From b64994b973ad8f4268bf785f25f92b66c8dced40 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 16 Sep 2022 19:24:48 +0300 Subject: [PATCH 01/20] added original negative prompt to img2img alt --- scripts/img2imgalt.py | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/scripts/img2imgalt.py b/scripts/img2imgalt.py index dbda3255..7f1f53a7 100644 --- a/scripts/img2imgalt.py +++ b/scripts/img2imgalt.py @@ -59,7 +59,7 @@ def find_noise_for_image(p, cond, uncond, cfg_scale, steps): return x / x.std() -Cached = namedtuple("Cached", ["noise", "cfg_scale", "steps", "latent", "original_prompt"]) +Cached = namedtuple("Cached", ["noise", "cfg_scale", "steps", "latent", "original_prompt", "original_negative_prompt"]) class Script(scripts.Script): @@ -74,19 +74,20 @@ class Script(scripts.Script): def ui(self, is_img2img): original_prompt = gr.Textbox(label="Original prompt", lines=1) + original_negative_prompt = gr.Textbox(label="Original negative prompt", lines=1) cfg = gr.Slider(label="Decode CFG scale", minimum=0.0, maximum=15.0, step=0.1, value=1.0) st = gr.Slider(label="Decode steps", minimum=1, maximum=150, step=1, value=50) randomness = gr.Slider(label="randomness", minimum=0.0, maximum=1.0, step=0.01, value=0.0) - return [original_prompt, cfg, st, randomness] + return [original_prompt, original_negative_prompt, cfg, st, randomness] - def run(self, p, original_prompt, cfg, st, randomness): + def run(self, p, original_prompt, original_negative_prompt, cfg, st, randomness): p.batch_size = 1 p.batch_count = 1 def sample_extra(x, conditioning, unconditional_conditioning): lat = (p.init_latent.cpu().numpy() * 10).astype(int) - same_params = self.cache is not None and self.cache.cfg_scale == cfg and self.cache.steps == st and self.cache.original_prompt == original_prompt + same_params = self.cache is not None and self.cache.cfg_scale == cfg and self.cache.steps == st and self.cache.original_prompt == original_prompt and self.cache.original_negative_prompt == original_negative_prompt same_everything = same_params and self.cache.latent.shape == lat.shape and np.abs(self.cache.latent-lat).sum() < 100 if same_everything: @@ -94,9 +95,9 @@ class Script(scripts.Script): else: shared.state.job_count += 1 cond = p.sd_model.get_learned_conditioning(p.batch_size * [original_prompt]) - uncond = p.sd_model.get_learned_conditioning(p.batch_size * [""]) + uncond = p.sd_model.get_learned_conditioning(p.batch_size * [original_negative_prompt]) rec_noise = find_noise_for_image(p, cond, uncond, cfg, st) - self.cache = Cached(rec_noise, cfg, st, lat, original_prompt) + self.cache = Cached(rec_noise, cfg, st, lat, original_prompt, original_negative_prompt) rand_noise = processing.create_random_tensors(p.init_latent.shape[1:], [p.seed + x + 1 for x in range(p.init_latent.shape[0])]) From 2ee9fc8eb84d5e1864dbabd8a8c6b279a6ae21ac Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 16 Sep 2022 22:18:30 +0300 Subject: [PATCH 02/20] new outpainting script --- scripts/outpainting_mk_2.py | 290 ++++++++++++++++++++++++++++++++++++ 1 file changed, 290 insertions(+) create mode 100644 scripts/outpainting_mk_2.py diff --git a/scripts/outpainting_mk_2.py b/scripts/outpainting_mk_2.py new file mode 100644 index 00000000..a42c1aed --- /dev/null +++ b/scripts/outpainting_mk_2.py @@ -0,0 +1,290 @@ +import math + +import numpy as np +import skimage + +import modules.scripts as scripts +import gradio as gr +from PIL import Image, ImageDraw + +from modules import images, processing, devices +from modules.processing import Processed, process_images +from modules.shared import opts, cmd_opts, state + + +def expand(x, dir, amount, power=0.75): + is_left = dir == 3 + is_right = dir == 1 + is_up = dir == 0 + is_down = dir == 2 + + if is_left or is_right: + noise = np.zeros((x.shape[0], amount, 3), dtype=float) + indexes = np.random.random((x.shape[0], amount)) ** power * (1 - np.arange(amount) / amount) + if is_right: + indexes = 1 - indexes + indexes = (indexes * (x.shape[1] - 1)).astype(int) + + for row in range(x.shape[0]): + if is_left: + noise[row] = x[row][indexes[row]] + else: + noise[row] = np.flip(x[row][indexes[row]], axis=0) + + x = np.concatenate([noise, x] if is_left else [x, noise], axis=1) + return x + + if is_up or is_down: + noise = np.zeros((amount, x.shape[1], 3), dtype=float) + indexes = np.random.random((x.shape[1], amount)) ** power * (1 - np.arange(amount) / amount) + if is_down: + indexes = 1 - indexes + indexes = (indexes * x.shape[0] - 1).astype(int) + + for row in range(x.shape[1]): + if is_up: + noise[:, row] = x[:, row][indexes[row]] + else: + noise[:, row] = np.flip(x[:, row][indexes[row]], axis=0) + + x = np.concatenate([noise, x] if is_up else [x, noise], axis=0) + return x + + +def get_matched_noise(_np_src_image, np_mask_rgb, noise_q=1, color_variation=0.05): + # helper fft routines that keep ortho normalization and auto-shift before and after fft + def _fft2(data): + if data.ndim > 2: # has channels + out_fft = np.zeros((data.shape[0], data.shape[1], data.shape[2]), dtype=np.complex128) + for c in range(data.shape[2]): + c_data = data[:, :, c] + out_fft[:, :, c] = np.fft.fft2(np.fft.fftshift(c_data), norm="ortho") + out_fft[:, :, c] = np.fft.ifftshift(out_fft[:, :, c]) + else: # one channel + out_fft = np.zeros((data.shape[0], data.shape[1]), dtype=np.complex128) + out_fft[:, :] = np.fft.fft2(np.fft.fftshift(data), norm="ortho") + out_fft[:, :] = np.fft.ifftshift(out_fft[:, :]) + + return out_fft + + def _ifft2(data): + if data.ndim > 2: # has channels + out_ifft = np.zeros((data.shape[0], data.shape[1], data.shape[2]), dtype=np.complex128) + for c in range(data.shape[2]): + c_data = data[:, :, c] + out_ifft[:, :, c] = np.fft.ifft2(np.fft.fftshift(c_data), norm="ortho") + out_ifft[:, :, c] = np.fft.ifftshift(out_ifft[:, :, c]) + else: # one channel + out_ifft = np.zeros((data.shape[0], data.shape[1]), dtype=np.complex128) + out_ifft[:, :] = np.fft.ifft2(np.fft.fftshift(data), norm="ortho") + out_ifft[:, :] = np.fft.ifftshift(out_ifft[:, :]) + + return out_ifft + + def _get_gaussian_window(width, height, std=3.14, mode=0): + window_scale_x = float(width / min(width, height)) + window_scale_y = float(height / min(width, height)) + + window = np.zeros((width, height)) + x = (np.arange(width) / width * 2. - 1.) * window_scale_x + for y in range(height): + fy = (y / height * 2. - 1.) * window_scale_y + if mode == 0: + window[:, y] = np.exp(-(x ** 2 + fy ** 2) * std) + else: + window[:, y] = (1 / ((x ** 2 + 1.) * (fy ** 2 + 1.))) ** (std / 3.14) # hey wait a minute that's not gaussian + + return window + + def _get_masked_window_rgb(np_mask_grey, hardness=1.): + np_mask_rgb = np.zeros((np_mask_grey.shape[0], np_mask_grey.shape[1], 3)) + if hardness != 1.: + hardened = np_mask_grey[:] ** hardness + else: + hardened = np_mask_grey[:] + for c in range(3): + np_mask_rgb[:, :, c] = hardened[:] + return np_mask_rgb + + width = _np_src_image.shape[0] + height = _np_src_image.shape[1] + num_channels = _np_src_image.shape[2] + + np_src_image = _np_src_image[:] * (1. - np_mask_rgb) + np_mask_grey = (np.sum(np_mask_rgb, axis=2) / 3.) + img_mask = np_mask_grey > 1e-6 + ref_mask = np_mask_grey < 1e-3 + + windowed_image = _np_src_image * (1. - _get_masked_window_rgb(np_mask_grey)) + windowed_image /= np.max(windowed_image) + windowed_image += np.average(_np_src_image) * np_mask_rgb # / (1.-np.average(np_mask_rgb)) # rather than leave the masked area black, we get better results from fft by filling the average unmasked color + + src_fft = _fft2(windowed_image) # get feature statistics from masked src img + src_dist = np.absolute(src_fft) + src_phase = src_fft / src_dist + + noise_window = _get_gaussian_window(width, height, mode=1) # start with simple gaussian noise + noise_rgb = np.random.random_sample((width, height, num_channels)) + noise_grey = (np.sum(noise_rgb, axis=2) / 3.) + noise_rgb *= color_variation # the colorfulness of the starting noise is blended to greyscale with a parameter + for c in range(num_channels): + noise_rgb[:, :, c] += (1. - color_variation) * noise_grey + + noise_fft = _fft2(noise_rgb) + for c in range(num_channels): + noise_fft[:, :, c] *= noise_window + noise_rgb = np.real(_ifft2(noise_fft)) + shaped_noise_fft = _fft2(noise_rgb) + shaped_noise_fft[:, :, :] = np.absolute(shaped_noise_fft[:, :, :]) ** 2 * (src_dist ** noise_q) * src_phase # perform the actual shaping + + brightness_variation = 0. # color_variation # todo: temporarily tieing brightness variation to color variation for now + contrast_adjusted_np_src = _np_src_image[:] * (brightness_variation + 1.) - brightness_variation * 2. + + # scikit-image is used for histogram matching, very convenient! + shaped_noise = np.real(_ifft2(shaped_noise_fft)) + shaped_noise -= np.min(shaped_noise) + shaped_noise /= np.max(shaped_noise) + shaped_noise[img_mask, :] = skimage.exposure.match_histograms(shaped_noise[img_mask, :] ** 1., contrast_adjusted_np_src[ref_mask, :], channel_axis=1) + shaped_noise = _np_src_image[:] * (1. - np_mask_rgb) + shaped_noise * np_mask_rgb + + matched_noise = shaped_noise[:] + + return np.clip(matched_noise, 0., 1.) + + + +class Script(scripts.Script): + def title(self): + return "Outpainting mk2" + + def show(self, is_img2img): + return is_img2img + + def ui(self, is_img2img): + if not is_img2img: + return None + + info = gr.HTML("

Recommended settings: Sampling Steps: 80-100, Sampler: Euler a, Denoising strength: 0.8

") + + pixels = gr.Slider(label="Pixels to expand", minimum=8, maximum=256, step=8, value=128) + mask_blur = gr.Slider(label='Mask blur', minimum=0, maximum=64, step=1, value=8, visible=False) + direction = gr.CheckboxGroup(label="Outpainting direction", choices=['left', 'right', 'up', 'down'], value=['left', 'right', 'up', 'down']) + noise_q = gr.Slider(label="Fall-off exponent (lower=higher detail)", minimum=0.0, maximum=4.0, step=0.01, value=1.0) + color_variation = gr.Slider(label="Color variation", minimum=0.0, maximum=1.0, step=0.01, value=0.05) + + return [info, pixels, mask_blur, direction, noise_q, color_variation] + + def run(self, p, _, pixels, mask_blur, direction, noise_q, color_variation): + initial_seed_and_info = [None, None] + + process_width = p.width + process_height = p.height + + p.mask_blur = mask_blur*4 + p.inpaint_full_res = False + p.inpainting_fill = 1 + p.do_not_save_samples = True + p.do_not_save_grid = True + + left = pixels if "left" in direction else 0 + right = pixels if "right" in direction else 0 + up = pixels if "up" in direction else 0 + down = pixels if "down" in direction else 0 + + init_img = p.init_images[0] + target_w = math.ceil((init_img.width + left + right) / 64) * 64 + target_h = math.ceil((init_img.height + up + down) / 64) * 64 + + if left > 0: + left = left * (target_w - init_img.width) // (left + right) + if right > 0: + right = target_w - init_img.width - left + + if up > 0: + up = up * (target_h - init_img.height) // (up + down) + + if down > 0: + down = target_h - init_img.height - up + + init_image = p.init_images[0] + + state.job_count = (1 if left > 0 else 0) + (1 if right > 0 else 0)+ (1 if up > 0 else 0)+ (1 if down > 0 else 0) + + def expand(init, expand_pixels, is_left=False, is_right=False, is_top=False, is_bottom=False): + is_horiz = is_left or is_right + is_vert = is_top or is_bottom + pixels_horiz = expand_pixels if is_horiz else 0 + pixels_vert = expand_pixels if is_vert else 0 + + img = Image.new("RGB", (init.width + pixels_horiz, init.height + pixels_vert)) + img.paste(init, (pixels_horiz if is_left else 0, pixels_vert if is_top else 0)) + mask = Image.new("RGB", (init.width + pixels_horiz, init.height + pixels_vert), "white") + draw = ImageDraw.Draw(mask) + draw.rectangle(( + expand_pixels + mask_blur if is_left else 0, + expand_pixels + mask_blur if is_top else 0, + mask.width - expand_pixels - mask_blur if is_right else mask.width, + mask.height - expand_pixels - mask_blur if is_bottom else mask.height, + ), fill="black") + + np_image = (np.asarray(img) / 255.0).astype(np.float64) + np_mask = (np.asarray(mask) / 255.0).astype(np.float64) + noised = get_matched_noise(np_image, np_mask, noise_q, color_variation) + out = Image.fromarray(np.clip(noised * 255., 0., 255.).astype(np.uint8), mode="RGB") + + target_width = min(process_width, init.width + pixels_horiz) if is_horiz else img.width + target_height = min(process_height, init.height + pixels_vert) if is_vert else img.height + + crop_region = ( + 0 if is_left else out.width - target_width, + 0 if is_top else out.height - target_height, + target_width if is_left else out.width, + target_height if is_top else out.height, + ) + + image_to_process = out.crop(crop_region) + mask = mask.crop(crop_region) + + p.width = target_width if is_horiz else img.width + p.height = target_height if is_vert else img.height + p.init_images = [image_to_process] + p.image_mask = mask + + latent_mask = Image.new("RGB", (p.width, p.height), "white") + draw = ImageDraw.Draw(latent_mask) + draw.rectangle(( + expand_pixels + mask_blur * 2 if is_left else 0, + expand_pixels + mask_blur * 2 if is_top else 0, + mask.width - expand_pixels - mask_blur * 2 if is_right else mask.width, + mask.height - expand_pixels - mask_blur * 2 if is_bottom else mask.height, + ), fill="black") + p.latent_mask = latent_mask + + proc = process_images(p) + proc_img = proc.images[0] + + if initial_seed_and_info[0] is None: + initial_seed_and_info[0] = proc.seed + initial_seed_and_info[1] = proc.info + + out.paste(proc_img, (0 if is_left else out.width - proc_img.width, 0 if is_top else out.height - proc_img.height)) + return out + + img = init_image + + if left > 0: + img = expand(img, left, is_left=True) + if right > 0: + img = expand(img, right, is_right=True) + if up > 0: + img = expand(img, up, is_top=True) + if down > 0: + img = expand(img, down, is_bottom=True) + + res = Processed(p, [img], initial_seed_and_info[0], initial_seed_and_info[1]) + + if opts.samples_save: + images.save_image(img, p.outpath_samples, "", res.seed, p.prompt, opts.grid_format, info=res.info, p=p) + + return res + From 449719b2fc9e8fd2e61f219e0979deb83c05177c Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 16 Sep 2022 23:17:10 +0300 Subject: [PATCH 03/20] added Noise generation for outpainting mk2 to credits --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 84a78da5..6c232e38 100644 --- a/README.md +++ b/README.md @@ -81,6 +81,7 @@ The documentation was moved from this README over to the project's [wiki](https: - Ideas for optimizations - https://github.com/basujindal/stable-diffusion - Doggettx - Cross Attention layer optimization - https://github.com/Doggettx/stable-diffusion, original idea for prompt editing. - Idea for SD upscale - https://github.com/jquesnelle/txt2imghd +- Noise generation for outpainting mk2 - https://github.com/parlance-zz/g-diffuser-bot - CLIP interrogator idea and borrowing some code - https://github.com/pharmapsychotic/clip-interrogator - Initial Gradio script - posted on 4chan by an Anonymous user. Thank you Anonymous user. - (You) From 6168d09218073c369fd08b7208f93805f624ec05 Mon Sep 17 00:00:00 2001 From: uservar <63248296+uservar@users.noreply.github.com> Date: Fri, 16 Sep 2022 19:07:14 +0000 Subject: [PATCH 04/20] Prevent uploading previous output from javascript As it is currently, txt2img and img2img send back the previous output args (txt2img_gallery, generation_info, html_info) whenever you generate a new image. This can lead to uploading a huge gallery of previously generated images, which leads to an unnecessary delay between submitting and beginning to generate. --- script.js | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/script.js b/script.js index 0852e421..8b10b53d 100644 --- a/script.js +++ b/script.js @@ -177,7 +177,7 @@ function submit(){ window.setTimeout(requestProgress, 500) res = [] - for(var i=0;i Date: Sat, 17 Sep 2022 08:03:47 +0300 Subject: [PATCH 05/20] add a comment and some checks for the functionality of the last PR --- script.js | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/script.js b/script.js index 8b10b53d..4a70e51d 100644 --- a/script.js +++ b/script.js @@ -177,9 +177,18 @@ function submit(){ window.setTimeout(requestProgress, 500) res = [] - for(var i=0;i Date: Fri, 16 Sep 2022 07:28:57 +1000 Subject: [PATCH 06/20] image.save parameter fix * image.save takes exif as a parameter * piexif takes the bytes as a parameter, not the exif_bytes function itself * reduce calls to create_exif_bytes --- modules/images.py | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/modules/images.py b/modules/images.py index f37f5f08..8cd7fe37 100644 --- a/modules/images.py +++ b/modules/images.py @@ -345,7 +345,7 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i if not os.path.exists(fullfn): break - def exif_bytes(): + def create_exif_bytes(): return piexif.dump({ "Exif": { piexif.ExifIFD.UserComment: piexif.helper.UserComment.dump(info or "", encoding="unicode") @@ -353,7 +353,8 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i }) if extension.lower() in ("jpg", "jpeg", "webp"): - image.save(fullfn, quality=opts.jpeg_quality, exif_bytes=exif_bytes()) + exif_bytes = create_exif_bytes() + image.save(fullfn, quality=opts.jpeg_quality, exif=exif_bytes) else: image.save(fullfn, quality=opts.jpeg_quality, pnginfo=pnginfo) @@ -370,7 +371,11 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i elif oversize: image = image.resize((image.width * target_side_length // image.height, target_side_length), LANCZOS) - image.save(fullfn_without_extension + ".jpg", quality=opts.jpeg_quality, exif_bytes=exif_bytes()) + if exif_bytes in locals(): + pass + else: + exif_bytes = create_exif_bytes() + image.save(fullfn_without_extension + ".jpg", quality=opts.jpeg_quality, exif=exif_bytes) if opts.save_txt and info is not None: with open(f"{fullfn_without_extension}.txt", "w", encoding="utf8") as file: From 3c665b8dd6da07c60af7783f0e0dd1dec714a9b4 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 17 Sep 2022 08:32:15 +0300 Subject: [PATCH 07/20] the last PR broke saving EXiF completely for me. I don't know if it was broken already or some condition changed, but it seems like the person who originally added EXIF said, saving it with PIL may not work. I switched to using piexif to add data after the file written. --- modules/images.py | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-) diff --git a/modules/images.py b/modules/images.py index 8cd7fe37..7e1e506c 100644 --- a/modules/images.py +++ b/modules/images.py @@ -346,6 +346,7 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i break def create_exif_bytes(): + def exif_bytes(): return piexif.dump({ "Exif": { piexif.ExifIFD.UserComment: piexif.helper.UserComment.dump(info or "", encoding="unicode") @@ -353,14 +354,12 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i }) if extension.lower() in ("jpg", "jpeg", "webp"): - exif_bytes = create_exif_bytes() - image.save(fullfn, quality=opts.jpeg_quality, exif=exif_bytes) + image.save(fullfn, quality=opts.jpeg_quality) + if opts.enable_pnginfo and info is not None: + piexif.insert(exif_bytes(), fullfn) else: image.save(fullfn, quality=opts.jpeg_quality, pnginfo=pnginfo) - if extension.lower() == "webp": - piexif.insert(exif_bytes, fullfn) - target_side_length = 4000 oversize = image.width > target_side_length or image.height > target_side_length if opts.export_for_4chan and (oversize or os.stat(fullfn).st_size > 4 * 1024 * 1024): @@ -371,11 +370,9 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i elif oversize: image = image.resize((image.width * target_side_length // image.height, target_side_length), LANCZOS) - if exif_bytes in locals(): - pass - else: - exif_bytes = create_exif_bytes() - image.save(fullfn_without_extension + ".jpg", quality=opts.jpeg_quality, exif=exif_bytes) + image.save(fullfn_without_extension + ".jpg", quality=opts.jpeg_quality) + if opts.enable_pnginfo and info is not None: + piexif.insert(exif_bytes(), fullfn) if opts.save_txt and info is not None: with open(f"{fullfn_without_extension}.txt", "w", encoding="utf8") as file: From 1fc1c537c7303be88e0da93c3a632c48acb101e9 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 17 Sep 2022 09:01:10 +0300 Subject: [PATCH 08/20] fix --- modules/images.py | 1 - 1 file changed, 1 deletion(-) diff --git a/modules/images.py b/modules/images.py index 7e1e506c..b62c48f8 100644 --- a/modules/images.py +++ b/modules/images.py @@ -345,7 +345,6 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i if not os.path.exists(fullfn): break - def create_exif_bytes(): def exif_bytes(): return piexif.dump({ "Exif": { From ed6787ca2fe950f633a925ccb0467eafd4ec0f43 Mon Sep 17 00:00:00 2001 From: EyeDeck Date: Sat, 17 Sep 2022 00:49:31 -0400 Subject: [PATCH 09/20] Add VRAM monitoring --- modules/memmon.py | 77 +++++++++++++++++++++++++++++++++++++++++++++++ modules/shared.py | 5 +++ modules/ui.py | 14 ++++++++- style.css | 18 ++++++++++- 4 files changed, 112 insertions(+), 2 deletions(-) create mode 100644 modules/memmon.py diff --git a/modules/memmon.py b/modules/memmon.py new file mode 100644 index 00000000..f2cac841 --- /dev/null +++ b/modules/memmon.py @@ -0,0 +1,77 @@ +import threading +import time +from collections import defaultdict + +import torch + + +class MemUsageMonitor(threading.Thread): + run_flag = None + device = None + disabled = False + opts = None + data = None + + def __init__(self, name, device, opts): + threading.Thread.__init__(self) + self.name = name + self.device = device + self.opts = opts + + self.daemon = True + self.run_flag = threading.Event() + self.data = defaultdict(int) + + def run(self): + if self.disabled: + return + + while True: + self.run_flag.wait() + + torch.cuda.reset_peak_memory_stats() + self.data.clear() + + if self.opts.memmon_poll_rate <= 0: + self.run_flag.clear() + continue + + self.data["min_free"] = torch.cuda.mem_get_info()[0] + + while self.run_flag.is_set(): + free, total = torch.cuda.mem_get_info() # calling with self.device errors, torch bug? + self.data["min_free"] = min(self.data["min_free"], free) + + time.sleep(1 / self.opts.memmon_poll_rate) + + def dump_debug(self): + print(self, 'recorded data:') + for k, v in self.read().items(): + print(k, -(v // -(1024 ** 2))) + + print(self, 'raw torch memory stats:') + tm = torch.cuda.memory_stats(self.device) + for k, v in tm.items(): + if 'bytes' not in k: + continue + print('\t' if 'peak' in k else '', k, -(v // -(1024 ** 2))) + + print(torch.cuda.memory_summary()) + + def monitor(self): + self.run_flag.set() + + def read(self): + free, total = torch.cuda.mem_get_info() + self.data["total"] = total + + torch_stats = torch.cuda.memory_stats(self.device) + self.data["active_peak"] = torch_stats["active_bytes.all.peak"] + self.data["reserved_peak"] = torch_stats["reserved_bytes.all.peak"] + self.data["system_peak"] = total - self.data["min_free"] + + return self.data + + def stop(self): + self.run_flag.clear() + return self.read() diff --git a/modules/shared.py b/modules/shared.py index da56b6ae..4f877036 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -12,6 +12,7 @@ from modules.paths import script_path, sd_path from modules.devices import get_optimal_device import modules.styles import modules.interrogate +import modules.memmon sd_model_file = os.path.join(script_path, 'model.ckpt') if not os.path.exists(sd_model_file): @@ -138,6 +139,7 @@ class Options: "show_progressbar": OptionInfo(True, "Show progressbar"), "show_progress_every_n_steps": OptionInfo(0, "Show show image creation progress every N sampling steps. Set 0 to disable.", gr.Slider, {"minimum": 0, "maximum": 32, "step": 1}), "multiple_tqdm": OptionInfo(True, "Add a second progress bar to the console that shows progress for an entire job. Broken in PyCharm console."), + "memmon_poll_rate": OptionInfo(8, "VRAM usage polls per second during generation. Set to 0 to disable.", gr.Slider, {"minimum": 0, "maximum": 40, "step":1}), "face_restoration_model": OptionInfo(None, "Face restoration model", gr.Radio, lambda: {"choices": [x.name() for x in face_restorers]}), "code_former_weight": OptionInfo(0.5, "CodeFormer weight parameter; 0 = maximum effect; 1 = minimum effect", gr.Slider, {"minimum": 0, "maximum": 1, "step": 0.01}), "save_images_before_face_restoration": OptionInfo(False, "Save a copy of image before doing face restoration."), @@ -217,3 +219,6 @@ class TotalTQDM: total_tqdm = TotalTQDM() + +mem_mon = modules.memmon.MemUsageMonitor("MemMon", device, opts) +mem_mon.start() diff --git a/modules/ui.py b/modules/ui.py index 738ac945..01b2ba85 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -119,6 +119,7 @@ def save_files(js_data, images, index): def wrap_gradio_call(func): def f(*args, **kwargs): + shared.mem_mon.monitor() t = time.perf_counter() try: @@ -135,8 +136,19 @@ def wrap_gradio_call(func): elapsed = time.perf_counter() - t + mem_stats = {k:-(v//-(1024*1024)) for k,v in shared.mem_mon.stop().items()} + active_peak = mem_stats['active_peak'] + reserved_peak = mem_stats['reserved_peak'] + sys_peak = '?' if opts.memmon_poll_rate <= 0 else mem_stats['system_peak'] + sys_total = mem_stats['total'] + sys_pct = '?' if opts.memmon_poll_rate <= 0 else round(sys_peak/sys_total * 100, 2) + vram_tooltip = "Torch active: Peak amount of VRAM used by Torch during generation, excluding cached data. " \ + "Torch reserved: Peak amount of VRAM allocated by Torch, including all active and cached data. " \ + "Sys VRAM: Peak amount of VRAM allocation across all applications / total GPU VRAM (peak utilization%)." + # last item is always HTML - res[-1] = res[-1] + f"

Time taken: {elapsed:.2f}s

" + res[-1] += f"

Time taken: {elapsed:.2f}s

" \ + f"

Torch active/reserved: {active_peak}/{reserved_peak} MiB, Sys VRAM: {sys_peak}/{sys_total} MiB ({sys_pct}%)

" shared.state.interrupted = False diff --git a/style.css b/style.css index d41c098c..67ce8550 100644 --- a/style.css +++ b/style.css @@ -1,5 +1,21 @@ .output-html p {margin: 0 0.5em;} -.performance { font-size: 0.85em; color: #444; } + +.performance { + font-size: 0.85em; + color: #444; + display: flex; + justify-content: space-between; + white-space: nowrap; +} + +.performance .time { + margin-right: 0; +} + +.performance .vram { + margin-left: 0; + text-align: right; +} #generate{ min-height: 4.5em; From b8be33dad13d4937c6ef8fbb49715d843c3dd586 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 17 Sep 2022 09:23:31 +0300 Subject: [PATCH 10/20] hide VRAM text if polling is disabled --- modules/ui.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/modules/ui.py b/modules/ui.py index 01b2ba85..437bce66 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -136,7 +136,7 @@ def wrap_gradio_call(func): elapsed = time.perf_counter() - t - mem_stats = {k:-(v//-(1024*1024)) for k,v in shared.mem_mon.stop().items()} + mem_stats = {k: -(v//-(1024*1024)) for k,v in shared.mem_mon.stop().items()} active_peak = mem_stats['active_peak'] reserved_peak = mem_stats['reserved_peak'] sys_peak = '?' if opts.memmon_poll_rate <= 0 else mem_stats['system_peak'] @@ -146,9 +146,10 @@ def wrap_gradio_call(func): "Torch reserved: Peak amount of VRAM allocated by Torch, including all active and cached data. " \ "Sys VRAM: Peak amount of VRAM allocation across all applications / total GPU VRAM (peak utilization%)." + vram_html = '' if opts.memmon_poll_rate == 0 else f"

Torch active/reserved: {active_peak}/{reserved_peak} MiB, Sys VRAM: {sys_peak}/{sys_total} MiB ({sys_pct}%)

" + # last item is always HTML - res[-1] += f"

Time taken: {elapsed:.2f}s

" \ - f"

Torch active/reserved: {active_peak}/{reserved_peak} MiB, Sys VRAM: {sys_peak}/{sys_total} MiB ({sys_pct}%)

" + res[-1] += f"

Time taken: {elapsed:.2f}s

{vram_html}
" shared.state.interrupted = False From 247f58a5e740a7bd3980815961425b778d77ec28 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 17 Sep 2022 12:05:04 +0300 Subject: [PATCH 11/20] add support for switching model checkpoints at runtime --- modules/images.py | 2 +- modules/processing.py | 2 +- modules/sd_models.py | 148 ++++++++++++++++++++++++++++++++++++++++++ modules/shared.py | 19 ++++-- modules/ui.py | 5 ++ webui.py | 67 ++++--------------- 6 files changed, 182 insertions(+), 61 deletions(-) create mode 100644 modules/sd_models.py diff --git a/modules/images.py b/modules/images.py index b62c48f8..a3064333 100644 --- a/modules/images.py +++ b/modules/images.py @@ -274,7 +274,7 @@ def apply_filename_pattern(x, p, seed, prompt): x = x.replace("[height]", str(p.height)) x = x.replace("[sampler]", sd_samplers.samplers[p.sampler_index].name) - x = x.replace("[model_hash]", shared.sd_model_hash) + x = x.replace("[model_hash]", shared.sd_model.sd_model_hash) x = x.replace("[date]", datetime.date.today().isoformat()) if cmd_opts.hide_ui_dir_config: diff --git a/modules/processing.py b/modules/processing.py index 81c83f06..3a4ff224 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -227,7 +227,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: "Seed": all_seeds[index], "Face restoration": (opts.face_restoration_model if p.restore_faces else None), "Size": f"{p.width}x{p.height}", - "Model hash": (None if not opts.add_model_hash_to_info or not shared.sd_model_hash else shared.sd_model_hash), + "Model hash": (None if not opts.add_model_hash_to_info or not shared.sd_model.sd_model_hash else shared.sd_model.sd_model_hash), "Batch size": (None if p.batch_size < 2 else p.batch_size), "Batch pos": (None if p.batch_size < 2 else position_in_batch), "Variation seed": (None if p.subseed_strength == 0 else all_subseeds[index]), diff --git a/modules/sd_models.py b/modules/sd_models.py new file mode 100644 index 00000000..036af0e4 --- /dev/null +++ b/modules/sd_models.py @@ -0,0 +1,148 @@ +import glob +import os.path +import sys +from collections import namedtuple +import torch +from omegaconf import OmegaConf + + +from ldm.util import instantiate_from_config + +from modules import shared + +CheckpointInfo = namedtuple("CheckpointInfo", ['filename', 'title', 'hash']) +checkpoints_list = {} + +try: + # this silences the annoying "Some weights of the model checkpoint were not used when initializing..." message at start. + + from transformers import logging + + logging.set_verbosity_error() +except Exception: + pass + + +def list_models(): + checkpoints_list.clear() + + model_dir = os.path.abspath(shared.cmd_opts.ckpt_dir) + + def modeltitle(path, h): + abspath = os.path.abspath(path) + + if abspath.startswith(model_dir): + name = abspath.replace(model_dir, '') + else: + name = os.path.basename(path) + + if name.startswith("\\") or name.startswith("/"): + name = name[1:] + + return f'{name} [{h}]' + + cmd_ckpt = shared.cmd_opts.ckpt + if os.path.exists(cmd_ckpt): + h = model_hash(cmd_ckpt) + title = modeltitle(cmd_ckpt, h) + checkpoints_list[title] = CheckpointInfo(cmd_ckpt, title, h) + elif cmd_ckpt is not None and cmd_ckpt != shared.default_sd_model_file: + print(f"Checkpoint in --ckpt argument not found: {cmd_ckpt}", file=sys.stderr) + + if os.path.exists(model_dir): + for filename in glob.glob(model_dir + '/**/*.ckpt', recursive=True): + h = model_hash(filename) + title = modeltitle(filename, h) + checkpoints_list[title] = CheckpointInfo(filename, title, h) + + +def model_hash(filename): + try: + with open(filename, "rb") as file: + import hashlib + m = hashlib.sha256() + + file.seek(0x100000) + m.update(file.read(0x10000)) + return m.hexdigest()[0:8] + except FileNotFoundError: + return 'NOFILE' + + +def select_checkpoint(): + model_checkpoint = shared.opts.sd_model_checkpoint + checkpoint_info = checkpoints_list.get(model_checkpoint, None) + if checkpoint_info is not None: + return checkpoint_info + + if len(checkpoints_list) == 0: + print(f"Checkpoint {model_checkpoint} not found and no other checkpoints found", file=sys.stderr) + return None + + checkpoint_info = next(iter(checkpoints_list.values())) + if model_checkpoint is not None: + print(f"Checkpoint {model_checkpoint} not found; loading fallback {checkpoint_info.title}", file=sys.stderr) + + return checkpoint_info + + +def load_model_weights(model, checkpoint_file, sd_model_hash): + print(f"Loading weights [{sd_model_hash}] from {checkpoint_file}") + + pl_sd = torch.load(checkpoint_file, map_location="cpu") + if "global_step" in pl_sd: + print(f"Global Step: {pl_sd['global_step']}") + sd = pl_sd["state_dict"] + + model.load_state_dict(sd, strict=False) + + if shared.cmd_opts.opt_channelslast: + model.to(memory_format=torch.channels_last) + + if not shared.cmd_opts.no_half: + model.half() + + model.sd_model_hash = sd_model_hash + model.sd_model_checkpint = checkpoint_file + + +def load_model(): + from modules import lowvram, sd_hijack + checkpoint_info = select_checkpoint() + + sd_config = OmegaConf.load(shared.cmd_opts.config) + sd_model = instantiate_from_config(sd_config.model) + load_model_weights(sd_model, checkpoint_info.filename, checkpoint_info.hash) + + if shared.cmd_opts.lowvram or shared.cmd_opts.medvram: + lowvram.setup_for_low_vram(sd_model, shared.cmd_opts.medvram) + else: + sd_model.to(shared.device) + + sd_hijack.model_hijack.hijack(sd_model) + + sd_model.eval() + + print(f"Model loaded.") + return sd_model + + +def reload_model_weights(sd_model): + from modules import lowvram, devices + checkpoint_info = select_checkpoint() + + if sd_model.sd_model_checkpint == checkpoint_info.filename: + return + + if shared.cmd_opts.lowvram or shared.cmd_opts.medvram: + lowvram.send_everything_to_cpu() + else: + sd_model.to(devices.cpu) + + load_model_weights(sd_model, checkpoint_info.filename, checkpoint_info.hash) + + if not shared.cmd_opts.lowvram and not shared.cmd_opts.medvram: + sd_model.to(devices.device) + + print(f"Weights loaded.") + return sd_model diff --git a/modules/shared.py b/modules/shared.py index 4f877036..3c3aa9b6 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -13,14 +13,15 @@ from modules.devices import get_optimal_device import modules.styles import modules.interrogate import modules.memmon +import modules.sd_models sd_model_file = os.path.join(script_path, 'model.ckpt') -if not os.path.exists(sd_model_file): - sd_model_file = "models/ldm/stable-diffusion-v1/model.ckpt" +default_sd_model_file = sd_model_file parser = argparse.ArgumentParser() parser.add_argument("--config", type=str, default=os.path.join(sd_path, "configs/stable-diffusion/v1-inference.yaml"), help="path to config which constructs model",) -parser.add_argument("--ckpt", type=str, default=os.path.join(sd_path, sd_model_file), help="path to checkpoint of model",) +parser.add_argument("--ckpt", type=str, default=sd_model_file, help="path to checkpoint of stable diffusion model; this checkpoint will be added to the list of checkpoints and loaded by default if you don't have a checkpoint selected in settings",) +parser.add_argument("--ckpt-dir", type=str, default=os.path.join(script_path, 'models'), help="path to directory with stable diffusion checkpoints",) parser.add_argument("--gfpgan-dir", type=str, help="GFPGAN directory", default=('./src/gfpgan' if os.path.exists('./src/gfpgan') else './GFPGAN')) parser.add_argument("--gfpgan-model", type=str, help="GFPGAN model file name", default='GFPGANv1.3.pth') parser.add_argument("--no-half", action='store_true', help="do not switch the model to 16-bit floats") @@ -88,13 +89,17 @@ interrogator = modules.interrogate.InterrogateModels("interrogate") face_restorers = [] +modules.sd_models.list_models() + + class Options: class OptionInfo: - def __init__(self, default=None, label="", component=None, component_args=None): + def __init__(self, default=None, label="", component=None, component_args=None, onchange=None): self.default = default self.label = label self.component = component self.component_args = component_args + self.onchange = onchange data = None hide_dirs = {"visible": False} if cmd_opts.hide_ui_dir_config else None @@ -150,6 +155,7 @@ class Options: "interrogate_clip_min_length": OptionInfo(24, "Interrogate: minimum description length (excluding artists, etc..)", gr.Slider, {"minimum": 1, "maximum": 128, "step": 1}), "interrogate_clip_max_length": OptionInfo(48, "Interrogate: maximum description length", gr.Slider, {"minimum": 1, "maximum": 256, "step": 1}), "interrogate_clip_dict_limit": OptionInfo(1500, "Interrogate: maximum number of lines in text file (0 = No limit)"), + "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Radio, lambda: {"choices": [x.title for x in modules.sd_models.checkpoints_list.values()]}), } def __init__(self): @@ -180,6 +186,10 @@ class Options: with open(filename, "r", encoding="utf8") as file: self.data = json.load(file) + def onchange(self, key, func): + item = self.data_labels.get(key) + item.onchange = func + opts = Options() if os.path.exists(config_filename): @@ -188,7 +198,6 @@ if os.path.exists(config_filename): sd_upscalers = [] sd_model = None -sd_model_hash = '' progress_print_out = sys.stdout diff --git a/modules/ui.py b/modules/ui.py index 437bce66..36e3c664 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -758,7 +758,12 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo): if comp_args and isinstance(comp_args, dict) and comp_args.get('visible') is False: continue + oldval = opts.data.get(key, None) opts.data[key] = value + + if oldval != value and opts.data_labels[key].onchange is not None: + opts.data_labels[key].onchange() + up.append(comp.update(value=value)) opts.save(shared.config_filename) diff --git a/webui.py b/webui.py index add72123..ff8997db 100644 --- a/webui.py +++ b/webui.py @@ -3,13 +3,8 @@ import threading from modules.paths import script_path -import torch -from omegaconf import OmegaConf - import signal -from ldm.util import instantiate_from_config - from modules.shared import opts, cmd_opts, state import modules.shared as shared import modules.ui @@ -24,6 +19,7 @@ import modules.extras import modules.lowvram import modules.txt2img import modules.img2img +import modules.sd_models modules.codeformer_model.setup_codeformer() @@ -33,31 +29,19 @@ shared.face_restorers.append(modules.face_restoration.FaceRestoration()) esrgan.load_models(cmd_opts.esrgan_models_path) realesrgan.setup_realesrgan() - -def load_model_from_config(config, ckpt, verbose=False): - print(f"Loading model [{shared.sd_model_hash}] from {ckpt}") - pl_sd = torch.load(ckpt, map_location="cpu") - if "global_step" in pl_sd: - print(f"Global Step: {pl_sd['global_step']}") - sd = pl_sd["state_dict"] - - model = instantiate_from_config(config.model) - m, u = model.load_state_dict(sd, strict=False) - if len(m) > 0 and verbose: - print("missing keys:") - print(m) - if len(u) > 0 and verbose: - print("unexpected keys:") - print(u) - if cmd_opts.opt_channelslast: - model = model.to(memory_format=torch.channels_last) - model.eval() - return model - - queue_lock = threading.Lock() +def wrap_queued_call(func): + def f(*args, **kwargs): + with queue_lock: + res = func(*args, **kwargs) + + return res + + return f + + def wrap_gradio_gpu_call(func): def f(*args, **kwargs): shared.state.sampling_step = 0 @@ -80,33 +64,8 @@ def wrap_gradio_gpu_call(func): modules.scripts.load_scripts(os.path.join(script_path, "scripts")) -try: - # this silences the annoying "Some weights of the model checkpoint were not used when initializing..." message at start. - - from transformers import logging - - logging.set_verbosity_error() -except Exception: - pass - -with open(cmd_opts.ckpt, "rb") as file: - import hashlib - m = hashlib.sha256() - - file.seek(0x100000) - m.update(file.read(0x10000)) - shared.sd_model_hash = m.hexdigest()[0:8] - -sd_config = OmegaConf.load(cmd_opts.config) -shared.sd_model = load_model_from_config(sd_config, cmd_opts.ckpt) -shared.sd_model = (shared.sd_model if cmd_opts.no_half else shared.sd_model.half()) - -if cmd_opts.lowvram or cmd_opts.medvram: - modules.lowvram.setup_for_low_vram(shared.sd_model, cmd_opts.medvram) -else: - shared.sd_model = shared.sd_model.to(shared.device) - -modules.sd_hijack.model_hijack.hijack(shared.sd_model) +shared.sd_model = modules.sd_models.load_model() +shared.opts.onchange("sd_model_checkpoint", wrap_queued_call(lambda: modules.sd_models.reload_model_weights(shared.sd_model))) def webui(): From f8f17e3b9e61f238dd32b6d1bab5db040c531559 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 17 Sep 2022 12:12:55 +0300 Subject: [PATCH 12/20] updated readme to reflect new model location --- README.md | 2 +- models/Put Stable Diffusion checkpoints here.txt | 0 2 files changed, 1 insertion(+), 1 deletion(-) create mode 100644 models/Put Stable Diffusion checkpoints here.txt diff --git a/README.md b/README.md index 6c232e38..d97ebc3f 100644 --- a/README.md +++ b/README.md @@ -51,7 +51,7 @@ Alternatively, use [Google Colab](https://colab.research.google.com/drive/1Iy-xW 1. Install [Python 3.10.6](https://www.python.org/downloads/windows/), checking "Add Python to PATH" 2. Install [git](https://git-scm.com/download/win). 3. Download the stable-diffusion-webui repository, for example by running `git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui.git`. -4. Place `model.ckpt` in the base directory, alongside `webui.py`. +4. Place `model.ckpt` in the `models` directory. 5. _*(Optional)*_ Place `GFPGANv1.3.pth` in the base directory, alongside `webui.py`. 6. Run `webui-user.bat` from Windows Explorer as normal, non-administrate, user. diff --git a/models/Put Stable Diffusion checkpoints here.txt b/models/Put Stable Diffusion checkpoints here.txt new file mode 100644 index 00000000..e69de29b From 99585b3514e2d7e987651d5c6a0806f933af012b Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 17 Sep 2022 12:38:15 +0300 Subject: [PATCH 13/20] moved progressbar to top by request --- modules/ui.py | 7 ++++--- style.css | 6 ++++++ 2 files changed, 10 insertions(+), 3 deletions(-) diff --git a/modules/ui.py b/modules/ui.py index 36e3c664..960f1e36 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -337,6 +337,8 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo): custom_inputs = modules.scripts.scripts_txt2img.setup_ui(is_img2img=False) with gr.Column(variant='panel'): + progressbar = gr.HTML(elem_id="progressbar") + with gr.Group(): txt2img_preview = gr.Image(elem_id='txt2img_preview', visible=False) txt2img_gallery = gr.Gallery(label='Output', elem_id='txt2img_gallery').style(grid=4) @@ -349,8 +351,6 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo): send_to_extras = gr.Button('Send to extras') interrupt = gr.Button('Interrupt') - progressbar = gr.HTML(elem_id="progressbar") - with gr.Group(): html_info = gr.HTML() generation_info = gr.Textbox(visible=False) @@ -474,6 +474,8 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo): custom_inputs = modules.scripts.scripts_img2img.setup_ui(is_img2img=True) with gr.Column(variant='panel'): + progressbar = gr.HTML(elem_id="progressbar") + with gr.Group(): img2img_preview = gr.Image(elem_id='img2img_preview', visible=False) img2img_gallery = gr.Gallery(label='Output', elem_id='img2img_gallery').style(grid=4) @@ -487,7 +489,6 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo): interrupt = gr.Button('Interrupt') img2img_save_style = gr.Button('Save prompt as style') - progressbar = gr.HTML(elem_id="progressbar") with gr.Group(): html_info = gr.HTML() diff --git a/style.css b/style.css index 67ce8550..752d2cf4 100644 --- a/style.css +++ b/style.css @@ -167,6 +167,12 @@ input[type="range"]{ #txt2img_negative_prompt, #img2img_negative_prompt{ } +#progressbar{ + position: absolute; + z-index: 1000; + right: 0; +} + .progressDiv{ width: 100%; height: 30px; From 304222ef94d1c3c60fab466a96c448868f391bce Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 17 Sep 2022 13:49:36 +0300 Subject: [PATCH 14/20] X/Y plot support for switching checkpoints. --- modules/sd_models.py | 4 ++-- script.js | 2 ++ scripts/xy_grid.py | 15 +++++++++++++++ 3 files changed, 19 insertions(+), 2 deletions(-) diff --git a/modules/sd_models.py b/modules/sd_models.py index 036af0e4..4bd70fc5 100644 --- a/modules/sd_models.py +++ b/modules/sd_models.py @@ -127,9 +127,9 @@ def load_model(): return sd_model -def reload_model_weights(sd_model): +def reload_model_weights(sd_model, info=None): from modules import lowvram, devices - checkpoint_info = select_checkpoint() + checkpoint_info = info or select_checkpoint() if sd_model.sd_model_checkpint == checkpoint_info.filename: return diff --git a/script.js b/script.js index 4a70e51d..e63e0695 100644 --- a/script.js +++ b/script.js @@ -66,6 +66,8 @@ titles = { "Style 2": "Style to apply; styles have components for both positive and negative prompts and apply to both", "Apply style": "Insert selected styles into prompt fields", "Create style": "Save current prompts as a style. If you add the token {prompt} to the text, the style use that as placeholder for your prompt when you use the style in the future.", + + "Checkpoint name": "Loads weights from checkpoint before making images. You can either use hash or a part of filename (as seen in settings) for checkpoint name. Recommended to use with Y axis for less switching.", } function gradioApp(){ diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index eccfda87..680dd702 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -10,7 +10,9 @@ import gradio as gr from modules import images from modules.processing import process_images, Processed from modules.shared import opts, cmd_opts, state +import modules.shared as shared import modules.sd_samplers +import modules.sd_models import re @@ -41,6 +43,15 @@ def apply_sampler(p, x, xs): p.sampler_index = sampler_index +def apply_checkpoint(p, x, xs): + applicable = [info for info in modules.sd_models.checkpoints_list.values() if x in info.title] + assert len(applicable) > 0, f'Checkpoint {x} for found' + + info = applicable[0] + + modules.sd_models.reload_model_weights(shared.sd_model, info) + + def format_value_add_label(p, opt, x): if type(x) == float: x = round(x, 8) @@ -74,6 +85,7 @@ axis_options = [ AxisOption("CFG Scale", float, apply_field("cfg_scale"), format_value_add_label), AxisOption("Prompt S/R", str, apply_prompt, format_value), AxisOption("Sampler", str, apply_sampler, format_value), + AxisOption("Checkpoint name", str, apply_checkpoint, format_value), AxisOptionImg2Img("Denoising", float, apply_field("denoising_strength"), format_value_add_label), # as it is now all AxisOptionImg2Img items must go after AxisOption ones ] @@ -215,4 +227,7 @@ class Script(scripts.Script): if opts.grid_save: images.save_image(processed.images[0], p.outpath_grids, "xy_grid", prompt=p.prompt, seed=processed.seed, grid=True, p=p) + # restore checkpoint in case it was changed by axes + modules.sd_models.reload_model_weights(shared.sd_model) + return processed From 140f89315380dbcc541f6e18e3d355a06ea3e2f0 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 17 Sep 2022 14:55:40 +0300 Subject: [PATCH 15/20] process all values for x/y plot right away to error out if any are bad before any processing begins --- scripts/xy_grid.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 680dd702..6a157722 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -90,11 +90,11 @@ axis_options = [ ] -def draw_xy_grid(p, xs, ys, x_label, y_label, cell, draw_legend): +def draw_xy_grid(p, xs, ys, x_labels, y_labels, cell, draw_legend): res = [] - ver_texts = [[images.GridAnnotation(y_label(y))] for y in ys] - hor_texts = [[images.GridAnnotation(x_label(x))] for x in xs] + ver_texts = [[images.GridAnnotation(y)] for y in y_labels] + hor_texts = [[images.GridAnnotation(x)] for x in x_labels] first_pocessed = None @@ -218,8 +218,8 @@ class Script(scripts.Script): p, xs=xs, ys=ys, - x_label=lambda x: x_opt.format_value(p, x_opt, x), - y_label=lambda y: y_opt.format_value(p, y_opt, y), + x_labels=[x_opt.format_value(p, x_opt, x) for x in xs], + y_labels=[y_opt.format_value(p, y_opt, y) for y in ys], cell=cell, draw_legend=draw_legend ) From ba295b32688629cf575d67f1750a7838b008858b Mon Sep 17 00:00:00 2001 From: Tony Beeman Date: Sat, 17 Sep 2022 01:34:33 -0700 Subject: [PATCH 16/20] * Fix process_images where the number of images is not a multiple of (batch_size * n_iter), which would cause us to throw an exception. * Add a textbox option to Prompts from file (ease of use and it makes it much easier to use on a mobile device) * Fix the fact that Prompts from file was sometimes passing an empty batch. --- modules/processing.py | 9 ++++++++- scripts/prompts_from_file.py | 34 ++++++++++++++++++++++++---------- 2 files changed, 32 insertions(+), 11 deletions(-) diff --git a/modules/processing.py b/modules/processing.py index 3a4ff224..6a99d383 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -188,7 +188,11 @@ def fix_seed(p): def process_images(p: StableDiffusionProcessing) -> Processed: """this is the main loop that both txt2img and img2img use; it calls func_init once inside all the scopes and func_sample once per batch""" - assert p.prompt is not None + if type(p.prompt) == list: + assert(len(p.prompt) > 0) + else: + assert p.prompt is not None + devices.torch_gc() fix_seed(p) @@ -265,6 +269,9 @@ def process_images(p: StableDiffusionProcessing) -> Processed: seeds = all_seeds[n * p.batch_size:(n + 1) * p.batch_size] subseeds = all_subseeds[n * p.batch_size:(n + 1) * p.batch_size] + if (len(prompts) == 0): + break + #uc = p.sd_model.get_learned_conditioning(len(prompts) * [p.negative_prompt]) #c = p.sd_model.get_learned_conditioning(prompts) uc = prompt_parser.get_learned_conditioning(len(prompts) * [p.negative_prompt], p.steps) diff --git a/scripts/prompts_from_file.py b/scripts/prompts_from_file.py index d9b01c81..513d9a1c 100644 --- a/scripts/prompts_from_file.py +++ b/scripts/prompts_from_file.py @@ -13,28 +13,42 @@ from modules.shared import opts, cmd_opts, state class Script(scripts.Script): def title(self): - return "Prompts from file" + return "Prompts from file or textbox" def ui(self, is_img2img): + # This checkbox would look nicer as two tabs, but there are two problems: + # 1) There is a bug in Gradio 3.3 that prevents visibility from working on Tabs + # 2) Even with Gradio 3.3.1, returning a control (like Tabs) that can't be used as input + # causes a AttributeError: 'Tabs' object has no attribute 'preprocess' assert, + # due to the way Script assumes all controls returned can be used as inputs. + # Therefore, there's no good way to use grouping components right now, + # so we will use a checkbox! :) + checkbox_txt = gr.Checkbox(label="Show Textbox", value=False) file = gr.File(label="File with inputs", type='bytes') + prompt_txt = gr.TextArea(label="Prompts") + checkbox_txt.change(fn=lambda x: [gr.File.update(visible = not x), gr.TextArea.update(visible = x)], inputs=[checkbox_txt], outputs=[file, prompt_txt]) + return [checkbox_txt, file, prompt_txt] - return [file] - - def run(self, p, data: bytes): - lines = [x.strip() for x in data.decode('utf8', errors='ignore').split("\n")] + def run(self, p, checkbox_txt, data: bytes, prompt_txt: str): + if (checkbox_txt): + lines = [x.strip() for x in prompt_txt.splitlines()] + else: + lines = [x.strip() for x in data.decode('utf8', errors='ignore').split("\n")] lines = [x for x in lines if len(x) > 0] - batch_count = math.ceil(len(lines) / p.batch_size) - print(f"Will process {len(lines) * p.n_iter} images in {batch_count * p.n_iter} batches.") + img_count = len(lines) * p.n_iter + batch_count = math.ceil(img_count / p.batch_size) + loop_count = math.ceil(batch_count / p.n_iter) + print(f"Will process {img_count} images in {batch_count} batches.") p.do_not_save_grid = True state.job_count = batch_count images = [] - for batch_no in range(batch_count): - state.job = f"{batch_no + 1} out of {batch_count * p.n_iter}" - p.prompt = lines[batch_no*p.batch_size:(batch_no+1)*p.batch_size] * p.n_iter + for loop_no in range(loop_count): + state.job = f"{loop_no + 1} out of {loop_count}" + p.prompt = lines[loop_no*p.batch_size:(loop_no+1)*p.batch_size] * p.n_iter proc = process_images(p) images += proc.images From 65be5312dc2b73e659299ea052d5484e6ae6c0ea Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 17 Sep 2022 01:00:45 +0100 Subject: [PATCH 17/20] Add modal css classes --- style.css | 37 +++++++++++++++++++++++++++++++++++++ 1 file changed, 37 insertions(+) diff --git a/style.css b/style.css index 752d2cf4..2de83a94 100644 --- a/style.css +++ b/style.css @@ -196,3 +196,40 @@ input[type="range"]{ border-radius: 8px; } +#lightboxModal{ + display: none; + position: fixed; + z-index: 900; + padding-top: 100px; + left: 0; + top: 0; + width: 100%; + height: 100%; + overflow: auto; + background-color: black; +} + +.modalClose { + color: white; + position: absolute; + top: 10px; + right: 25px; + font-size: 35px; + font-weight: bold; +} + +.modalClose:hover, +.modalClose:focus { + color: #999; + text-decoration: none; + cursor: pointer; +} + +#modalImage { + display: block; + margin-left: auto; + margin-right: auto; + margin-top: auto; + width: auto; +} + From 1a513370774ccb4cd9562f1b40048adc2ab7c896 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 17 Sep 2022 01:03:03 +0100 Subject: [PATCH 18/20] Add modal creation and functions --- script.js | 60 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 60 insertions(+) diff --git a/script.js b/script.js index e63e0695..7c27da74 100644 --- a/script.js +++ b/script.js @@ -76,6 +76,35 @@ function gradioApp(){ global_progressbar = null +function closeModal() { + gradioApp().getElementById("lightboxModal").style.display = "none"; +} + +function showModal(elem) { + gradioApp().getElementById("modalImage").src = elem.src + gradioApp().getElementById("lightboxModal").style.display = "block"; +} + +function showGalleryImage(){ + setTimeout(function() { + fullImg_preview = gradioApp().querySelectorAll('img.w-full.object-contain') + + if(fullImg_preview != null){ + fullImg_preview.forEach(function function_name(e) { + if(e && e.parentElement.tagName == 'DIV'){ + e.style.cursor='pointer' + + elemfunc = function(elem){ + elem.onclick = function(){showModal(elem)}; + } + elemfunc(e) + } + }); + } + + }, 100); +} + function addTitles(root){ root.querySelectorAll('span, button, select').forEach(function(span){ tooltip = titles[span.textContent]; @@ -117,8 +146,18 @@ function addTitles(root){ img2img_preview.style.width = img2img_gallery.clientWidth + "px" img2img_preview.style.height = img2img_gallery.clientHeight + "px" } + + fullImg_preview = gradioApp().querySelectorAll('img.w-full') + if(fullImg_preview != null){ + fullImg_preview.forEach(function function_name(e) { + if(e && e.parentElement.tagName == 'BUTTON'){ + e.onclick = showGalleryImage; + } + }); + } + window.setTimeout(requestProgress, 500) }); mutationObserver.observe( progressbar, { childList:true, subtree:true }) @@ -131,6 +170,27 @@ document.addEventListener("DOMContentLoaded", function() { addTitles(gradioApp()); }); mutationObserver.observe( gradioApp(), { childList:true, subtree:true }) + + const modalFragment = document.createDocumentFragment(); + const modal = document.createElement('div') + modal.onclick = closeModal; + + const modalClose = document.createElement('span') + modalClose.className = 'modalClose cursor'; + modalClose.innerHTML = '×' + modalClose.onclick = closeModal; + modal.id = "lightboxModal"; + modal.appendChild(modalClose) + + const modalImage = document.createElement('img') + modalImage.id = 'modalImage'; + modalImage.onclick = closeModal; + modal.appendChild(modalImage) + + gradioApp().getRootNode().appendChild(modal) + + document.body.appendChild(modalFragment); + }); function selected_gallery_index(){ From a66d857345c090674430c21fba1256c76d769635 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 17 Sep 2022 01:13:28 +0100 Subject: [PATCH 19/20] make background semi-transparent not black; --- style.css | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/style.css b/style.css index 2de83a94..2bdd1e0e 100644 --- a/style.css +++ b/style.css @@ -206,7 +206,7 @@ input[type="range"]{ width: 100%; height: 100%; overflow: auto; - background-color: black; + background-color: rgba(20, 20, 20, 0.95); } .modalClose { From 1ef79f926e6314b3ef9308b12ff7ad482afd790a Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 17 Sep 2022 03:26:47 +0100 Subject: [PATCH 20/20] generalise to work on all non-masked images on all tabs --- script.js | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) diff --git a/script.js b/script.js index 7c27da74..a016eb4e 100644 --- a/script.js +++ b/script.js @@ -105,6 +105,12 @@ function showGalleryImage(){ }, 100); } +function galleryImageHandler(e){ + if(e && e.parentElement.tagName == 'BUTTON'){ + e.onclick = showGalleryImage; + } +} + function addTitles(root){ root.querySelectorAll('span, button, select').forEach(function(span){ tooltip = titles[span.textContent]; @@ -147,22 +153,17 @@ function addTitles(root){ img2img_preview.style.height = img2img_gallery.clientHeight + "px" } - fullImg_preview = gradioApp().querySelectorAll('img.w-full') - - if(fullImg_preview != null){ - - fullImg_preview.forEach(function function_name(e) { - if(e && e.parentElement.tagName == 'BUTTON'){ - e.onclick = showGalleryImage; - } - }); - } - window.setTimeout(requestProgress, 500) }); mutationObserver.observe( progressbar, { childList:true, subtree:true }) } + + fullImg_preview = gradioApp().querySelectorAll('img.w-full') + if(fullImg_preview != null){ + fullImg_preview.forEach(galleryImageHandler); + } + } document.addEventListener("DOMContentLoaded", function() {