Merge branch 'main_upstream' into hotfix/prevent-skipping-and-stopping-by-other-users

# Conflicts:
#	webui.py

modules/advanced_parameters.py

@@ -5,7 +5,8 @@ disable_preview, adm_scaler_positive, adm_scaler_negative, adm_scaler_end, adapt
     debugging_cn_preprocessor, skipping_cn_preprocessor, controlnet_softness, canny_low_threshold, canny_high_threshold, \
     refiner_swap_method, \
     freeu_enabled, freeu_b1, freeu_b2, freeu_s1, freeu_s2, \
-    debugging_inpaint_preprocessor, inpaint_disable_initial_latent, inpaint_engine, inpaint_strength, inpaint_respective_field = [None] * 32
+    debugging_inpaint_preprocessor, inpaint_disable_initial_latent, inpaint_engine, inpaint_strength, inpaint_respective_field, \
+    inpaint_mask_upload_checkbox, invert_mask_checkbox, inpaint_erode_or_dilate = [None] * 35
 
 
 def set_all_advanced_parameters(*args):
@@ -16,7 +17,8 @@ def set_all_advanced_parameters(*args):
         debugging_cn_preprocessor, skipping_cn_preprocessor, controlnet_softness, canny_low_threshold, canny_high_threshold, \
         refiner_swap_method, \
         freeu_enabled, freeu_b1, freeu_b2, freeu_s1, freeu_s2, \
-        debugging_inpaint_preprocessor, inpaint_disable_initial_latent, inpaint_engine, inpaint_strength, inpaint_respective_field
+        debugging_inpaint_preprocessor, inpaint_disable_initial_latent, inpaint_engine, inpaint_strength, inpaint_respective_field, \
+        inpaint_mask_upload_checkbox, invert_mask_checkbox, inpaint_erode_or_dilate
 
     disable_preview, adm_scaler_positive, adm_scaler_negative, adm_scaler_end, adaptive_cfg, sampler_name, \
         scheduler_name, generate_image_grid, overwrite_step, overwrite_switch, overwrite_width, overwrite_height, \
@@ -25,6 +27,7 @@ def set_all_advanced_parameters(*args):
         debugging_cn_preprocessor, skipping_cn_preprocessor, controlnet_softness, canny_low_threshold, canny_high_threshold, \
         refiner_swap_method, \
         freeu_enabled, freeu_b1, freeu_b2, freeu_s1, freeu_s2, \
-        debugging_inpaint_preprocessor, inpaint_disable_initial_latent, inpaint_engine, inpaint_strength, inpaint_respective_field = args
+        debugging_inpaint_preprocessor, inpaint_disable_initial_latent, inpaint_engine, inpaint_strength, inpaint_respective_field, \
+        inpaint_mask_upload_checkbox, invert_mask_checkbox, inpaint_erode_or_dilate = args
 
     return

modules/async_worker.py

@@ -42,7 +42,7 @@ def worker():
     from modules.private_logger import log
     from extras.expansion import safe_str
     from modules.util import remove_empty_str, HWC3, resize_image, \
-        get_image_shape_ceil, set_image_shape_ceil, get_shape_ceil, resample_image
+        get_image_shape_ceil, set_image_shape_ceil, get_shape_ceil, resample_image, erode_or_dilate
     from modules.upscaler import perform_upscale
 
     try:
@@ -142,6 +142,7 @@ def worker():
         outpaint_selections = args.pop()
         inpaint_input_image = args.pop()
         inpaint_additional_prompt = args.pop()
+        inpaint_mask_image_upload = args.pop()
 
         cn_tasks = {x: [] for x in flags.ip_list}
         for _ in range(4):
@@ -277,6 +278,22 @@ def worker():
                     and isinstance(inpaint_input_image, dict):
                 inpaint_image = inpaint_input_image['image']
                 inpaint_mask = inpaint_input_image['mask'][:, :, 0]
+                
+                if advanced_parameters.inpaint_mask_upload_checkbox:
+                    if isinstance(inpaint_mask_image_upload, np.ndarray):
+                        if inpaint_mask_image_upload.ndim == 3:
+                            H, W, C = inpaint_image.shape
+                            inpaint_mask_image_upload = resample_image(inpaint_mask_image_upload, width=W, height=H)
+                            inpaint_mask_image_upload = np.mean(inpaint_mask_image_upload, axis=2)
+                            inpaint_mask_image_upload = (inpaint_mask_image_upload > 127).astype(np.uint8) * 255
+                            inpaint_mask = np.maximum(inpaint_mask, inpaint_mask_image_upload)
+
+                if int(advanced_parameters.inpaint_erode_or_dilate) != 0:
+                    inpaint_mask = erode_or_dilate(inpaint_mask, advanced_parameters.inpaint_erode_or_dilate)
+
+                if advanced_parameters.invert_mask_checkbox:
+                    inpaint_mask = 255 - inpaint_mask
+
                 inpaint_image = HWC3(inpaint_image)
                 if isinstance(inpaint_image, np.ndarray) and isinstance(inpaint_mask, np.ndarray) \
                         and (np.any(inpaint_mask > 127) or len(outpaint_selections) > 0):

modules/config.py

@@ -243,10 +243,15 @@ default_advanced_checkbox = get_config_item_or_set_default(
     default_value=False,
     validator=lambda x: isinstance(x, bool)
 )
+default_max_image_number = get_config_item_or_set_default(
+    key='default_max_image_number',
+    default_value=32,
+    validator=lambda x: isinstance(x, int) and x >= 1
+)
 default_image_number = get_config_item_or_set_default(
     key='default_image_number',
     default_value=2,
-    validator=lambda x: isinstance(x, int) and 1 <= x <= 32
+    validator=lambda x: isinstance(x, int) and 1 <= x <= default_max_image_number
 )
 checkpoint_downloads = get_config_item_or_set_default(
     key='checkpoint_downloads',

modules/meta_parser.py

@@ -3,7 +3,7 @@ import gradio as gr
 import modules.config
 
 
-def load_parameter_button_click(raw_prompt_txt):
+def load_parameter_button_click(raw_prompt_txt, is_generating):
     loaded_parameter_dict = json.loads(raw_prompt_txt)
     assert isinstance(loaded_parameter_dict, dict)
 
@@ -128,7 +128,11 @@ def load_parameter_button_click(raw_prompt_txt):
         results.append(gr.update())
         results.append(gr.update())
 
-    results.append(gr.update(visible=True))
+    if is_generating:
+        results.append(gr.update())
+    else:
+        results.append(gr.update(visible=True))
+    
     results.append(gr.update(visible=False))
 
     for i in range(1, 6):

modules/ops.py

@@ -0,0 +1,19 @@
+import torch
+import contextlib
+
+
+@contextlib.contextmanager
+def use_patched_ops(operations):
+    op_names = ['Linear', 'Conv2d', 'Conv3d', 'GroupNorm', 'LayerNorm']
+    backups = {op_name: getattr(torch.nn, op_name) for op_name in op_names}
+
+    try:
+        for op_name in op_names:
+            setattr(torch.nn, op_name, getattr(operations, op_name))
+
+        yield
+
+    finally:
+        for op_name in op_names:
+            setattr(torch.nn, op_name, backups[op_name])
+    return

modules/patch.py

@@ -218,7 +218,7 @@ def compute_cfg(uncond, cond, cfg_scale, t):
 def patched_sampling_function(model, x, timestep, uncond, cond, cond_scale, model_options=None, seed=None):
     global eps_record
 
-    if math.isclose(cond_scale, 1.0):
+    if math.isclose(cond_scale, 1.0) and not model_options.get("disable_cfg1_optimization", False):
         final_x0 = calc_cond_uncond_batch(model, cond, None, x, timestep, model_options)[0]
 
         if eps_record is not None:
@@ -480,6 +480,10 @@ def build_loaded(module, loader_name):
 
 
 def patch_all():
+    if ldm_patched.modules.model_management.directml_enabled:
+        ldm_patched.modules.model_management.lowvram_available = True
+        ldm_patched.modules.model_management.OOM_EXCEPTION = Exception
+    
     patch_all_precision()
     patch_all_clip()
 

modules/patch_clip.py

@@ -16,30 +16,12 @@ import ldm_patched.modules.samplers
 import ldm_patched.modules.sd
 import ldm_patched.modules.sd1_clip
 import ldm_patched.modules.clip_vision
-import ldm_patched.modules.model_management as model_management
 import ldm_patched.modules.ops as ops
-import contextlib
 
+from modules.ops import use_patched_ops
 from transformers import CLIPTextModel, CLIPTextConfig, modeling_utils, CLIPVisionConfig, CLIPVisionModelWithProjection
 
 
-@contextlib.contextmanager
-def use_patched_ops(operations):
-    op_names = ['Linear', 'Conv2d', 'Conv3d', 'GroupNorm', 'LayerNorm']
-    backups = {op_name: getattr(torch.nn, op_name) for op_name in op_names}
-
-    try:
-        for op_name in op_names:
-            setattr(torch.nn, op_name, getattr(operations, op_name))
-
-        yield
-
-    finally:
-        for op_name in op_names:
-            setattr(torch.nn, op_name, backups[op_name])
-    return
-
-
 def patched_encode_token_weights(self, token_weight_pairs):
     to_encode = list()
     max_token_len = 0

modules/private_logger.py

@@ -44,13 +44,28 @@ def log(img, dic):
     )
 
     js = (
-        "<script>"
-        "function to_clipboard(txt) { "
-        "txt = decodeURIComponent(txt);"
-        "navigator.clipboard.writeText(txt);"
-        "alert('Copied to Clipboard!\\nPaste to prompt area to load parameters.\\nCurrent clipboard content is:\\n\\n' + txt);"
-        "}"
-        "</script>"
+        """<script>
+        function to_clipboard(txt) { 
+        txt = decodeURIComponent(txt);
+        if (navigator.clipboard && navigator.permissions) {
+            navigator.clipboard.writeText(txt)
+        } else {
+            const textArea = document.createElement('textArea')
+            textArea.value = txt
+            textArea.style.width = 0
+            textArea.style.position = 'fixed'
+            textArea.style.left = '-999px'
+            textArea.style.top = '10px'
+            textArea.setAttribute('readonly', 'readonly')
+            document.body.appendChild(textArea)
+
+            textArea.select()
+            document.execCommand('copy')
+            document.body.removeChild(textArea)
+        }
+        alert('Copied to Clipboard!\\nPaste to prompt area to load parameters.\\nCurrent clipboard content is:\\n\\n' + txt);
+        }
+        </script>"""
     )
 
     begin_part = f"<html><head><title>Fooocus Log {date_string}</title>{css_styles}</head><body>{js}<p>Fooocus Log {date_string} (private)</p>\n<p>All images are clean, without any hidden data/meta, and safe to share with others.</p><!--fooocus-log-split-->\n\n"

modules/sample_hijack.py

@@ -99,6 +99,13 @@ def sample_hacked(model, noise, positive, negative, cfg, device, sampler, sigmas
     calculate_start_end_timesteps(model, negative)
     calculate_start_end_timesteps(model, positive)
 
+    if latent_image is not None:
+        latent_image = model.process_latent_in(latent_image)
+
+    if hasattr(model, 'extra_conds'):
+        positive = encode_model_conds(model.extra_conds, positive, noise, device, "positive", latent_image=latent_image, denoise_mask=denoise_mask)
+        negative = encode_model_conds(model.extra_conds, negative, noise, device, "negative", latent_image=latent_image, denoise_mask=denoise_mask)
+
     #make sure each cond area has an opposite one with the same area
     for c in positive:
         create_cond_with_same_area_if_none(negative, c)
@@ -111,13 +118,6 @@ def sample_hacked(model, noise, positive, negative, cfg, device, sampler, sigmas
     apply_empty_x_to_equal_area(list(filter(lambda c: c.get('control_apply_to_uncond', False) == True, positive)), negative, 'control', lambda cond_cnets, x: cond_cnets[x])
     apply_empty_x_to_equal_area(positive, negative, 'gligen', lambda cond_cnets, x: cond_cnets[x])
 
-    if latent_image is not None:
-        latent_image = model.process_latent_in(latent_image)
-
-    if hasattr(model, 'extra_conds'):
-        positive = encode_model_conds(model.extra_conds, positive, noise, device, "positive", latent_image=latent_image, denoise_mask=denoise_mask)
-        negative = encode_model_conds(model.extra_conds, negative, noise, device, "negative", latent_image=latent_image, denoise_mask=denoise_mask)
-
     extra_args = {"cond":positive, "uncond":negative, "cond_scale": cfg, "model_options": model_options, "seed":seed}
 
     if current_refiner is not None and hasattr(current_refiner.model, 'extra_conds'):
@@ -174,7 +174,7 @@ def calculate_sigmas_scheduler_hacked(model, scheduler_name, steps):
     elif scheduler_name == "sgm_uniform":
         sigmas = normal_scheduler(model, steps, sgm=True)
     elif scheduler_name == "turbo":
-        sigmas = SDTurboScheduler().get_sigmas(namedtuple('Patcher', ['model'])(model=model), steps)[0]
+        sigmas = SDTurboScheduler().get_sigmas(namedtuple('Patcher', ['model'])(model=model), steps=steps, denoise=1.0)[0]
     else:
         raise TypeError("error invalid scheduler")
     return sigmas

modules/util.py

@@ -3,6 +3,7 @@ import datetime
 import random
 import math
 import os
+import cv2
 
 from PIL import Image
 
@@ -10,6 +11,15 @@ from PIL import Image
 LANCZOS = (Image.Resampling.LANCZOS if hasattr(Image, 'Resampling') else Image.LANCZOS)
 
 
+def erode_or_dilate(x, k):
+    k = int(k)
+    if k > 0:
+        return cv2.dilate(x, kernel=np.ones(shape=(3, 3), dtype=np.uint8), iterations=k)
+    if k < 0:
+        return cv2.erode(x, kernel=np.ones(shape=(3, 3), dtype=np.uint8), iterations=-k)
+    return x
+
+
 def resample_image(im, width, height):
     im = Image.fromarray(im)
     im = im.resize((int(width), int(height)), resample=LANCZOS)