Merge branch 'master' into enable-override-hypernet

modules/api/api.py

@@ -1,14 +1,19 @@
+import base64
+import io
 import time
 import uvicorn
+from threading import Lock
 from gradio.processing_utils import encode_pil_to_base64, decode_base64_to_file, decode_base64_to_image
-from fastapi import APIRouter, Depends, HTTPException
+from fastapi import APIRouter, Depends, FastAPI, HTTPException
 import modules.shared as shared
-from modules import devices
 from modules.api.models import *
 from modules.processing import StableDiffusionProcessingTxt2Img, StableDiffusionProcessingImg2Img, process_images
 from modules.sd_samplers import all_samplers
 from modules.extras import run_extras, run_pnginfo
-
+from PIL import PngImagePlugin
+from modules.sd_models import checkpoints_list
+from modules.realesrgan_model import get_realesrgan_models
+from typing import List
 
 def upscaler_to_index(name: str):
     try:
@@ -29,8 +34,26 @@ def setUpscalers(req: dict):
     return reqDict
 
 
+def encode_pil_to_base64(image):
+    with io.BytesIO() as output_bytes:
+
+        # Copy any text-only metadata
+        use_metadata = False
+        metadata = PngImagePlugin.PngInfo()
+        for key, value in image.info.items():
+            if isinstance(key, str) and isinstance(value, str):
+                metadata.add_text(key, value)
+                use_metadata = True
+
+        image.save(
+            output_bytes, "PNG", pnginfo=(metadata if use_metadata else None)
+        )
+        bytes_data = output_bytes.getvalue()
+    return base64.b64encode(bytes_data)
+
+
 class Api:
-    def __init__(self, app, queue_lock):
+    def __init__(self, app: FastAPI, queue_lock: Lock):
         self.router = APIRouter()
         self.app = app
         self.queue_lock = queue_lock
@@ -40,7 +63,20 @@ class Api:
         self.app.add_api_route("/sdapi/v1/extra-batch-images", self.extras_batch_images_api, methods=["POST"], response_model=ExtrasBatchImagesResponse)
         self.app.add_api_route("/sdapi/v1/png-info", self.pnginfoapi, methods=["POST"], response_model=PNGInfoResponse)
         self.app.add_api_route("/sdapi/v1/progress", self.progressapi, methods=["GET"], response_model=ProgressResponse)
+        self.app.add_api_route("/sdapi/v1/interrogate", self.interrogateapi, methods=["POST"])
         self.app.add_api_route("/sdapi/v1/interrupt", self.interruptapi, methods=["POST"])
+        self.app.add_api_route("/sdapi/v1/options", self.get_config, methods=["GET"], response_model=OptionsModel)
+        self.app.add_api_route("/sdapi/v1/options", self.set_config, methods=["POST"])
+        self.app.add_api_route("/sdapi/v1/cmd-flags", self.get_cmd_flags, methods=["GET"], response_model=FlagsModel)
+        self.app.add_api_route("/sdapi/v1/samplers", self.get_samplers, methods=["GET"], response_model=List[SamplerItem])
+        self.app.add_api_route("/sdapi/v1/upscalers", self.get_upscalers, methods=["GET"], response_model=List[UpscalerItem])
+        self.app.add_api_route("/sdapi/v1/sd-models", self.get_sd_models, methods=["GET"], response_model=List[SDModelItem])
+        self.app.add_api_route("/sdapi/v1/hypernetworks", self.get_hypernetworks, methods=["GET"], response_model=List[HypernetworkItem])
+        self.app.add_api_route("/sdapi/v1/face-restorers", self.get_face_restorers, methods=["GET"], response_model=List[FaceRestorerItem])
+        self.app.add_api_route("/sdapi/v1/realesrgan-models", self.get_realesrgan_models, methods=["GET"], response_model=List[RealesrganItem])
+        self.app.add_api_route("/sdapi/v1/prompt-styles", self.get_promp_styles, methods=["GET"], response_model=List[PromptStyleItem])
+        self.app.add_api_route("/sdapi/v1/artist-categories", self.get_artists_categories, methods=["GET"], response_model=List[str])
+        self.app.add_api_route("/sdapi/v1/artists", self.get_artists, methods=["GET"], response_model=List[ArtistItem])
 
     def text2imgapi(self, txt2imgreq: StableDiffusionTxt2ImgProcessingAPI):
         sampler_index = sampler_to_index(txt2imgreq.sampler_index)
@@ -171,17 +207,96 @@ class Api:
 
         progress = min(progress, 1)
 
+        shared.state.set_current_image()
+
         current_image = None
         if shared.state.current_image and not req.skip_current_image:
             current_image = encode_pil_to_base64(shared.state.current_image)
 
         return ProgressResponse(progress=progress, eta_relative=eta_relative, state=shared.state.dict(), current_image=current_image)
 
+    def interrogateapi(self, interrogatereq: InterrogateRequest):
+        image_b64 = interrogatereq.image
+        if image_b64 is None:
+            raise HTTPException(status_code=404, detail="Image not found") 
+
+        img = self.__base64_to_image(image_b64)
+
+        # Override object param
+        with self.queue_lock:
+            processed = shared.interrogator.interrogate(img)
+        
+        return InterrogateResponse(caption=processed)
+
     def interruptapi(self):
         shared.state.interrupt()
 
         return {}
 
+    def get_config(self):
+        options = {}
+        for key in shared.opts.data.keys():
+            metadata = shared.opts.data_labels.get(key)
+            if(metadata is not None):
+                options.update({key: shared.opts.data.get(key, shared.opts.data_labels.get(key).default)})
+            else:
+                options.update({key: shared.opts.data.get(key, None)})
+
+        return options
+
+    def set_config(self, req: OptionsModel):
+        # currently req has all options fields even if you send a dict like { "send_seed": false }, which means it will
+        # overwrite all options with default values.
+        raise RuntimeError('Setting options via API is not supported')
+
+        reqDict = vars(req)
+        for o in reqDict:
+            setattr(shared.opts, o, reqDict[o])
+
+        shared.opts.save(shared.config_filename)
+        return
+
+    def get_cmd_flags(self):
+        return vars(shared.cmd_opts)
+
+    def get_samplers(self):
+        return [{"name":sampler[0], "aliases":sampler[2], "options":sampler[3]} for sampler in all_samplers]
+
+    def get_upscalers(self):
+        upscalers = []
+
+        for upscaler in shared.sd_upscalers:
+            u = upscaler.scaler
+            upscalers.append({"name":u.name, "model_name":u.model_name, "model_path":u.model_path, "model_url":u.model_url})
+
+        return upscalers
+
+    def get_sd_models(self):
+        return [{"title":x.title, "model_name":x.model_name, "hash":x.hash, "filename": x.filename, "config": x.config} for x in checkpoints_list.values()]
+
+    def get_hypernetworks(self):
+        return [{"name": name, "path": shared.hypernetworks[name]} for name in shared.hypernetworks]
+
+    def get_face_restorers(self):
+        return [{"name":x.name(), "cmd_dir": getattr(x, "cmd_dir", None)} for x in shared.face_restorers]
+
+    def get_realesrgan_models(self):
+        return [{"name":x.name,"path":x.data_path, "scale":x.scale} for x in get_realesrgan_models(None)]
+
+    def get_promp_styles(self):
+        styleList = []
+        for k in shared.prompt_styles.styles:
+            style = shared.prompt_styles.styles[k]
+            styleList.append({"name":style[0], "prompt": style[1], "negative_prompr": style[2]})
+
+        return styleList
+
+    def get_artists_categories(self):
+        return shared.artist_db.cats
+
+    def get_artists(self):
+        return [{"name":x[0], "score":x[1], "category":x[2]} for x in shared.artist_db.artists]
+
     def launch(self, server_name, port):
         self.app.include_router(self.router)
         uvicorn.run(self.app, host=server_name, port=port)

modules/api/models.py

@@ -1,11 +1,11 @@
 import inspect
-from click import prompt
 from pydantic import BaseModel, Field, create_model
 from typing import Any, Optional
 from typing_extensions import Literal
 from inflection import underscore
 from modules.processing import StableDiffusionProcessingTxt2Img, StableDiffusionProcessingImg2Img
-from modules.shared import sd_upscalers
+from modules.shared import sd_upscalers, opts, parser
+from typing import Dict, List
 
 API_NOT_ALLOWED = [
     "self",
@@ -65,6 +65,7 @@ class PydanticModelGenerator:
 
         self._model_name = model_name
         self._class_data = merge_class_params(class_instance)
+
         self._model_def = [
             ModelDef(
                 field=underscore(k),
@@ -109,12 +110,12 @@ StableDiffusionImg2ImgProcessingAPI = PydanticModelGenerator(
 ).generate_model()
 
 class TextToImageResponse(BaseModel):
-    images: list[str] = Field(default=None, title="Image", description="The generated image in base64 format.")
+    images: List[str] = Field(default=None, title="Image", description="The generated image in base64 format.")
     parameters: dict
     info: str
 
 class ImageToImageResponse(BaseModel):
-    images: list[str] = Field(default=None, title="Image", description="The generated image in base64 format.")
+    images: List[str] = Field(default=None, title="Image", description="The generated image in base64 format.")
     parameters: dict
     info: str
 
@@ -131,6 +132,7 @@ class ExtrasBaseRequest(BaseModel):
     upscaler_1: str = Field(default="None", title="Main upscaler", description=f"The name of the main upscaler to use, it has to be one of this list: {' , '.join([x.name for x in sd_upscalers])}")
     upscaler_2: str = Field(default="None", title="Secondary upscaler", description=f"The name of the secondary upscaler to use, it has to be one of this list: {' , '.join([x.name for x in sd_upscalers])}")
     extras_upscaler_2_visibility: float = Field(default=0, title="Secondary upscaler visibility", ge=0, le=1, allow_inf_nan=False, description="Sets the visibility of secondary upscaler, values should be between 0 and 1.")
+    upscale_first: bool = Field(default=False, title="Upscale first", description="Should the upscaler run before restoring faces?")
 
 class ExtraBaseResponse(BaseModel):
     html_info: str = Field(title="HTML info", description="A series of HTML tags containing the process info.")
@@ -146,10 +148,10 @@ class FileData(BaseModel):
     name: str = Field(title="File name")
 
 class ExtrasBatchImagesRequest(ExtrasBaseRequest):
-    imageList: list[FileData] = Field(title="Images", description="List of images to work on. Must be Base64 strings")
+    imageList: List[FileData] = Field(title="Images", description="List of images to work on. Must be Base64 strings")
 
 class ExtrasBatchImagesResponse(ExtraBaseResponse):
-    images: list[str] = Field(title="Images", description="The generated images in base64 format.")
+    images: List[str] = Field(title="Images", description="The generated images in base64 format.")
 
 class PNGInfoRequest(BaseModel):
     image: str = Field(title="Image", description="The base64 encoded PNG image")
@@ -165,3 +167,75 @@ class ProgressResponse(BaseModel):
     eta_relative: float = Field(title="ETA in secs")
     state: dict = Field(title="State", description="The current state snapshot")
     current_image: str = Field(default=None, title="Current image", description="The current image in base64 format. opts.show_progress_every_n_steps is required for this to work.")
+
+class InterrogateRequest(BaseModel):
+    image: str = Field(default="", title="Image", description="Image to work on, must be a Base64 string containing the image's data.")
+
+class InterrogateResponse(BaseModel):
+    caption: str = Field(default=None, title="Caption", description="The generated caption for the image.")
+
+fields = {}
+for key, value in opts.data.items():
+    metadata = opts.data_labels.get(key)
+    optType = opts.typemap.get(type(value), type(value))
+
+    if (metadata is not None):
+        fields.update({key: (Optional[optType], Field(
+            default=metadata.default ,description=metadata.label))})
+    else:
+        fields.update({key: (Optional[optType], Field())})
+
+OptionsModel = create_model("Options", **fields)
+
+flags = {}
+_options = vars(parser)['_option_string_actions']
+for key in _options:
+    if(_options[key].dest != 'help'):
+        flag = _options[key]
+        _type = str
+        if _options[key].default is not None: _type = type(_options[key].default)
+        flags.update({flag.dest: (_type,Field(default=flag.default, description=flag.help))})
+
+FlagsModel = create_model("Flags", **flags)
+
+class SamplerItem(BaseModel):
+    name: str = Field(title="Name")
+    aliases: List[str] = Field(title="Aliases")
+    options: Dict[str, str] = Field(title="Options")
+
+class UpscalerItem(BaseModel):
+    name: str = Field(title="Name")
+    model_name: Optional[str] = Field(title="Model Name")
+    model_path: Optional[str] = Field(title="Path")
+    model_url: Optional[str] = Field(title="URL")
+
+class SDModelItem(BaseModel):
+    title: str = Field(title="Title")
+    model_name: str = Field(title="Model Name")
+    hash: str = Field(title="Hash")
+    filename: str = Field(title="Filename")
+    config: str = Field(title="Config file")
+
+class HypernetworkItem(BaseModel):
+    name: str = Field(title="Name")
+    path: Optional[str] = Field(title="Path")
+
+class FaceRestorerItem(BaseModel):
+    name: str = Field(title="Name")
+    cmd_dir: Optional[str] = Field(title="Path")
+
+class RealesrganItem(BaseModel):
+    name: str = Field(title="Name")
+    path: Optional[str] = Field(title="Path")
+    scale: Optional[int] = Field(title="Scale")
+
+class PromptStyleItem(BaseModel):
+    name: str = Field(title="Name")
+    prompt: Optional[str] = Field(title="Prompt")
+    negative_prompt: Optional[str] = Field(title="Negative Prompt")
+
+class ArtistItem(BaseModel):
+    name: str = Field(title="Name")
+    score: float = Field(title="Score")
+    category: str = Field(title="Category")
+

modules/esrgan_model.py

@@ -50,6 +50,7 @@ def mod2normal(state_dict):
 def resrgan2normal(state_dict, nb=23):
     # this code is copied from https://github.com/victorca25/iNNfer
     if "conv_first.weight" in state_dict and "body.0.rdb1.conv1.weight" in state_dict:
+        re8x = 0
         crt_net = {}
         items = []
         for k, v in state_dict.items():
@@ -75,10 +76,18 @@ def resrgan2normal(state_dict, nb=23):
         crt_net['model.3.bias'] = state_dict['conv_up1.bias']
         crt_net['model.6.weight'] = state_dict['conv_up2.weight']
         crt_net['model.6.bias'] = state_dict['conv_up2.bias']
-        crt_net['model.8.weight'] = state_dict['conv_hr.weight']
-        crt_net['model.8.bias'] = state_dict['conv_hr.bias']
-        crt_net['model.10.weight'] = state_dict['conv_last.weight']
-        crt_net['model.10.bias'] = state_dict['conv_last.bias']
+
+        if 'conv_up3.weight' in state_dict:
+            # modification supporting: https://github.com/ai-forever/Real-ESRGAN/blob/main/RealESRGAN/rrdbnet_arch.py
+            re8x = 3
+            crt_net['model.9.weight'] = state_dict['conv_up3.weight']
+            crt_net['model.9.bias'] = state_dict['conv_up3.bias']
+
+        crt_net[f'model.{8+re8x}.weight'] = state_dict['conv_hr.weight']
+        crt_net[f'model.{8+re8x}.bias'] = state_dict['conv_hr.bias']
+        crt_net[f'model.{10+re8x}.weight'] = state_dict['conv_last.weight']
+        crt_net[f'model.{10+re8x}.bias'] = state_dict['conv_last.bias']
+
         state_dict = crt_net
     return state_dict
 

modules/extensions.py

@@ -34,8 +34,11 @@ class Extension:
         if repo is None or repo.bare:
             self.remote = None
         else:
-            self.remote = next(repo.remote().urls, None)
-            self.status = 'unknown'
+            try:
+                self.remote = next(repo.remote().urls, None)
+                self.status = 'unknown'
+            except Exception:
+                self.remote = None
 
     def list_files(self, subdir, extension):
         from modules import scripts

modules/extras.py

@@ -136,12 +136,13 @@ def run_extras(extras_mode, resize_mode, image, image_folder, input_dir, output_
 
     def run_upscalers_blend(params: List[UpscaleParams], image: Image.Image, info: str) -> Tuple[Image.Image, str]:
         blended_result: Image.Image = None
+        image_hash: str = hash(np.array(image.getdata()).tobytes())
         for upscaler in params:
             upscale_args = (upscaler.upscaler_idx, upscaling_resize, resize_mode,
                             upscaling_resize_w, upscaling_resize_h, upscaling_crop)
-            cache_key = LruCache.Key(image_hash=hash(np.array(image.getdata()).tobytes()),
+            cache_key = LruCache.Key(image_hash=image_hash,
                                      info_hash=hash(info),
-                                     args_hash=hash((upscale_args, upscale_first)))
+                                     args_hash=hash(upscale_args))
             cached_entry = cached_images.get(cache_key)
             if cached_entry is None:
                 res = upscale(image, *upscale_args)

modules/hypernetworks/hypernetwork.py

@@ -22,6 +22,8 @@ from collections import defaultdict, deque
 from statistics import stdev, mean
 
 
+optimizer_dict = {optim_name : cls_obj for optim_name, cls_obj in inspect.getmembers(torch.optim, inspect.isclass) if optim_name != "Optimizer"}
+
 class HypernetworkModule(torch.nn.Module):
     multiplier = 1.0
     activation_dict = {
@@ -35,7 +37,8 @@ class HypernetworkModule(torch.nn.Module):
     }
     activation_dict.update({cls_name.lower(): cls_obj for cls_name, cls_obj in inspect.getmembers(torch.nn.modules.activation) if inspect.isclass(cls_obj) and cls_obj.__module__ == 'torch.nn.modules.activation'})
 
-    def __init__(self, dim, state_dict=None, layer_structure=None, activation_func=None, weight_init='Normal', add_layer_norm=False, use_dropout=False):
+    def __init__(self, dim, state_dict=None, layer_structure=None, activation_func=None, weight_init='Normal',
+                 add_layer_norm=False, use_dropout=False, activate_output=False, last_layer_dropout=True):
         super().__init__()
 
         assert layer_structure is not None, "layer_structure must not be None"
@@ -48,8 +51,8 @@ class HypernetworkModule(torch.nn.Module):
             # Add a fully-connected layer
             linears.append(torch.nn.Linear(int(dim * layer_structure[i]), int(dim * layer_structure[i+1])))
 
-            # Add an activation func
-            if activation_func == "linear" or activation_func is None:
+            # Add an activation func except last layer
+            if activation_func == "linear" or activation_func is None or (i >= len(layer_structure) - 2 and not activate_output):
                 pass
             elif activation_func in self.activation_dict:
                 linears.append(self.activation_dict[activation_func]())
@@ -60,8 +63,8 @@ class HypernetworkModule(torch.nn.Module):
             if add_layer_norm:
                 linears.append(torch.nn.LayerNorm(int(dim * layer_structure[i+1])))
 
-            # Add dropout expect last layer
-            if use_dropout and i < len(layer_structure) - 3:
+            # Add dropout except last layer
+            if use_dropout and (i < len(layer_structure) - 3 or last_layer_dropout and i < len(layer_structure) - 2):
                 linears.append(torch.nn.Dropout(p=0.3))
 
         self.linear = torch.nn.Sequential(*linears)
@@ -75,7 +78,7 @@ class HypernetworkModule(torch.nn.Module):
                     w, b = layer.weight.data, layer.bias.data
                     if weight_init == "Normal" or type(layer) == torch.nn.LayerNorm:
                         normal_(w, mean=0.0, std=0.01)
-                        normal_(b, mean=0.0, std=0.005)
+                        normal_(b, mean=0.0, std=0)
                     elif weight_init == 'XavierUniform':
                         xavier_uniform_(w)
                         zeros_(b)
@@ -127,7 +130,7 @@ class Hypernetwork:
     filename = None
     name = None
 
-    def __init__(self, name=None, enable_sizes=None, layer_structure=None, activation_func=None, weight_init=None, add_layer_norm=False, use_dropout=False):
+    def __init__(self, name=None, enable_sizes=None, layer_structure=None, activation_func=None, weight_init=None, add_layer_norm=False, use_dropout=False, activate_output=False, **kwargs):
         self.filename = None
         self.name = name
         self.layers = {}
@@ -139,11 +142,17 @@ class Hypernetwork:
         self.weight_init = weight_init
         self.add_layer_norm = add_layer_norm
         self.use_dropout = use_dropout
+        self.activate_output = activate_output
+        self.last_layer_dropout = kwargs['last_layer_dropout'] if 'last_layer_dropout' in kwargs else True
+        self.optimizer_name = None
+        self.optimizer_state_dict = None
 
         for size in enable_sizes or []:
             self.layers[size] = (
-                HypernetworkModule(size, None, self.layer_structure, self.activation_func, self.weight_init, self.add_layer_norm, self.use_dropout),
-                HypernetworkModule(size, None, self.layer_structure, self.activation_func, self.weight_init, self.add_layer_norm, self.use_dropout),
+                HypernetworkModule(size, None, self.layer_structure, self.activation_func, self.weight_init,
+                                   self.add_layer_norm, self.use_dropout, self.activate_output, last_layer_dropout=self.last_layer_dropout),
+                HypernetworkModule(size, None, self.layer_structure, self.activation_func, self.weight_init,
+                                   self.add_layer_norm, self.use_dropout, self.activate_output, last_layer_dropout=self.last_layer_dropout),
             )
 
     def weights(self):
@@ -158,6 +167,7 @@ class Hypernetwork:
 
     def save(self, filename):
         state_dict = {}
+        optimizer_saved_dict = {}
 
         for k, v in self.layers.items():
             state_dict[k] = (v[0].state_dict(), v[1].state_dict())
@@ -171,8 +181,17 @@ class Hypernetwork:
         state_dict['use_dropout'] = self.use_dropout
         state_dict['sd_checkpoint'] = self.sd_checkpoint
         state_dict['sd_checkpoint_name'] = self.sd_checkpoint_name
+        state_dict['activate_output'] = self.activate_output
+        state_dict['last_layer_dropout'] = self.last_layer_dropout
+
+        if self.optimizer_name is not None:
+            optimizer_saved_dict['optimizer_name'] = self.optimizer_name
 
         torch.save(state_dict, filename)
+        if shared.opts.save_optimizer_state and self.optimizer_state_dict:
+            optimizer_saved_dict['hash'] = sd_models.model_hash(filename)
+            optimizer_saved_dict['optimizer_state_dict'] = self.optimizer_state_dict
+            torch.save(optimizer_saved_dict, filename + '.optim')
 
     def load(self, filename):
         self.filename = filename
@@ -191,12 +210,29 @@ class Hypernetwork:
         print(f"Layer norm is set to {self.add_layer_norm}")
         self.use_dropout = state_dict.get('use_dropout', False)
         print(f"Dropout usage is set to {self.use_dropout}" )
+        self.activate_output = state_dict.get('activate_output', True)
+        print(f"Activate last layer is set to {self.activate_output}")
+        self.last_layer_dropout = state_dict.get('last_layer_dropout', False)
+
+        optimizer_saved_dict = torch.load(self.filename + '.optim', map_location = 'cpu') if os.path.exists(self.filename + '.optim') else {}
+        self.optimizer_name = optimizer_saved_dict.get('optimizer_name', 'AdamW')
+        print(f"Optimizer name is {self.optimizer_name}")
+        if sd_models.model_hash(filename) == optimizer_saved_dict.get('hash', None):
+            self.optimizer_state_dict = optimizer_saved_dict.get('optimizer_state_dict', None)
+        else:
+            self.optimizer_state_dict = None
+        if self.optimizer_state_dict:
+            print("Loaded existing optimizer from checkpoint")
+        else:
+            print("No saved optimizer exists in checkpoint")
 
         for size, sd in state_dict.items():
             if type(size) == int:
                 self.layers[size] = (
-                    HypernetworkModule(size, sd[0], self.layer_structure, self.activation_func, self.weight_init, self.add_layer_norm, self.use_dropout),
-                    HypernetworkModule(size, sd[1], self.layer_structure, self.activation_func, self.weight_init, self.add_layer_norm, self.use_dropout),
+                    HypernetworkModule(size, sd[0], self.layer_structure, self.activation_func, self.weight_init,
+                                       self.add_layer_norm, self.use_dropout, self.activate_output, last_layer_dropout=self.last_layer_dropout),
+                    HypernetworkModule(size, sd[1], self.layer_structure, self.activation_func, self.weight_init,
+                                       self.add_layer_norm, self.use_dropout, self.activate_output, last_layer_dropout=self.last_layer_dropout),
                 )
 
         self.name = state_dict.get('name', self.name)
@@ -207,11 +243,11 @@ class Hypernetwork:
 
 def list_hypernetworks(path):
     res = {}
-    for filename in glob.iglob(os.path.join(path, '**/*.pt'), recursive=True):
+    for filename in sorted(glob.iglob(os.path.join(path, '**/*.pt'), recursive=True)):
         name = os.path.splitext(os.path.basename(filename))[0]
         # Prevent a hypothetical "None.pt" from being listed.
         if name != "None":
-            res[name] = filename
+            res[name + f"({sd_models.model_hash(filename)})"] = filename
     return res
 
 
@@ -346,6 +382,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log
     shared.state.textinfo = "Initializing hypernetwork training..."
     shared.state.job_count = steps
 
+    hypernetwork_name = hypernetwork_name.rsplit('(', 1)[0]
     filename = os.path.join(shared.cmd_opts.hypernetwork_dir, f'{hypernetwork_name}.pt')
 
     log_directory = os.path.join(log_directory, datetime.datetime.now().strftime("%Y-%m-%d"), hypernetwork_name)
@@ -392,8 +429,22 @@ def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log
     weights = hypernetwork.weights()
     for weight in weights:
         weight.requires_grad = True
-    # if optimizer == "AdamW": or else Adam / AdamW / SGD, etc...
-    optimizer = torch.optim.AdamW(weights, lr=scheduler.learn_rate)
+
+    # Here we use optimizer from saved HN, or we can specify as UI option.
+    if hypernetwork.optimizer_name in optimizer_dict:
+        optimizer = optimizer_dict[hypernetwork.optimizer_name](params=weights, lr=scheduler.learn_rate)
+        optimizer_name = hypernetwork.optimizer_name
+    else:
+        print(f"Optimizer type {hypernetwork.optimizer_name} is not defined!")
+        optimizer = torch.optim.AdamW(params=weights, lr=scheduler.learn_rate)
+        optimizer_name = 'AdamW'
+
+    if hypernetwork.optimizer_state_dict:  # This line must be changed if Optimizer type can be different from saved optimizer.
+        try:
+            optimizer.load_state_dict(hypernetwork.optimizer_state_dict)
+        except RuntimeError as e:
+            print("Cannot resume from saved optimizer!")
+            print(e)
 
     steps_without_grad = 0
 
@@ -455,7 +506,11 @@ def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log
             # Before saving, change name to match current checkpoint.
             hypernetwork_name_every = f'{hypernetwork_name}-{steps_done}'
             last_saved_file = os.path.join(hypernetwork_dir, f'{hypernetwork_name_every}.pt')
+            hypernetwork.optimizer_name = optimizer_name
+            if shared.opts.save_optimizer_state:
+                hypernetwork.optimizer_state_dict = optimizer.state_dict()
             save_hypernetwork(hypernetwork, checkpoint, hypernetwork_name, last_saved_file)
+            hypernetwork.optimizer_state_dict = None  # dereference it after saving, to save memory.
 
         textual_inversion.write_loss(log_directory, "hypernetwork_loss.csv", hypernetwork.step, len(ds), {
             "loss": f"{previous_mean_loss:.7f}",
@@ -518,8 +573,12 @@ Last saved image: {html.escape(last_saved_image)}<br/>
     report_statistics(loss_dict)
 
     filename = os.path.join(shared.cmd_opts.hypernetwork_dir, f'{hypernetwork_name}.pt')
+    hypernetwork.optimizer_name = optimizer_name
+    if shared.opts.save_optimizer_state:
+        hypernetwork.optimizer_state_dict = optimizer.state_dict()
     save_hypernetwork(hypernetwork, checkpoint, hypernetwork_name, filename)
-
+    del optimizer
+    hypernetwork.optimizer_state_dict = None  # dereference it after saving, to save memory.
     return hypernetwork, filename
 
 def save_hypernetwork(hypernetwork, checkpoint, hypernetwork_name, filename):

modules/hypernetworks/ui.py

@@ -9,7 +9,7 @@ from modules import devices, sd_hijack, shared
 from modules.hypernetworks import hypernetwork
 
 not_available = ["hardswish", "multiheadattention"]
-keys = ["linear"] + list(x for x in hypernetwork.HypernetworkModule.activation_dict.keys() if x not in not_available)
+keys = list(x for x in hypernetwork.HypernetworkModule.activation_dict.keys() if x not in not_available)
 
 def create_hypernetwork(name, enable_sizes, overwrite_old, layer_structure=None, activation_func=None, weight_init=None, add_layer_norm=False, use_dropout=False):
     # Remove illegal characters from name.

modules/img2img.py

@@ -81,7 +81,8 @@ def img2img(mode: int, prompt: str, negative_prompt: str, prompt_style: str, pro
         mask = None
 
     # Use the EXIF orientation of photos taken by smartphones.
-    image = ImageOps.exif_transpose(image) 
+    if image is not None:
+        image = ImageOps.exif_transpose(image) 
 
     assert 0. <= denoising_strength <= 1., 'can only work with strength in [0.0, 1.0]'
 
@@ -137,6 +138,8 @@ def img2img(mode: int, prompt: str, negative_prompt: str, prompt_style: str, pro
         if processed is None:
             processed = process_images(p)
 
+    p.close()
+
     shared.total_tqdm.clear()
 
     generation_info_js = processed.js()

modules/ldsr_model_arch.py

@@ -101,8 +101,8 @@ class LDSR:
         down_sample_rate = target_scale / 4
         wd = width_og * down_sample_rate
         hd = height_og * down_sample_rate
-        width_downsampled_pre = int(wd)
-        height_downsampled_pre = int(hd)
+        width_downsampled_pre = int(np.ceil(wd))
+        height_downsampled_pre = int(np.ceil(hd))
 
         if down_sample_rate != 1:
             print(
@@ -110,7 +110,12 @@ class LDSR:
             im_og = im_og.resize((width_downsampled_pre, height_downsampled_pre), Image.LANCZOS)
         else:
             print(f"Down sample rate is 1 from {target_scale} / 4 (Not downsampling)")
-        logs = self.run(model["model"], im_og, diffusion_steps, eta)
+        
+        # pad width and height to multiples of 64, pads with the edge values of image to avoid artifacts
+        pad_w, pad_h = np.max(((2, 2), np.ceil(np.array(im_og.size) / 64).astype(int)), axis=0) * 64 - im_og.size
+        im_padded = Image.fromarray(np.pad(np.array(im_og), ((0, pad_h), (0, pad_w), (0, 0)), mode='edge'))
+        
+        logs = self.run(model["model"], im_padded, diffusion_steps, eta)
 
         sample = logs["sample"]
         sample = sample.detach().cpu()
@@ -120,6 +125,9 @@ class LDSR:
         sample = np.transpose(sample, (0, 2, 3, 1))
         a = Image.fromarray(sample[0])
 
+        # remove padding
+        a = a.crop((0, 0) + tuple(np.array(im_og.size) * 4))
+
         del model
         gc.collect()
         torch.cuda.empty_cache()

modules/localization.py

@@ -3,6 +3,7 @@ import os
 import sys
 import traceback
 
+
 localizations = {}
 
 
@@ -16,6 +17,11 @@ def list_localizations(dirname):
 
         localizations[fn] = os.path.join(dirname, file)
 
+    from modules import scripts
+    for file in scripts.list_scripts("localizations", ".json"):
+        fn, ext = os.path.splitext(file.filename)
+        localizations[fn] = file.path
+
 
 def localization_js(current_localization_name):
     fn = localizations.get(current_localization_name, None)

modules/masking.py

@@ -49,7 +49,7 @@ def expand_crop_region(crop_region, processing_width, processing_height, image_w
     ratio_processing = processing_width / processing_height
 
     if ratio_crop_region > ratio_processing:
-        desired_height = (x2 - x1) * ratio_processing
+        desired_height = (x2 - x1) / ratio_processing
         desired_height_diff = int(desired_height - (y2-y1))
         y1 -= desired_height_diff//2
         y2 += desired_height_diff - desired_height_diff//2

modules/modelloader.py

@@ -85,6 +85,9 @@ def cleanup_models():
     src_path = os.path.join(root_path, "ESRGAN")
     dest_path = os.path.join(models_path, "ESRGAN")
     move_files(src_path, dest_path)
+    src_path = os.path.join(models_path, "BSRGAN")
+    dest_path = os.path.join(models_path, "ESRGAN")
+    move_files(src_path, dest_path, ".pth")
     src_path = os.path.join(root_path, "gfpgan")
     dest_path = os.path.join(models_path, "GFPGAN")
     move_files(src_path, dest_path)

modules/processing.py

@@ -134,11 +134,7 @@ class StableDiffusionProcessing():
             # Dummy zero conditioning if we're not using inpainting model.
             # Still takes up a bit of memory, but no encoder call.
             # Pretty sure we can just make this a 1x1 image since its not going to be used besides its batch size.
-            return torch.zeros(
-                x.shape[0], 5, 1, 1, 
-                dtype=x.dtype, 
-                device=x.device
-            )
+            return x.new_zeros(x.shape[0], 5, 1, 1)
 
         height = height or self.height
         width = width or self.width
@@ -156,11 +152,7 @@ class StableDiffusionProcessing():
     def img2img_image_conditioning(self, source_image, latent_image, image_mask = None):
         if self.sampler.conditioning_key not in {'hybrid', 'concat'}:
             # Dummy zero conditioning if we're not using inpainting model.
-            return torch.zeros(
-                latent_image.shape[0], 5, 1, 1,
-                dtype=latent_image.dtype,
-                device=latent_image.device
-            )
+            return latent_image.new_zeros(latent_image.shape[0], 5, 1, 1)
 
         # Handle the different mask inputs
         if image_mask is not None:
@@ -174,11 +166,11 @@ class StableDiffusionProcessing():
                 # Inpainting model uses a discretized mask as input, so we round to either 1.0 or 0.0
                 conditioning_mask = torch.round(conditioning_mask)
         else:
-            conditioning_mask = torch.ones(1, 1, *source_image.shape[-2:])
+            conditioning_mask = source_image.new_ones(1, 1, *source_image.shape[-2:])
 
         # Create another latent image, this time with a masked version of the original input.
         # Smoothly interpolate between the masked and unmasked latent conditioning image using a parameter.
-        conditioning_mask = conditioning_mask.to(source_image.device)
+        conditioning_mask = conditioning_mask.to(source_image.device).to(source_image.dtype)
         conditioning_image = torch.lerp(
             source_image,
             source_image * (1.0 - conditioning_mask),
@@ -199,9 +191,13 @@ class StableDiffusionProcessing():
     def init(self, all_prompts, all_seeds, all_subseeds):
         pass
 
-    def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength):
+    def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts):
         raise NotImplementedError()
 
+    def close(self):
+        self.sd_model = None
+        self.sampler = None
+
 
 class Processed:
     def __init__(self, p: StableDiffusionProcessing, images_list, seed=-1, info="", subseed=None, all_prompts=None, all_seeds=None, all_subseeds=None, index_of_first_image=0, infotexts=None):
@@ -507,6 +503,9 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
             if len(prompts) == 0:
                 break
 
+            if p.scripts is not None:
+                p.scripts.process_batch(p, batch_number=n, prompts=prompts, seeds=seeds, subseeds=subseeds)
+
             with devices.autocast():
                 uc = prompt_parser.get_learned_conditioning(shared.sd_model, len(prompts) * [p.negative_prompt], p.steps)
                 c = prompt_parser.get_multicond_learned_conditioning(shared.sd_model, prompts, p.steps)
@@ -519,7 +518,7 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
                 shared.state.job = f"Batch {n+1} out of {p.n_iter}"
 
             with devices.autocast():
-                samples_ddim = p.sample(conditioning=c, unconditional_conditioning=uc, seeds=seeds, subseeds=subseeds, subseed_strength=p.subseed_strength)
+                samples_ddim = p.sample(conditioning=c, unconditional_conditioning=uc, seeds=seeds, subseeds=subseeds, subseed_strength=p.subseed_strength, prompts=prompts)
 
             samples_ddim = samples_ddim.to(devices.dtype_vae)
             x_samples_ddim = decode_first_stage(p.sd_model, samples_ddim)
@@ -599,9 +598,6 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
     if p.scripts is not None:
         p.scripts.postprocess(p, res)
 
-    p.sd_model = None
-    p.sampler = None
-
     return res
 
 
@@ -650,7 +646,7 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing):
             self.truncate_x = int(self.firstphase_width - firstphase_width_truncated) // opt_f
             self.truncate_y = int(self.firstphase_height - firstphase_height_truncated) // opt_f
 
-    def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength):
+    def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts):
         self.sampler = sd_samplers.create_sampler_with_index(sd_samplers.samplers, self.sampler_index, self.sd_model)
 
         if not self.enable_hr:
@@ -663,9 +659,28 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing):
 
         samples = samples[:, :, self.truncate_y//2:samples.shape[2]-self.truncate_y//2, self.truncate_x//2:samples.shape[3]-self.truncate_x//2]
 
+        """saves image before applying hires fix, if enabled in options; takes as an arguyment either an image or batch with latent space images"""
+        def save_intermediate(image, index):
+            if not opts.save or self.do_not_save_samples or not opts.save_images_before_highres_fix:
+                return
+
+            if not isinstance(image, Image.Image):
+                image = sd_samplers.sample_to_image(image, index)
+
+            images.save_image(image, self.outpath_samples, "", seeds[index], prompts[index], opts.samples_format, suffix="-before-highres-fix")
+
         if opts.use_scale_latent_for_hires_fix:
+            for i in range(samples.shape[0]):
+                save_intermediate(samples, i)
+
             samples = torch.nn.functional.interpolate(samples, size=(self.height // opt_f, self.width // opt_f), mode="bilinear")
 
+            # Avoid making the inpainting conditioning unless necessary as 
+            # this does need some extra compute to decode / encode the image again.
+            if getattr(self, "inpainting_mask_weight", shared.opts.inpainting_mask_weight) < 1.0:
+                image_conditioning = self.img2img_image_conditioning(decode_first_stage(self.sd_model, samples), samples)
+            else:
+                image_conditioning = self.txt2img_image_conditioning(samples)
         else:
             decoded_samples = decode_first_stage(self.sd_model, samples)
             lowres_samples = torch.clamp((decoded_samples + 1.0) / 2.0, min=0.0, max=1.0)
@@ -675,6 +690,9 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing):
                 x_sample = 255. * np.moveaxis(x_sample.cpu().numpy(), 0, 2)
                 x_sample = x_sample.astype(np.uint8)
                 image = Image.fromarray(x_sample)
+
+                save_intermediate(image, i)
+
                 image = images.resize_image(0, image, self.width, self.height)
                 image = np.array(image).astype(np.float32) / 255.0
                 image = np.moveaxis(image, 2, 0)
@@ -686,14 +704,14 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing):
 
             samples = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(decoded_samples))
 
+            image_conditioning = self.img2img_image_conditioning(decoded_samples, samples)
+
         shared.state.nextjob()
 
         self.sampler = sd_samplers.create_sampler_with_index(sd_samplers.samplers, self.sampler_index, self.sd_model)
 
         noise = create_random_tensors(samples.shape[1:], seeds=seeds, subseeds=subseeds, subseed_strength=subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self)
 
-        image_conditioning = self.txt2img_image_conditioning(x)
-
         # GC now before running the next img2img to prevent running out of memory
         x = None
         devices.torch_gc()
@@ -832,8 +850,7 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing):
 
         self.image_conditioning = self.img2img_image_conditioning(image, self.init_latent, self.image_mask)
 
-
-    def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength):
+    def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts):
         x = create_random_tensors([opt_C, self.height // opt_f, self.width // opt_f], seeds=seeds, subseeds=subseeds, subseed_strength=self.subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self)
 
         samples = self.sampler.sample_img2img(self, self.init_latent, x, conditioning, unconditional_conditioning, image_conditioning=self.image_conditioning)
@@ -844,4 +861,4 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing):
         del x
         devices.torch_gc()
 
-        return samples
+        return samples

modules/safe.py

@@ -23,11 +23,18 @@ def encode(*args):
 
 
 class RestrictedUnpickler(pickle.Unpickler):
+    extra_handler = None
+
     def persistent_load(self, saved_id):
         assert saved_id[0] == 'storage'
         return TypedStorage()
 
     def find_class(self, module, name):
+        if self.extra_handler is not None:
+            res = self.extra_handler(module, name)
+            if res is not None:
+                return res
+
         if module == 'collections' and name == 'OrderedDict':
             return getattr(collections, name)
         if module == 'torch._utils' and name in ['_rebuild_tensor_v2', '_rebuild_parameter']:
@@ -52,7 +59,7 @@ class RestrictedUnpickler(pickle.Unpickler):
             return set
 
         # Forbid everything else.
-        raise pickle.UnpicklingError(f"global '{module}/{name}' is forbidden")
+        raise Exception(f"global '{module}/{name}' is forbidden")
 
 
 allowed_zip_names = ["archive/data.pkl", "archive/version"]
@@ -69,7 +76,7 @@ def check_zip_filenames(filename, names):
         raise Exception(f"bad file inside {filename}: {name}")
 
 
-def check_pt(filename):
+def check_pt(filename, extra_handler):
     try:
 
         # new pytorch format is a zip file
@@ -78,6 +85,7 @@ def check_pt(filename):
 
             with z.open('archive/data.pkl') as file:
                 unpickler = RestrictedUnpickler(file)
+                unpickler.extra_handler = extra_handler
                 unpickler.load()
 
     except zipfile.BadZipfile:
@@ -85,16 +93,42 @@ def check_pt(filename):
         # if it's not a zip file, it's an olf pytorch format, with five objects written to pickle
         with open(filename, "rb") as file:
             unpickler = RestrictedUnpickler(file)
+            unpickler.extra_handler = extra_handler
             for i in range(5):
                 unpickler.load()
 
 
 def load(filename, *args, **kwargs):
+    return load_with_extra(filename, *args, **kwargs)
+
+
+def load_with_extra(filename, extra_handler=None, *args, **kwargs):
+    """
+    this functon is intended to be used by extensions that want to load models with
+    some extra classes in them that the usual unpickler would find suspicious.
+
+    Use the extra_handler argument to specify a function that takes module and field name as text,
+    and returns that field's value:
+
+    ```python
+    def extra(module, name):
+        if module == 'collections' and name == 'OrderedDict':
+            return collections.OrderedDict
+
+        return None
+
+    safe.load_with_extra('model.pt', extra_handler=extra)
+    ```
+
+    The alternative to this is just to use safe.unsafe_torch_load('model.pt'), which as the name implies is
+    definitely unsafe.
+    """
+
     from modules import shared
 
     try:
         if not shared.cmd_opts.disable_safe_unpickle:
-            check_pt(filename)
+            check_pt(filename, extra_handler)
 
     except pickle.UnpicklingError:
         print(f"Error verifying pickled file from {filename}:", file=sys.stderr)

modules/script_callbacks.py

@@ -2,6 +2,7 @@ import sys
 import traceback
 from collections import namedtuple
 import inspect
+from typing import Optional
 
 from fastapi import FastAPI
 from gradio import Blocks
@@ -26,25 +27,42 @@ class ImageSaveParams:
         """dictionary with parameters for image's PNG info data; infotext will have the key 'parameters'"""
 
 
+class CFGDenoiserParams:
+    def __init__(self, x, image_cond, sigma, sampling_step, total_sampling_steps):
+        self.x = x
+        """Latent image representation in the process of being denoised"""
+        
+        self.image_cond = image_cond
+        """Conditioning image"""
+        
+        self.sigma = sigma
+        """Current sigma noise step value"""
+        
+        self.sampling_step = sampling_step
+        """Current Sampling step number"""
+        
+        self.total_sampling_steps = total_sampling_steps
+        """Total number of sampling steps planned"""
+
+
 ScriptCallback = namedtuple("ScriptCallback", ["script", "callback"])
-callbacks_app_started = []
-callbacks_model_loaded = []
-callbacks_ui_tabs = []
-callbacks_ui_settings = []
-callbacks_before_image_saved = []
-callbacks_image_saved = []
+callback_map = dict(
+    callbacks_app_started=[],
+    callbacks_model_loaded=[],
+    callbacks_ui_tabs=[],
+    callbacks_ui_settings=[],
+    callbacks_before_image_saved=[],
+    callbacks_image_saved=[],
+    callbacks_cfg_denoiser=[]
+)
 
 
 def clear_callbacks():
-    callbacks_model_loaded.clear()
-    callbacks_ui_tabs.clear()
-    callbacks_ui_settings.clear()
-    callbacks_before_image_saved.clear()
-    callbacks_image_saved.clear()
+    for callback_list in callback_map.values():
+        callback_list.clear()
 
-
-def app_started_callback(demo: Blocks, app: FastAPI):
-    for c in callbacks_app_started:
+def app_started_callback(demo: Optional[Blocks], app: FastAPI):
+    for c in callback_map['callbacks_app_started']:
         try:
             c.callback(demo, app)
         except Exception:
@@ -52,7 +70,7 @@ def app_started_callback(demo: Blocks, app: FastAPI):
 
 
 def model_loaded_callback(sd_model):
-    for c in callbacks_model_loaded:
+    for c in callback_map['callbacks_model_loaded']:
         try:
             c.callback(sd_model)
         except Exception:
@@ -62,7 +80,7 @@ def model_loaded_callback(sd_model):
 def ui_tabs_callback():
     res = []
     
-    for c in callbacks_ui_tabs:
+    for c in callback_map['callbacks_ui_tabs']:
         try:
             res += c.callback() or []
         except Exception:
@@ -72,7 +90,7 @@ def ui_tabs_callback():
 
 
 def ui_settings_callback():
-    for c in callbacks_ui_settings:
+    for c in callback_map['callbacks_ui_settings']:
         try:
             c.callback()
         except Exception:
@@ -80,7 +98,7 @@ def ui_settings_callback():
 
 
 def before_image_saved_callback(params: ImageSaveParams):
-    for c in callbacks_before_image_saved:
+    for c in callback_map['callbacks_before_image_saved']:
         try:
             c.callback(params)
         except Exception:
@@ -88,30 +106,54 @@ def before_image_saved_callback(params: ImageSaveParams):
 
 
 def image_saved_callback(params: ImageSaveParams):
-    for c in callbacks_image_saved:
+    for c in callback_map['callbacks_image_saved']:
         try:
             c.callback(params)
         except Exception:
             report_exception(c, 'image_saved_callback')
 
 
+def cfg_denoiser_callback(params: CFGDenoiserParams):
+    for c in callback_map['callbacks_cfg_denoiser']:
+        try:
+            c.callback(params)
+        except Exception:
+            report_exception(c, 'cfg_denoiser_callback')
+
+
 def add_callback(callbacks, fun):
     stack = [x for x in inspect.stack() if x.filename != __file__]
     filename = stack[0].filename if len(stack) > 0 else 'unknown file'
 
     callbacks.append(ScriptCallback(filename, fun))
 
+    
+def remove_current_script_callbacks():
+    stack = [x for x in inspect.stack() if x.filename != __file__]
+    filename = stack[0].filename if len(stack) > 0 else 'unknown file'
+    if filename == 'unknown file':
+        return
+    for callback_list in callback_map.values():
+        for callback_to_remove in [cb for cb in callback_list if cb.script == filename]:
+            callback_list.remove(callback_to_remove)
+
+
+def remove_callbacks_for_function(callback_func):
+    for callback_list in callback_map.values():
+        for callback_to_remove in [cb for cb in callback_list if cb.callback == callback_func]:
+            callback_list.remove(callback_to_remove)
+
 
 def on_app_started(callback):
     """register a function to be called when the webui started, the gradio `Block` component and
     fastapi `FastAPI` object are passed as the arguments"""
-    add_callback(callbacks_app_started, callback)
+    add_callback(callback_map['callbacks_app_started'], callback)
 
 
 def on_model_loaded(callback):
     """register a function to be called when the stable diffusion model is created; the model is
     passed as an argument"""
-    add_callback(callbacks_model_loaded, callback)
+    add_callback(callback_map['callbacks_model_loaded'], callback)
 
 
 def on_ui_tabs(callback):
@@ -124,13 +166,13 @@ def on_ui_tabs(callback):
     title is tab text displayed to user in the UI
     elem_id is HTML id for the tab
     """
-    add_callback(callbacks_ui_tabs, callback)
+    add_callback(callback_map['callbacks_ui_tabs'], callback)
 
 
 def on_ui_settings(callback):
     """register a function to be called before UI settings are populated; add your settings
     by using shared.opts.add_option(shared.OptionInfo(...)) """
-    add_callback(callbacks_ui_settings, callback)
+    add_callback(callback_map['callbacks_ui_settings'], callback)
 
 
 def on_before_image_saved(callback):
@@ -138,7 +180,7 @@ def on_before_image_saved(callback):
     The callback is called with one argument:
         - params: ImageSaveParams - parameters the image is to be saved with. You can change fields in this object.
     """
-    add_callback(callbacks_before_image_saved, callback)
+    add_callback(callback_map['callbacks_before_image_saved'], callback)
 
 
 def on_image_saved(callback):
@@ -146,4 +188,12 @@ def on_image_saved(callback):
     The callback is called with one argument:
         - params: ImageSaveParams - parameters the image was saved with. Changing fields in this object does nothing.
     """
-    add_callback(callbacks_image_saved, callback)
+    add_callback(callback_map['callbacks_image_saved'], callback)
+
+
+def on_cfg_denoiser(callback):
+    """register a function to be called in the kdiffussion cfg_denoiser method after building the inner model inputs.
+    The callback is called with one argument:
+        - params: CFGDenoiserParams - parameters to be passed to the inner model and sampling state details.
+    """
+    add_callback(callback_map['callbacks_cfg_denoiser'], callback)

modules/scripts.py

@@ -3,7 +3,6 @@ import sys
 import traceback
 from collections import namedtuple
 
-import modules.ui as ui
 import gradio as gr
 
 from modules.processing import StableDiffusionProcessing
@@ -18,6 +17,9 @@ class Script:
     args_to = None
     alwayson = False
 
+    """A gr.Group component that has all script's UI inside it"""
+    group = None
+
     infotext_fields = None
     """if set in ui(), this is a list of pairs of gradio component + text; the text will be used when
     parsing infotext to set the value for the component; see ui.py's txt2img_paste_fields for an example
@@ -70,6 +72,19 @@ class Script:
 
         pass
 
+    def process_batch(self, p, *args, **kwargs):
+        """
+        Same as process(), but called for every batch.
+
+        **kwargs will have those items:
+          - batch_number - index of current batch, from 0 to number of batches-1
+          - prompts - list of prompts for current batch; you can change contents of this list but changing the number of entries will likely break things
+          - seeds - list of seeds for current batch
+          - subseeds - list of subseeds for current batch
+        """
+
+        pass
+
     def postprocess(self, p, processed, *args):
         """
         This function is called after processing ends for AlwaysVisible scripts.
@@ -218,8 +233,6 @@ class ScriptRunner:
 
             for control in controls:
                 control.custom_script_source = os.path.basename(script.filename)
-                if not script.alwayson:
-                    control.visible = False
 
             if script.infotext_fields is not None:
                 self.infotext_fields += script.infotext_fields
@@ -229,40 +242,41 @@ class ScriptRunner:
             script.args_to = len(inputs)
 
         for script in self.alwayson_scripts:
-            with gr.Group():
+            with gr.Group() as group:
                 create_script_ui(script, inputs, inputs_alwayson)
 
+            script.group = group
+
         dropdown = gr.Dropdown(label="Script", elem_id="script_list", choices=["None"] + self.titles, value="None", type="index")
         dropdown.save_to_config = True
         inputs[0] = dropdown
 
         for script in self.selectable_scripts:
-            create_script_ui(script, inputs, inputs_alwayson)
+            with gr.Group(visible=False) as group:
+                create_script_ui(script, inputs, inputs_alwayson)
+
+            script.group = group
 
         def select_script(script_index):
-            if 0 < script_index <= len(self.selectable_scripts):
-                script = self.selectable_scripts[script_index-1]
-                args_from = script.args_from
-                args_to = script.args_to
-            else:
-                args_from = 0
-                args_to = 0
+            selected_script = self.selectable_scripts[script_index - 1] if script_index>0 else None
 
-            return [ui.gr_show(True if i == 0 else args_from <= i < args_to or is_alwayson) for i, is_alwayson in enumerate(inputs_alwayson)]
+            return [gr.update(visible=selected_script == s) for s in self.selectable_scripts]
 
         def init_field(title):
+            """called when an initial value is set from ui-config.json to show script's UI components"""
+
             if title == 'None':
                 return
+
             script_index = self.titles.index(title)
-            script = self.selectable_scripts[script_index]
-            for i in range(script.args_from, script.args_to):
-                inputs[i].visible = True
+            self.selectable_scripts[script_index].group.visible = True
 
         dropdown.init_field = init_field
+
         dropdown.change(
             fn=select_script,
             inputs=[dropdown],
-            outputs=inputs
+            outputs=[script.group for script in self.selectable_scripts]
         )
 
         return inputs
@@ -294,6 +308,15 @@ class ScriptRunner:
                 print(f"Error running process: {script.filename}", file=sys.stderr)
                 print(traceback.format_exc(), file=sys.stderr)
 
+    def process_batch(self, p, **kwargs):
+        for script in self.alwayson_scripts:
+            try:
+                script_args = p.script_args[script.args_from:script.args_to]
+                script.process_batch(p, *script_args, **kwargs)
+            except Exception:
+                print(f"Error running process_batch: {script.filename}", file=sys.stderr)
+                print(traceback.format_exc(), file=sys.stderr)
+
     def postprocess(self, p, processed):
         for script in self.alwayson_scripts:
             try:

modules/sd_models.py

@@ -9,7 +9,7 @@ from omegaconf import OmegaConf
 
 from ldm.util import instantiate_from_config
 
-from modules import shared, modelloader, devices, script_callbacks
+from modules import shared, modelloader, devices, script_callbacks, sd_vae
 from modules.paths import models_path
 from modules.sd_hijack_inpainting import do_inpainting_hijack, should_hijack_inpainting
 
@@ -159,13 +159,16 @@ def get_state_dict_from_checkpoint(pl_sd):
     return pl_sd
 
 
-vae_ignore_keys = {"model_ema.decay", "model_ema.num_updates"}
-
-
-def load_model_weights(model, checkpoint_info):
+def load_model_weights(model, checkpoint_info, vae_file="auto"):
     checkpoint_file = checkpoint_info.filename
     sd_model_hash = checkpoint_info.hash
 
+    if shared.opts.sd_checkpoint_cache > 0 and hasattr(model, "sd_checkpoint_info"):
+        sd_vae.restore_base_vae(model)
+        checkpoints_loaded[model.sd_checkpoint_info] = model.state_dict().copy()
+
+    vae_file = sd_vae.resolve_vae(checkpoint_file, vae_file=vae_file)
+
     if checkpoint_info not in checkpoints_loaded:
         print(f"Loading weights [{sd_model_hash}] from {checkpoint_file}")
 
@@ -182,37 +185,36 @@ def load_model_weights(model, checkpoint_info):
             model.to(memory_format=torch.channels_last)
 
         if not shared.cmd_opts.no_half:
+            vae = model.first_stage_model
+
+            # with --no-half-vae, remove VAE from model when doing half() to prevent its weights from being converted to float16
+            if shared.cmd_opts.no_half_vae:
+                model.first_stage_model = None
+
             model.half()
+            model.first_stage_model = vae
 
         devices.dtype = torch.float32 if shared.cmd_opts.no_half else torch.float16
         devices.dtype_vae = torch.float32 if shared.cmd_opts.no_half or shared.cmd_opts.no_half_vae else torch.float16
 
-        vae_file = os.path.splitext(checkpoint_file)[0] + ".vae.pt"
-
-        if not os.path.exists(vae_file) and shared.cmd_opts.vae_path is not None:
-            vae_file = shared.cmd_opts.vae_path
-
-        if os.path.exists(vae_file):
-            print(f"Loading VAE weights from: {vae_file}")
-            vae_ckpt = torch.load(vae_file, map_location=shared.weight_load_location)
-            vae_dict = {k: v for k, v in vae_ckpt["state_dict"].items() if k[0:4] != "loss" and k not in vae_ignore_keys}
-            model.first_stage_model.load_state_dict(vae_dict)
-
         model.first_stage_model.to(devices.dtype_vae)
 
-        if shared.opts.sd_checkpoint_cache > 0:
-            checkpoints_loaded[checkpoint_info] = model.state_dict().copy()
-            while len(checkpoints_loaded) > shared.opts.sd_checkpoint_cache:
-                checkpoints_loaded.popitem(last=False)  # LRU
     else:
-        print(f"Loading weights [{sd_model_hash}] from cache")
-        checkpoints_loaded.move_to_end(checkpoint_info)
+        vae_name = sd_vae.get_filename(vae_file) if vae_file else None
+        vae_message = f" with {vae_name} VAE" if vae_name else ""
+        print(f"Loading weights [{sd_model_hash}]{vae_message} from cache")
         model.load_state_dict(checkpoints_loaded[checkpoint_info])
 
+    if shared.opts.sd_checkpoint_cache > 0:
+        while len(checkpoints_loaded) > shared.opts.sd_checkpoint_cache:
+            checkpoints_loaded.popitem(last=False)  # LRU
+
     model.sd_model_hash = sd_model_hash
     model.sd_model_checkpoint = checkpoint_file
     model.sd_checkpoint_info = checkpoint_info
 
+    sd_vae.load_vae(model, vae_file)
+
 
 def load_model(checkpoint_info=None):
     from modules import lowvram, sd_hijack
@@ -263,7 +265,7 @@ def load_model(checkpoint_info=None):
 def reload_model_weights(sd_model=None, info=None):
     from modules import lowvram, devices, sd_hijack
     checkpoint_info = info or select_checkpoint()
-
+ 
     if not sd_model:
         sd_model = shared.sd_model
 

modules/sd_samplers.py

@@ -12,6 +12,7 @@ from modules import prompt_parser, devices, processing, images
 
 from modules.shared import opts, cmd_opts, state
 import modules.shared as shared
+from modules.script_callbacks import CFGDenoiserParams, cfg_denoiser_callback
 
 
 SamplerData = namedtuple('SamplerData', ['name', 'constructor', 'aliases', 'options'])
@@ -23,11 +24,15 @@ samplers_k_diffusion = [
     ('Heun', 'sample_heun', ['k_heun'], {}),
     ('DPM2', 'sample_dpm_2', ['k_dpm_2'], {}),
     ('DPM2 a', 'sample_dpm_2_ancestral', ['k_dpm_2_a'], {}),
+    ('DPM++ 2S a', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a'], {}),
+    ('DPM++ 2M', 'sample_dpmpp_2m', ['k_dpmpp_2m'], {}),
     ('DPM fast', 'sample_dpm_fast', ['k_dpm_fast'], {}),
     ('DPM adaptive', 'sample_dpm_adaptive', ['k_dpm_ad'], {}),
     ('LMS Karras', 'sample_lms', ['k_lms_ka'], {'scheduler': 'karras'}),
     ('DPM2 Karras', 'sample_dpm_2', ['k_dpm_2_ka'], {'scheduler': 'karras'}),
     ('DPM2 a Karras', 'sample_dpm_2_ancestral', ['k_dpm_2_a_ka'], {'scheduler': 'karras'}),
+    ('DPM++ 2S a Karras', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a_ka'], {'scheduler': 'karras'}),
+    ('DPM++ 2M Karras', 'sample_dpmpp_2m', ['k_dpmpp_2m_ka'], {'scheduler': 'karras'}),
 ]
 
 samplers_data_k_diffusion = [
@@ -92,8 +97,8 @@ def single_sample_to_image(sample):
     return Image.fromarray(x_sample)
 
 
-def sample_to_image(samples):
-    return single_sample_to_image(samples[0])
+def sample_to_image(samples, index=0):
+    return single_sample_to_image(samples[index])
 
 
 def samples_to_image_grid(samples):
@@ -280,6 +285,12 @@ class CFGDenoiser(torch.nn.Module):
         image_cond_in = torch.cat([torch.stack([image_cond[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [image_cond])
         sigma_in = torch.cat([torch.stack([sigma[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [sigma])
 
+        denoiser_params = CFGDenoiserParams(x_in, image_cond_in, sigma_in, state.sampling_step, state.sampling_steps)
+        cfg_denoiser_callback(denoiser_params)
+        x_in = denoiser_params.x
+        image_cond_in = denoiser_params.image_cond
+        sigma_in = denoiser_params.sigma
+
         if tensor.shape[1] == uncond.shape[1]:
             cond_in = torch.cat([tensor, uncond])
 

modules/sd_vae.py

@@ -0,0 +1,207 @@
+import torch
+import os
+from collections import namedtuple
+from modules import shared, devices, script_callbacks
+from modules.paths import models_path
+import glob
+
+
+model_dir = "Stable-diffusion"
+model_path = os.path.abspath(os.path.join(models_path, model_dir))
+vae_dir = "VAE"
+vae_path = os.path.abspath(os.path.join(models_path, vae_dir))
+
+
+vae_ignore_keys = {"model_ema.decay", "model_ema.num_updates"}
+
+
+default_vae_dict = {"auto": "auto", "None": "None"}
+default_vae_list = ["auto", "None"]
+
+
+default_vae_values = [default_vae_dict[x] for x in default_vae_list]
+vae_dict = dict(default_vae_dict)
+vae_list = list(default_vae_list)
+first_load = True
+
+
+base_vae = None
+loaded_vae_file = None
+checkpoint_info = None
+
+
+def get_base_vae(model):
+    if base_vae is not None and checkpoint_info == model.sd_checkpoint_info and model:
+        return base_vae
+    return None
+
+
+def store_base_vae(model):
+    global base_vae, checkpoint_info
+    if checkpoint_info != model.sd_checkpoint_info:
+        base_vae = model.first_stage_model.state_dict().copy()
+        checkpoint_info = model.sd_checkpoint_info
+
+
+def delete_base_vae():
+    global base_vae, checkpoint_info
+    base_vae = None
+    checkpoint_info = None
+
+
+def restore_base_vae(model):
+    global base_vae, checkpoint_info
+    if base_vae is not None and checkpoint_info == model.sd_checkpoint_info:
+        load_vae_dict(model, base_vae)
+    delete_base_vae()
+
+
+def get_filename(filepath):
+    return os.path.splitext(os.path.basename(filepath))[0]
+
+
+def refresh_vae_list(vae_path=vae_path, model_path=model_path):
+    global vae_dict, vae_list
+    res = {}
+    candidates = [
+        *glob.iglob(os.path.join(model_path, '**/*.vae.ckpt'), recursive=True),
+        *glob.iglob(os.path.join(model_path, '**/*.vae.pt'), recursive=True),
+        *glob.iglob(os.path.join(vae_path, '**/*.ckpt'), recursive=True),
+        *glob.iglob(os.path.join(vae_path, '**/*.pt'), recursive=True)
+    ]
+    if shared.cmd_opts.vae_path is not None and os.path.isfile(shared.cmd_opts.vae_path):
+        candidates.append(shared.cmd_opts.vae_path)
+    for filepath in candidates:
+        name = get_filename(filepath)
+        res[name] = filepath
+    vae_list.clear()
+    vae_list.extend(default_vae_list)
+    vae_list.extend(list(res.keys()))
+    vae_dict.clear()
+    vae_dict.update(res)
+    vae_dict.update(default_vae_dict)
+    return vae_list
+
+
+def resolve_vae(checkpoint_file, vae_file="auto"):
+    global first_load, vae_dict, vae_list
+
+    # if vae_file argument is provided, it takes priority, but not saved
+    if vae_file and vae_file not in default_vae_list:
+        if not os.path.isfile(vae_file):
+            vae_file = "auto"
+            print("VAE provided as function argument doesn't exist")
+    # for the first load, if vae-path is provided, it takes priority, saved, and failure is reported
+    if first_load and shared.cmd_opts.vae_path is not None:
+        if os.path.isfile(shared.cmd_opts.vae_path):
+            vae_file = shared.cmd_opts.vae_path
+            shared.opts.data['sd_vae'] = get_filename(vae_file)
+        else:
+            print("VAE provided as command line argument doesn't exist")
+    # else, we load from settings
+    if vae_file == "auto" and shared.opts.sd_vae is not None:
+        # if saved VAE settings isn't recognized, fallback to auto
+        vae_file = vae_dict.get(shared.opts.sd_vae, "auto")
+        # if VAE selected but not found, fallback to auto
+        if vae_file not in default_vae_values and not os.path.isfile(vae_file):
+            vae_file = "auto"
+            print("Selected VAE doesn't exist")
+    # vae-path cmd arg takes priority for auto
+    if vae_file == "auto" and shared.cmd_opts.vae_path is not None:
+        if os.path.isfile(shared.cmd_opts.vae_path):
+            vae_file = shared.cmd_opts.vae_path
+            print("Using VAE provided as command line argument")
+    # if still not found, try look for ".vae.pt" beside model
+    model_path = os.path.splitext(checkpoint_file)[0]
+    if vae_file == "auto":
+        vae_file_try = model_path + ".vae.pt"
+        if os.path.isfile(vae_file_try):
+            vae_file = vae_file_try
+            print("Using VAE found beside selected model")
+    # if still not found, try look for ".vae.ckpt" beside model
+    if vae_file == "auto":
+        vae_file_try = model_path + ".vae.ckpt"
+        if os.path.isfile(vae_file_try):
+            vae_file = vae_file_try
+            print("Using VAE found beside selected model")
+    # No more fallbacks for auto
+    if vae_file == "auto":
+        vae_file = None
+    # Last check, just because
+    if vae_file and not os.path.exists(vae_file):
+        vae_file = None
+
+    return vae_file
+
+
+def load_vae(model, vae_file=None):
+    global first_load, vae_dict, vae_list, loaded_vae_file
+    # save_settings = False
+
+    if vae_file:
+        print(f"Loading VAE weights from: {vae_file}")
+        vae_ckpt = torch.load(vae_file, map_location=shared.weight_load_location)
+        vae_dict_1 = {k: v for k, v in vae_ckpt["state_dict"].items() if k[0:4] != "loss" and k not in vae_ignore_keys}
+        load_vae_dict(model, vae_dict_1)
+
+        # If vae used is not in dict, update it
+        # It will be removed on refresh though
+        vae_opt = get_filename(vae_file)
+        if vae_opt not in vae_dict:
+            vae_dict[vae_opt] = vae_file
+            vae_list.append(vae_opt)
+
+    loaded_vae_file = vae_file
+
+    """
+    # Save current VAE to VAE settings, maybe? will it work?
+    if save_settings:
+        if vae_file is None:
+            vae_opt = "None"
+
+        # shared.opts.sd_vae = vae_opt
+    """
+
+    first_load = False
+
+
+# don't call this from outside
+def load_vae_dict(model, vae_dict_1=None):
+    if vae_dict_1:
+        store_base_vae(model)
+        model.first_stage_model.load_state_dict(vae_dict_1)
+    else:
+        restore_base_vae()
+    model.first_stage_model.to(devices.dtype_vae)
+
+
+def reload_vae_weights(sd_model=None, vae_file="auto"):
+    from modules import lowvram, devices, sd_hijack
+
+    if not sd_model:
+        sd_model = shared.sd_model
+
+    checkpoint_info = sd_model.sd_checkpoint_info
+    checkpoint_file = checkpoint_info.filename
+    vae_file = resolve_vae(checkpoint_file, vae_file=vae_file)
+
+    if loaded_vae_file == vae_file:
+        return
+
+    if shared.cmd_opts.lowvram or shared.cmd_opts.medvram:
+        lowvram.send_everything_to_cpu()
+    else:
+        sd_model.to(devices.cpu)
+
+    sd_hijack.model_hijack.undo_hijack(sd_model)
+
+    load_vae(sd_model, vae_file)
+
+    sd_hijack.model_hijack.hijack(sd_model)
+    script_callbacks.model_loaded_callback(sd_model)
+
+    if not shared.cmd_opts.lowvram and not shared.cmd_opts.medvram:
+        sd_model.to(devices.device)
+
+    print(f"VAE Weights loaded.")
+    return sd_model

modules/shared.py

@@ -4,6 +4,7 @@ import json
 import os
 import sys
 from collections import OrderedDict
+import time
 
 import gradio as gr
 import tqdm
@@ -14,7 +15,7 @@ import modules.memmon
 import modules.sd_models
 import modules.styles
 import modules.devices as devices
-from modules import sd_samplers, sd_models, localization
+from modules import sd_samplers, sd_models, localization, sd_vae
 from modules.hypernetworks import hypernetwork
 from modules.paths import models_path, script_path, sd_path
 
@@ -43,6 +44,7 @@ parser.add_argument("--precision", type=str, help="evaluate at this precision",
 parser.add_argument("--share", action='store_true', help="use share=True for gradio and make the UI accessible through their site")
 parser.add_argument("--ngrok", type=str, help="ngrok authtoken, alternative to gradio --share", default=None)
 parser.add_argument("--ngrok-region", type=str, help="The region in which ngrok should start.", default="us")
+parser.add_argument("--enable-insecure-extension-access", action='store_true', help="enable extensions tab regardless of other options")
 parser.add_argument("--codeformer-models-path", type=str, help="Path to directory with codeformer model file(s).", default=os.path.join(models_path, 'Codeformer'))
 parser.add_argument("--gfpgan-models-path", type=str, help="Path to directory with GFPGAN model file(s).", default=os.path.join(models_path, 'GFPGAN'))
 parser.add_argument("--esrgan-models-path", type=str, help="Path to directory with ESRGAN model file(s).", default=os.path.join(models_path, 'ESRGAN'))
@@ -84,6 +86,10 @@ parser.add_argument("--nowebui", action='store_true', help="use api=True to laun
 parser.add_argument("--ui-debug-mode", action='store_true', help="Don't load model to quickly launch UI")
 parser.add_argument("--device-id", type=str, help="Select the default CUDA device to use (export CUDA_VISIBLE_DEVICES=0,1,etc might be needed before)", default=None)
 parser.add_argument("--administrator", action='store_true', help="Administrator rights", default=False)
+parser.add_argument("--cors-allow-origins", type=str, help="Allowed CORS origins", default=None)
+parser.add_argument("--tls-keyfile", type=str, help="Partially enables TLS, requires --tls-certfile to fully function", default=None)
+parser.add_argument("--tls-certfile", type=str, help="Partially enables TLS, requires --tls-keyfile to fully function", default=None)
+parser.add_argument("--server-name", type=str, help="Sets hostname of server", default=None)
 
 cmd_opts = parser.parse_args()
 restricted_opts = {
@@ -98,7 +104,7 @@ restricted_opts = {
     "outdir_save",
 }
 
-cmd_opts.disable_extension_access = cmd_opts.share or cmd_opts.listen
+cmd_opts.disable_extension_access = (cmd_opts.share or cmd_opts.listen) and not cmd_opts.enable_insecure_extension_access
 
 devices.device, devices.device_interrogate, devices.device_gfpgan, devices.device_swinir, devices.device_esrgan, devices.device_scunet, devices.device_codeformer = \
 (devices.cpu if any(y in cmd_opts.use_cpu for y in [x, 'all']) else devices.get_optimal_device() for x in ['sd', 'interrogate', 'gfpgan', 'swinir', 'esrgan', 'scunet', 'codeformer'])
@@ -135,6 +141,7 @@ class State:
     current_image = None
     current_image_sampling_step = 0
     textinfo = None
+    time_start = None
     need_restart = False
 
     def skip(self):
@@ -144,6 +151,9 @@ class State:
         self.interrupted = True
 
     def nextjob(self):
+        if opts.show_progress_every_n_steps == -1:
+            self.do_set_current_image()
+
         self.job_no += 1
         self.sampling_step = 0
         self.current_image_sampling_step = 0
@@ -172,6 +182,7 @@ class State:
         self.skipped = False
         self.interrupted = False
         self.textinfo = None
+        self.time_start = time.time()
 
         devices.torch_gc()
 
@@ -181,6 +192,24 @@ class State:
 
         devices.torch_gc()
 
+    """sets self.current_image from self.current_latent if enough sampling steps have been made after the last call to this"""
+    def set_current_image(self):
+        if self.sampling_step - self.current_image_sampling_step >= opts.show_progress_every_n_steps and opts.show_progress_every_n_steps > 0:
+            self.do_set_current_image()
+
+    def do_set_current_image(self):
+        if not parallel_processing_allowed:
+            return
+        if self.current_latent is None:
+            return
+
+        if opts.show_progress_grid:
+            self.current_image = sd_samplers.samples_to_image_grid(self.current_latent)
+        else:
+            self.current_image = sd_samplers.sample_to_image(self.current_latent)
+
+        self.current_image_sampling_step = self.sampling_step
+
 state = State()
 
 artist_db = modules.artists.ArtistsDatabase(os.path.join(script_path, 'artists.csv'))
@@ -192,8 +221,6 @@ interrogator = modules.interrogate.InterrogateModels("interrogate")
 
 face_restorers = []
 
-localization.list_localizations(cmd_opts.localizations_dir)
-
 
 def realesrgan_models_names():
     import modules.realesrgan_model
@@ -238,6 +265,8 @@ options_templates.update(options_section(('saving-images', "Saving images/grids"
     "enable_pnginfo": OptionInfo(True, "Save text information about generation parameters as chunks to png files"),
     "save_txt": OptionInfo(False, "Create a text file next to every image with generation parameters."),
     "save_images_before_face_restoration": OptionInfo(False, "Save a copy of image before doing face restoration."),
+    "save_images_before_highres_fix": OptionInfo(False, "Save a copy of image before applying highres fix."),
+    "save_images_before_color_correction": OptionInfo(False, "Save a copy of image before applying color correction to img2img results"),
     "jpeg_quality": OptionInfo(80, "Quality for saved jpeg images", gr.Slider, {"minimum": 1, "maximum": 100, "step": 1}),
     "export_for_4chan": OptionInfo(True, "If PNG image is larger than 4MB or any dimension is larger than 4000, downscale and save copy as JPG"),
 
@@ -290,6 +319,7 @@ options_templates.update(options_section(('system', "System"), {
 
 options_templates.update(options_section(('training', "Training"), {
     "unload_models_when_training": OptionInfo(False, "Move VAE and CLIP to RAM when training if possible. Saves VRAM."),
+    "save_optimizer_state": OptionInfo(False, "Saves Optimizer state as separate *.optim file. Training can be resumed with HN itself and matching optim file."),
     "dataset_filename_word_regex": OptionInfo("", "Filename word regex"),
     "dataset_filename_join_string": OptionInfo(" ", "Filename join string"),
     "training_image_repeats_per_epoch": OptionInfo(1, "Number of repeats for a single input image per epoch; used only for displaying epoch number", gr.Number, {"precision": 0}),
@@ -300,11 +330,11 @@ options_templates.update(options_section(('training', "Training"), {
 options_templates.update(options_section(('sd', "Stable Diffusion"), {
     "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": modules.sd_models.checkpoint_tiles()}, refresh=sd_models.list_models),
     "sd_checkpoint_cache": OptionInfo(0, "Checkpoints to cache in RAM", gr.Slider, {"minimum": 0, "maximum": 10, "step": 1}),
+    "sd_vae": OptionInfo("auto", "SD VAE", gr.Dropdown, lambda: {"choices": list(sd_vae.vae_list)}, refresh=sd_vae.refresh_vae_list),
     "sd_hypernetwork": OptionInfo("None", "Hypernetwork", gr.Dropdown, lambda: {"choices": ["None"] + [x for x in hypernetworks.keys()]}, refresh=reload_hypernetworks),
     "sd_hypernetwork_strength": OptionInfo(1.0, "Hypernetwork strength", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.001}),
     "inpainting_mask_weight": OptionInfo(1.0, "Inpainting conditioning mask strength", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}),
     "img2img_color_correction": OptionInfo(False, "Apply color correction to img2img results to match original colors."),
-    "save_images_before_color_correction": OptionInfo(False, "Save a copy of image before applying color correction to img2img results"),
     "img2img_fix_steps": OptionInfo(False, "With img2img, do exactly the amount of steps the slider specifies (normally you'd do less with less denoising)."),
     "enable_quantization": OptionInfo(False, "Enable quantization in K samplers for sharper and cleaner results. This may change existing seeds. Requires restart to apply."),
     "enable_emphasis": OptionInfo(True, "Emphasis: use (text) to make model pay more attention to text and [text] to make it pay less attention"),
@@ -332,7 +362,7 @@ options_templates.update(options_section(('interrogate', "Interrogate Options"),
 
 options_templates.update(options_section(('ui', "User interface"), {
     "show_progressbar": OptionInfo(True, "Show progressbar"),
-    "show_progress_every_n_steps": OptionInfo(0, "Show image creation progress every N sampling steps. Set 0 to disable.", gr.Slider, {"minimum": 0, "maximum": 32, "step": 1}),
+    "show_progress_every_n_steps": OptionInfo(0, "Show image creation progress every N sampling steps. Set to 0 to disable. Set to -1 to show after completion of batch.", gr.Slider, {"minimum": -1, "maximum": 32, "step": 1}),
     "show_progress_grid": OptionInfo(True, "Show previews of all images generated in a batch as a grid"),
     "return_grid": OptionInfo(True, "Show grid in results for web"),
     "do_not_show_images": OptionInfo(False, "Do not show any images in results for web"),
@@ -377,6 +407,16 @@ class Options:
     def __setattr__(self, key, value):
         if self.data is not None:
             if key in self.data or key in self.data_labels:
+                assert not cmd_opts.freeze_settings, "changing settings is disabled"
+
+                info = opts.data_labels.get(key, None)
+                comp_args = info.component_args if info else None
+                if isinstance(comp_args, dict) and comp_args.get('visible', True) is False:
+                    raise RuntimeError(f"not possible to set {key} because it is restricted")
+
+                if cmd_opts.hide_ui_dir_config and key in restricted_opts:
+                    raise RuntimeError(f"not possible to set {key} because it is restricted")
+
                 self.data[key] = value
                 return
 
@@ -393,6 +433,8 @@ class Options:
         return super(Options, self).__getattribute__(item)
 
     def save(self, filename):
+        assert not cmd_opts.freeze_settings, "saving settings is disabled"
+
         with open(filename, "w", encoding="utf8") as file:
             json.dump(self.data, file, indent=4)
 
@@ -419,11 +461,12 @@ class Options:
         if bad_settings > 0:
             print(f"The program is likely to not work with bad settings.\nSettings file: {filename}\nEither fix the file, or delete it and restart.", file=sys.stderr)
 
-    def onchange(self, key, func):
+    def onchange(self, key, func, call=True):
         item = self.data_labels.get(key)
         item.onchange = func
 
-        func()
+        if call:
+            func()
 
     def dumpjson(self):
         d = {k: self.data.get(k, self.data_labels.get(k).default) for k in self.data_labels.keys()}

modules/txt2img.py

@@ -47,6 +47,8 @@ def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2:
     if processed is None:
         processed = process_images(p)
 
+    p.close()
+
     shared.total_tqdm.clear()
 
     generation_info_js = processed.js()

modules/ui.py

@@ -174,9 +174,9 @@ def save_pil_to_file(pil_image, dir=None):
 gr.processing_utils.save_pil_to_file = save_pil_to_file
 
 
-def wrap_gradio_call(func, extra_outputs=None):
+def wrap_gradio_call(func, extra_outputs=None, add_stats=False):
     def f(*args, extra_outputs_array=extra_outputs, **kwargs):
-        run_memmon = opts.memmon_poll_rate > 0 and not shared.mem_mon.disabled
+        run_memmon = opts.memmon_poll_rate > 0 and not shared.mem_mon.disabled and add_stats
         if run_memmon:
             shared.mem_mon.monitor()
         t = time.perf_counter()
@@ -203,11 +203,18 @@ def wrap_gradio_call(func, extra_outputs=None):
 
             res = extra_outputs_array + [f"<div class='error'>{plaintext_to_html(type(e).__name__+': '+str(e))}</div>"]
 
+        shared.state.skipped = False
+        shared.state.interrupted = False
+        shared.state.job_count = 0
+
+        if not add_stats:
+            return tuple(res)
+
         elapsed = time.perf_counter() - t
         elapsed_m = int(elapsed // 60)
         elapsed_s = elapsed % 60
         elapsed_text = f"{elapsed_s:.2f}s"
-        if (elapsed_m > 0):
+        if elapsed_m > 0:
             elapsed_text = f"{elapsed_m}m "+elapsed_text
 
         if run_memmon:
@@ -225,10 +232,6 @@ def wrap_gradio_call(func, extra_outputs=None):
         # last item is always HTML
         res[-1] += f"<div class='performance'><p class='time'>Time taken: <wbr>{elapsed_text}</p>{vram_html}</div>"
 
-        shared.state.skipped = False
-        shared.state.interrupted = False
-        shared.state.job_count = 0
-
         return tuple(res)
 
     return f
@@ -276,16 +279,8 @@ def check_progress_call(id_part):
     image = gr_show(False)
     preview_visibility = gr_show(False)
 
-    if opts.show_progress_every_n_steps > 0:
-        if shared.parallel_processing_allowed:
-
-            if shared.state.sampling_step - shared.state.current_image_sampling_step >= opts.show_progress_every_n_steps and shared.state.current_latent is not None:
-                if opts.show_progress_grid:
-                    shared.state.current_image = modules.sd_samplers.samples_to_image_grid(shared.state.current_latent)
-                else:
-                    shared.state.current_image = modules.sd_samplers.sample_to_image(shared.state.current_latent)
-                shared.state.current_image_sampling_step = shared.state.sampling_step
-
+    if opts.show_progress_every_n_steps != 0:
+        shared.state.set_current_image()
         image = shared.state.current_image
 
         if image is None:
@@ -671,6 +666,8 @@ def create_ui(wrap_gradio_gpu_call):
     import modules.img2img
     import modules.txt2img
 
+    reload_javascript()
+
     parameters_copypaste.reset()
 
     with gr.Blocks(analytics_enabled=False) as txt2img_interface:
@@ -1058,9 +1055,11 @@ def create_ui(wrap_gradio_gpu_call):
                         extras_batch_output_dir = gr.Textbox(label="Output directory", **shared.hide_dirs, placeholder="Leave blank to save images to the default path.")
                         show_extras_results = gr.Checkbox(label='Show result images', value=True)
 
+                submit = gr.Button('Generate', elem_id="extras_generate", variant='primary')
+
                 with gr.Tabs(elem_id="extras_resize_mode"):
                     with gr.TabItem('Scale by'):
-                        upscaling_resize = gr.Slider(minimum=1.0, maximum=4.0, step=0.05, label="Resize", value=2)
+                        upscaling_resize = gr.Slider(minimum=1.0, maximum=8.0, step=0.05, label="Resize", value=4)
                     with gr.TabItem('Scale to'):
                         with gr.Group():
                             with gr.Row():
@@ -1085,8 +1084,6 @@ def create_ui(wrap_gradio_gpu_call):
                 with gr.Group():
                     upscale_before_face_fix = gr.Checkbox(label='Upscale Before Restoring Faces', value=False)
 
-                submit = gr.Button('Generate', elem_id="extras_generate", variant='primary')
-
             result_images, html_info_x, html_info = create_output_panel("extras", opts.outdir_extras_samples)
 
         submit.click(
@@ -1188,8 +1185,8 @@ def create_ui(wrap_gradio_gpu_call):
                     new_hypernetwork_name = gr.Textbox(label="Name")
                     new_hypernetwork_sizes = gr.CheckboxGroup(label="Modules", value=["768", "320", "640", "1280"], choices=["768", "320", "640", "1280"])
                     new_hypernetwork_layer_structure = gr.Textbox("1, 2, 1", label="Enter hypernetwork layer structure", placeholder="1st and last digit must be 1. ex:'1, 2, 1'")
-                    new_hypernetwork_activation_func = gr.Dropdown(value="linear", label="Select activation function of hypernetwork", choices=modules.hypernetworks.ui.keys)
-                    new_hypernetwork_initialization_option = gr.Dropdown(value = "Normal", label="Select Layer weights initialization. relu-like - Kaiming, sigmoid-like - Xavier is recommended", choices=["Normal", "KaimingUniform", "KaimingNormal", "XavierUniform", "XavierNormal"])
+                    new_hypernetwork_activation_func = gr.Dropdown(value="linear", label="Select activation function of hypernetwork. Recommended : Swish / Linear(none)", choices=modules.hypernetworks.ui.keys)
+                    new_hypernetwork_initialization_option = gr.Dropdown(value = "Normal", label="Select Layer weights initialization. Recommended: Kaiming for relu-like, Xavier for sigmoid-like, Normal otherwise", choices=["Normal", "KaimingUniform", "KaimingNormal", "XavierUniform", "XavierNormal"])
                     new_hypernetwork_add_layer_norm = gr.Checkbox(label="Add layer normalization")
                     new_hypernetwork_use_dropout = gr.Checkbox(label="Use dropout")
                     overwrite_old_hypernetwork = gr.Checkbox(value=False, label="Overwrite Old Hypernetwork")
@@ -1442,50 +1439,41 @@ def create_ui(wrap_gradio_gpu_call):
     opts.reorder()
 
     def run_settings(*args):
-        changed = 0
-
-        assert not shared.cmd_opts.freeze_settings, "changing settings is disabled"
+        changed = []
 
         for key, value, comp in zip(opts.data_labels.keys(), args, components):
-            if comp != dummy_component and not opts.same_type(value, opts.data_labels[key].default):
-                return f"Bad value for setting {key}: {value}; expecting {type(opts.data_labels[key].default).__name__}", opts.dumpjson()
+            assert comp == dummy_component or opts.same_type(value, opts.data_labels[key].default), f"Bad value for setting {key}: {value}; expecting {type(opts.data_labels[key].default).__name__}"
 
         for key, value, comp in zip(opts.data_labels.keys(), args, components):
             if comp == dummy_component:
                 continue
 
-            comp_args = opts.data_labels[key].component_args
-            if comp_args and isinstance(comp_args, dict) and comp_args.get('visible') is False:
-                continue
-
-            if cmd_opts.hide_ui_dir_config and key in restricted_opts:
-                continue
-
             oldval = opts.data.get(key, None)
-            opts.data[key] = value
-
+            try:
+                setattr(opts, key, value)
+            except RuntimeError:
+                continue
             if oldval != value:
                 if opts.data_labels[key].onchange is not None:
                     opts.data_labels[key].onchange()
 
-                changed += 1
-
-        opts.save(shared.config_filename)
-
-        return f'{changed} settings changed.', opts.dumpjson()
+                changed.append(key)
+        try:
+            opts.save(shared.config_filename)
+        except RuntimeError:
+            return opts.dumpjson(), f'{len(changed)} settings changed without save: {", ".join(changed)}.'
+        return opts.dumpjson(), f'{len(changed)} settings changed: {", ".join(changed)}.'
 
     def run_settings_single(value, key):
-        assert not shared.cmd_opts.freeze_settings, "changing settings is disabled"
-
         if not opts.same_type(value, opts.data_labels[key].default):
             return gr.update(visible=True), opts.dumpjson()
 
         oldval = opts.data.get(key, None)
-        if cmd_opts.hide_ui_dir_config and key in restricted_opts:
+        try:
+            setattr(opts, key, value)
+        except Exception:
             return gr.update(value=oldval), opts.dumpjson()
 
-        opts.data[key] = value
-
         if oldval != value:
             if opts.data_labels[key].onchange is not None:
                 opts.data_labels[key].onchange()
@@ -1570,8 +1558,7 @@ def create_ui(wrap_gradio_gpu_call):
         reload_script_bodies.click(
             fn=reload_scripts,
             inputs=[],
-            outputs=[],
-            _js='function(){}'
+            outputs=[]
         )
 
         def request_restart():
@@ -1579,11 +1566,10 @@ def create_ui(wrap_gradio_gpu_call):
             shared.state.need_restart = True
 
         restart_gradio.click(
-
             fn=request_restart,
+            _js='restart_reload',
             inputs=[],
             outputs=[],
-            _js='function(){restart_reload()}'
         )
 
         if column is not None:
@@ -1639,9 +1625,9 @@ def create_ui(wrap_gradio_gpu_call):
 
         text_settings = gr.Textbox(elem_id="settings_json", value=lambda: opts.dumpjson(), visible=False)
         settings_submit.click(
-            fn=run_settings,
+            fn=wrap_gradio_call(run_settings, extra_outputs=[gr.update()]),
             inputs=components,
-            outputs=[result, text_settings],
+            outputs=[text_settings, result],
         )
 
         for i, k, item in quicksettings_list:
@@ -1782,4 +1768,3 @@ def load_javascript(raw_response):
 
 
 reload_javascript = partial(load_javascript, gradio.routes.templates.TemplateResponse)
-reload_javascript()

modules/ui_extensions.py

@@ -86,7 +86,7 @@ def extension_table():
         code += f"""
             <tr>
                 <td><label><input class="gr-check-radio gr-checkbox" name="enable_{html.escape(ext.name)}" type="checkbox" {'checked="checked"' if ext.enabled else ''}>{html.escape(ext.name)}</label></td>
-                <td><a href="{html.escape(ext.remote or '')}">{html.escape(ext.remote or '')}</a></td>
+                <td><a href="{html.escape(ext.remote or '')}" target="_blank">{html.escape(ext.remote or '')}</a></td>
                 <td{' class="extension_status"' if ext.remote is not None else ''}>{ext_status}</td>
             </tr>
     """
@@ -140,13 +140,15 @@ def install_extension_from_url(dirname, url):
         shutil.rmtree(tmpdir, True)
 
 
-def install_extension_from_index(url):
+def install_extension_from_index(url, hide_tags):
     ext_table, message = install_extension_from_url(None, url)
 
-    return refresh_available_extensions_from_data(), ext_table, message
+    code, _ = refresh_available_extensions_from_data(hide_tags)
+
+    return code, ext_table, message
 
 
-def refresh_available_extensions(url):
+def refresh_available_extensions(url, hide_tags):
     global available_extensions
 
     import urllib.request
@@ -155,13 +157,25 @@ def refresh_available_extensions(url):
 
     available_extensions = json.loads(text)
 
-    return url, refresh_available_extensions_from_data(), ''
+    code, tags = refresh_available_extensions_from_data(hide_tags)
+
+    return url, code, gr.CheckboxGroup.update(choices=tags), ''
 
 
-def refresh_available_extensions_from_data():
+def refresh_available_extensions_for_tags(hide_tags):
+    code, _ = refresh_available_extensions_from_data(hide_tags)
+
+    return code, ''
+
+
+def refresh_available_extensions_from_data(hide_tags):
     extlist = available_extensions["extensions"]
     installed_extension_urls = {normalize_git_url(extension.remote): extension.name for extension in extensions.extensions}
 
+    tags = available_extensions.get("tags", {})
+    tags_to_hide = set(hide_tags)
+    hidden = 0
+
     code = f"""<!-- {time.time()} -->
     <table id="available_extensions">
         <thead>
@@ -178,17 +192,24 @@ def refresh_available_extensions_from_data():
         name = ext.get("name", "noname")
         url = ext.get("url", None)
         description = ext.get("description", "")
+        extension_tags = ext.get("tags", [])
 
         if url is None:
             continue
 
+        if len([x for x in extension_tags if x in tags_to_hide]) > 0:
+            hidden += 1
+            continue
+
         existing = installed_extension_urls.get(normalize_git_url(url), None)
 
         install_code = f"""<input onclick="install_extension_from_index(this, '{html.escape(url)}')" type="button" value="{"Install" if not existing else "Installed"}" {"disabled=disabled" if existing else ""} class="gr-button gr-button-lg gr-button-secondary">"""
 
+        tags_text = ", ".join([f"<span class='extension-tag' title='{tags.get(x, '')}'>{x}</span>" for x in extension_tags])
+
         code += f"""
             <tr>
-                <td><a href="{html.escape(url)}">{html.escape(name)}</a></td>
+                <td><a href="{html.escape(url)}" target="_blank">{html.escape(name)}</a><br />{tags_text}</td>
                 <td>{html.escape(description)}</td>
                 <td>{install_code}</td>
             </tr>
@@ -199,7 +220,10 @@ def refresh_available_extensions_from_data():
     </table>
     """
 
-    return code
+    if hidden > 0:
+        code += f"<p>Extension hidden: {hidden}</p>"
+
+    return code, list(tags)
 
 
 def create_ui():
@@ -238,21 +262,30 @@ def create_ui():
                     extension_to_install = gr.Text(elem_id="extension_to_install", visible=False)
                     install_extension_button = gr.Button(elem_id="install_extension_button", visible=False)
 
+                with gr.Row():
+                    hide_tags = gr.CheckboxGroup(value=["ads", "localization"], label="Hide extensions with tags", choices=["script", "ads", "localization"])
+
                 install_result = gr.HTML()
                 available_extensions_table = gr.HTML()
 
                 refresh_available_extensions_button.click(
-                    fn=modules.ui.wrap_gradio_call(refresh_available_extensions, extra_outputs=[gr.update(), gr.update()]),
-                    inputs=[available_extensions_index],
-                    outputs=[available_extensions_index, available_extensions_table, install_result],
+                    fn=modules.ui.wrap_gradio_call(refresh_available_extensions, extra_outputs=[gr.update(), gr.update(), gr.update()]),
+                    inputs=[available_extensions_index, hide_tags],
+                    outputs=[available_extensions_index, available_extensions_table, hide_tags, install_result],
                 )
 
                 install_extension_button.click(
                     fn=modules.ui.wrap_gradio_call(install_extension_from_index, extra_outputs=[gr.update(), gr.update()]),
-                    inputs=[extension_to_install],
+                    inputs=[extension_to_install, hide_tags],
                     outputs=[available_extensions_table, extensions_table, install_result],
                 )
 
+                hide_tags.change(
+                    fn=modules.ui.wrap_gradio_call(refresh_available_extensions_for_tags, extra_outputs=[gr.update()]),
+                    inputs=[hide_tags],
+                    outputs=[available_extensions_table, install_result]
+                )
+
             with gr.TabItem("Install from URL"):
                 install_url = gr.Text(label="URL for extension's git repository")
                 install_dirname = gr.Text(label="Local directory name", placeholder="Leave empty for auto")

modules/upscaler.py

@@ -10,6 +10,7 @@ import modules.shared
 from modules import modelloader, shared
 
 LANCZOS = (Image.Resampling.LANCZOS if hasattr(Image, 'Resampling') else Image.LANCZOS)
+NEAREST = (Image.Resampling.NEAREST if hasattr(Image, 'Resampling') else Image.NEAREST)
 from modules.paths import models_path
 
 
@@ -56,10 +57,18 @@ class Upscaler:
         self.scale = scale
         dest_w = img.width * scale
         dest_h = img.height * scale
+
         for i in range(3):
+            shape = (img.width, img.height)
+
+            img = self.do_upscale(img, selected_model)
+
+            if shape == (img.width, img.height):
+                break
+
             if img.width >= dest_w and img.height >= dest_h:
                 break
-            img = self.do_upscale(img, selected_model)
+
         if img.width != dest_w or img.height != dest_h:
             img = img.resize((int(dest_w), int(dest_h)), resample=LANCZOS)
 
@@ -120,3 +129,17 @@ class UpscalerLanczos(Upscaler):
         self.name = "Lanczos"
         self.scalers = [UpscalerData("Lanczos", None, self)]
 
+
+class UpscalerNearest(Upscaler):
+    scalers = []
+
+    def do_upscale(self, img, selected_model=None):
+        return img.resize((int(img.width * self.scale), int(img.height * self.scale)), resample=NEAREST)
+
+    def load_model(self, _):
+        pass
+
+    def __init__(self, dirname=None):
+        super().__init__(False)
+        self.name = "Nearest"
+        self.scalers = [UpscalerData("Nearest", None, self)]