From 8b0763a228ef0fd6d076d3f2024130fe613c5753 Mon Sep 17 00:00:00 2001 From: Jaylen Lee Date: Tue, 19 Mar 2024 19:30:21 +0800 Subject: [PATCH 1/8] ui&bugs fix --- README.md | 23 +++++++++- README_CN.md | 23 +++++++++- scripts/tileglobal.py | 90 +++++++++++++++----------------------- tile_methods/demofusion.py | 10 ++--- 4 files changed, 84 insertions(+), 62 deletions(-) diff --git a/README.md b/README.md index e4db6af..44a47e1 100644 --- a/README.md +++ b/README.md @@ -18,7 +18,7 @@ If you like the project, please give me a star! ⭐ The extension enables **large image drawing & upscaling with limited VRAM** via the following techniques: -1. Two SOTA diffusion tiling algorithms: [Mixture of Diffusers](https://github.com/albarji/mixture-of-diffusers) and [MultiDiffusion](https://multidiffusion.github.io), add [Demofusion](https://github.com/PRIS-CV/DemoFusion) +1. SOTA diffusion tiling algorithms: [Mixture of Diffusers](https://github.com/albarji/mixture-of-diffusers) and [MultiDiffusion](https://multidiffusion.github.io), add [Demofusion](https://github.com/PRIS-CV/DemoFusion) 2. My original Tiled VAE algorithm. 3. My original TIled Noise Inversion for better upscaling. @@ -32,6 +32,7 @@ The extension enables **large image drawing & upscaling with limited VRAM** via - [x] [Regional Prompt Control](#region-prompt-control) - [x] [Img2img upscale](#img2img-upscale) - [x] [Ultra-Large image generation](#ultra-large-image-generation) +- [x] [Demofusion available](#Demofusion available) => Quickstart Tutorial: [Tutorial for multidiffusion upscaler for automatic1111](https://civitai.com/models/34726), thanks to [@PotatoBananaApple](https://github.com/pkuliyi2015/multidiffusion-upscaler-for-automatic1111/discussions/120) 🎉 @@ -186,6 +187,26 @@ The extension enables **large image drawing & upscaling with limited VRAM** via **** +### Demofusion available + +ℹ Please set seed in for a better result. + +ℹ Ticking the random jitter usually yields better results. + +ℹ Recommend using higher steps, such as 30 or more, for better results + +ℹ If you set the image size to 512 * 512, the appropriate window size and overlap are 64 and 32 or smaller. If it is 1024, it is recommended to double it. + +ℹ Recommend using a higher denoising strength in img2img，and try to use the original model, seeds, and prompt as much as possible + +ℹ Do not enable it together with tilediffusion. It supports operations such as tilevae, noise inversion, etc. + +ℹFor parameters such as c1, c2, and c3, please refer to the demofusion. You don't have to adjust it. + +**** + +### + ## Installation ⚪ Method 1: Official Market diff --git a/README_CN.md b/README_CN.md index 7a807db..e993662 100644 --- a/README_CN.md +++ b/README_CN.md @@ -1,4 +1,4 @@ -# 用 Tiled Diffusion & VAE 生成大型图像 +用 Tiled Diffusion & VAE 生成大型图像 [![CC 署名-非商用-相同方式共享 4.0][cc-by-nc-sa-shield]][cc-by-nc-sa] @@ -28,6 +28,7 @@ - [x] [区域提示控制](#区域提示控制) - [x] [Img2img 放大](#img2img-放大) - [x] [生成超大图像](#生成超大图像) +- [x] [Demofusion现已可用](#Demofusion现已可用) => 快速入门教程: [Tutorial for multidiffusion upscaler for automatic1111](https://civitai.com/models/34726), 感谢由 [@PotatoBananaApple](https://github.com/pkuliyi2015/multidiffusion-upscaler-for-automatic1111/discussions/120) 提供 🎉 @@ -181,6 +182,26 @@ **** +### Demofusion现已可用 + +ℹ 建议设定随机种子以获取更好的图像 + +ℹ 勾选随机抖动通常会得到更好的结果 + +ℹ推荐使用较高的步数，例如30步以上，往往会有更好的效果 + +ℹ如果你设定的生图大小是512*512， 合适的window size和overlap是64和32或者更小。如果是1024则推荐翻倍 + +ℹimg2img推荐使用较高的重绘幅度，并尽可能地使用原图的生图模型、随机种子以及prompt等 + +ℹ不要同时开启tilediffusion. 但该组件支持tilevae、noise inversion等常用功能 + +ℹc1，c2，c3等参数可以参考demofusion. 你不必调整这些参数 + +**** + +### + ## 安装 ⚪ 方法 1: 官方市场 diff --git a/scripts/tileglobal.py b/scripts/tileglobal.py index 26c2cc2..7c23680 100644 --- a/scripts/tileglobal.py +++ b/scripts/tileglobal.py @@ -42,54 +42,45 @@ class Script(scripts.Script): with gr.Accordion('DemoFusion', open=False, elem_id=f'MD-{tab}'): with gr.Row(variant='compact') as tab_enable: enabled = gr.Checkbox(label='Enable DemoFusion(Do not open it with tilediffusion)', value=False, elem_id=uid('enabled')) - # overwrite_size = gr.Checkbox(label='Overwrite image size', value=False, visible=not is_img2img, elem_id=uid('overwrite-image-size')) - keep_input_size = gr.Checkbox(label='Keep input image size', value=True, visible=is_img2img, elem_id=uid('keep-input-size')) + + with gr.Row(variant='compact') as tab_default: random_jitter = gr.Checkbox(label='Random jitter windows', value=True, elem_id=uid('random-jitter')) + keep_input_size = gr.Checkbox(label='Keep input image size', value=True, visible=is_img2img, elem_id=uid('keep-input-size')) + gaussian_filter = gr.Checkbox(label='Gaussian filter', value=False, elem_id=uid('gaussian')) - # with gr.Row(variant='compact', visible=False) as tab_size: - # image_width = gr.Slider(minimum=256, maximum=16384, step=16, label='Image width', value=1024, elem_id=f'MD-overwrite-width-{tab}') - # image_height = gr.Slider(minimum=256, maximum=16384, step=16, label='Image height', value=1024, elem_id=f'MD-overwrite-height-{tab}') - # overwrite_size.change(fn=gr_show, inputs=overwrite_size, outputs=tab_size, show_progress=False) - - # with gr.Row(variant='compact', visible=True) as tab_size: - # c1 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c1', value=3, elem_id=f'c1-{tab}') - # c2 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c2', value=1, elem_id=f'c2-{tab}') - # c3 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c3', value=1, elem_id=f'c3-{tab}') with gr.Row(variant='compact') as tab_param: - method = gr.Dropdown(label='Method', choices=[Method_2.DEMO_FU.value], value=Method_2.DEMO_FU.value, elem_id=uid('method-2')) - control_tensor_cpu = gr.Checkbox(label='Move ControlNet tensor to CPU (if applicable)', value=False, elem_id=uid('control-tensor-cpu-2')) + method = gr.Dropdown(label='Method', choices=[Method_2.DEMO_FU.value], value=Method_2.DEMO_FU.value, elem_id=uid('method')) + control_tensor_cpu = gr.Checkbox(label='Move ControlNet tensor to CPU (if applicable)', value=False, elem_id=uid('control-tensor-cpu')) reset_status = gr.Button(value='Free GPU', variant='tool') reset_status.click(fn=self.reset_and_gc, show_progress=False) with gr.Group() as tab_tile: with gr.Row(variant='compact'): window_size = gr.Slider(minimum=16, maximum=256, step=16, label='Latent window size', value=128, elem_id=uid('latent-window-size')) - # tile_height = gr.Slider(minimum=16, maximum=256, step=16, label='Latent tile height', value=96, elem_id=uid('latent-tile-height')) with gr.Row(variant='compact'): - overlap = gr.Slider(minimum=0, maximum=256, step=4, label='Latent window overlap', value=64, elem_id=uid('latent-tile-overlap-2')) - batch_size = gr.Slider(minimum=1, maximum=8, step=1, label='Latent window batch size', value=4, elem_id=uid('latent-tile-batch-size-2')) + overlap = gr.Slider(minimum=0, maximum=256, step=4, label='Latent window overlap', value=64, elem_id=uid('latent-tile-overlap')) + batch_size = gr.Slider(minimum=1, maximum=8, step=1, label='Latent window batch size', value=4, elem_id=uid('latent-tile-batch-size')) with gr.Row(variant='compact', visible=True) as tab_size: c1 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c1', value=3, elem_id=f'c1-{tab}') c2 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c2', value=1, elem_id=f'c2-{tab}') - c3 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c3', value=1, visible=False, elem_id=f'c3-{tab}') #XXX:this parameter is useless in current version - + c3 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c3', value=1, elem_id=f'c3-{tab}') with gr.Row(variant='compact') as tab_upscale: - # upscaler_name = gr.Dropdown(label='Upscaler', choices=[x.name for x in shared.sd_upscalers], value='None', elem_id=uid('upscaler-index')) - scale_factor = gr.Slider(minimum=1.0, maximum=8.0, step=1, label='Scale_Factor', value=2.0, elem_id=uid('upscaler-factor-2')) - # scale_factor = gr.Slider(minimum=1.0, maximum=8.0, step=1, label='Overwrite Scale Factor', value=2.0,value=is_img2img, elem_id=uid('upscaler-factor')) + + scale_factor = gr.Slider(minimum=1.0, maximum=8.0, step=1, label='Scale Factor', value=2.0, elem_id=uid('upscaler-factor')) + with gr.Accordion('Noise Inversion', open=True, visible=is_img2img) as tab_noise_inv: with gr.Row(variant='compact'): - noise_inverse = gr.Checkbox(label='Enable Noise Inversion', value=False, elem_id=uid('noise-inverse-2')) - noise_inverse_steps = gr.Slider(minimum=1, maximum=200, step=1, label='Inversion steps', value=10, elem_id=uid('noise-inverse-steps-2')) + noise_inverse = gr.Checkbox(label='Enable Noise Inversion', value=False, elem_id=uid('noise-inverse')) + noise_inverse_steps = gr.Slider(minimum=1, maximum=200, step=1, label='Inversion steps', value=10, elem_id=uid('noise-inverse-steps')) gr.HTML('

Please test on small images before actual upscale. Default params require denoise <= 0.6

') with gr.Row(variant='compact'): - noise_inverse_retouch = gr.Slider(minimum=1, maximum=100, step=0.1, label='Retouch', value=1, elem_id=uid('noise-inverse-retouch-2')) - noise_inverse_renoise_strength = gr.Slider(minimum=0, maximum=2, step=0.01, label='Renoise strength', value=1, elem_id=uid('noise-inverse-renoise-strength-2')) - noise_inverse_renoise_kernel = gr.Slider(minimum=2, maximum=512, step=1, label='Renoise kernel size', value=64, elem_id=uid('noise-inverse-renoise-kernel-2')) + noise_inverse_retouch = gr.Slider(minimum=1, maximum=100, step=0.1, label='Retouch', value=1, elem_id=uid('noise-inverse-retouch')) + noise_inverse_renoise_strength = gr.Slider(minimum=0, maximum=2, step=0.01, label='Renoise strength', value=1, elem_id=uid('noise-inverse-renoise-strength')) + noise_inverse_renoise_kernel = gr.Slider(minimum=2, maximum=512, step=1, label='Renoise kernel size', value=64, elem_id=uid('noise-inverse-renoise-kernel')) # The control includes txt2img and img2img, we use t2i and i2i to distinguish them @@ -101,7 +92,7 @@ class Script(scripts.Script): noise_inverse, noise_inverse_steps, noise_inverse_retouch, noise_inverse_renoise_strength, noise_inverse_renoise_kernel, control_tensor_cpu, random_jitter, - c1,c2,c3 + c1,c2,c3,gaussian_filter ] @@ -113,7 +104,7 @@ class Script(scripts.Script): noise_inverse: bool, noise_inverse_steps: int, noise_inverse_retouch: float, noise_inverse_renoise_strength: float, noise_inverse_renoise_kernel: int, control_tensor_cpu: bool, random_jitter:bool, - c1,c2,c3 + c1,c2,c3,gaussian_filter ): # unhijack & unhook, in case it broke at last time @@ -129,6 +120,7 @@ class Script(scripts.Script): p.width_original_md = p.width p.height_original_md = p.height p.current_scale_num = 1 + p.gaussian_filter = gaussian_filter p.scale_factor = int(scale_factor) is_img2img = hasattr(p, "init_images") and len(p.init_images) > 0 @@ -136,16 +128,17 @@ class Script(scripts.Script): init_img = p.init_images[0] init_img = images.flatten(init_img, opts.img2img_background_color) image = init_img - if keep_input_size: #若 scale factor为1则为真 - p.scale_factor = 1 + if keep_input_size: p.width = image.width p.height = image.height + p.width_original_md = p.width + p.height_original_md = p.height else: #XXX:To adapt to noise inversion, we do not multiply the scale factor here p.width = p.width_original_md p.height = p.height_original_md else: # txt2img - p.width = p.width*(p.scale_factor) - p.height = p.height*(p.scale_factor) + p.width = p.width_original_md + p.height = p.height_original_md if 'png info': info = {} @@ -227,31 +220,23 @@ class Script(scripts.Script): ''' ↓↓↓ inner API hijack ↓↓↓ ''' @torch.no_grad() def sample_hijack(self, conditioning, unconditional_conditioning,seeds, subseeds, subseed_strength, prompts,p,image_ori,window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3): - - if self.delegate==None: - p.denoising_strength=1 - # p.sampler = Script.create_sampler_original_md(p.sampler_name, p.sd_model) - p.sampler = sd_samplers.create_sampler(p.sampler_name, p.sd_model) #NOTE:Wrong but very useful. If corrected, please replace with the content from the previous line - # 3. Encode input prompts - shared.state.sampling_step = 0 - noise = p.rng.next() - - if hasattr(p,'initial_noise_multiplier'): - if p.initial_noise_multiplier != 1.0: - p.extra_generation_params["Noise multiplier"] = p.initial_noise_multiplier - noise *= p.initial_noise_multiplier - ################################################## Phase Initialization ###################################################### if not image_ori: - latents = p.rng.next() #Same with line 233. Replaced with the following lines - # latents = p.sampler.sample(p, x, conditioning, unconditional_conditioning, image_conditioning=p.txt2img_image_conditioning(x)) - # del x - # p.denoising_strength=1 - # p.sampler = sd_samplers.create_sampler(p.sampler_name, p.sd_model) + + # p.sampler = sd_samplers.create_sampler(p.sampler_name, p.sd_model) #NOTE:Wrong but very useful. If corrected, please replace with the content with the following lines + # latents = p.rng.next() + p.denoising_strength = 1 + p.sampler = Script.create_sampler_original_md(p.sampler_name, p.sd_model) + x = p.rng.next() + latents = p.sampler.sample(p, x, conditioning, unconditional_conditioning, image_conditioning=p.txt2img_image_conditioning(x)) + del x + p.sampler = sd_samplers.create_sampler(p.sampler_name, p.sd_model) + starting_scale = 2 else: # img2img print("### Encoding Real Image ###") latents = p.init_latent + starting_scale = 1 anchor_mean = latents.mean() @@ -260,7 +245,6 @@ class Script(scripts.Script): devices.torch_gc() ####################################################### Phase Upscaling ##################################################### - starting_scale = 1 p.cosine_scale_1 = c1 # 3 p.cosine_scale_2 = c2 # 1 p.cosine_scale_3 = c3 # 1 @@ -424,8 +408,6 @@ class Script(scripts.Script): org_random_tensors = torch.where(background_noise_count > 0, background_noise, org_random_tensors) org_random_tensors = torch.where(foreground_noise_count > 0, foreground_noise, org_random_tensors) return org_random_tensors - # p.sd_model.sd_model_hash改为p.sd_model_hash - ''' ↓↓↓ helper methods ↓↓↓ ''' ''' ↓↓↓ helper methods ↓↓↓ ''' diff --git a/tile_methods/demofusion.py b/tile_methods/demofusion.py index 291bc1f..c727ac9 100644 --- a/tile_methods/demofusion.py +++ b/tile_methods/demofusion.py @@ -108,10 +108,6 @@ class DemoFusion(AbstractDiffusion): cols=1 dx = (w_l - tile_w) / (cols - 1) if cols > 1 else 0 dy = (h_l - tile_h) / (rows - 1) if rows > 1 else 0 - if self.p.random_jitter: - self.jitter_range = max((min(self.w, self.h)-self.stride)//4,0) - else: - self.jitter_range=0 bbox_list: List[BBox] = [] for row in range(rows): for col in range(cols): @@ -232,8 +228,10 @@ class DemoFusion(AbstractDiffusion): w,h = bbox ###### - - x_global_i = self.sampler.model_wrap_cfg.forward_ori(self.x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) # x_in_tmp could be changed to latents_gaussian + if self.gaussian_filter: + x_global_i = self.sampler.model_wrap_cfg.forward_ori(self.latents_gaussian[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) + else: + x_global_i = self.sampler.model_wrap_cfg.forward_ori(self.x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) # x_in_tmp could be changed to latents_gaussian x_global[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num] += x_global_i ###### From 361f55d9109326a4f67c03a6dff6f5f4537a9375 Mon Sep 17 00:00:00 2001 From: Jaylen Lee Date: Tue, 19 Mar 2024 19:35:05 +0800 Subject: [PATCH 2/8] fix --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 44a47e1..a0cceb5 100644 --- a/README.md +++ b/README.md @@ -32,7 +32,7 @@ The extension enables **large image drawing & upscaling with limited VRAM** via - [x] [Regional Prompt Control](#region-prompt-control) - [x] [Img2img upscale](#img2img-upscale) - [x] [Ultra-Large image generation](#ultra-large-image-generation) -- [x] [Demofusion available](#Demofusion available) +- [x] [Demofusion available](#Demofusion-available) => Quickstart Tutorial: [Tutorial for multidiffusion upscaler for automatic1111](https://civitai.com/models/34726), thanks to [@PotatoBananaApple](https://github.com/pkuliyi2015/multidiffusion-upscaler-for-automatic1111/discussions/120) 🎉 From 62902a09ca36f6692db5f8c6c658b749c29c2396 Mon Sep 17 00:00:00 2001 From: Jaylen Lee Date: Tue, 19 Mar 2024 19:41:36 +0800 Subject: [PATCH 3/8] fix --- README.md | 4 +++- README_CN.md | 2 ++ 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index a0cceb5..f366ae0 100644 --- a/README.md +++ b/README.md @@ -201,7 +201,9 @@ The extension enables **large image drawing & upscaling with limited VRAM** via ℹ Do not enable it together with tilediffusion. It supports operations such as tilevae, noise inversion, etc. -ℹFor parameters such as c1, c2, and c3, please refer to the demofusion. You don't have to adjust it. +ℹ more suitable for realistic styles + +ℹ For parameters such as c1, c2, and c3, please refer to the demofusion. You don't have to adjust it. **** diff --git a/README_CN.md b/README_CN.md index e993662..a76ba43 100644 --- a/README_CN.md +++ b/README_CN.md @@ -196,6 +196,8 @@ ℹ不要同时开启tilediffusion. 但该组件支持tilevae、noise inversion等常用功能 +ℹ更适合写实风格的图像 + ℹc1，c2，c3等参数可以参考demofusion. 你不必调整这些参数 **** From f36c4ad6b1615ed78323f9091f0f446cbf86aab5 Mon Sep 17 00:00:00 2001 From: Jaylen Lee Date: Tue, 19 Mar 2024 19:42:29 +0800 Subject: [PATCH 4/8] fix --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index f366ae0..da3db5d 100644 --- a/README.md +++ b/README.md @@ -201,7 +201,7 @@ The extension enables **large image drawing & upscaling with limited VRAM** via ℹ Do not enable it together with tilediffusion. It supports operations such as tilevae, noise inversion, etc. -ℹ more suitable for realistic styles +ℹ More suitable for realistic styles ℹ For parameters such as c1, c2, and c3, please refer to the demofusion. You don't have to adjust it. From 3beff8152624461cad928071409a7e530c8e9a57 Mon Sep 17 00:00:00 2001 From: Jaylen Lee Date: Mon, 25 Mar 2024 04:40:45 +0800 Subject: [PATCH 5/8] fix --- README.md | 6 ++---- README_CN.md | 6 ++---- scripts/tileglobal.py | 40 ++++++++++++++++++++++++++------------ tile_methods/demofusion.py | 29 +++++++++++++-------------- 4 files changed, 45 insertions(+), 36 deletions(-) diff --git a/README.md b/README.md index da3db5d..38ff1aa 100644 --- a/README.md +++ b/README.md @@ -189,8 +189,6 @@ The extension enables **large image drawing & upscaling with limited VRAM** via ### Demofusion available -ℹ Please set seed in for a better result. - ℹ Ticking the random jitter usually yields better results. ℹ Recommend using higher steps, such as 30 or more, for better results @@ -201,10 +199,10 @@ The extension enables **large image drawing & upscaling with limited VRAM** via ℹ Do not enable it together with tilediffusion. It supports operations such as tilevae, noise inversion, etc. -ℹ More suitable for realistic styles - ℹ For parameters such as c1, c2, and c3, please refer to the demofusion. You don't have to adjust it. +ℹ Euler sampler performs better + **** ### diff --git a/README_CN.md b/README_CN.md index a76ba43..93c1ef1 100644 --- a/README_CN.md +++ b/README_CN.md @@ -184,8 +184,6 @@ ### Demofusion现已可用 -ℹ 建议设定随机种子以获取更好的图像 - ℹ 勾选随机抖动通常会得到更好的结果 ℹ推荐使用较高的步数，例如30步以上，往往会有更好的效果 @@ -196,10 +194,10 @@ ℹ不要同时开启tilediffusion. 但该组件支持tilevae、noise inversion等常用功能 -ℹ更适合写实风格的图像 - ℹc1，c2，c3等参数可以参考demofusion. 你不必调整这些参数 +ℹ Euler采样器表现得更好 + **** ### diff --git a/scripts/tileglobal.py b/scripts/tileglobal.py index 7c23680..aa456fa 100644 --- a/scripts/tileglobal.py +++ b/scripts/tileglobal.py @@ -13,13 +13,14 @@ from modules.ui import gr_show from tile_methods.abstractdiffusion import AbstractDiffusion from tile_methods.demofusion import DemoFusion from tile_utils.utils import * +from modules.sd_samplers_common import InterruptedException +# import k_diffusion.sampling CFG_PATH = os.path.join(scripts.basedir(), 'region_configs') BBOX_MAX_NUM = min(getattr(shared.cmd_opts, 'md_max_regions', 8), 16) - class Script(scripts.Script): def __init__(self): self.controlnet_script: ModuleType = None @@ -42,10 +43,8 @@ class Script(scripts.Script): with gr.Accordion('DemoFusion', open=False, elem_id=f'MD-{tab}'): with gr.Row(variant='compact') as tab_enable: enabled = gr.Checkbox(label='Enable DemoFusion(Do not open it with tilediffusion)', value=False, elem_id=uid('enabled')) - - with gr.Row(variant='compact') as tab_default: - random_jitter = gr.Checkbox(label='Random jitter windows', value=True, elem_id=uid('random-jitter')) - keep_input_size = gr.Checkbox(label='Keep input image size', value=True, visible=is_img2img, elem_id=uid('keep-input-size')) + random_jitter = gr.Checkbox(label='Random jitter', value = True, elem_id=uid('random-jitter')) + keep_input_size = gr.Checkbox(label='Keep input-image size', value=False,visible=is_img2img, elem_id=uid('keep-input-size')) gaussian_filter = gr.Checkbox(label='Gaussian filter', value=False, elem_id=uid('gaussian')) @@ -64,9 +63,9 @@ class Script(scripts.Script): batch_size = gr.Slider(minimum=1, maximum=8, step=1, label='Latent window batch size', value=4, elem_id=uid('latent-tile-batch-size')) with gr.Row(variant='compact', visible=True) as tab_size: - c1 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c1', value=3, elem_id=f'c1-{tab}') - c2 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c2', value=1, elem_id=f'c2-{tab}') - c3 = gr.Slider(minimum=0.5, maximum=3, step=0.1, label='c3', value=1, elem_id=f'c3-{tab}') + c1 = gr.Slider(minimum=0, maximum=5, step=0.1, label='c1', value=3, elem_id=f'c1-{tab}') + c2 = gr.Slider(minimum=0, maximum=5, step=0.1, label='c2', value=1, elem_id=f'c2-{tab}') + c3 = gr.Slider(minimum=0, maximum=5, step=0.1, label='c3', value=1, elem_id=f'c3-{tab}') with gr.Row(variant='compact') as tab_upscale: scale_factor = gr.Slider(minimum=1.0, maximum=8.0, step=1, label='Scale Factor', value=2.0, elem_id=uid('upscaler-factor')) @@ -223,11 +222,12 @@ class Script(scripts.Script): ################################################## Phase Initialization ###################################################### if not image_ori: - + p.current_step = 0 + p.denoising_strength = 1 # p.sampler = sd_samplers.create_sampler(p.sampler_name, p.sd_model) #NOTE:Wrong but very useful. If corrected, please replace with the content with the following lines # latents = p.rng.next() - p.denoising_strength = 1 - p.sampler = Script.create_sampler_original_md(p.sampler_name, p.sd_model) + + p.sampler = Script.create_sampler_original_md(p.sampler_name, p.sd_model) #scale x = p.rng.next() latents = p.sampler.sample(p, x, conditioning, unconditional_conditioning, image_conditioning=p.txt2img_image_conditioning(x)) del x @@ -285,7 +285,7 @@ class Script(scripts.Script): p.noise = noise p.x = p.latents.clone() - p.current_step=-1 + p.current_step=0 p.latents = p.sampler.sample_img2img(p,p.latents, noise , conditioning, unconditional_conditioning, image_conditioning=p.image_conditioning) if self.flag_noise_inverse: @@ -297,6 +297,17 @@ class Script(scripts.Script): p.width = p.width*p.scale_factor p.height = p.height*p.scale_factor return p.latents + + @staticmethod + def callback_hijack(self_sampler,d,p): + p.current_step = d['i'] + + if self_sampler.stop_at is not None and p.current_step > self_sampler.stop_at: + raise InterruptedException + + state.sampling_step = p.current_step + shared.total_tqdm.update() + p.current_step += 1 def create_sampler_hijack( @@ -313,6 +324,8 @@ class Script(scripts.Script): return self.delegate.sampler_raw else: self.reset() + sd_samplers_common.Sampler.callback_ori = sd_samplers_common.Sampler.callback_state + sd_samplers_common.Sampler.callback_state = lambda self_sampler,d:Script.callback_hijack(self_sampler,d,p) self.flag_noise_inverse = hasattr(p, "init_images") and len(p.init_images) > 0 and noise_inverse flag_noise_inverse = self.flag_noise_inverse @@ -467,6 +480,9 @@ class Script(scripts.Script): if hasattr(Script, "create_random_tensors_original_md"): processing.create_random_tensors = Script.create_random_tensors_original_md del Script.create_random_tensors_original_md + if hasattr(sd_samplers_common.Sampler, "callback_ori"): + sd_samplers_common.Sampler.callback_state = sd_samplers_common.Sampler.callback_ori + del sd_samplers_common.Sampler.callback_ori DemoFusion.unhook() self.delegate = None diff --git a/tile_methods/demofusion.py b/tile_methods/demofusion.py index c727ac9..c81da03 100644 --- a/tile_methods/demofusion.py +++ b/tile_methods/demofusion.py @@ -17,10 +17,6 @@ class DemoFusion(AbstractDiffusion): super().__init__(p, *args, **kwargs) assert p.sampler_name != 'UniPC', 'Demofusion is not compatible with UniPC!' - def add_one(self): - self.p.current_step += 1 - return - def hook(self): steps, self.t_enc = sd_samplers_common.setup_img2img_steps(self.p, None) @@ -177,13 +173,13 @@ class DemoFusion(AbstractDiffusion): blurred_latents = F.conv2d(latents, kernel, padding=kernel_size//2, groups=channels) return blurred_latents + ''' ↓↓↓ kernel hijacks ↓↓↓ ''' @torch.no_grad() @keep_signature def forward_one_step(self, x_in, sigma, **kwarg): - self.add_one() if self.is_kdiff: self.xi = self.p.x + self.p.noise * self.p.sigmas[self.p.current_step] else: @@ -197,23 +193,24 @@ class DemoFusion(AbstractDiffusion): self.c1 = self.cosine_factor ** self.p.cosine_scale_1 - self.x_in_tmp = x_in*(1 - self.c1) + self.xi * self.c1 + x_in_tmp = x_in*(1 - self.c1) + self.xi * self.c1 if self.p.random_jitter: jitter_range = self.jitter_range else: jitter_range = 0 - self.x_in_tmp_ = F.pad(self.x_in_tmp,(jitter_range, jitter_range, jitter_range, jitter_range),'constant',value=0) - _,_,H,W = self.x_in_tmp.shape + x_in_tmp_ = F.pad(x_in_tmp,(jitter_range, jitter_range, jitter_range, jitter_range),'constant',value=0) + _,_,H,W = x_in_tmp.shape - std_, mean_ = self.x_in_tmp.std(), self.x_in_tmp.mean() + std_, mean_ = x_in_tmp.std(), x_in_tmp.mean() c3 = 0.99 * self.cosine_factor ** self.p.cosine_scale_3 + 1e-2 - latents_gaussian = self.gaussian_filter(self.x_in_tmp, kernel_size=(2*self.p.current_scale_num-1), sigma=0.8*c3) - self.latents_gaussian = (latents_gaussian - latents_gaussian.mean()) / latents_gaussian.std() * std_ + mean_ + latents_gaussian = self.gaussian_filter(x_in_tmp, kernel_size=(2*self.p.current_scale_num-1), sigma=0.8*c3) + latents_gaussian = (latents_gaussian - latents_gaussian.mean()) / latents_gaussian.std() * std_ + mean_ self.jitter_range = jitter_range self.sampler.model_wrap_cfg.inner_model.forward = self.sample_one_step_local self.repeat_3 = False - x_local = self.sampler.model_wrap_cfg.forward_ori(self.x_in_tmp_,sigma, **kwarg) + + x_local = self.sampler.model_wrap_cfg.forward_ori(x_in_tmp_,sigma, **kwarg) self.sampler.model_wrap_cfg.inner_model.forward = self.sampler_forward x_local = x_local[:,:,jitter_range:jitter_range+H,jitter_range:jitter_range+W] @@ -229,9 +226,9 @@ class DemoFusion(AbstractDiffusion): ###### if self.gaussian_filter: - x_global_i = self.sampler.model_wrap_cfg.forward_ori(self.latents_gaussian[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) + x_global_i = self.sampler.model_wrap_cfg.forward_ori(latents_gaussian[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) else: - x_global_i = self.sampler.model_wrap_cfg.forward_ori(self.x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) # x_in_tmp could be changed to latents_gaussian + x_global_i = self.sampler.model_wrap_cfg.forward_ori(x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) # x_in_tmp could be changed to latents_gaussian x_global[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num] += x_global_i ###### @@ -241,9 +238,9 @@ class DemoFusion(AbstractDiffusion): # self.x_out_list = [] # self.x_out_idx = -1 # self.flag = 1 - # self.sampler.model_wrap_cfg.forward_ori(self.latents_gaussian[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma,**kwarg) + # self.sampler.model_wrap_cfg.forward_ori(x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma,**kwarg) # self.flag = 0 - # x_global_i = self.sampler.model_wrap_cfg.forward_ori(self.x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma,**kwarg) + # x_global_i = self.sampler.model_wrap_cfg.forward_ori(x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma,**kwarg) # x_global[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num] += x_global_i self.p.sd_model.apply_model = self.p.sd_model.apply_model_ori From 01037629188d7d5ed66158c31e597a3339dab2de Mon Sep 17 00:00:00 2001 From: Jaylen Lee Date: Mon, 25 Mar 2024 05:31:05 +0800 Subject: [PATCH 6/8] update --- README.md | 2 +- README_CN.md | 2 +- scripts/tileglobal.py | 27 ++++++++++++++------------- 3 files changed, 16 insertions(+), 15 deletions(-) diff --git a/README.md b/README.md index 38ff1aa..7add1e0 100644 --- a/README.md +++ b/README.md @@ -195,7 +195,7 @@ The extension enables **large image drawing & upscaling with limited VRAM** via ℹ If you set the image size to 512 * 512, the appropriate window size and overlap are 64 and 32 or smaller. If it is 1024, it is recommended to double it. -ℹ Recommend using a higher denoising strength in img2img，and try to use the original model, seeds, and prompt as much as possible +ℹ Recommend using a higher denoising strength in img2img, maybe 0.8-1，and try to use the original model, seeds, and prompt as much as possible ℹ Do not enable it together with tilediffusion. It supports operations such as tilevae, noise inversion, etc. diff --git a/README_CN.md b/README_CN.md index 93c1ef1..164ff64 100644 --- a/README_CN.md +++ b/README_CN.md @@ -190,7 +190,7 @@ ℹ如果你设定的生图大小是512*512， 合适的window size和overlap是64和32或者更小。如果是1024则推荐翻倍 -ℹimg2img推荐使用较高的重绘幅度，并尽可能地使用原图的生图模型、随机种子以及prompt等 +ℹimg2img推荐使用较高的重绘幅度,比如0.8-1，并尽可能地使用原图的生图模型、随机种子以及prompt等 ℹ不要同时开启tilediffusion. 但该组件支持tilevae、noise inversion等常用功能 diff --git a/scripts/tileglobal.py b/scripts/tileglobal.py index aa456fa..f2e5794 100644 --- a/scripts/tileglobal.py +++ b/scripts/tileglobal.py @@ -61,11 +61,12 @@ class Script(scripts.Script): with gr.Row(variant='compact'): overlap = gr.Slider(minimum=0, maximum=256, step=4, label='Latent window overlap', value=64, elem_id=uid('latent-tile-overlap')) batch_size = gr.Slider(minimum=1, maximum=8, step=1, label='Latent window batch size', value=4, elem_id=uid('latent-tile-batch-size')) - - with gr.Row(variant='compact', visible=True) as tab_size: - c1 = gr.Slider(minimum=0, maximum=5, step=0.1, label='c1', value=3, elem_id=f'c1-{tab}') - c2 = gr.Slider(minimum=0, maximum=5, step=0.1, label='c2', value=1, elem_id=f'c2-{tab}') - c3 = gr.Slider(minimum=0, maximum=5, step=0.1, label='c3', value=1, elem_id=f'c3-{tab}') + with gr.Row(variant='compact', visible=True) as tab_c: + c1 = gr.Slider(minimum=0, maximum=5, step=0.1, label='C1', value=3, elem_id=f'C1-{tab}') + c2 = gr.Slider(minimum=0, maximum=5, step=0.1, label='C2', value=1, elem_id=f'C2-{tab}') + c3 = gr.Slider(minimum=0, maximum=5, step=0.1, label='C3', value=1, elem_id=f'C3-{tab}') + with gr.Group() as tab_denoise: + strength = gr.Slider(minimum=0, maximum=1, step=0.01, value = 0.9, label='Denoising strength substage',visible= not is_img2img, elem_id=f'strength-{tab}') with gr.Row(variant='compact') as tab_upscale: scale_factor = gr.Slider(minimum=1.0, maximum=8.0, step=1, label='Scale Factor', value=2.0, elem_id=uid('upscaler-factor')) @@ -91,7 +92,7 @@ class Script(scripts.Script): noise_inverse, noise_inverse_steps, noise_inverse_retouch, noise_inverse_renoise_strength, noise_inverse_renoise_kernel, control_tensor_cpu, random_jitter, - c1,c2,c3,gaussian_filter + c1,c2,c3,gaussian_filter,strength ] @@ -103,7 +104,7 @@ class Script(scripts.Script): noise_inverse: bool, noise_inverse_steps: int, noise_inverse_retouch: float, noise_inverse_renoise_strength: float, noise_inverse_renoise_kernel: int, control_tensor_cpu: bool, random_jitter:bool, - c1,c2,c3,gaussian_filter + c1,c2,c3,gaussian_filter,strength ): # unhijack & unhook, in case it broke at last time @@ -190,7 +191,7 @@ class Script(scripts.Script): p.sample = lambda conditioning, unconditional_conditioning,seeds, subseeds, subseed_strength, prompts: self.sample_hijack( conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts,p, is_img2img, - window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3) + window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3,strength) ## end @@ -218,12 +219,12 @@ class Script(scripts.Script): ''' ↓↓↓ inner API hijack ↓↓↓ ''' @torch.no_grad() - def sample_hijack(self, conditioning, unconditional_conditioning,seeds, subseeds, subseed_strength, prompts,p,image_ori,window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3): + def sample_hijack(self, conditioning, unconditional_conditioning,seeds, subseeds, subseed_strength, prompts,p,image_ori,window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3,strength): ################################################## Phase Initialization ###################################################### if not image_ori: p.current_step = 0 - p.denoising_strength = 1 + p.denoising_strength = strength # p.sampler = sd_samplers.create_sampler(p.sampler_name, p.sd_model) #NOTE:Wrong but very useful. If corrected, please replace with the content with the following lines # latents = p.rng.next() @@ -245,9 +246,9 @@ class Script(scripts.Script): devices.torch_gc() ####################################################### Phase Upscaling ##################################################### - p.cosine_scale_1 = c1 # 3 - p.cosine_scale_2 = c2 # 1 - p.cosine_scale_3 = c3 # 1 + p.cosine_scale_1 = c1 + p.cosine_scale_2 = c2 + p.cosine_scale_3 = c3 p.latents = latents for current_scale_num in range(starting_scale, p.scale_factor+1): p.current_scale_num = current_scale_num From b629e087a094d65cb5fcfa50e75d650dbafa71b0 Mon Sep 17 00:00:00 2001 From: Jaylen Lee Date: Wed, 27 Mar 2024 05:27:22 +0800 Subject: [PATCH 7/8] fix & update --- scripts/tilediffusion.py | 2 +- scripts/tileglobal.py | 89 +++++++++++++++++----- tile_methods/abstractdiffusion.py | 4 +- tile_methods/demofusion.py | 120 +++++++++++++++--------------- tile_methods/multidiffusion.py | 4 +- tile_utils/utils.py | 2 +- 6 files changed, 134 insertions(+), 87 deletions(-) diff --git a/scripts/tilediffusion.py b/scripts/tilediffusion.py index 5a4f1c2..e4de28d 100644 --- a/scripts/tilediffusion.py +++ b/scripts/tilediffusion.py @@ -369,7 +369,7 @@ class Script(scripts.Script): info["Region control"] = region_info Script.create_random_tensors_original_md = processing.create_random_tensors processing.create_random_tensors = lambda *args, **kwargs: self.create_random_tensors_hijack( - bbox_settings, region_info, + bbox_settings, region_info, *args, **kwargs, ) diff --git a/scripts/tileglobal.py b/scripts/tileglobal.py index f2e5794..fac770c 100644 --- a/scripts/tileglobal.py +++ b/scripts/tileglobal.py @@ -21,6 +21,20 @@ CFG_PATH = os.path.join(scripts.basedir(), 'region_configs') BBOX_MAX_NUM = min(getattr(shared.cmd_opts, 'md_max_regions', 8), 16) +def create_infotext_hijack(p, all_prompts, all_seeds, all_subseeds, comments=None, iteration=0, position_in_batch=0, use_main_prompt=False, index=-1, all_negative_prompts=None): + idx = index + if index == -1: + idx = None + text = processing.create_infotext_ori(p, all_prompts, all_seeds, all_subseeds, comments, iteration, position_in_batch, use_main_prompt, idx, all_negative_prompts) + start_index = text.find("Size") + if start_index != -1: + r_text = f"Size:{p.width_list[index]}x{p.height_list[index]}" + end_index = text.find(",", start_index) + if end_index != -1: + replaced_string = text[:start_index] + r_text + text[end_index:] + return replaced_string + return text + class Script(scripts.Script): def __init__(self): self.controlnet_script: ModuleType = None @@ -42,14 +56,14 @@ class Script(scripts.Script): with gr.Accordion('DemoFusion', open=False, elem_id=f'MD-{tab}'): with gr.Row(variant='compact') as tab_enable: - enabled = gr.Checkbox(label='Enable DemoFusion(Do not open it with tilediffusion)', value=False, elem_id=uid('enabled')) - random_jitter = gr.Checkbox(label='Random jitter', value = True, elem_id=uid('random-jitter')) + enabled = gr.Checkbox(label='Enable DemoFusion(Dont open with tilediffusion)', value=False, elem_id=uid('enabled')) + random_jitter = gr.Checkbox(label='Random Jitter', value = True, elem_id=uid('random-jitter')) + gaussian_filter = gr.Checkbox(label='Gaussian Filter', value=True, visible=False, elem_id=uid('gaussian')) keep_input_size = gr.Checkbox(label='Keep input-image size', value=False,visible=is_img2img, elem_id=uid('keep-input-size')) - gaussian_filter = gr.Checkbox(label='Gaussian filter', value=False, elem_id=uid('gaussian')) with gr.Row(variant='compact') as tab_param: - method = gr.Dropdown(label='Method', choices=[Method_2.DEMO_FU.value], value=Method_2.DEMO_FU.value, elem_id=uid('method')) + method = gr.Dropdown(label='Method', choices=[Method_2.DEMO_FU.value], value=Method_2.DEMO_FU.value, visible= False, elem_id=uid('method')) control_tensor_cpu = gr.Checkbox(label='Move ControlNet tensor to CPU (if applicable)', value=False, elem_id=uid('control-tensor-cpu')) reset_status = gr.Button(value='Free GPU', variant='tool') reset_status.click(fn=self.reset_and_gc, show_progress=False) @@ -62,13 +76,13 @@ class Script(scripts.Script): overlap = gr.Slider(minimum=0, maximum=256, step=4, label='Latent window overlap', value=64, elem_id=uid('latent-tile-overlap')) batch_size = gr.Slider(minimum=1, maximum=8, step=1, label='Latent window batch size', value=4, elem_id=uid('latent-tile-batch-size')) with gr.Row(variant='compact', visible=True) as tab_c: - c1 = gr.Slider(minimum=0, maximum=5, step=0.1, label='C1', value=3, elem_id=f'C1-{tab}') - c2 = gr.Slider(minimum=0, maximum=5, step=0.1, label='C2', value=1, elem_id=f'C2-{tab}') - c3 = gr.Slider(minimum=0, maximum=5, step=0.1, label='C3', value=1, elem_id=f'C3-{tab}') + c1 = gr.Slider(minimum=0, maximum=5, step=0.01, label='Cosine Scale 1', value=3, elem_id=f'C1-{tab}') + c2 = gr.Slider(minimum=0, maximum=5, step=0.01, label='Cosine Scale 2', value=1, elem_id=f'C2-{tab}') + c3 = gr.Slider(minimum=0, maximum=5, step=0.01, label='Cosine Scale 3', value=1, elem_id=f'C3-{tab}') + sigma = gr.Slider(minimum=0, maximum=1, step=0.01, label='Sigma', value=0.4, elem_id=f'Sigma-{tab}') with gr.Group() as tab_denoise: - strength = gr.Slider(minimum=0, maximum=1, step=0.01, value = 0.9, label='Denoising strength substage',visible= not is_img2img, elem_id=f'strength-{tab}') + strength = gr.Slider(minimum=0, maximum=1, step=0.01, value = 0.9,label='Denoising Strength for Substage',visible=not is_img2img, elem_id=f'strength-{tab}') with gr.Row(variant='compact') as tab_upscale: - scale_factor = gr.Slider(minimum=1.0, maximum=8.0, step=1, label='Scale Factor', value=2.0, elem_id=uid('upscaler-factor')) @@ -92,7 +106,7 @@ class Script(scripts.Script): noise_inverse, noise_inverse_steps, noise_inverse_retouch, noise_inverse_renoise_strength, noise_inverse_renoise_kernel, control_tensor_cpu, random_jitter, - c1,c2,c3,gaussian_filter,strength + c1,c2,c3,gaussian_filter,strength,sigma ] @@ -104,7 +118,7 @@ class Script(scripts.Script): noise_inverse: bool, noise_inverse_steps: int, noise_inverse_retouch: float, noise_inverse_renoise_strength: float, noise_inverse_renoise_kernel: int, control_tensor_cpu: bool, random_jitter:bool, - c1,c2,c3,gaussian_filter,strength + c1,c2,c3,gaussian_filter,strength,sigma ): # unhijack & unhook, in case it broke at last time @@ -191,7 +205,14 @@ class Script(scripts.Script): p.sample = lambda conditioning, unconditional_conditioning,seeds, subseeds, subseed_strength, prompts: self.sample_hijack( conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts,p, is_img2img, - window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3,strength) + window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3,strength,sigma) + + processing.create_infotext_ori = processing.create_infotext + + p.width_list = [p.height] + p.height_list = [p.height] + + processing.create_infotext = create_infotext_hijack ## end @@ -200,6 +221,20 @@ class Script(scripts.Script): if self.delegate is not None: self.delegate.reset_controlnet_tensors() + def postprocess_batch_list(self, p, pp, *args, **kwargs): + for idx,image in enumerate(pp.images): + idx_b = idx//p.batch_size + pp.images[idx] = image[:,:image.shape[1]//(p.scale_factor)*(idx_b+1),:image.shape[2]//(p.scale_factor)*(idx_b+1)] + p.seeds = [item for _ in range(p.scale_factor) for item in p.seeds] + p.prompts = [item for _ in range(p.scale_factor) for item in p.prompts] + p.all_negative_prompts = [item for _ in range(p.scale_factor) for item in p.all_negative_prompts] + p.negative_prompts = [item for _ in range(p.scale_factor) for item in p.negative_prompts] + if p.color_corrections != None: + p.color_corrections = [item for _ in range(p.scale_factor) for item in p.color_corrections] + p.width_list = [item*(idx+1) for idx in range(p.scale_factor) for item in [p.width for _ in range(p.batch_size)]] + p.height_list = [item*(idx+1) for idx in range(p.scale_factor) for item in [p.height for _ in range(p.batch_size)]] + return + def postprocess(self, p: Processing, processed, enabled, *args): if not enabled: return # unhijack & unhook @@ -219,8 +254,10 @@ class Script(scripts.Script): ''' ↓↓↓ inner API hijack ↓↓↓ ''' @torch.no_grad() - def sample_hijack(self, conditioning, unconditional_conditioning,seeds, subseeds, subseed_strength, prompts,p,image_ori,window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3,strength): + def sample_hijack(self, conditioning, unconditional_conditioning,seeds, subseeds, subseed_strength, prompts,p,image_ori,window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3,strength,sigma): ################################################## Phase Initialization ###################################################### + # p.width = p.width_original_md + # p.height = p.height_original_md if not image_ori: p.current_step = 0 @@ -230,7 +267,10 @@ class Script(scripts.Script): p.sampler = Script.create_sampler_original_md(p.sampler_name, p.sd_model) #scale x = p.rng.next() + print("### Phase 1 Denoising ###") latents = p.sampler.sample(p, x, conditioning, unconditional_conditioning, image_conditioning=p.txt2img_image_conditioning(x)) + latents_ = F.pad(latents, (0, latents.shape[3]*(p.scale_factor-1), 0, latents.shape[2]*(p.scale_factor-1))) + res = latents_ del x p.sampler = sd_samplers.create_sampler(p.sampler_name, p.sd_model) starting_scale = 2 @@ -249,6 +289,7 @@ class Script(scripts.Script): p.cosine_scale_1 = c1 p.cosine_scale_2 = c2 p.cosine_scale_3 = c3 + self.delegate.sig = sigma p.latents = latents for current_scale_num in range(starting_scale, p.scale_factor+1): p.current_scale_num = current_scale_num @@ -268,7 +309,7 @@ class Script(scripts.Script): info = ', '.join([ # f"{method.value} hooked into {name!r} sampler", - f"Tile size: {window_size}", + f"Tile size: {self.delegate.window_size}", f"Tile count: {self.delegate.num_tiles}", f"Batch size: {self.delegate.tile_bs}", f"Tile batches: {len(self.delegate.batched_bboxes)}", @@ -294,10 +335,17 @@ class Script(scripts.Script): self.flag_noise_inverse = False p.latents = (p.latents - p.latents.mean()) / p.latents.std() * anchor_std + anchor_mean + latents_ = F.pad(p.latents, (0, p.latents.shape[3]//current_scale_num*(p.scale_factor-current_scale_num), 0, p.latents.shape[2]//current_scale_num*(p.scale_factor-current_scale_num))) + if current_scale_num==1: + res = latents_ + else: + res = torch.concatenate((res,latents_),axis=0) + ######################################################################################################################################### - p.width = p.width*p.scale_factor - p.height = p.height*p.scale_factor - return p.latents + # p.width = p.width*p.scale_factor + # p.height = p.height*p.scale_factor + + return res @staticmethod def callback_hijack(self_sampler,d,p): @@ -341,7 +389,7 @@ class Script(scripts.Script): else: raise NotImplementedError(f"Method {method} not implemented.") delegate = delegate_cls(p, sampler) - delegate.window_size = window_size + delegate.window_size = min(min(window_size,p.width//8),p.height//8) p.random_jitter = random_jitter if flag_noise_inverse: @@ -363,7 +411,7 @@ class Script(scripts.Script): info = ', '.join([ f"{method.value} hooked into {name!r} sampler", - f"Tile size: {window_size}", + f"Tile size: {delegate.window_size}", f"Tile count: {delegate.num_tiles}", f"Batch size: {delegate.tile_bs}", f"Tile batches: {len(delegate.batched_bboxes)}", @@ -484,6 +532,9 @@ class Script(scripts.Script): if hasattr(sd_samplers_common.Sampler, "callback_ori"): sd_samplers_common.Sampler.callback_state = sd_samplers_common.Sampler.callback_ori del sd_samplers_common.Sampler.callback_ori + if hasattr(processing, "create_infotext_ori"): + processing.create_infotext = processing.create_infotext_ori + del processing.create_infotext_ori DemoFusion.unhook() self.delegate = None diff --git a/tile_methods/abstractdiffusion.py b/tile_methods/abstractdiffusion.py index 9b153ed..88916f3 100644 --- a/tile_methods/abstractdiffusion.py +++ b/tile_methods/abstractdiffusion.py @@ -104,7 +104,7 @@ class AbstractDiffusion: def init_done(self): ''' Call this after all `init_*`, settings are done, now perform: - - settings sanity check + - settings sanity check - pre-computations, cache init - anything thing needed before denoising starts ''' @@ -214,7 +214,7 @@ class AbstractDiffusion: h = min(self.h - y, h) self.custom_bboxes.append(CustomBBox(x, y, w, h, p, n, blend_mode, feather_ratio, seed)) - if len(self.custom_bboxes) == 0: + if len(self.custom_bboxes) == 0: self.enable_custom_bbox = False return diff --git a/tile_methods/demofusion.py b/tile_methods/demofusion.py index c81da03..4cdb6cb 100644 --- a/tile_methods/demofusion.py +++ b/tile_methods/demofusion.py @@ -20,18 +20,16 @@ class DemoFusion(AbstractDiffusion): def hook(self): steps, self.t_enc = sd_samplers_common.setup_img2img_steps(self.p, None) - # print("ENC",self.t_enc) self.sampler.model_wrap_cfg.forward_ori = self.sampler.model_wrap_cfg.forward - self.sampler.model_wrap_cfg.forward = self.forward_one_step self.sampler_forward = self.sampler.model_wrap_cfg.inner_model.forward + self.sampler.model_wrap_cfg.forward = self.forward_one_step if self.is_kdiff: self.sampler: KDiffusionSampler self.sampler.model_wrap_cfg: CFGDenoiserKDiffusion self.sampler.model_wrap_cfg.inner_model: Union[CompVisDenoiser, CompVisVDenoiser] - sigmas = self.sampler.get_sigmas(self.p, steps) - # print("SIGMAS:",sigmas) - self.p.sigmas = sigmas[steps - self.t_enc - 1:] + # sigmas = self.sampler.get_sigmas(self.p, steps) + # self.p.sigmas = sigmas[steps - self.t_enc - 1:] else: self.sampler: CompVisSampler self.sampler.model_wrap_cfg: CFGDenoiserTimesteps @@ -105,12 +103,13 @@ class DemoFusion(AbstractDiffusion): dx = (w_l - tile_w) / (cols - 1) if cols > 1 else 0 dy = (h_l - tile_h) / (rows - 1) if rows > 1 else 0 bbox_list: List[BBox] = [] + self.jitter_range = 0 for row in range(rows): for col in range(cols): h = min(int(row * dy), h_l - tile_h) w = min(int(col * dx), w_l - tile_w) if self.p.random_jitter: - self.jitter_range = min(max((min(self.w, self.h)-self.stride)//4,0),int(self.stride/2)) + self.jitter_range = min(max((min(self.w, self.h)-self.stride)//4,0),min(int(self.window_size/2),int(self.overlap/2))) jitter_range = self.jitter_range w_jitter = 0 h_jitter = 0 @@ -141,12 +140,11 @@ class DemoFusion(AbstractDiffusion): self.overlap = max(0, min(overlap, self.window_size - 4)) - self.stride = max(1,self.window_size - self.overlap) + self.stride = max(4,self.window_size - self.overlap) # split the latent into overlapped tiles, then batching # weights basically indicate how many times a pixel is painted - bboxes, _ = self.split_bboxes_jitter(self.w, self.h, self.tile_w, self.tile_h, overlap, self.get_tile_weights()) - print("BBOX:",len(bboxes)) + bboxes, _ = self.split_bboxes_jitter(self.w, self.h, self.tile_w, self.tile_h, self.overlap, self.get_tile_weights()) self.num_tiles = len(bboxes) self.num_batches = math.ceil(self.num_tiles / tile_bs) self.tile_bs = math.ceil(len(bboxes) / self.num_batches) # optimal_batch_size @@ -181,77 +179,39 @@ class DemoFusion(AbstractDiffusion): @keep_signature def forward_one_step(self, x_in, sigma, **kwarg): if self.is_kdiff: - self.xi = self.p.x + self.p.noise * self.p.sigmas[self.p.current_step] + x_noisy = self.p.x + self.p.noise * sigma[0] else: alphas_cumprod = self.p.sd_model.alphas_cumprod sqrt_alpha_cumprod = torch.sqrt(alphas_cumprod[self.timesteps[self.t_enc-self.p.current_step]]) sqrt_one_minus_alpha_cumprod = torch.sqrt(1 - alphas_cumprod[self.timesteps[self.t_enc-self.p.current_step]]) - self.xi = self.p.x*sqrt_alpha_cumprod + self.p.noise * sqrt_one_minus_alpha_cumprod + x_noisy = self.p.x*sqrt_alpha_cumprod + self.p.noise * sqrt_one_minus_alpha_cumprod self.cosine_factor = 0.5 * (1 + torch.cos(torch.pi *torch.tensor(((self.p.current_step + 1) / (self.t_enc+1))))) - c2 = self.cosine_factor**self.p.cosine_scale_2 - self.c1 = self.cosine_factor ** self.p.cosine_scale_1 + c1 = self.cosine_factor ** self.p.cosine_scale_1 - x_in_tmp = x_in*(1 - self.c1) + self.xi * self.c1 + x_in = x_in*(1 - c1) + x_noisy * c1 if self.p.random_jitter: jitter_range = self.jitter_range else: jitter_range = 0 - x_in_tmp_ = F.pad(x_in_tmp,(jitter_range, jitter_range, jitter_range, jitter_range),'constant',value=0) - _,_,H,W = x_in_tmp.shape + x_in_ = F.pad(x_in,(jitter_range, jitter_range, jitter_range, jitter_range),'constant',value=0) + _,_,H,W = x_in.shape - std_, mean_ = x_in_tmp.std(), x_in_tmp.mean() - c3 = 0.99 * self.cosine_factor ** self.p.cosine_scale_3 + 1e-2 - latents_gaussian = self.gaussian_filter(x_in_tmp, kernel_size=(2*self.p.current_scale_num-1), sigma=0.8*c3) - latents_gaussian = (latents_gaussian - latents_gaussian.mean()) / latents_gaussian.std() * std_ + mean_ - self.jitter_range = jitter_range - self.sampler.model_wrap_cfg.inner_model.forward = self.sample_one_step_local + self.sampler.model_wrap_cfg.inner_model.forward = self.sample_one_step self.repeat_3 = False - x_local = self.sampler.model_wrap_cfg.forward_ori(x_in_tmp_,sigma, **kwarg) + x_out = self.sampler.model_wrap_cfg.forward_ori(x_in_,sigma, **kwarg) self.sampler.model_wrap_cfg.inner_model.forward = self.sampler_forward - x_local = x_local[:,:,jitter_range:jitter_range+H,jitter_range:jitter_range+W] + x_out = x_out[:,:,jitter_range:jitter_range+H,jitter_range:jitter_range+W] - ############################################# Dilated Sampling ############################################# - if not hasattr(self.p.sd_model, 'apply_model_ori'): - self.p.sd_model.apply_model_ori = self.p.sd_model.apply_model - self.p.sd_model.apply_model = self.apply_model_hijack - x_global = torch.zeros_like(x_local) - - for batch_id, bboxes in enumerate(self.global_batched_bboxes): - for bbox in bboxes: - w,h = bbox - - ###### - if self.gaussian_filter: - x_global_i = self.sampler.model_wrap_cfg.forward_ori(latents_gaussian[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) - else: - x_global_i = self.sampler.model_wrap_cfg.forward_ori(x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma, **kwarg) # x_in_tmp could be changed to latents_gaussian - x_global[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num] += x_global_i - - ###### - - #NOTE: Predicting Noise on Gaussian Latent and Obtaining Denoised on Original Latent - - # self.x_out_list = [] - # self.x_out_idx = -1 - # self.flag = 1 - # self.sampler.model_wrap_cfg.forward_ori(x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma,**kwarg) - # self.flag = 0 - # x_global_i = self.sampler.model_wrap_cfg.forward_ori(x_in_tmp[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num],sigma,**kwarg) - # x_global[:,:,h::self.p.current_scale_num,w::self.p.current_scale_num] += x_global_i - - self.p.sd_model.apply_model = self.p.sd_model.apply_model_ori - - x_out= x_local*(1-c2)+ x_global*c2 return x_out @torch.no_grad() @keep_signature - def sample_one_step_local(self, x_in, sigma, cond): + def sample_one_step(self, x_in, sigma, cond): assert LatentDiffusion.apply_model def repeat_func_1(x_tile:Tensor, bboxes:List[CustomBBox]) -> Tensor: sigma_tile = self.repeat_tensor(sigma, len(bboxes)) @@ -282,11 +242,12 @@ class DemoFusion(AbstractDiffusion): repeat_func = repeat_func_2 N,_,_,_ = x_in.shape - H = self.h - W = self.w + # H = self.h + # W = self.w self.x_buffer = torch.zeros_like(x_in) self.weights = torch.zeros_like(x_in) + for batch_id, bboxes in enumerate(self.batched_bboxes): if state.interrupted: return x_in x_tile = torch.cat([x_in[bbox.slicer] for bbox in bboxes], dim=0) @@ -297,9 +258,44 @@ class DemoFusion(AbstractDiffusion): self.weights[bbox.slicer] += 1 self.weights = torch.where(self.weights == 0, torch.tensor(1), self.weights) #Prevent NaN from appearing in random_jitter mode - x_buffer = self.x_buffer/self.weights + x_local = self.x_buffer/self.weights - return x_buffer + self.x_buffer = torch.zeros_like(self.x_buffer) + self.weights = torch.zeros_like(self.weights) + + std_, mean_ = x_in.std(), x_in.mean() + c3 = 0.99 * self.cosine_factor ** self.p.cosine_scale_3 + 1e-2 + if self.p.gaussian_filter: + x_in_g = self.gaussian_filter(x_in, kernel_size=(2*self.p.current_scale_num-1), sigma=self.sig*c3) + x_in_g = (x_in_g - x_in_g.mean()) / x_in_g.std() * std_ + mean_ + + if not hasattr(self.p.sd_model, 'apply_model_ori'): + self.p.sd_model.apply_model_ori = self.p.sd_model.apply_model + self.p.sd_model.apply_model = self.apply_model_hijack + x_global = torch.zeros_like(x_local) + jitter_range = self.jitter_range + end = x_global.shape[3]-jitter_range + + for batch_id, bboxes in enumerate(self.global_batched_bboxes): + for bbox in bboxes: + w,h = bbox + # self.x_out_list = [] + # self.x_out_idx = -1 + # self.flag = 1 + x_global_i0 = self.sampler_forward(x_in_g[:,:,h+jitter_range:end:self.p.current_scale_num,w+jitter_range:end:self.p.current_scale_num],sigma,cond = cond) + # self.flag = 0 + x_global_i1 = self.sampler_forward(x_in[:,:,h+jitter_range:end:self.p.current_scale_num,w+jitter_range:end:self.p.current_scale_num],sigma,cond = cond) #NOTE According to the original execution process, it would be very strange to use the predicted noise of gaussian latents to predict the denoised data in non Gaussian latents. Why? + self.x_buffer[:,:,h+jitter_range:end:self.p.current_scale_num,w+jitter_range:end:self.p.current_scale_num] += (x_global_i0 + x_global_i1)/2 + self.weights[:,:,h+jitter_range:end:self.p.current_scale_num,w+jitter_range:end:self.p.current_scale_num] += 1 + + self.p.sd_model.apply_model = self.p.sd_model.apply_model_ori + self.weights = torch.where(self.weights == 0, torch.tensor(1), self.weights) #Prevent NaN from appearing in random_jitter mode + + x_global = self.x_buffer/self.weights + c2 = self.cosine_factor**self.p.cosine_scale_2 + self.x_buffer= x_local*(1-c2)+ x_global*c2 + + return self.x_buffer @@ -310,7 +306,7 @@ class DemoFusion(AbstractDiffusion): x_tile_out = self.p.sd_model.apply_model_ori(x_in,t_in,cond) return x_tile_out - #NOTE: Using Gaussian Latent to Predict Noise on the Original Latent + # NOTE: Using Gaussian Latent to Predict Noise on the Original Latent # if self.flag == 1: # x_tile_out = self.p.sd_model.apply_model_ori(x_in,t_in,cond) # self.x_out_list.append(x_tile_out) diff --git a/tile_methods/multidiffusion.py b/tile_methods/multidiffusion.py index 5f230cf..026fc18 100644 --- a/tile_methods/multidiffusion.py +++ b/tile_methods/multidiffusion.py @@ -36,7 +36,7 @@ class MultiDiffusion(AbstractDiffusion): def reset_buffer(self, x_in:Tensor): super().reset_buffer(x_in) - + @custom_bbox def init_custom_bbox(self, *args): super().init_custom_bbox(*args) @@ -229,7 +229,7 @@ class MultiDiffusion(AbstractDiffusion): cond_out = self.repeat_cond_dict(cond_in_original, bboxes) x_tile_out = shared.sd_model.apply_model(x_tile, sigma_in_tile, cond=cond_out) return x_tile_out - + def custom_func(x:Tensor, bbox_id:int, bbox:CustomBBox): # The negative prompt in custom bbox should not be used for noise inversion # otherwise the result will be astonishingly bad. diff --git a/tile_utils/utils.py b/tile_utils/utils.py index 4bf9b41..ef8497e 100644 --- a/tile_utils/utils.py +++ b/tile_utils/utils.py @@ -23,7 +23,7 @@ class ComparableEnum(Enum): def __eq__(self, other: Any) -> bool: if isinstance(other, str): return self.value == other elif isinstance(other, ComparableEnum): return self.value == other.value - else: raise TypeError(f'unsupported type: {type(other)}') + else: raise TypeError(f'unsupported type: {type(other)}') class Method(ComparableEnum): From f0cd2ad381438e156101a8e39d8f47e75ece57af Mon Sep 17 00:00:00 2001 From: Jaylen Lee Date: Thu, 28 Mar 2024 15:08:11 +0800 Subject: [PATCH 8/8] update --- README.md | 21 +++++++++++++++++---- README_CN.md | 20 +++++++++++++++----- scripts/tileglobal.py | 8 ++------ tile_methods/demofusion.py | 4 ---- 4 files changed, 34 insertions(+), 19 deletions(-) diff --git a/README.md b/README.md index 7add1e0..416cf0b 100644 --- a/README.md +++ b/README.md @@ -189,19 +189,32 @@ The extension enables **large image drawing & upscaling with limited VRAM** via ### Demofusion available -ℹ Ticking the random jitter usually yields better results. +ℹ The execution time of Demofusion will be relatively long, but it can obtain multiple images of different resolutions at once. Just like tilediffusion, please enable tilevae when an OOM error occurs. ℹ Recommend using higher steps, such as 30 or more, for better results -ℹ If you set the image size to 512 * 512, the appropriate window size and overlap are 64 and 32 or smaller. If it is 1024, it is recommended to double it. +ℹ If you set the image size to 512 * 512, the appropriate window size and overlap are 64 and 32 or smaller. If it is 1024, it is recommended to double it, and so on. ℹ Recommend using a higher denoising strength in img2img, maybe 0.8-1，and try to use the original model, seeds, and prompt as much as possible ℹ Do not enable it together with tilediffusion. It supports operations such as tilevae, noise inversion, etc. -ℹ For parameters such as c1, c2, and c3, please refer to the demofusion. You don't have to adjust it. +ℹ Due to differences in implementation details, parameters such as c1, c2, c3 and sigma can refer to the [demofusion](https://ruoyidu.github.io/demofusion/demofusion.html), but may not be entirely effective sometimes. If there are blurred images, it is recommended to increase c3 and reduce Sigma. -ℹ Euler sampler performs better +![demo-example](https://github.com/Jaylen-Lee/image-demo/blob/main/example.png?raw=true) + +#### Example: txt2img, 1024 * 1024 image 3x upscale + +- Params： + + - Step=45, sampler=Euler, same prompt as official demofusion, random seed 35, model SDXL1.0 + + - Denoising Strength for Substage = 0.85， Sigma=0.5，use default values for the rest，enable tilevae + + - Device： 4060ti 16GB + - The following are the images obtained at a resolution of 1024, 2048, and 3072 + + ![demo-result](https://github.com/Jaylen-Lee/image-demo/blob/main/3.png?raw=true) **** diff --git a/README_CN.md b/README_CN.md index 164ff64..3c33819 100644 --- a/README_CN.md +++ b/README_CN.md @@ -184,19 +184,29 @@ ### Demofusion现已可用 -ℹ 勾选随机抖动通常会得到更好的结果 +ℹ Demofusion的执行时间会比较长，不过可以一次得到多张不同分辨率的图片。与tilediffusion一样，出现OOM错误时请开启tilevae -ℹ推荐使用较高的步数，例如30步以上，往往会有更好的效果 +ℹ推荐使用较高的步数，例如30步以上，效果会更好 -ℹ如果你设定的生图大小是512*512， 合适的window size和overlap是64和32或者更小。如果是1024则推荐翻倍 +ℹ如果你设定的生图大小是512*512， 合适的window size和overlap是64和32或者更小。如果是1024则推荐翻倍，以此类推 ℹimg2img推荐使用较高的重绘幅度,比如0.8-1，并尽可能地使用原图的生图模型、随机种子以及prompt等 ℹ不要同时开启tilediffusion. 但该组件支持tilevae、noise inversion等常用功能 -ℹc1，c2，c3等参数可以参考demofusion. 你不必调整这些参数 +ℹ由于实现细节的差异，c1，c2，c3等参数可以参考[demofusion](https://ruoyidu.github.io/demofusion/demofusion.html)但不一定完全奏效. 如果出现模糊图像建议提高c3并降低Sigma -ℹ Euler采样器表现得更好 +![demo-example](https://github.com/Jaylen-Lee/image-demo/blob/main/example.png?raw=true) + +#### 示例 txt2img 1024*1024图片 3倍放大 + +- 参数： + - 步数 = 45，采样器 = Euler，与demofusion示例同样的提示语，随机种子35，模型为SDXL1.0 + - 子阶段降噪 = 0.85， Sigma=0.5，其余采用默认值，开启tilevae + - 设备 4060ti 16GB + - 以下为获得的1024，2048，3072分辨率的图片 + +![demo-result](https://github.com/Jaylen-Lee/image-demo/blob/main/3.png?raw=true) **** diff --git a/scripts/tileglobal.py b/scripts/tileglobal.py index fac770c..ebf8efc 100644 --- a/scripts/tileglobal.py +++ b/scripts/tileglobal.py @@ -79,9 +79,9 @@ class Script(scripts.Script): c1 = gr.Slider(minimum=0, maximum=5, step=0.01, label='Cosine Scale 1', value=3, elem_id=f'C1-{tab}') c2 = gr.Slider(minimum=0, maximum=5, step=0.01, label='Cosine Scale 2', value=1, elem_id=f'C2-{tab}') c3 = gr.Slider(minimum=0, maximum=5, step=0.01, label='Cosine Scale 3', value=1, elem_id=f'C3-{tab}') - sigma = gr.Slider(minimum=0, maximum=1, step=0.01, label='Sigma', value=0.4, elem_id=f'Sigma-{tab}') + sigma = gr.Slider(minimum=0, maximum=1, step=0.01, label='Sigma', value=0.5, elem_id=f'Sigma-{tab}') with gr.Group() as tab_denoise: - strength = gr.Slider(minimum=0, maximum=1, step=0.01, value = 0.9,label='Denoising Strength for Substage',visible=not is_img2img, elem_id=f'strength-{tab}') + strength = gr.Slider(minimum=0, maximum=1, step=0.01, value = 0.85,label='Denoising Strength for Substage',visible=not is_img2img, elem_id=f'strength-{tab}') with gr.Row(variant='compact') as tab_upscale: scale_factor = gr.Slider(minimum=1.0, maximum=8.0, step=1, label='Scale Factor', value=2.0, elem_id=uid('upscaler-factor')) @@ -256,8 +256,6 @@ class Script(scripts.Script): @torch.no_grad() def sample_hijack(self, conditioning, unconditional_conditioning,seeds, subseeds, subseed_strength, prompts,p,image_ori,window_size, overlap, tile_batch_size,random_jitter,c1,c2,c3,strength,sigma): ################################################## Phase Initialization ###################################################### - # p.width = p.width_original_md - # p.height = p.height_original_md if not image_ori: p.current_step = 0 @@ -342,8 +340,6 @@ class Script(scripts.Script): res = torch.concatenate((res,latents_),axis=0) ######################################################################################################################################### - # p.width = p.width*p.scale_factor - # p.height = p.height*p.scale_factor return res diff --git a/tile_methods/demofusion.py b/tile_methods/demofusion.py index 4cdb6cb..cecc26e 100644 --- a/tile_methods/demofusion.py +++ b/tile_methods/demofusion.py @@ -28,8 +28,6 @@ class DemoFusion(AbstractDiffusion): self.sampler: KDiffusionSampler self.sampler.model_wrap_cfg: CFGDenoiserKDiffusion self.sampler.model_wrap_cfg.inner_model: Union[CompVisDenoiser, CompVisVDenoiser] - # sigmas = self.sampler.get_sigmas(self.p, steps) - # self.p.sigmas = sigmas[steps - self.t_enc - 1:] else: self.sampler: CompVisSampler self.sampler.model_wrap_cfg: CFGDenoiserTimesteps @@ -242,8 +240,6 @@ class DemoFusion(AbstractDiffusion): repeat_func = repeat_func_2 N,_,_,_ = x_in.shape - # H = self.h - # W = self.w self.x_buffer = torch.zeros_like(x_in) self.weights = torch.zeros_like(x_in)