Upload sd-webui-chunk-weights using SD-Hub

.gitattributes

@@ -181,3 +181,7 @@ sd-webui-llul/images/sample3.jpg filter=lfs diff=lfs merge=lfs -text
 sd-webui-llul/images/sample4.jpg filter=lfs diff=lfs merge=lfs -text
 asymmetric-tiling-sd-webui-2.0/Руководство[[:space:]]Advanced[[:space:]]Tiling[[:space:]]v3.0[[:space:]]-[[:space:]]Новые[[:space:]]функции.pdf filter=lfs diff=lfs merge=lfs -text
 asymmetric-tiling-sd-webui-2.0/Руководство[[:space:]]по[[:space:]]новым[[:space:]]функциям.pdf filter=lfs diff=lfs merge=lfs -text
+sd-webui-chunk-weights/example/1.jpg filter=lfs diff=lfs merge=lfs -text
+sd-webui-chunk-weights/example/2.jpg filter=lfs diff=lfs merge=lfs -text
+sd-webui-chunk-weights/example/3.jpg filter=lfs diff=lfs merge=lfs -text
+sd-webui-chunk-weights/example/4.jpg filter=lfs diff=lfs merge=lfs -text

sd-webui-chunk-weights/.gitignore

@@ -0,0 +1 @@
+__pycache__

sd-webui-chunk-weights/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Haoming
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

sd-webui-chunk-weights/README.md

@@ -0,0 +1,41 @@
+# SD Forge Chunk Weights
+This is an Extension for [Forge Classic](https://github.com/Haoming02/sd-webui-forge-classic), which allows you to control the weighting for each chunk of prompts *(**i.e.** every 75 tokens)*.
+
+> [!Tip]
+> In the WebUI, you can use the keyword **`BREAK`** to manually separate prompts into different chunks to group similar concepts together
+
+## How to Use
+In the `Weighting` text field, enter a list of **comma-separated floats**, corresponding to the weights of each chunk in order *(the default weight is `1.0`)*
+
+## Examples
+
+<table>
+    <tr align="center">
+        <td>
+            <img src="./example/1.jpg" width=384><br>
+            a photo of a dog, a house<br>
+            <b>Extension:</b> <code>Disabled</code>
+        </td>
+        <td>
+            <img src="./example/2.jpg" width=384><br>
+            a photo of a dog, BREAK, a house<br>
+            <b>Extension:</b> <code>Disabled</code>
+        </td>
+    </tr>
+    <tr align="center">
+        <td>
+            <img src="./example/3.jpg" width=384><br>
+            a photo of a dog, BREAK, a house<br>
+            <b>Weights:</b> <code>1.5, 1.0</code>
+        </td>
+        <td>
+            <img src="./example/4.jpg" width=384><br>
+            a photo of a dog, BREAK, a house<br>
+            <b>Weights:</b> <code>1.0, 1.5</code>
+        </td>
+    </tr>
+</table>
+
+<hr>
+
+- Idea by. **[@jeanhadrien](https://github.com/jeanhadrien)** in [#89](https://github.com/Haoming02/sd-webui-forge-classic/issues/89), based on this [Extension](https://github.com/klimaleksus/stable-diffusion-webui-embedding-merge/)

sd-webui-chunk-weights/scripts/chunk_weighting.py

@@ -0,0 +1,242 @@
+import logging
+import sys
+from functools import wraps
+from typing import List
+
+import gradio as gr
+from modules import scripts
+from modules.infotext_utils import PasteField
+from modules.processing import StableDiffusionProcessing, StableDiffusionProcessingTxt2Img
+from modules.script_callbacks import on_app_started, on_script_unloaded
+from modules.ui_components import InputAccordion
+
+# ==============================================================================
+# ЧАСТЬ 1: ЛОГГЕР (Встроено, чтобы избежать ошибок импорта)
+# ==============================================================================
+class ColorCode:
+    RESET = "\033[0m"
+    BLACK = "\033[0;90m"
+    CYAN = "\033[0;36m"
+    YELLOW = "\033[0;33m"
+    RED = "\033[0;31m"
+
+    MAP = {
+        "DEBUG": BLACK,
+        "INFO": CYAN,
+        "WARNING": YELLOW,
+        "ERROR": RED,
+    }
+
+class ColoredFormatter(logging.Formatter):
+    def format(self, record):
+        levelname = record.levelname
+        if levelname in ColorCode.MAP:
+            record.levelname = f"{ColorCode.MAP[levelname]}{levelname}{ColorCode.RESET}"
+        return super().format(record)
+
+logger = logging.getLogger("ChunkWeight")
+logger.setLevel(logging.INFO)
+logger.propagate = False
+
+if not logger.handlers:
+    handler = logging.StreamHandler(sys.stdout)
+    handler.setFormatter(ColoredFormatter("[%(name)s] %(levelname)s - %(message)s"))
+    logger.addHandler(handler)
+
+# ==============================================================================
+# ЧАСТЬ 2: СКРИПТ (Логика весов)
+# ==============================================================================
+
+# Поиск целевого класса для патчинга.
+# Мы ищем базовый класс, от которого наследуются и SD1.5, и SDXL эмбеддеры.
+target_classes = []
+
+try:
+    from modules.sd_hijack_clip import FrozenCLIPEmbedderWithCustomWordsBase
+    target_classes.append(FrozenCLIPEmbedderWithCustomWordsBase)
+except ImportError:
+    # Фолбэк для очень старых версий или если архитектура изменится
+    logger.warning("Base class FrozenCLIPEmbedderWithCustomWordsBase not found. Trying individual classes.")
+    try:
+        from modules.sd_hijack_clip import FrozenCLIPEmbedderWithCustomWords
+        target_classes.append(FrozenCLIPEmbedderWithCustomWords)
+    except ImportError: pass
+    
+    try:
+        from modules.sd_hijack_open_clip import FrozenOpenCLIPEmbedderWithCustomWords
+        target_classes.append(FrozenOpenCLIPEmbedderWithCustomWords)
+    except ImportError: pass
+
+IS_NEGATIVE_PROMPT: bool = False
+WEIGHTS: List[float] = []
+INDEX: int = 0
+PATCHED = False
+
+class ChunkWeight(scripts.Script):
+    _error_logged: bool = False
+
+    def title(self):
+        return "Chunk Weight"
+
+    def show(self, is_img2img):
+        return scripts.AlwaysVisible
+
+    def ui(self, is_img2img):
+        with InputAccordion(False, label=self.title()) as enable:
+            weights = gr.Textbox(
+                label="Weighting (comma separated floats)",
+                placeholder="1.0, 1.5, 0.8",
+                value="",
+                lines=1,
+                max_lines=1,
+            )
+            weights.do_not_save_to_config = True
+
+        self.infotext_fields = [PasteField(weights, "Chunk Weights")]
+        return [enable, weights]
+
+    def setup(self, p: "StableDiffusionProcessing", enable: bool, weights: str):
+        WEIGHTS.clear()
+        ChunkWeight._error_logged = False
+
+        if not enable:
+            return
+
+        # Парсинг весов
+        for v in weights.split(","):
+            v = v.strip()
+            if not v: continue
+            try:
+                WEIGHTS.append(float(v))
+            except ValueError:
+                logger.error(f'Failed to parse "{v}" as number...')
+                continue
+
+        p.extra_generation_params["Chunk Weights"] = ", ".join(str(v) for v in WEIGHTS)
+
+        # Сброс кэшей A1111. Это критически важно, иначе изменение весов
+        # не применится к повторным генерациям с тем же промптом.
+        p.cached_c = [None, None]
+        p.cached_uc = [None, None]
+        
+        # Поддержка Hires Fix (Highres. fix кэширует свои кондишены отдельно)
+        if hasattr(p, 'cached_hr_c'):
+            p.cached_hr_c = [None, None]
+            p.cached_hr_uc = [None, None]
+
+    def postprocess(self, *args):
+        # Очистка глобальных кэшей класса после генерации для надежности
+        StableDiffusionProcessing.cached_c = [None, None]
+        StableDiffusionProcessing.cached_uc = [None, None]
+        StableDiffusionProcessingTxt2Img.cached_hr_c = [None, None]
+        StableDiffusionProcessingTxt2Img.cached_hr_uc = [None, None]
+
+
+# ==============================================================================
+# ЧАСТЬ 3: ПАТЧИНГ (Внедрение в ядро)
+# ==============================================================================
+
+def patch_embedder(cls):
+    """Патчит класс Embedder, добавляя умножение весов."""
+    
+    if hasattr(cls, '_chunk_weight_patched'):
+        return
+
+    original_process_texts = cls.process_texts
+    original_process_tokens = cls.process_tokens
+
+    @wraps(original_process_texts)
+    def patched_process_texts(self, texts: List[str]):
+        global IS_NEGATIVE_PROMPT, INDEX
+
+        # A1111 передает флаг is_negative_prompt как атрибут списка texts.
+        # Это работает благодаря классу SdConditioning в modules/prompt_parser.py
+        if hasattr(texts, "is_negative_prompt"):
+            IS_NEGATIVE_PROMPT = texts.is_negative_prompt
+        else:
+            # Если атрибут не найден (редкий случай), считаем это позитивным промптом
+            # чтобы веса применились.
+            IS_NEGATIVE_PROMPT = False
+
+        # Сбрасываем индекс чанка перед обработкой нового текста
+        INDEX = 0
+        return original_process_texts(self, texts)
+
+    @wraps(original_process_tokens)
+    def patched_process_tokens(self, remade_batch_tokens: list, batch_multipliers: list):
+        global INDEX, WEIGHTS, IS_NEGATIVE_PROMPT
+
+        # Если веса не заданы пользователем, работаем как оригинальный метод
+        if not WEIGHTS:
+            return original_process_tokens(self, remade_batch_tokens, batch_multipliers)
+
+        # Применяем веса только к позитивным промптам
+        if INDEX >= 0 and not IS_NEGATIVE_PROMPT:
+            batches: int = len(batch_multipliers)
+            
+            if INDEX < len(WEIGHTS):
+                current_weight = WEIGHTS[INDEX]
+                logger.debug(f"Applying weight {current_weight}x to Chunk {INDEX}")
+                
+                for b in range(batches):
+                    # batch_multipliers[b] - это список множителей (по умолчанию 1.0) для каждого токена
+                    for i in range(len(batch_multipliers[b])):
+                         batch_multipliers[b][i] *= current_weight
+            else:
+                 # Если весов меньше, чем чанков, остальные получают 1.0 (стандарт)
+                 if not ChunkWeight._error_logged:
+                    logger.warning(f"Not enough weights provided! Chunk {INDEX} uses default weight 1.0.")
+                    ChunkWeight._error_logged = True
+            
+            # Увеличиваем индекс, так как process_tokens вызывается для каждого чанка (BREAK) по очереди
+            INDEX += 1
+
+        return original_process_tokens(self, remade_batch_tokens, batch_multipliers)
+
+    # Применяем подмену методов
+    cls.process_texts = patched_process_texts
+    cls.process_tokens = patched_process_tokens
+    cls._chunk_weight_patched = True
+    
+    # Сохраняем оригиналы для возможности отключения
+    cls._original_process_texts = original_process_texts
+    cls._original_process_tokens = original_process_tokens
+
+
+def apply_patches():
+    global PATCHED
+    if PATCHED: return
+    
+    patched_count = 0
+    for cls in target_classes:
+        try:
+            patch_embedder(cls)
+            patched_count += 1
+            logger.info(f"Chunk Weight logic attached to {cls.__name__}.")
+        except Exception as e:
+            logger.error(f"Failed to patch {cls}: {e}")
+            
+    if patched_count > 0:
+        PATCHED = True
+    else:
+        logger.error("COULD NOT FIND EMBEDDER CLASSES. Chunk Weights will not work.")
+
+
+def remove_patches():
+    global PATCHED
+    if not PATCHED: return
+    
+    for cls in target_classes:
+        if hasattr(cls, '_chunk_weight_patched'):
+            cls.process_texts = cls._original_process_texts
+            cls.process_tokens = cls._original_process_tokens
+            del cls._chunk_weight_patched
+            del cls._original_process_texts
+            del cls._original_process_tokens
+    
+    PATCHED = False
+    logger.info("Chunk Weight patches removed.")
+
+# Регистрируем коллбэки для автозапуска
+on_app_started(apply_patches)
+on_script_unloaded(remove_patches)