIVE_PRIVATE_REPO, HF_SDXL_EMBEDS_POSITIVE_PRIVATE_REPO,
DIRECTORY_MODELS, DIRECTORY_LORAS, DIRECTORY_VAES, DIRECTORY_EMBEDS, DIRECTORY_EMBEDS_SDXL,
DIRECTORY_EMBEDS_POSITIVE_SDXL, LOAD_DIFFUSERS_FORMAT_MODEL,
DOWNLOAD_MODEL_LIST, DOWNLOAD_LORA_LIST, DOWNLOAD_VAE_LIST, DOWNLOAD_EMBEDS)
download_private_repo(HF_LORA_ESSENTIAL_PRIVATE_REPO, DIRECTORY_LORAS, True)
download_private_repo(HF_VAE_PRIVATE_REPO, DIRECTORY_VAES, False)
## END MOD
directories = [DIRECTORY_MODELS, DIRECTORY_LORAS, DIRECTORY_VAES, DIRECTORY_EMBEDS, DIRECTORY_UPSCALERS]
for directory in directories:
os.makedirs(directory, exist_ok=True)
# - **Download Models**
DOWNLOAD_MODEL = ", ".join(DOWNLOAD_MODEL_LIST)
# - **Download VAEs**
DOWNLOAD_VAE = ", ".join(DOWNLOAD_VAE_LIST)
# - **Download LoRAs**
DOWNLOAD_LORA = ", ".join(DOWNLOAD_LORA_LIST)
# Download stuffs
for url in [url.strip() for url in DOWNLOAD_MODEL.split(',')]:
if not os.path.exists(f"./models/{url.split('/')[-1]}"):
download_things(DIRECTORY_MODELS, url, HF_TOKEN, CIVITAI_API_KEY)
for url in [url.strip() for url in DOWNLOAD_VAE.split(',')]:
if not os.path.exists(f"./vaes/{url.split('/')[-1]}"):
download_things(DIRECTORY_VAES, url, HF_TOKEN, CIVITAI_API_KEY)
for url in [url.strip() for url in DOWNLOAD_LORA.split(',')]:
if not os.path.exists(f"./loras/{url.split('/')[-1]}"):
download_things(DIRECTORY_LORAS, url, HF_TOKEN, CIVITAI_API_KEY)
# Download Embeddings
for url_embed in DOWNLOAD_EMBEDS:
if not os.path.exists(f"./embedings/{url_embed.split('/')[-1]}"):
download_things(DIRECTORY_EMBEDS, url_embed, HF_TOKEN, CIVITAI_API_KEY)
# Build list models
embed_list = get_model_list(DIRECTORY_EMBEDS)
lora_model_list = get_lora_model_list()
vae_model_list = get_model_list(DIRECTORY_VAES)
vae_model_list.insert(0, "BakedVAE")
vae_model_list.insert(0, "None")
## BEGIN MOD
single_file_model_list = get_model_list(DIRECTORY_MODELS)
model_list = list_uniq(get_model_id_list() + LOAD_DIFFUSERS_FORMAT_MODEL + single_file_model_list)
download_private_repo(HF_SDXL_EMBEDS_NEGATIVE_PRIVATE_REPO, DIRECTORY_EMBEDS_SDXL, False)
download_private_repo(HF_SDXL_EMBEDS_POSITIVE_PRIVATE_REPO, DIRECTORY_EMBEDS_POSITIVE_SDXL, False)
embed_sdxl_list = get_model_list(DIRECTORY_EMBEDS_SDXL) + get_model_list(DIRECTORY_EMBEDS_POSITIVE_SDXL)
def get_embed_list(pipeline_name):
return get_tupled_embed_list(embed_sdxl_list if pipeline_name == "StableDiffusionXLPipeline" else embed_list)
## END MOD
print('\033[33m🏁 Download and listing of valid models completed.\033[0m')
flux_repo = "camenduru/FLUX.1-dev-diffusers"
flux_pipe = FluxPipeline.from_pretrained(
flux_repo,
transformer=None,
torch_dtype=torch.bfloat16,
)#.to("cuda")
components = flux_pipe.components
components.pop("transformer", None)
components.pop("scheduler", None)
delete_model(flux_repo)
# components = None
#######################
# GUI
#######################
logging.getLogger("diffusers").setLevel(logging.ERROR)
diffusers.utils.logging.set_verbosity(40)
warnings.filterwarnings(action="ignore", category=FutureWarning, module="diffusers")
warnings.filterwarnings(action="ignore", category=UserWarning, module="diffusers")
warnings.filterwarnings(action="ignore", category=FutureWarning, module="transformers")
## BEGIN MOD
#logger.setLevel(logging.CRITICAL)
logger.setLevel(logging.DEBUG)
from tagger.v2 import V2_ALL_MODELS, v2_random_prompt, v2_upsampling_prompt
from tagger.utils import (gradio_copy_text, COPY_ACTION_JS, gradio_copy_prompt,
V2_ASPECT_RATIO_OPTIONS, V2_RATING_OPTIONS, V2_LENGTH_OPTIONS, V2_IDENTITY_OPTIONS)
from tagger.tagger import (predict_tags_wd, convert_danbooru_to_e621_prompt,
remove_specific_prompt, insert_recom_prompt, insert_model_recom_prompt,
compose_prompt_to_copy, translate_prompt, select_random_character)
def description_ui():
gr.Markdown(
"""
## Danbooru Tags Transformer V2 Demo with WD Tagger
(Image =>) Prompt => Upsampled longer prompt
- Mod of p1atdev's [Danbooru Tags Transformer V2 Demo](https://huggingface.co/spaces/p1atdev/danbooru-tags-transformer-v2) and [WD Tagger with 🤗 transformers](https://huggingface.co/spaces/p1atdev/wd-tagger-transformers).
- Models: p1atdev's [wd-swinv2-tagger-v3-hf](https://huggingface.co/p1atdev/wd-swinv2-tagger-v3-hf), [dart-v2-moe-sft](https://huggingface.co/p1atdev/dart-v2-moe-sft)
"""
)
## END MOD
class GuiSD:
def __init__(self, stream=True):
self.model = None
self.status_loading = False
self.sleep_loading = 4
self.last_load = datetime.now()
self.inventory = []
def update_storage_models(self, storage_floor_gb=24, required_inventory_for_purge=3):
while get_used_storage_gb() > storage_floor_gb:
if len(self.inventory) < required_inventory_for_purge:
break
removal_candidate = self.inventory.pop(0)
delete_model(removal_candidate)
def update_inventory(self, model_name):
if model_name not in single_file_model_list:
self.inventory = [
m for m in self.inventory if m != model_name
] + [model_name]
print(self.inventory)
def load_new_model(self, model_name, vae_model, task, controlnet_model, progress=gr.Progress(track_tqdm=True)):
# download link model > model_name
if "http" in model_name: #
model_name, model_type = download_link_model(model_name, DIRECTORY_MODELS) #
is_link_model = True #
else: is_link_model = False #
self.update_storage_models()
vae_model = vae_model if vae_model != "None" else None
model_type = get_model_type(model_name) if not is_link_model else model_type #
dtype_model = torch.bfloat16 if model_type == "FLUX" else torch.float16
if not os.path.exists(model_name):
_ = download_diffuser_repo(
repo_name=model_name,
model_type=model_type,
revision="main",
token=True,
)
self.update_inventory(model_name)
for i in range(68):
if not self.status_loading:
self.status_loading = True
if i > 0:
time.sleep(self.sleep_loading)
print("Previous model ops...")
break
time.sleep(0.5)
print(f"Waiting queue {i}")
yield "Waiting queue"
self.status_loading = True
yield f"Loading model: {model_name}"
if vae_model == "BakedVAE":
vae_model = model_name
elif vae_model:
vae_type = "SDXL" if "sdxl" in vae_model.lower() else "SD 1.5"
if model_type != vae_type:
gr.Warning(WARNING_MSG_VAE)
print("Loading model...")
try:
start_time = time.time()
if self.model is None:
self.model = Model_Diffusers(
base_model_id=model_name,
task_name=TASK_STABLEPY[task],
vae_model=vae_model,
type_model_precision=dtype_model,
retain_task_model_in_cache=False,
controlnet_model=controlnet_model,
device="cpu",
env_components=components,
)
self.model.advanced_params(image_preprocessor_cuda_active=True)
else:
if self.model.base_model_id != model_name:
load_now_time = datetime.now()
elapsed_time = max((load_now_time - self.last_load).total_seconds(), 0)
if elapsed_time <= 9:
print("Waiting for the previous model's time ops...")
time.sleep(9 - elapsed_time)
self.model.device = torch.device("cpu")
self.model.load_pipe(
model_name,
task_name=TASK_STABLEPY[task],
vae_model=vae_model,
type_model_precision=dtype_model,
retain_task_model_in_cache=False,
controlnet_model=controlnet_model,
)
end_time = time.time()
self.sleep_loading = max(min(int(end_time - start_time), 10), 4)
except Exception as e:
self.last_load = datetime.now()
self.status_loading = False
self.sleep_loading = 4
raise e
self.last_load = datetime.now()
self.status_loading = False
yield f"Model loaded: {model_name}"
#@spaces.GPU
@torch.inference_mode()
def generate_pipeline(
self,
prompt,
neg_prompt,
num_images,
steps,
cfg,
clip_skip,
seed,
lora1,
lora_scale1,
lora2,
lora_scale2,
lora3,
lora_scale3,
lora4,
lora_scale4,
lora5,
lora_scale5,
lora6,
lora_scale6,
lora7,
lora_scale7,
sampler,
schedule_type,
schedule_prediction_type,
img_height,
img_width,
model_name,
vae_model,
task,
image_control,
preprocessor_name,
preprocess_resolution,
image_resolution,
style_prompt, # list []
style_json_file,
image_mask,
strength,
low_threshold,
high_threshold,
value_threshold,
distance_threshold,
recolor_gamma_correction,
tile_blur_sigma,
controlnet_output_scaling_in_unet,
controlnet_start_threshold,
controlnet_stop_threshold,
textual_inversion,
syntax_weights,
upscaler_model_path,
upscaler_increases_size,
upscaler_tile_size,
upscaler_tile_overlap,
hires_steps,
hires_denoising_strength,
hires_sampler,
hires_prompt,
hires_negative_prompt,
hires_before_adetailer,
hires_after_adetailer,
hires_schedule_type,
hires_guidance_scale,
controlnet_model,
loop_generation,
leave_progress_bar,
disable_progress_bar,
image_previews,
display_images,
save_generated_images,
filename_pattern,
image_storage_location,
retain_compel_previous_load,
retain_detailfix_model_previous_load,
retain_hires_model_previous_load,
t2i_adapter_preprocessor,
t2i_adapter_conditioning_scale,
t2i_adapter_conditioning_factor,
xformers_memory_efficient_attention,
freeu,
generator_in_cpu,
adetailer_inpaint_only,
adetailer_verbose,
adetailer_sampler,
adetailer_active_a,
prompt_ad_a,
negative_prompt_ad_a,
strength_ad_a,
face_detector_ad_a,
person_detector_ad_a,
hand_detector_ad_a,
mask_dilation_a,
mask_blur_a,
mask_padding_a,
adetailer_active_b,
prompt_ad_b,
negative_prompt_ad_b,
strength_ad_b,
face_detector_ad_b,
person_detector_ad_b,
hand_detector_ad_b,
mask_dilation_b,
mask_blur_b,
mask_padding_b,
retain_task_cache_gui,
guidance_rescale,
image_ip1,
mask_ip1,
model_ip1,
mode_ip1,
scale_ip1,
image_ip2,
mask_ip2,
model_ip2,
mode_ip2,
scale_ip2,
pag_scale,
face_restoration_model,
face_restoration_visibility,
face_restoration_weight,
):
info_state = html_template_message("Navigating latent space...")
yield info_state, gr.update(), gr.update()
vae_model = vae_model if vae_model != "None" else None
loras_list = [lora1, lora2, lora3, lora4, lora5, lora6, lora7]
vae_msg = f"VAE: {vae_model}" if vae_model else ""
msg_lora = ""
## BEGIN MOD
loras_list = [s if s else "None" for s in loras_list]
global lora_model_list
lora_model_list = get_lora_model_list()
lora1, lora_scale1, lora2, lora_scale2, lora3, lora_scale3, lora4, lora_scale4, lora5, lora_scale5, lora6, lora_scale6, lora7, lora_scale7 = \
set_prompt_loras(prompt, syntax_weights, model_name, lora1, lora_scale1, lora2, lora_scale2, lora3,
lora_scale3, lora4, lora_scale4, lora5, lora_scale5, lora6, lora_scale6, lora7, lora_scale7)
## END MOD
print("Config model:", model_name, vae_model, loras_list)
task = TASK_STABLEPY[task]
params_ip_img = []
params_ip_msk = []
params_ip_model = []
params_ip_mode = []
params_ip_scale = []
all_adapters = [
(image_ip1, mask_ip1, model_ip1, mode_ip1, scale_ip1),
(image_ip2, mask_ip2, model_ip2, mode_ip2, scale_ip2),
]
if not hasattr(self.model.pipe, "transformer"):
for imgip, mskip, modelip, modeip, scaleip in all_adapters:
if imgip:
params_ip_img.append(imgip)
if mskip:
params_ip_msk.append(mskip)
params_ip_model.append(modelip)
params_ip_mode.append(modeip)
params_ip_scale.append(scaleip)
concurrency = 5
self.model.stream_config(concurrency=concurrency, latent_resize_by=1, vae_decoding=False)
if task != "txt2img" and not image_control:
raise ValueError("Reference image is required. Please upload one in 'Image ControlNet/Inpaint/Img2img'.")
if task in ["inpaint", "repaint"] and not image_mask:
raise ValueError("Mask image not found. Upload one in 'Image Mask' to proceed.")
if "https://" not in str(UPSCALER_DICT_GUI[upscaler_model_path]):
upscaler_model = upscaler_model_path
else:
url_upscaler = UPSCALER_DICT_GUI[upscaler_model_path]
if not os.path.exists(f"./{DIRECTORY_UPSCALERS}/{url_upscaler.split('/')[-1]}"):
download_things(DIRECTORY_UPSCALERS, url_upscaler, HF_TOKEN)
upscaler_model = f"./{DIRECTORY_UPSCALERS}/{url_upscaler.split('/')[-1]}"
logging.getLogger("ultralytics").setLevel(logging.INFO if adetailer_verbose else logging.ERROR)
adetailer_params_A = {
"face_detector_ad": face_detector_ad_a,
"person_detector_ad": person_detector_ad_a,
"hand_detector_ad": hand_detector_ad_a,
"prompt": prompt_ad_a,
"negative_prompt": negative_prompt_ad_a,
"strength": strength_ad_a,
# "image_list_task" : None,
"mask_dilation": mask_dilation_a,
"mask_blur": mask_blur_a,
"mask_padding": mask_padding_a,
"inpaint_only": adetailer_inpaint_only,
"sampler": adetailer_sampler,
}
adetailer_params_B = {
"face_detector_ad": face_detector_ad_b,
"person_detector_ad": person_detector_ad_b,
"hand_detector_ad": hand_detector_ad_b,
"prompt": prompt_ad_b,
"negative_prompt": negative_prompt_ad_b,
"strength": strength_ad_b,
# "image_list_task" : None,
"mask_dilation": mask_dilation_b,
"mask_blur": mask_blur_b,
"mask_padding": mask_padding_b,
}
pipe_params = {
"prompt": prompt,
"negative_prompt": neg_prompt,
"img_height": img_height,
"img_width": img_width,
"num_images": num_images,
"num_steps": steps,
"guidance_scale": cfg,
"clip_skip": clip_skip,
"pag_scale": float(pag_scale),
"seed": seed,
"image": image_control,
"preprocessor_name": preprocessor_name,
"preprocess_resolution": preprocess_resolution,
"image_resolution": image_resolution,
"style_prompt": style_prompt if style_prompt else "",
"style_json_file": "",
"image_mask": image_mask, # only for Inpaint
"strength": strength, # only for Inpaint or ...
"low_threshold": low_threshold,
"high_threshold": high_threshold,
"value_threshold": value_threshold,
"distance_threshold": distance_threshold,
"recolor_gamma_correction": float(recolor_gamma_correction),
"tile_blur_sigma": int(tile_blur_sigma),
"lora_A": lora1 if lora1 != "None" else None,
"lora_scale_A": lora_scale1,
"lora_B": lora2 if lora2 != "None" else None,
"lora_scale_B": lora_scale2,
"lora_C": lora3 if lora3 != "None" else None,
"lora_scale_C": lora_scale3,
"lora_D": lora4 if lora4 != "None" else None,
"lora_scale_D": lora_scale4,
"lora_E": lora5 if lora5 != "None" else None,
"lora_scale_E": lora_scale5,
"lora_F": lora6 if lora6 != "None" else None,
"lora_scale_F": lora_scale6,
"lora_G": lora7 if lora7 != "None" else None,
"lora_scale_G": lora_scale7,
## BEGIN MOD
"textual_inversion": get_embed_list(self.model.class_name) if textual_inversion else [],
## END MOD
"syntax_weights": syntax_weights, # "Classic"
"sampler": sampler,
"schedule_type": schedule_type,
"schedule_prediction_type": schedule_prediction_type,
"xformers_memory_efficient_attention": xformers_memory_efficient_attention,
"gui_active": True,
"loop_generation": loop_generation,
"controlnet_conditioning_scale": float(controlnet_output_scaling_in_unet),
"control_guidance_start": float(controlnet_start_threshold),
"control_guidance_end": float(controlnet_stop_threshold),
"generator_in_cpu": generator_in_cpu,
"FreeU": freeu,
"adetailer_A": adetailer_active_a,
"adetailer_A_params": adetailer_params_A,
"adetailer_B": adetailer_active_b,
"adetailer_B_params": adetailer_params_B,
"leave_progress_bar": leave_progress_bar,
"disable_progress_bar": disable_progress_bar,
"image_previews": image_previews,
"display_images": display_images,
"save_generated_images": save_generated_images,
"filename_pattern": filename_pattern,
"image_storage_location": image_storage_location,
"retain_compel_previous_load": retain_compel_previous_load,
"retain_detailfix_model_previous_load": retain_detailfix_model_previous_load,
"retain_hires_model_previous_load": retain_hires_model_previous_load,
"t2i_adapter_preprocessor": t2i_adapter_preprocessor,
"t2i_adapter_conditioning_scale": float(t2i_adapter_conditioning_scale),
"t2i_adapter_conditioning_factor": float(t2i_adapter_conditioning_factor),
"upscaler_model_path": upscaler_model,
"upscaler_increases_size": upscaler_increases_size,
"upscaler_tile_size": upscaler_tile_size,
"upscaler_tile_overlap": upscaler_tile_overlap,
"hires_steps": hires_steps,
"hires_denoising_strength": hires_denoising_strength,
"hires_prompt": hires_prompt,
"hires_negative_prompt": hires_negative_prompt,
"hires_sampler": hires_sampler,
"hires_before_adetailer": hires_before_adetailer,
"hires_after_adetailer": hires_after_adetailer,
"hires_schedule_type": hires_schedule_type,
"hires_guidance_scale": hires_guidance_scale,
"ip_adapter_image": params_ip_img,
"ip_adapter_mask": params_ip_msk,
"ip_adapter_model": params_ip_model,
"ip_adapter_mode": params_ip_mode,
"ip_adapter_scale": params_ip_scale,
"face_restoration_model": face_restoration_model,
"face_restoration_visibility": face_restoration_visibility,
"face_restoration_weight": face_restoration_weight,
}
# kwargs for diffusers pipeline
if guidance_rescale:
pipe_params["guidance_rescale"] = guidance_rescale
self.model.device = torch.device("cuda:0")
if hasattr(self.model.pipe, "transformer") and loras_list != ["None"] * self.model.num_loras:
self.model.pipe.transformer.to(self.model.device)
print("transformer to cuda")
actual_progress = 0
info_images = gr.update()
for img, [seed, image_path, metadata] in self.model(**pipe_params):
info_state = progress_step_bar(actual_progress, steps)
actual_progress += concurrency
if image_path:
info_images = f"Seeds: {str(seed)}"
if vae_msg:
info_images = info_images + "<br>" + vae_msg
if "Cannot copy out of meta tensor; no data!" in self.model.last_lora_error:
msg_ram = "Unable to process the LoRAs due to high RAM usage; please try again later."
print(msg_ram)
msg_lora += f"<br>{msg_ram}"
for status, lora in zip(self.model.lora_status, self.model.lora_memory):
if status:
msg_lora += f"<br>Loaded: {lora}"
elif status is not None:
msg_lora += f"<br>Error with: {lora}"
if msg_lora:
info_images += msg_lora
info_images = info_images + "<br>" + "GENERATION DATA:<br>" + escape_html(metadata[-1]) + "<br>-------<br>"
download_links = "<br>".join(
[
f'<a href="{path.replace("/images/", "/file=/home/user/app/images/")}" download="{os.path.basename(path)}">Download Image {i + 1}</a>'
for i, path in enumerate(image_path)
]
)
if save_generated_images:
info_images += f"<br>{download_links}"
## BEGIN MOD
img = save_images(img, metadata)
## END MOD
info_state = "COMPLETE"
yield info_state, img, info_images
def dynamic_gpu_duration(func, duration, *args):
@spaces.GPU(duration=duration)
def wrapped_func():
yield from func(*args)
return wrapped_func()
@spaces.GPU
def dummy_gpu():
return None
def sd_gen_generate_pipeline(*args):
gpu_duration_arg = int(args[-1]) if args[-1] else 59
verbose_arg = int(args[-2])
load_lora_cpu = args[-3]
generation_args = args[:-3]
lora_list = [
None if item == "None" or item == "" else item # MOD
for item in [args[7], args[9], args[11], args[13], args[15], args[17], args[19]]
]
lora_status = [None] * sd_gen.model.num_loras
msg_load_lora = "Updating LoRAs in GPU..."
if load_lora_cpu:
msg_load_lora = "Updating LoRAs in CPU..."
if lora_list != sd_gen.model.lora_memory and lora_list != [None] * sd_gen.model.num_loras:
yield msg_load_lora, gr.update(), gr.update()
# Load lora in CPU
if load_lora_cpu:
lora_status = sd_gen.model.load_lora_on_the_fly(
lora_A=lora_list[0], lora_scale_A=args[8],
lora_B=lora_list[1], lora_scale_B=args[10],
lora_C=lora_list[2], lora_scale_C=args[12],
lora_D=lora_list[3], lora_scale_D=args[14],
lora_E=lora_list[4], lora_scale_E=args[16],
lora_F=lora_list[5], lora_scale_F=args[18],
lora_G=lora_list[6], lora_scale_G=args[20],
)
print(lora_status)
sampler_name = args[21]
schedule_type_name = args[22]
_, _, msg_sampler = check_scheduler_compatibility(
sd_gen.model.class_name, sampler_name, schedule_type_name
)
if msg_sampler:
gr.Warning(msg_sampler)
if verbose_arg:
for status, lora in zip(lora_status, lora_list):
if status:
gr.Info(f"LoRA loaded in CPU: {lora}")
elif status is not None:
gr.Warning(f"Failed to load LoRA: {lora}")
if lora_status == [None] * sd_gen.model.num_loras and sd_gen.model.lora_memory != [None] * sd_gen.model.num_loras and load_lora_cpu:
lora_cache_msg = ", ".join(
str(x) for x in sd_gen.model.lora_memory if x is not None
)
gr.Info(f"LoRAs in cache: {lora_cache_msg}")
msg_request = f"Requesting {gpu_duration_arg}s. of GPU time.\nModel: {sd_gen.model.base_model_id}"
if verbose_arg:
gr.Info(msg_request)
print(msg_request)
yield msg_request.replace("\n", "<br>"), gr.update(), gr.update()
start_time = time.time()
# yield from sd_gen.generate_pipeline(*generation_args)
yield from dynamic_gpu_duration(
sd_gen.generate_pipeline,
gpu_duration_arg,
*generation_args,
)
end_time = time.time()
execution_time = end_time - start_time
msg_task_complete = (
f"GPU task complete in: {int(round(execution_time, 0) + 1)} seconds"
)
if verbose_arg:
gr.Info(msg_task_complete)
print(msg_task_complete)
yield msg_task_complete, gr.update(), gr.update()
@spaces.GPU(duration=15)
def process_upscale(image, upscaler_name, upscaler_size):
if image is None: return None
from stablepy.diffusers_vanilla.utils import save_pil_image_with_metadata
from stablepy import load_upscaler_model
image = image.convert("RGB")
exif_image = extract_exif_data(image)
name_upscaler = UPSCALER_DICT_GUI[upscaler_name]
if "https://" in str(name_upscaler):
if not os.path.exists(f"./{DIRECTORY_UPSCALERS}/{name_upscaler.split('/')[-1]}"):
download_things(DIRECTORY_UPSCALERS, name_upscaler, HF_TOKEN)
name_upscaler = f"./{DIRECTORY_UPSCALERS}/{name_upscaler.split('/')[-1]}"
scaler_beta = load_upscaler_model(model=name_upscaler, tile=0, tile_overlap=8, device="cuda", half=True)
image_up = scaler_beta.upscale(image, upscaler_size, True)
image_path = save_pil_image_with_metadata(image_up, f'{os.getcwd()}/up_images', exif_image)
return image_path
# https://huggingface.co/spaces/BestWishYsh/ConsisID-preview-Space/discussions/1#674969a022b99c122af5d407
dynamic_gpu_duration.zerogpu = True
sd_gen_generate_pipeline.zerogpu = True
sd_gen = GuiSD()
## BEGIN MOD
CSS ="""
.gradio-container, #main { width:100%; height:100%; max-width:100%; padding-left:0; padding-right:0; margin-left:0; margin-right:0; }
.contain { display:flex; flex-direction:column; }
#component-0 { width:100%; height:100%; }
#gallery { flex-grow:1; }
#load_model { height: 50px; }
.lora { min-width:480px; }
#model-info { text-align:center; }
.title { font-size: 3em; align-items: center; text-align: center; }
.info { align-items: center; text-align: center; }
.desc [src$='#float'] { float: right; margin: 20px; }
"""
with gr.Blocks(theme='NoCrypt/miku@>=1.2.2', elem_id="main", fill_width=True, css=CSS, delete_cache=(60, 3600)) as app:
gr.Markdown("# 🧩 DiffuseCraft Mod", elem_classes="title")
gr.Markdown("This space is a modification of [r3gm's DiffuseCraft](https://huggingface.co/spaces/r3gm/DiffuseCraft).", elem_classes="info")
with gr.Column():
with gr.Tab("Generation"):
with gr.Row():
with gr.Column(scale=1):
def update_task_options(model_name, task_name):
new_choices = MODEL_TYPE_TASK[get_model_type(valid_model_name(model_name))]
if task_name not in new_choices:
task_name = "txt2img"
return gr.update(value=task_name, choices=new_choices)
interface_mode_gui = gr.Radio(label="Quick settings", choices=["Simple", "Standard", "Fast", "LoRA"], value="Standard")
with gr.Accordion("Model and Task", open=False) as menu_model:
task_gui = gr.Dropdown(label="Task", choices=SDXL_TASK, value=TASK_MODEL_LIST[0])
with gr.Group():
model_name_gui = gr.Dropdown(label="Model", info="You can enter a huggingface model repo_id to want to use.", choices=get_tupled_model_list(model_list), value="votepurchase/animagine-xl-3.1", allow_custom_value=True)
model_info_gui = gr.Markdown(elem_classes="info")
with gr.Row():
quick_model_type_gui = gr.Radio(label="Model Type", choices=["None", "Auto", "Animagine", "Pony"], value="Auto", interactive=True)
quick_genre_gui = gr.Radio(label="Genre", choices=["Anime", "Photo"], value="Anime", interactive=True)
quick_speed_gui = gr.Radio(label="Speed", choices=["Fast", "Standard", "Heavy"], value="Standard", interactive=True)
quick_aspect_gui = gr.Radio(label="Aspect Ratio", choices=["1:1", "3:4"], value="1:1", interactive=True)
with gr.Row():
quality_selector_gui = gr.Dropdown(label="Quality Tags Presets", interactive=True, choices=list(preset_quality.keys()), value="None")
style_selector_gui = gr.Dropdown(label="Style Preset", interactive=True, choices=list(preset_styles.keys()), value="None")
sampler_selector_gui = gr.Dropdown(label="Sampler Quick Settings", interactive=True, choices=list(preset_sampler_setting.keys()), value="None")
optimization_gui = gr.Dropdown(label="Optimization for SDXL", choices=list(optimization_list.keys()), value="None", interactive=True)
with gr.Group():
with gr.Accordion("Prompt from Image", open=False) as menu_from_image:
input_image_gui = gr.Image(label="Input image", type="pil", sources=["upload", "clipboard"], height=256)
with gr.Accordion(label="Advanced options", open=False):
with gr.Row():
general_threshold_gui = gr.Slider(label="Threshold", minimum=0.0, maximum=1.0, value=0.3, step=0.01, interactive=True)
character_threshold_gui = gr.Slider(label="Character threshold", minimum=0.0, maximum=1.0, value=0.8, step=0.01, interactive=True)
with gr.Row():
tag_type_gui = gr.Radio(label="Convert tags to", info="danbooru for Animagine, e621 for Pony.", choices=["danbooru", "e621"], value="danbooru")
recom_prompt_gui = gr.Radio(label="Insert reccomended prompt", choices=["None", "Animagine", "Pony"], value="None", interactive=True)
keep_tags_gui = gr.Radio(label="Remove tags leaving only the following", choices=["body", "dress", "all"], value="all")
image_algorithms = gr.CheckboxGroup(["Use WD Tagger"], label="Algorithms", value=["Use WD Tagger"], visible=False)
generate_from_image_btn_gui = gr.Button(value="GENERATE TAGS FROM IMAGE")
prompt_gui = gr.Textbox(lines=6, placeholder="1girl, solo, ...", label="Prompt", show_copy_button=True)
with gr.Accordion("Negative prompt, etc.", open=False) as menu_negative:
neg_prompt_gui = gr.Textbox(lines=3, placeholder="Enter Neg prompt", label="Negative prompt", value="lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, worst quality, low quality, very displeasing, (bad)", show_copy_button=True)
translate_prompt_button = gr.Button(value="Translate prompt to English", size="sm", variant="secondary")
with gr.Row():
insert_prompt_gui = gr.Radio(label="Insert reccomended positive / negative prompt", choices=["None", "Auto", "Animagine", "Pony"], value="Auto", interactive=True)
prompt_type_gui = gr.Radio(label="Convert tags to", choices=["danbooru", "e621"], value="e621", visible=False)
prompt_type_button = gr.Button(value="Convert prompt to Pony e621 style", size="sm", variant="secondary")
with gr.Row():
character_dbt = gr.Textbox(lines=1, placeholder="kafuu chino, ...", label="Character names")
series_dbt = gr.Textbox(lines=1, placeholder="Is the order a rabbit?, ...", label="Series names")
random_character_gui = gr.Button(value="Random character 🎲", size="sm", variant="secondary")
model_name_dbt = gr.Dropdown(label="Model", choices=list(V2_ALL_MODELS.keys()), value=list(V2_ALL_MODELS.keys())[0], visible=False)
aspect_ratio_dbt = gr.Radio(label="Aspect ratio", choices=list(V2_ASPECT_RATIO_OPTIONS), value="square", visible=False)
length_dbt = gr.Radio(label="Length", choices=list(V2_LENGTH_OPTIONS), value="very_long", visible=False)
identity_dbt = gr.Radio(label="Keep identity", choices=list(V2_IDENTITY_OPTIONS), value="lax", visible=False)
ban_tags_dbt = gr.Textbox(label="Ban tags", placeholder="alternate costumen, ...", value="futanari, censored, furry, furrification", visible=False)
copy_button_dbt = gr.Button(value="Copy to clipboard", visible=False)
rating_dbt = gr.Radio(label="Rating", choices=list(V2_RATING_OPTIONS), value="sfw")
generate_db_random_button = gr.Button(value="EXTEND PROMPT 🎲")
with gr.Row():
translate_prompt_gui = gr.Button(value="Translate Prompt 📝", variant="secondary", size="sm")
set_random_seed = gr.Button(value="Seed 🎲", variant="secondary", size="sm")
set_params_gui = gr.Button(value="Params ↙️", variant="secondary", size="sm")
clear_prompt_gui = gr.Button(value="Clear 🗑️", variant="secondary", size="sm")
generate_button = gr.Button(value="GENERATE IMAGE", size="lg", variant="primary")
model_name_gui.change(
update_task_options,
[model_name_gui, task_gui],
[task_gui],
)
load_model_gui = gr.HTML(elem_id="load_model", elem_classes="contain")
result_images = gr.Gallery(
label="Generated images",
show_label=False,
elem_id="gallery",
#columns=[2],
columns=[1],
#rows=[2],
rows=[1],
object_fit="contain",
# height="auto",
interactive=False,
preview=False,
show_share_button=False,
show_download_button=True,
selected_index=50,
format="png",
)
result_images_files = gr.Files(interactive=False, visible=False)
actual_task_info = gr.HTML()
with gr.Accordion("History", open=False):
history_files = gr.Files(interactive=False, visible=False)
history_gallery = gr.Gallery(label="History", columns=6, object_fit="contain", format="png", interactive=False, show_share_button=False, show_download_button=True)
history_clear_button = gr.Button(value="Clear History", variant="secondary")
history_clear_button.click(lambda: ([], []), None, [history_gallery, history_files], queue=False, show_api=False)
with gr.Row(equal_height=False, variant="default"):
gpu_duration_gui = gr.Number(minimum=5, maximum=240, value=59, show_label=False, container=False, info="GPU time duration (seconds)")
with gr.Column():
verbose_info_gui = gr.Checkbox(value=False, container=False, label="Status info")
load_lora_cpu_gui = gr.Checkbox(value=False, container=False, label="Load LoRAs on CPU")
with gr.Column(scale=1):
with gr.Accordion("Generation settings", open=False, visible=True) as menu_gen:
with gr.Row():
img_width_gui = gr.Slider(minimum=64, maximum=4096, step=8, value=1024, label="Img Width")
img_height_gui = gr.Slider(minimum=64, maximum=4096, step=8, value=1024, label="Img Height")
steps_gui = gr.Slider(minimum=1, maximum=100, step=1, value=28, label="Steps")
cfg_gui = gr.Slider(minimum=0, maximum=30, step=0.5, value=7.0, label="CFG")
guidance_rescale_gui = gr.Slider(label="CFG rescale:", value=0., step=0.01, minimum=0., maximum=1.5)
with gr.Row():
seed_gui = gr.Number(minimum=-1, maximum=2**32-1, value=-1, label="Seed")
pag_scale_gui = gr.Slider(minimum=0.0, maximum=10.0, step=0.1, value=0.0, label="PAG Scale")
num_images_gui = gr.Slider(minimum=1, maximum=5, step=1, value=1, label="Images")
clip_skip_gui = gr.Checkbox(value=False, label="Layer 2 Clip Skip")
free_u_gui = gr.Checkbox(value=False, label="FreeU")
with gr.Row():
sampler_gui = gr.Dropdown(label="Sampler", choices=scheduler_names, value="Euler")
schedule_type_gui = gr.Dropdown(label="Schedule type", choices=SCHEDULE_TYPE_OPTIONS, value=SCHEDULE_TYPE_OPTIONS[0])
schedule_prediction_type_gui = gr.Dropdown(label="Discrete Sampling Type", choices=SCHEDULE_PREDICTION_TYPE_OPTIONS, value=SCHEDULE_PREDICTION_TYPE_OPTIONS[0])
vae_model_gui = gr.Dropdown(label="VAE Model", choices=vae_model_list, value=vae_model_list[0])
prompt_syntax_gui = gr.Dropdown(label="Prompt Syntax", choices=PROMPT_W_OPTIONS, value=PROMPT_W_OPTIONS[1][1])
with gr.Row(equal_height=False):
def run_set_params_gui(base_prompt, name_model):
valid_receptors = { # default values
"prompt": gr.update(value=base_prompt),
"neg_prompt": gr.update(value=""),
"Steps": gr.update(value=30),
"width": gr.update(value=1024),
"height": gr.update(value=1024),
"Seed": gr.update(value=-1),
"Sampler": gr.update(value="Euler"),
"CFG scale": gr.update(value=7.), # cfg
"Clip skip": gr.update(value=True),
"Model": gr.update(value=name_model),
"Schedule type": gr.update(value="Automatic"),
"PAG": gr.update(value=.0),
"FreeU": gr.update(value=False),
}
valid_keys = list(valid_receptors.keys())
parameters = extract_parameters(base_prompt)
# print(parameters)
if "Sampler" in parameters:
value_sampler = parameters["Sampler"]
for s_type in SCHEDULE_TYPE_OPTIONS:
if s_type in value_sampler:
value_sampler = value_sampler.replace(s_type, "").strip()
parameters["Sampler"] = value_sampler
parameters["Schedule type"] = s_type
for key, val in parameters.items():
# print(val)
if key in valid_keys:
try:
if key == "Sampler":
if val not in scheduler_names:
continue
if key == "Schedule type":
if val not in SCHEDULE_TYPE_OPTIONS:
val = "Automatic"
elif key == "Clip skip":
if "," in str(val):
val = val.replace(",", "")
if int(val) >= 2:
val = True
if key == "prompt":
if ">" in val and "<" in val:
val = re.sub(r'<[^>]+>', '', val)
print("Removed LoRA written in the prompt")
if key in ["prompt", "neg_prompt"]:
val = re.sub(r'\s+', ' ', re.sub(r',+', ',', val)).strip()
if key in ["Steps", "width", "height", "Seed"]:
val = int(val)
if key == "FreeU":
val = True
if key in ["CFG scale", "PAG"]:
val = float(val)
if key == "Model":
filtered_models = [m for m in model_list if val in m]
if filtered_models:
val = filtered_models[0]
else:
val = name_model
if key == "Seed":
continue
valid_receptors[key] = gr.update(value=val)
# print(val, type(val))
# print(valid_receptors)
except Exception as e:
print(str(e))
return [value for value in valid_receptors.values()]
set_params_gui.click(
run_set_params_gui, [prompt_gui, model_name_gui], [
prompt_gui,
neg_prompt_gui,
steps_gui,
img_width_gui,
img_height_gui,
seed_gui,
sampler_gui,
cfg_gui,
clip_skip_gui,
model_name_gui,
schedule_type_gui,
pag_scale_gui,
free_u_gui,
],
)
def run_clear_prompt_gui():
return gr.update(value=""), gr.update(value="")
clear_prompt_gui.click(
run_clear_prompt_gui, [], [prompt_gui, neg_prompt_gui]
)
def run_set_random_seed():
return -1
set_random_seed.click(
run_set_random_seed, [], seed_gui
)
with gr.Accordion("LoRA", open=False, visible=True) as menu_lora:
def lora_dropdown(label, visible=True):
return gr.Dropdown(label=label, choices=get_all_lora_tupled_list(), value="", allow_custom_value=True, elem_classes="lora", min_width=320, visible=visible)
def lora_scale_slider(label, visible=True):
return gr.Slider(minimum=-2, maximum=2, step=0.01, value=1.00, label=label, visible=visible)
def lora_textbox(label):
return gr.Textbox(label=label, info="Example of prompt:", value="None", show_copy_button=True, interactive=False, visible=False)
with gr.Row():
with gr.Column():
lora1_gui = lora_dropdown("LoRA1")
lora_scale_1_gui = lora_scale_slider("LoRA Scale 1")
with gr.Row():
with gr.Group():
lora1_info_gui = lora_textbox("LoRA1 prompts")
lora1_copy_gui = gr.Button(value="Copy example to prompt", visible=False)
lora1_desc_gui = gr.Markdown(value="", visible=False)
with gr.Column():
lora2_gui = lora_dropdown("LoRA2")
lora_scale_2_gui = lora_scale_slider("LoRA Scale 2")
with gr.Row():
with gr.Group():
lora2_info_gui = lora_textbox("LoRA2 prompts")
lora2_copy_gui = gr.Button(value="Copy example to prompt", visible=False)
lora2_desc_gui = gr.Markdown(value="", visible=False)
with gr.Column():
lora3_gui = lora_dropdown("LoRA3")
lora_scale_3_gui = lora_scale_slider("LoRA Scale 3")
with gr.Row():
with gr.Group():
lora3_info_gui = lora_textbox("LoRA3 prompts")
lora3_copy_gui = gr.Button(value="Copy example to prompt", visible=False)
lora3_desc_gui = gr.Markdown(value="", visible=False)
with gr.Column():
lora4_gui = lora_dropdown("LoRA4")
lora_scale_4_gui = lora_scale_slider("LoRA Scale 4")
with gr.Row():
with gr.Group():
lora4_info_gui = lora_textbox("LoRA4 prompts")
lora4_copy_gui = gr.Button(value="Copy example to prompt", visible=False)
lora4_desc_gui = gr.Markdown(value="", visible=False)
with gr.Column():
lora5_gui = lora_dropdown("LoRA5")
lora_scale_5_gui = lora_scale_slider("LoRA Scale 5")
with gr.Row():
with gr.Group():
lora5_info_gui = lora_textbox("LoRA5 prompts")
lora5_copy_gui = gr.Button(value="Copy example to prompt", visible=False)
lora5_desc_gui = gr.Markdown(value="", visible=False)
with gr.Column():
lora6_gui = lora_dropdown("LoRA6", visible=False)
lora_scale_6_gui = lora_scale_slider("LoRA Scale 6", visible=False)
with gr.Row():
with gr.Group():
lora6_info_gui = lora_textbox("LoRA6 prompts")
lora6_copy_gui = gr.Button(value="Copy example to prompt", visible=False)
lora6_desc_gui = gr.Markdown(value="", visible=False)
with gr.Column():
lora7_gui = lora_dropdown("LoRA7", visible=False)
lora_scale_7_gui = lora_scale_slider("LoRA Scale 7", visible=False)
with gr.Row():
with gr.Group():
lora7_info_gui = lora_textbox("LoRA7 prompts")
lora7_copy_gui = gr.Button(value="Copy example to prompt", visible=False)
lora7_desc_gui = gr.Markdown(value="", visible=False)
with gr.Accordion("From URL", open=True, visible=True):
with gr.Row():
search_civitai_basemodel_lora = gr.CheckboxGroup(label="Search LoRA for", choices=CIVITAI_BASEMODEL, value=["Pony", "Illustrious", "SDXL 1.0"])
search_civitai_sort_lora = gr.Radio(label="Sort", choices=CIVITAI_SORT, value="Highest Rated")
search_civitai_period_lora = gr.Radio(label="Period", choices=CIVITAI_PERIOD, value="AllTime")
with gr.Row():
search_civitai_query_lora = gr.Textbox(label="Query", placeholder="oomuro sakurako...", lines=1)
search_civitai_tag_lora = gr.Dropdown(label="Tag", choices=get_civitai_tag(), value=get_civitai_tag()[0], allow_custom_value=True)
search_civitai_user_lora = gr.Textbox(label="Username", lines=1)
search_civitai_button_lora = gr.Button("Search on Civitai")
search_civitai_desc_lora = gr.Markdown(value="", visible=False, elem_classes="desc")
with gr.Accordion("Select from Gallery", open=False):
search_civitai_gallery_lora = gr.Gallery([], label="Results", allow_preview=False, columns=5, show_share_button=False, interactive=False)
search_civitai_result_lora = gr.Dropdown(label="Search Results", choices=[("", "")], value="", allow_custom_value=True, visible=False)
with gr.Row():
text_lora = gr.Textbox(label="LoRA's download URL", placeholder="https://civitai.com/api/download/models/28907", info="It has to be .safetensors files, and you can also download them from Hugging Face.", lines=1, scale=4)
romanize_text = gr.Checkbox(value=False, label="Transliterate name", scale=1, visible=False)
button_lora = gr.Button("Get and Refresh the LoRA Lists")
new_lora_status = gr.HTML()
with gr.Accordion("From Local", open=True, visible=True):
file_output_lora = gr.File(label="Uploaded LoRA", file_types=['.ckpt', '.pt', '.pth', '.safetensors', '.bin'], file_count="multiple", interactive=False, visible=False)
upload_button_lora = gr.UploadButton(label="Upload LoRA from your disk (very slow)", file_types=['.ckpt', '.pt', '.pth', '.safetensors', '.bin'], file_count="multiple")
with gr.Column() as menu_advanced:
with gr.Accordion("Hires fix", open=False, visible=True) as menu_hires:
upscaler_model_path_gui = gr.Dropdown(label="Upscaler", choices=UPSCALER_KEYS, value=UPSCALER_KEYS[0])
with gr.Row():
upscaler_increases_size_gui = gr.Slider(minimum=1.1, maximum=6., step=0.1, value=1.0, label="Upscale by")
upscaler_tile_size_gui = gr.Slider(minimum=0, maximum=512, step=16, value=0, label="Upscaler Tile Size", info="0 = no tiling")
upscaler_tile_overlap_gui = gr.Slider(minimum=0, maximum=48, step=1, value=8, label="Upscaler Tile Overlap")
with gr.Row():
hires_steps_gui = gr.Slider(minimum=0, value=30, maximum=100, step=1, label="Hires Steps")
hires_denoising_strength_gui = gr.Slider(minimum=0.1, maximum=1.0, step=0.01, value=0.55, label="Hires Denoising Strength")
hires_sampler_gui = gr.Dropdown(label="Hires Sampler", choices=POST_PROCESSING_SAMPLER, value=POST_PROCESSING_SAMPLER[0])
hires_schedule_list = ["Use same schedule type"] + SCHEDULE_TYPE_OPTIONS
hires_schedule_type_gui = gr.Dropdown(label="Hires Schedule type", choices=hires_schedule_list, value=hires_schedule_list[0])
hires_guidance_scale_gui = gr.Slider(minimum=-1., maximum=30., step=0.5, value=-1., label="Hires CFG", info="If the value is -1, the main CFG will be used")
hires_prompt_gui = gr.Textbox(label="Hires Prompt", placeholder="Main prompt will be use", lines=3)
hires_negative_prompt_gui = gr.Textbox(label="Hires Negative Prompt", placeholder="Main negative prompt will be use", lines=3)
with gr.Accordion("Detailfix", open=False, visible=True) as menu_detail:
with gr.Row():
# Adetailer Inpaint Only
adetailer_inpaint_only_gui = gr.Checkbox(label="Inpaint only", value=True)
# Adetailer Verbose
adetailer_verbose_gui = gr.Checkbox(label="Verbose", value=False)
# Adetailer Sampler
adetailer_sampler_gui = gr.Dropdown(label="Adetailer sampler:", choices=POST_PROCESSING_SAMPLER, value=POST_PROCESSING_SAMPLER[0])
with gr.Accordion("Detailfix A", open=True, visible=True):
# Adetailer A
adetailer_active_a_gui = gr.Checkbox(label="Enable Adetailer A", value=False)
prompt_ad_a_gui = gr.Textbox(label="Main prompt", placeholder="Main prompt will be use", lines=3)
negative_prompt_ad_a_gui = gr.Textbox(label="Negative prompt", placeholder="Main negative prompt will be use", lines=3)
with gr.Row():
strength_ad_a_gui = gr.Number(label="Strength:", value=0.35, step=0.01, minimum=0.01, maximum=1.0)
face_detector_ad_a_gui = gr.Checkbox(label="Face detector", value=False)
person_detector_ad_a_gui = gr.Checkbox(label="Person detector", value=True)
hand_detector_ad_a_gui = gr.Checkbox(label="Hand detector", value=False)
with gr.Row():
mask_dilation_a_gui = gr.Number(label="Mask dilation:", value=4, minimum=1)
mask_blur_a_gui = gr.Number(label="Mask blur:", value=4, minimum=1)
mask_padding_a_gui = gr.Number(label="Mask padding:", value=32, minimum=1)
with gr.Accordion("Detailfix B", open=True, visible=True):
# Adetailer B
adetailer_active_b_gui = gr.Checkbox(label="Enable Adetailer B", value=False)
prompt_ad_b_gui = gr.Textbox(label="Main prompt", placeholder="Main prompt will be use", lines=3)
negative_prompt_ad_b_gui = gr.Textbox(label="Negative prompt", placeholder="Main negative prompt will be use", lines=3)
with gr.Row():
strength_ad_b_gui = gr.Number(label="Strength:", value=0.35, step=0.01, minimum=0.01, maximum=1.0)
face_detector_ad_b_gui = gr.Checkbox(label="Face detector", value=False)
person_detector_ad_b_gui = gr.Checkbox(label="Person detector", value=True)
hand_detector_ad_b_gui = gr.Checkbox(label="Hand detector", value=False)
with gr.Row():
mask_dilation_b_gui = gr.Number(label="Mask dilation:", value=4, minimum=1)
mask_blur_b_gui = gr.Number(label="Mask blur:", value=4, minimum=1)
mask_padding_b_gui = gr.Number(label="Mask padding:", value=32, minimum=1)
with gr.Accordion("Face restoration", open=False, visible=True):
face_rest_options = [None] + FACE_RESTORATION_MODELS
face_restoration_model_gui = gr.Dropdown(label="Face restoration model", choices=face_rest_options, value=face_rest_options[0])
with gr.Row():
face_restoration_visibility_gui = gr.Slider(minimum=0., maximum=1., step=0.001, value=1., label="Visibility")
face_restoration_weight_gui = gr.Slider(minimum=0., maximum=1., step=0.001, value=.5, label="Weight", info="(0 = maximum effect, 1 = minimum effect)")
with gr.Accordion("Textual inversion", open=False, visible=True) as menu_ti:
active_textual_inversion_gui = gr.Checkbox(value=False, label="Active Textual Inversion in prompt")
use_textual_inversion_gui = gr.CheckboxGroup(choices=get_embed_list(get_model_pipeline(model_name_gui.value)) if active_textual_inversion_gui.value else [], value=None, label="Use Textual Invertion in prompt")
def update_textual_inversion_gui(active_textual_inversion_gui, model_name_gui):
return gr.update(choices=get_embed_list(get_model_pipeline(model_name_gui)) if active_textual_inversion_gui else [])
active_textual_inversion_gui.change(update_textual_inversion_gui, [active_textual_inversion_gui, model_name_gui], [use_textual_inversion_gui])
model_name_gui.change(update_textual_inversion_gui, [active_textual_inversion_gui, model_name_gui], [use_textual_inversion_gui])
with gr.Accordion("ControlNet / Img2img / Inpaint", open=False, visible=True) as menu_i2i:
with gr.Row():
image_control = gr.Image(label="Image ControlNet/Inpaint/Img2img", type="filepath")
image_mask_gui = gr.Image(label="Image Mask", type="filepath")
with gr.Row():
strength_gui = gr.Slider(
minimum=0.01, maximum=1.0, step=0.01, value=0.55, label="Strength",
info="This option adjusts the level of changes for img2img, repaint and inpaint."
)
image_resolution_gui = gr.Slider(
minimum=64, maximum=2048, step=64, value=1024, label="Image Resolution",
info="The maximum proportional size of the generated image based on the uploaded image."
)
with gr.Row():
controlnet_model_gui = gr.Dropdown(label="ControlNet model", choices=DIFFUSERS_CONTROLNET_MODEL, value=DIFFUSERS_CONTROLNET_MODEL[0], allow_custom_value=True)
control_net_output_scaling_gui = gr.Slider(minimum=0, maximum=5.0, step=0.1, value=1, label="ControlNet Output Scaling in UNet")
control_net_start_threshold_gui = gr.Slider(minimum=0, maximum=1, step=0.01, value=0, label="ControlNet Start Threshold (%)")
control_net_stop_threshold_gui = gr.Slider(minimum=0, maximum=1, step=0.01, value=1, label="ControlNet Stop Threshold (%)")
with gr.Row():
preprocessor_name_gui = gr.Dropdown(label="Preprocessor Name", choices=TASK_AND_PREPROCESSORS["canny"])
preprocess_resolution_gui = gr.Slider(minimum=64, maximum=2048, step=64, value=512, label="Preprocessor Resolution")
low_threshold_gui = gr.Slider(minimum=1, maximum=255, step=1, value=100, label="'CANNY' low threshold")
high_threshold_gui = gr.Slider(minimum=1, maximum=255, step=1, value=200, label="'CANNY' high threshold")
def change_preprocessor_choices(task):
task = TASK_STABLEPY[task]
if task in TASK_AND_PREPROCESSORS.keys():
choices_task = TASK_AND_PREPROCESSORS[task]
else:
choices_task = TASK_AND_PREPROCESSORS["canny"]
return gr.update(choices=choices_task, value=choices_task[0])
task_gui.change(
change_preprocessor_choices,
[task_gui],
[preprocessor_name_gui],
)
with gr.Row():
value_threshold_gui = gr.Slider(minimum=1, maximum=2.0, step=0.01, value=0.1, label="'MLSD' Hough value threshold")
distance_threshold_gui = gr.Slider(minimum=1, maximum=20.0, step=0.01, value=0.1, label="'MLSD' Hough distance threshold")
recolor_gamma_correction_gui = gr.Number(minimum=0., maximum=25., value=1., step=0.001, label="'RECOLOR' gamma correction")
tile_blur_sigma_gui = gr.Number(minimum=0, maximum=100, value=9, step=1, label="'TILE' blur sigma")
with gr.Accordion("IP-Adapter", open=False, visible=True) as menu_ipa:
with gr.Accordion("IP-Adapter 1", open=True, visible=True):
with gr.Row():
image_ip1 = gr.Image(label="IP Image", type="filepath")
mask_ip1 = gr.Image(label="IP Mask", type="filepath")
with gr.Row():
model_ip1 = gr.Dropdown(value="plus_face", label="Model", choices=IP_MODELS)
mode_ip1 = gr.Dropdown(value="original", label="Mode", choices=MODE_IP_OPTIONS)
scale_ip1 = gr.Slider(minimum=0., maximum=2., step=0.01, value=0.7, label="Scale")
with gr.Accordion("IP-Adapter 2", open=True, visible=True):
with gr.Row():
image_ip2 = gr.Image(label="IP Image", type="filepath")
mask_ip2 = gr.Image(label="IP Mask (optional)", type="filepath")
with gr.Row():
model_ip2 = gr.Dropdown(value="base", label="Model", choices=IP_MODELS)
mode_ip2 = gr.Dropdown(value="style", label="Mode", choices=MODE_IP_OPTIONS)
scale_ip2 = gr.Slider(minimum=0., maximum=2., step=0.01, value=0.7, label="Scale")
with gr.Accordion("T2I adapter", open=False, visible=False) as menu_t2i:
t2i_adapter_preprocessor_gui = gr.Checkbox(value=True, label="T2i Adapter Preprocessor")
with gr.Row():
adapter_conditioning_scale_gui = gr.Slider(minimum=0, maximum=5., step=0.1, value=1, label="Adapter Conditioning Scale")
adapter_conditioning_factor_gui = gr.Slider(minimum=0, maximum=1., step=0.01, value=0.55, label="Adapter Conditioning Factor (%)")
with gr.Accordion("Styles", open=False, visible=True) as menu_styles:
try:
style_names_found = sd_gen.model.STYLE_NAMES
except Exception:
style_names_found = STYLE_NAMES
style_prompt_gui = gr.Dropdown(
style_names_found,
multiselect=True,
value=None,
label="Style Prompt",
interactive=True,
)
style_json_gui = gr.File(label="Style JSON File")
style_button = gr.Button("Load styles")
def load_json_style_file(json):
if not sd_gen.model:
gr.Info("First load the model")
return gr.update(value=None, choices=STYLE_NAMES)
sd_gen.model.load_style_file(json)
gr.Info(f"{len(sd_gen.model.STYLE_NAMES)} styles loaded")
return gr.update(value=None, choices=sd_gen.model.STYLE_NAMES)
style_button.click(load_json_style_file, [style_json_gui], [style_prompt_gui])
with gr.Accordion("Other settings", open=False, visible=True) as menu_other:
with gr.Row():
save_generated_images_gui = gr.Checkbox(value=False, label="Save Generated Images")
filename_pattern_gui = gr.Textbox(label="Filename pattern", value="model,seed", placeholder="model,seed,sampler,schedule_type,img_width,img_height,guidance_scale,num_steps,vae,prompt_section,neg_prompt_section", lines=1)
with gr.Row():
hires_before_adetailer_gui = gr.Checkbox(value=False, label="Hires Before Adetailer")
hires_after_adetailer_gui = gr.Checkbox(value=True, label="Hires After Adetailer")
generator_in_cpu_gui = gr.Checkbox(value=False, label="Generator in CPU")
with gr.Accordion("More settings", open=False, visible=False):
loop_generation_gui = gr.Slider(minimum=1, value=1, label="Loop Generation")
retain_task_cache_gui = gr.Checkbox(value=True, label="Retain task model in cache")
leave_progress_bar_gui = gr.Checkbox(value=True, label="Leave Progress Bar")
disable_progress_bar_gui = gr.Checkbox(value=False, label="Disable Progress Bar")
display_images_gui = gr.Checkbox(value=False, label="Display Images")
image_previews_gui = gr.Checkbox(value=True, label="Image Previews")
image_storage_location_gui = gr.Textbox(value="./images", label="Image Storage Location")
retain_compel_previous_load_gui = gr.Checkbox(value=False, label="Retain Compel Previous Load")
retain_detailfix_model_previous_load_gui = gr.Checkbox(value=False, label="Retain Detailfix Model Previous Load")
retain_hires_model_previous_load_gui = gr.Checkbox(value=False, label="Retain Hires Model Previous Load")
xformers_memory_efficient_attention_gui = gr.Checkbox(value=False, label="Xformers Memory Efficient Attention")
with gr.Accordion("Examples and help", open=True, visible=True) as menu_example:
gr.Examples(
examples=EXAMPLES_GUI,
fn=sd_gen.generate_pipeline,
inputs=[
prompt_gui,
neg_prompt_gui,
num_images_gui,
steps_gui,
cfg_gui,
clip_skip_gui,
seed_gui,
sampler_gui,
img_height_gui,
img_width_gui,
model_name_gui,
],
outputs=[load_model_gui, result_images, actual_task_info],
cache_examples=False,
#elem_id="examples",
)
gr.Markdown(RESOURCES)
## END MOD
with gr.Tab("Inpaint mask maker", render=True):
def create_mask_now(img, invert):
import numpy as np
import time
time.sleep(0.5)
transparent_image = img["layers"][0]
# Extract the alpha channel
alpha_channel = np.array(transparent_image)[:, :, 3]
# Create a binary mask by thresholding the alpha channel
binary_mask = alpha_channel > 1
if invert:
print("Invert")
# Invert the binary mask so that the drawn shape is white and the rest is black
binary_mask = np.invert(binary_mask)
# Convert the binary mask to a 3-channel RGB mask
rgb_mask = np.stack((binary_mask,) * 3, axis=-1)
# Convert the mask to uint8
rgb_mask = rgb_mask.astype(np.uint8) * 255
return img["background"], rgb_mask
with gr.Row():
with gr.Column(scale=2):
# image_base = gr.ImageEditor(label="Base image", show_label=True, brush=gr.Brush(colors=["#000000"]))
image_base = gr.ImageEditor(
sources=["upload", "clipboard"],
# crop_size="1:1",
# enable crop (or disable it)
# transforms=["crop"],
brush=gr.Brush(
default_size="16", # or leave it as 'auto'
color_mode="fixed", # 'fixed' hides the user swatches and colorpicker, 'defaults' shows it
# default_color="black", # html names are supported
colors=[
"rgba(0, 0, 0, 1)", # rgb(a)
"rgba(0, 0, 0, 0.1)",
"rgba(255, 255, 255, 0.1)",
# "hsl(360, 120, 120)" # in fact any valid colorstring
]
),
eraser=gr.Eraser(default_size="16")
)
invert_mask = gr.Checkbox(value=False, label="Invert mask")
btn = gr.Button("Create mask")
with gr.Column(scale=1):
img_source = gr.Image(interactive=False)
img_result = gr.Image(label="Mask image", show_label=True, interactive=False)
btn_send = gr.Button("Send to the first tab")
btn.click(create_mask_now, [image_base, invert_mask], [img_source, img_result])
def send_img(img_source, img_result):
return img_source, img_result
btn_send.click(send_img, [img_source, img_result], [image_control, image_mask_gui])
with gr.Tab("PNG Info"):
with gr.Row():
with gr.Column():
image_metadata = gr.Image(label="Image with metadata", type="pil", sources=["upload"])
with gr.Column():
result_metadata = gr.Textbox(label="Metadata", show_label=True, show_copy_button=True, interactive=False, container=True, max_lines=99)
image_metadata.change(
fn=extract_exif_data,
inputs=[image_metadata],
outputs=[result_metadata],
)
with gr.Tab("Upscaler"):
with gr.Row():
with gr.Column():
USCALER_TAB_KEYS = [name for name in UPSCALER_KEYS[9:]]
image_up_tab = gr.Image(label="Image", type="pil", sources=["upload"])
upscaler_tab = gr.Dropdown(label="Upscaler", choices=USCALER_TAB_KEYS, value=USCALER_TAB_KEYS[5])
upscaler_size_tab = gr.Slider(minimum=1., maximum=4., step=0.1, value=1.1, label="Upscale by")
generate_button_up_tab = gr.Button(value="START UPSCALE", variant="primary")
with gr.Column():
result_up_tab = gr.Image(label="Result", type="pil", interactive=False, format="png")
generate_button_up_tab.click(
fn=process_upscale,
inputs=[image_up_tab, upscaler_tab, upscaler_size_tab],
outputs=[result_up_tab],
)
with gr.Tab("Preprocessor", render=True):
preprocessor_tab()
## BEGIN MOD
interface_mode_gui.change(
change_interface_mode,
[interface_mode_gui],
[menu_model, menu_from_image, menu_negative, menu_gen, menu_hires, menu_lora, menu_advanced,
menu_example, task_gui, quick_speed_gui],
queue=False,
)
model_name_gui.change(get_t2i_model_info, [model_name_gui], [model_info_gui], queue=False)
translate_prompt_gui.click(translate_to_en, [prompt_gui], [prompt_gui], queue=False)\
.then(translate_to_en, [neg_prompt_gui], [neg_prompt_gui], queue=False)
gr.on(
triggers=[quick_model_type_gui.change, quick_genre_gui.change, quick_speed_gui.change, quick_aspect_gui.change],
fn=set_quick_presets,
inputs=[quick_genre_gui, quick_model_type_gui, quick_speed_gui, quick_aspect_gui],
outputs=[quality_selector_gui, style_selector_gui, sampler_selector_gui, optimization_gui, insert_prompt_gui],
queue=False,
trigger_mode="once",
)
gr.on(
triggers=[quality_selector_gui.change, style_selector_gui.change, insert_prompt_gui.change],
fn=process_style_prompt,
inputs=[prompt_gui, neg_prompt_gui, style_selector_gui, quality_selector_gui, insert_prompt_gui],
outputs=[prompt_gui, neg_prompt_gui, quick_model_type_gui],
queue=False,
trigger_mode="once",
)
sampler_selector_gui.change(set_sampler_settings, [sampler_selector_gui], [sampler_gui, steps_gui, cfg_gui, clip_skip_gui, img_width_gui, img_height_gui, optimization_gui], queue=False)
optimization_gui.change(set_optimization, [optimization_gui, steps_gui, cfg_gui, sampler_gui, clip_skip_gui, lora5_gui, lora_scale_5_gui], [steps_gui, cfg_gui, sampler_gui, clip_skip_gui, lora5_gui, lora_scale_5_gui], queue=False)
gr.on(
triggers=[lora1_gui.change, lora_scale_1_gui.change, lora2_gui.change, lora_scale_2_gui.change,
lora3_gui.change, lora_scale_3_gui.change, lora4_gui.change, lora_scale_4_gui.change,
lora5_gui.change, lora_scale_5_gui.change, lora6_gui.change, lora_scale_6_gui.change,
lora7_gui.change, lora_scale_7_gui.change, prompt_syntax_gui.change],
fn=update_loras,
inputs=[prompt_gui, prompt_syntax_gui, lora1_gui, lora_scale_1_gui, lora2_gui, lora_scale_2_gui,
lora3_gui, lora_scale_3_gui, lora4_gui, lora_scale_4_gui, lora5_gui, lora_scale_5_gui,
lora6_gui, lora_scale_6_gui, lora7_gui, lora_scale_7_gui],
outputs=[prompt_gui, lora1_gui, lora_scale_1_gui, lora1_info_gui, lora1_copy_gui, lora1_desc_gui,
lora2_gui, lora_scale_2_gui, lora2_info_gui, lora2_copy_gui, lora2_desc_gui,
lora3_gui, lora_scale_3_gui, lora3_info_gui, lora3_copy_gui, lora3_desc_gui,
lora4_gui, lora_scale_4_gui, lora4_info_gui, lora4_copy_gui, lora4_desc_gui,
lora5_gui, lora_scale_5_gui, lora5_info_gui, lora5_copy_gui, lora5_desc_gui,
lora6_gui, lora_scale_6_gui, lora6_info_gui, lora6_copy_gui, lora6_desc_gui,
lora7_gui, lora_scale_7_gui, lora7_info_gui, lora7_copy_gui, lora7_desc_gui],
queue=False,
trigger_mode="once",
)
lora1_copy_gui.click(apply_lora_prompt, [prompt_gui, lora1_info_gui], [prompt_gui], queue=False)
lora2_copy_gui.click(apply_lora_prompt, [prompt_gui, lora2_info_gui], [prompt_gui], queue=False)
lora3_copy_gui.click(apply_lora_prompt, [prompt_gui, lora3_info_gui], [prompt_gui], queue=False)
lora4_copy_gui.click(apply_lora_prompt, [prompt_gui, lora4_info_gui], [prompt_gui], queue=False)
lora5_copy_gui.click(apply_lora_prompt, [prompt_gui, lora5_info_gui], [prompt_gui], queue=False)
lora6_copy_gui.click(apply_lora_prompt, [prompt_gui, lora6_info_gui], [prompt_gui], queue=False)
lora7_copy_gui.click(apply_lora_prompt, [prompt_gui, lora7_info_gui], [prompt_gui], queue=False)
gr.on(
triggers=[search_civitai_button_lora.click, search_civitai_query_lora.submit],
fn=search_civitai_lora,
inputs=[search_civitai_query_lora, search_civitai_basemodel_lora, search_civitai_sort_lora, search_civitai_period_lora,
search_civitai_tag_lora, search_civitai_user_lora, search_civitai_gallery_lora],
outputs=[search_civitai_result_lora, search_civitai_desc_lora, search_civitai_button_lora, search_civitai_query_lora, search_civitai_gallery_lora],
queue=True,
scroll_to_output=True,
)
search_civitai_result_lora.change(select_civitai_lora, [search_civitai_result_lora], [text_lora, search_civitai_desc_lora], queue=False, scroll_to_output=True)
search_civitai_gallery_lora.select(update_civitai_selection, None, [search_civitai_result_lora], queue=False, show_api=False)
button_lora.click(get_my_lora, [text_lora, romanize_text], [lora1_gui, lora2_gui, lora3_gui, lora4_gui, lora5_gui, lora6_gui, lora7_gui, new_lora_status], scroll_to_output=True)
upload_button_lora.upload(upload_file_lora, [upload_button_lora], [file_output_lora, upload_button_lora]).success(
move_file_lora, [file_output_lora], [lora1_gui, lora2_gui, lora3_gui, lora4_gui, lora5_gui, lora6_gui, lora7_gui], scroll_to_output=True)
use_textual_inversion_gui.change(set_textual_inversion_prompt, [use_textual_inversion_gui, prompt_gui, neg_prompt_gui, prompt_syntax_gui], [prompt_gui, neg_prompt_gui])
generate_from_image_btn_gui.click(
lambda: ("", "", ""), None, [series_dbt, character_dbt, prompt_gui], queue=False,
).success(
predict_tags_wd,
[input_image_gui, prompt_gui, image_algorithms, general_threshold_gui, character_threshold_gui],
[series_dbt, character_dbt, prompt_gui, copy_button_dbt],
).success(
compose_prompt_to_copy, [character_dbt, series_dbt, prompt_gui], [prompt_gui], queue=False,
).success(
remove_specific_prompt, [prompt_gui, keep_tags_gui], [prompt_gui], queue=False,
).success(
convert_danbooru_to_e621_prompt, [prompt_gui, tag_type_gui], [prompt_gui], queue=False,
).success(
insert_recom_prompt, [prompt_gui, neg_prompt_gui, recom_prompt_gui], [prompt_gui, neg_prompt_gui], queue=False,
)
prompt_type_button.click(convert_danbooru_to_e621_prompt, [prompt_gui, prompt_type_gui], [prompt_gui], queue=False)
random_character_gui.click(select_random_character, [series_dbt, character_dbt], [series_dbt, character_dbt], queue=False)
generate_db_random_button.click(
v2_random_prompt,
[prompt_gui, series_dbt, character_dbt,
rating_dbt, aspect_ratio_dbt, length_dbt, identity_dbt, ban_tags_dbt, model_name_dbt],
[prompt_gui, series_dbt, character_dbt],
).success(
convert_danbooru_to_e621_prompt, [prompt_gui, tag_type_gui], [prompt_gui], queue=False,
)
translate_prompt_button.click(translate_prompt, [prompt_gui], [prompt_gui], queue=False)
translate_prompt_button.click(translate_prompt, [character_dbt], [character_dbt], queue=False)
translate_prompt_button.click(translate_prompt, [series_dbt], [series_dbt], queue=False)
generate_button.click(
fn=insert_model_recom_prompt,
inputs=[prompt_gui, neg_prompt_gui, model_name_gui, recom_prompt_gui],
outputs=[prompt_gui, neg_prompt_gui],
queue=False,
).success(
fn=sd_gen.load_new_model,
inputs=[
model_name_gui,
vae_model_gui,
task_gui,
controlnet_model_gui,
],
outputs=[load_model_gui],
queue=True,
show_progress="minimal",
).success(
fn=sd_gen_generate_pipeline,
inputs=[
prompt_gui,
neg_prompt_gui,
num_images_gui,
steps_gui,
cfg_gui,
clip_skip_gui,
seed_gui,
lora1_gui,
lora_scale_1_gui,
lora2_gui,
lora_scale_2_gui,
lora3_gui,
lora_scale_3_gui,
lora4_gui,
lora_scale_4_gui,
lora5_gui,
lora_scale_5_gui,
lora6_gui,
lora_scale_6_gui,
lora7_gui,
lora_scale_7_gui,
sampler_gui,
schedule_type_gui,
schedule_prediction_type_gui,
img_height_gui,
img_width_gui,
model_name_gui,
vae_model_gui,
task_gui,
image_control,
preprocessor_name_gui,
preprocess_resolution_gui,
image_resolution_gui,
style_prompt_gui,
style_json_gui,
image_mask_gui,
strength_gui,
low_threshold_gui,
high_threshold_gui,
value_threshold_gui,
distance_threshold_gui,
recolor_gamma_correction_gui,
tile_blur_sigma_gui,
control_net_output_scaling_gui,
control_net_start_threshold_gui,
control_net_stop_threshold_gui,
active_textual_inversion_gui,
prompt_syntax_gui,
upscaler_model_path_gui,
upscaler_increases_size_gui,
upscaler_tile_size_gui,
upscaler_tile_overlap_gui,
hires_steps_gui,
hires_denoising_strength_gui,
hires_sampler_gui,
hires_prompt_gui,
hires_negative_prompt_gui,
hires_before_adetailer_gui,
hires_after_adetailer_gui,
hires_schedule_type_gui,
hires_guidance_scale_gui,
controlnet_model_gui,
loop_generation_gui,
leave_progress_bar_gui,
disable_progress_bar_gui,
image_previews_gui,
display_images_gui,
save_generated_images_gui,
filename_pattern_gui,
image_storage_location_gui,
retain_compel_previous_load_gui,
retain_detailfix_model_previous_load_gui,
retain_hires_model_previous_load_gui,
t2i_adapter_preprocessor_gui,
adapter_conditioning_scale_gui,
adapter_conditioning_factor_gui,
xformers_memory_efficient_attention_gui,
free_u_gui,
generator_in_cpu_gui,
adetailer_inpaint_only_gui,
adetailer_verbose_gui,
adetailer_sampler_gui,
adetailer_active_a_gui,
prompt_ad_a_gui,
negative_prompt_ad_a_gui,
strength_ad_a_gui,
face_detector_ad_a_gui,
person_detector_ad_a_gui,
hand_detector_ad_a_gui,
mask_dilation_a_gui,
mask_blur_a_gui,
mask_padding_a_gui,
adetailer_active_b_gui,
prompt_ad_b_gui,
negative_prompt_ad_b_gui,
strength_ad_b_gui,
face_detector_ad_b_gui,
person_detector_ad_b_gui,
hand_detector_ad_b_gui,
mask_dilation_b_gui,
mask_blur_b_gui,
mask_padding_b_gui,
retain_task_cache_gui,
guidance_rescale_gui,
image_ip1,
mask_ip1,
model_ip1,
mode_ip1,
scale_ip1,
image_ip2,
mask_ip2,
model_ip2,
mode_ip2,
scale_ip2,
pag_scale_gui,
face_restoration_model_gui,
face_restoration_visibility_gui,
face_restoration_weight_gui,
load_lora_cpu_gui,
verbose_info_gui,
gpu_duration_gui,
],
outputs=[load_model_gui, result_images, actual_task_info],
queue=True,
show_progress="full",
).success(save_gallery_images, [result_images, model_name_gui], [result_images, result_images_files], queue=False, show_api=False)\
.success(save_gallery_history, [result_images, result_images_files, history_gallery, history_files], [history_gallery, history_files], queue=False, show_api=False)
with gr.Tab("Danbooru Tags Transformer with WD Tagger", render=True):
with gr.Column(scale=2):
with gr.Group():
input_image = gr.Image(label="Input image", type="pil", sources=["upload", "clipboard"], height=256)
with gr.Accordion(label="Advanced options", open=False):
general_threshold = gr.Slider(label="Threshold", minimum=0.0, maximum=1.0, value=0.3, step=0.01, interactive=True)
character_threshold = gr.Slider(label="Character threshold", minimum=0.0, maximum=1.0, value=0.8, step=0.01, interactive=True)
input_tag_type = gr.Radio(label="Convert tags to", info="danbooru for Animagine, e621 for Pony.", choices=["danbooru", "e621"], value="danbooru")
recom_prompt = gr.Radio(label="Insert reccomended prompt", choices=["None", "Animagine", "Pony"], value="None", interactive=True)
image_algorithms = gr.CheckboxGroup(["Use WD Tagger"], label="Algorithms", value=["Use WD Tagger"], visible=False)
keep_tags = gr.Radio(label="Remove tags leaving only the following", choices=["body", "dress", "all"], value="all")
generate_from_image_btn = gr.Button(value="GENERATE TAGS FROM IMAGE", size="lg", variant="primary")
with gr.Group():
with gr.Row():
input_character = gr.Textbox(label="Character tags", placeholder="hatsune miku")
input_copyright = gr.Textbox(label="Copyright tags", placeholder="vocaloid")
pick_random_character = gr.Button(value="Random character 🎲", size="sm")
input_general = gr.TextArea(label="General tags", lines=4, placeholder="1girl, ...", value="")
input_tags_to_copy = gr.Textbox(value="", visible=False)
with gr.Row():
copy_input_btn = gr.Button(value="Copy to clipboard", size="sm", interactive=False)
copy_prompt_btn_input = gr.Button(value="Copy to primary prompt", size="sm", interactive=False)
translate_input_prompt_button = gr.Button(value="Translate prompt to English", size="sm", variant="secondary")
tag_type = gr.Radio(label="Output tag conversion", info="danbooru for Animagine, e621 for Pony.", choices=["danbooru", "e621"], value="e621", visible=False)
input_rating = gr.Radio(label="Rating", choices=list(V2_RATING_OPTIONS), value="explicit")
with gr.Accordion(label="Advanced options", open=False):
input_aspect_ratio = gr.Radio(label="Aspect ratio", info="The aspect ratio of the image.", choices=list(V2_ASPECT_RATIO_OPTIONS), value="square")
input_length = gr.Radio(label="Length", info="The total length of the tags.", choices=list(V2_LENGTH_OPTIONS), value="very_long")
input_identity = gr.Radio(label="Keep identity", info="How strictly to keep the identity of the character or subject. If you specify the detail of subject in the prompt, you should choose `strict`. Otherwise, choose `none` or `lax`. `none` is very creative but sometimes ignores the input prompt.", choices=list(V2_IDENTITY_OPTIONS), value="lax")
input_ban_tags = gr.Textbox(label="Ban tags", info="Tags to ban from the output.", placeholder="alternate costumen, ...", value="censored")
model_name = gr.Dropdown(label="Model", choices=list(V2_ALL_MODELS.keys()), value=list(V2_ALL_MODELS.keys())[0])
dummy_np = gr.Textbox(label="Negative prompt", value="", visible=False)
recom_animagine = gr.Textbox(label="Animagine reccomended prompt", value="Animagine", visible=False)
recom_pony = gr.Textbox(label="Pony reccomended prompt", value="Pony", visible=False)
generate_btn = gr.Button(value="GENERATE TAGS", size="lg", variant="primary")
with gr.Row():
with gr.Group():
output_text = gr.TextArea(label="Output tags", interactive=False, show_copy_button=True)
with gr.Row():
copy_btn = gr.Button(value="Copy to clipboard", size="sm", interactive=False)
copy_prompt_btn = gr.Button(value="Copy to primary prompt", size="sm", interactive=False)
with gr.Group():
output_text_pony = gr.TextArea(label="Output tags (Pony e621 style)", interactive=False, show_copy_button=True)
with gr.Row():
copy_btn_pony = gr.Button(value="Copy to clipboard", size="sm", interactive=False)
copy_prompt_btn_pony = gr.Button(value="Copy to primary prompt", size="sm", interactive=False)
description_ui()
translate_input_prompt_button.click(translate_prompt, inputs=[input_general], outputs=[input_general], queue=False)
translate_input_prompt_button.click(translate_prompt, inputs=[input_character], outputs=[input_character], queue=False)
translate_input_prompt_button.click(translate_prompt, inputs=[input_copyright], outputs=[input_copyright], queue=False)
generate_from_image_btn.click(
lambda: ("", "", ""), None, [input_copyright, input_character, input_general], queue=False,
).success(
predict_tags_wd,
[input_image, input_general, image_algorithms, general_threshold, character_threshold],
[input_copyright, input_character, input_general, copy_input_btn],
).success(
remove_specific_prompt, inputs=[input_general, keep_tags], outputs=[input_general], queue=False,
).success(
convert_danbooru_to_e621_prompt, inputs=[input_general, input_tag_type], outputs=[input_general], queue=False,
).success(
insert_recom_prompt, inputs=[input_general, dummy_np, recom_prompt], outputs=[input_general, dummy_np], queue=False,
).success(lambda: gr.update(interactive=True), None, [copy_prompt_btn_input], queue=False)
copy_input_btn.click(compose_prompt_to_copy, inputs=[input_character, input_copyright, input_general], outputs=[input_tags_to_copy])\
.success(gradio_copy_text, inputs=[input_tags_to_copy], js=COPY_ACTION_JS)
copy_prompt_btn_input.click(compose_prompt_to_copy, inputs=[input_character, input_copyright, input_general], outputs=[input_tags_to_copy])\
.success(gradio_copy_prompt, inputs=[input_tags_to_copy], outputs=[prompt_gui])
pick_random_character.click(select_random_character, [input_copyright, input_character], [input_copyright, input_character])
generate_btn.click(
v2_upsampling_prompt,
[model_name, input_copyright, input_character, input_general,
input_rating, input_aspect_ratio, input_length, input_identity, input_ban_tags],
[output_text],
).success(
convert_danbooru_to_e621_prompt, inputs=[output_text, tag_type], outputs=[output_text_pony], queue=False,
).success(
insert_recom_prompt, inputs=[output_text, dummy_np, recom_animagine], outputs=[output_text, dummy_np], queue=False,
).success(
insert_recom_prompt, inputs=[output_text_pony, dummy_np, recom_pony], outputs=[output_text_pony, dummy_np], queue=False,
).success(lambda: (gr.update(interactive=True), gr.update(interactive=True), gr.update(interactive=True), gr.update(interactive=True)),
None, [copy_btn, copy_btn_pony, copy_prompt_btn, copy_prompt_btn_pony], queue=False)
copy_btn.click(gradio_copy_text, inputs=[output_text], js=COPY_ACTION_JS)
copy_btn_pony.click(gradio_copy_text, inputs=[output_text_pony], js=COPY_ACTION_JS)
copy_prompt_btn.click(gradio_copy_prompt, inputs=[output_text], outputs=[prompt_gui])
copy_prompt_btn_pony.click(gradio_copy_prompt, inputs=[output_text_pony], outputs=[prompt_gui])
gr.LoginButton()
gr.DuplicateButton(value="Duplicate Space for private use (This demo does not work on CPU. Requires GPU Space)")
app.queue()
app.launch(show_error=True, debug=True) # allowed_paths=["./images/"], show_error=True, debug=True
## END MOD