From 9e737cbadcdc89c23b119701815275e7c209ff00 Mon Sep 17 00:00:00 2001 From: Alexandre Simard Date: Mon, 26 Sep 2022 17:18:57 -0400 Subject: [PATCH 001/138] Solve issue #962 Fix by @MrAcademy --- .gitignore | 3 ++- javascript/ui.js | 5 ++--- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/.gitignore b/.gitignore index 9d78853af..fa1ab43e7 100644 --- a/.gitignore +++ b/.gitignore @@ -19,4 +19,5 @@ __pycache__ /webui-user.sh /interrogate /user.css -/.idea \ No newline at end of file +/.idea +/SwinIR diff --git a/javascript/ui.js b/javascript/ui.js index 076e9436c..7db4db48d 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -1,9 +1,8 @@ // various functions for interation with ui.py not large enough to warrant putting them in separate files function selected_gallery_index(){ - var gr = gradioApp() - var buttons = gradioApp().querySelectorAll(".gallery-item") - var button = gr.querySelector(".gallery-item.\\!ring-2") + var buttons = gradioApp().querySelectorAll('[style="display: block;"].tabitem .gallery-item') + var button = gradioApp().querySelector('[style="display: block;"].tabitem .gallery-item.\\!ring-2') var result = -1 buttons.forEach(function(v, i){ if(v==button) { result = i } }) From 03ee67bfd34b9e872b33eb05fef5db83410b16f3 Mon Sep 17 00:00:00 2001 From: WDevelopsWebApps <97454358+WDevelopsWebApps@users.noreply.github.com> Date: Wed, 28 Sep 2022 10:53:40 +0200 Subject: [PATCH 002/138] add advanced saving for save button --- modules/images.py | 5 ++++- modules/ui.py | 35 ++++++++++++++++++++++++++++------- 2 files changed, 32 insertions(+), 8 deletions(-) diff --git a/modules/images.py b/modules/images.py index 9458bf8d4..923f81dfb 100644 --- a/modules/images.py +++ b/modules/images.py @@ -290,7 +290,10 @@ def apply_filename_pattern(x, p, seed, prompt): x = x.replace("[cfg]", str(p.cfg_scale)) x = x.replace("[width]", str(p.width)) x = x.replace("[height]", str(p.height)) - x = x.replace("[styles]", sanitize_filename_part(", ".join(p.styles), replace_spaces=False)) + #currently disabled if using the save button, will work otherwise + # if enabled it will cause a bug because styles is not included in the save_files data dictionary + if hasattr(p, "styles"): + x = x.replace("[styles]", sanitize_filename_part(", ".join(p.styles), replace_spaces=False)) x = x.replace("[sampler]", sanitize_filename_part(sd_samplers.samplers[p.sampler_index].name, replace_spaces=False)) x = x.replace("[model_hash]", shared.sd_model.sd_model_hash) diff --git a/modules/ui.py b/modules/ui.py index 7db8edbd8..87a86a45d 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -28,6 +28,7 @@ import modules.gfpgan_model import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste +from modules.images import apply_filename_pattern, get_next_sequence_number # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI mimetypes.init() @@ -90,13 +91,26 @@ def send_gradio_gallery_to_image(x): def save_files(js_data, images, index): - import csv - - os.makedirs(opts.outdir_save, exist_ok=True) - + import csv filenames = [] + #quick dictionary to class object conversion. Its neccesary due apply_filename_pattern requiring it + class MyObject: + def __init__(self, d=None): + if d is not None: + for key, value in d.items(): + setattr(self, key, value) + data = json.loads(js_data) + p = MyObject(data) + path = opts.outdir_save + save_to_dirs = opts.save_to_dirs + + if save_to_dirs: + dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, p.seed, p.prompt) + path = os.path.join(opts.outdir_save, dirname) + + os.makedirs(path, exist_ok=True) if index > -1 and opts.save_selected_only and (index > 0 or not opts.return_grid): # ensures we are looking at a specific non-grid picture, and we have save_selected_only images = [images[index]] @@ -107,11 +121,18 @@ def save_files(js_data, images, index): writer = csv.writer(file) if at_start: writer.writerow(["prompt", "seed", "width", "height", "sampler", "cfgs", "steps", "filename", "negative_prompt"]) + file_decoration = opts.samples_filename_pattern or "[seed]-[prompt_spaces]" + if file_decoration != "": + file_decoration = "-" + file_decoration.lower() + file_decoration = apply_filename_pattern(file_decoration, p, p.seed, p.prompt) + truncated = (file_decoration[:240] + '..') if len(file_decoration) > 240 else file_decoration + filename_base = truncated - filename_base = str(int(time.time() * 1000)) + basecount = get_next_sequence_number(path, "") for i, filedata in enumerate(images): - filename = filename_base + ("" if len(images) == 1 else "-" + str(i + 1)) + ".png" - filepath = os.path.join(opts.outdir_save, filename) + file_number = f"{basecount+i:05}" + filename = file_number + filename_base + ".png" + filepath = os.path.join(path, filename) if filedata.startswith("data:image/png;base64,"): filedata = filedata[len("data:image/png;base64,"):] From c938679de7b87b4f14894d9f57fe0f40dd6e3c06 Mon Sep 17 00:00:00 2001 From: Jairo Correa Date: Wed, 28 Sep 2022 22:14:13 -0300 Subject: [PATCH 003/138] Fix memory leak and reduce memory usage --- modules/codeformer_model.py | 6 ++++-- modules/devices.py | 3 ++- modules/extras.py | 2 ++ modules/gfpgan_model.py | 11 +++++------ modules/processing.py | 33 ++++++++++++++++++++++++++------- webui.py | 3 +++ 6 files changed, 42 insertions(+), 16 deletions(-) diff --git a/modules/codeformer_model.py b/modules/codeformer_model.py index 8fbdea249..2177291a7 100644 --- a/modules/codeformer_model.py +++ b/modules/codeformer_model.py @@ -89,7 +89,7 @@ def setup_codeformer(): output = self.net(cropped_face_t, w=w if w is not None else shared.opts.code_former_weight, adain=True)[0] restored_face = tensor2img(output, rgb2bgr=True, min_max=(-1, 1)) del output - torch.cuda.empty_cache() + devices.torch_gc() except Exception as error: print(f'\tFailed inference for CodeFormer: {error}', file=sys.stderr) restored_face = tensor2img(cropped_face_t, rgb2bgr=True, min_max=(-1, 1)) @@ -106,7 +106,9 @@ def setup_codeformer(): restored_img = cv2.resize(restored_img, (0, 0), fx=original_resolution[1]/restored_img.shape[1], fy=original_resolution[0]/restored_img.shape[0], interpolation=cv2.INTER_LINEAR) if shared.opts.face_restoration_unload: - self.net.to(devices.cpu) + self.net = None + self.face_helper = None + devices.torch_gc() return restored_img diff --git a/modules/devices.py b/modules/devices.py index 07bb23397..df63dd88e 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -1,4 +1,5 @@ import torch +import gc # has_mps is only available in nightly pytorch (for now), `getattr` for compatibility from modules import errors @@ -17,8 +18,8 @@ def get_optimal_device(): return cpu - def torch_gc(): + gc.collect() if torch.cuda.is_available(): torch.cuda.empty_cache() torch.cuda.ipc_collect() diff --git a/modules/extras.py b/modules/extras.py index 9a825530f..38b861675 100644 --- a/modules/extras.py +++ b/modules/extras.py @@ -98,6 +98,8 @@ def run_extras(extras_mode, image, image_folder, gfpgan_visibility, codeformer_v outputs.append(image) + devices.torch_gc() + return outputs, plaintext_to_html(info), '' diff --git a/modules/gfpgan_model.py b/modules/gfpgan_model.py index 44c5dc6ca..b1288f0ca 100644 --- a/modules/gfpgan_model.py +++ b/modules/gfpgan_model.py @@ -49,6 +49,7 @@ def gfpgan(): def gfpgan_fix_faces(np_image): + global loaded_gfpgan_model model = gfpgan() np_image_bgr = np_image[:, :, ::-1] @@ -56,7 +57,9 @@ def gfpgan_fix_faces(np_image): np_image = gfpgan_output_bgr[:, :, ::-1] if shared.opts.face_restoration_unload: - model.gfpgan.to(devices.cpu) + del model + loaded_gfpgan_model = None + devices.torch_gc() return np_image @@ -83,11 +86,7 @@ def setup_gfpgan(): return "GFPGAN" def restore(self, np_image): - np_image_bgr = np_image[:, :, ::-1] - cropped_faces, restored_faces, gfpgan_output_bgr = gfpgan().enhance(np_image_bgr, has_aligned=False, only_center_face=False, paste_back=True) - np_image = gfpgan_output_bgr[:, :, ::-1] - - return np_image + return gfpgan_fix_faces(np_image) shared.face_restorers.append(FaceRestorerGFPGAN()) except Exception: diff --git a/modules/processing.py b/modules/processing.py index 4ecdfcd2d..de5cda793 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -12,7 +12,7 @@ import cv2 from skimage import exposure import modules.sd_hijack -from modules import devices, prompt_parser, masking +from modules import devices, prompt_parser, masking, lowvram from modules.sd_hijack import model_hijack from modules.sd_samplers import samplers, samplers_for_img2img from modules.shared import opts, cmd_opts, state @@ -335,7 +335,8 @@ def process_images(p: StableDiffusionProcessing) -> Processed: if state.job_count == -1: state.job_count = p.n_iter - for n in range(p.n_iter): + for n in range(p.n_iter): + with torch.no_grad(), precision_scope("cuda"), ema_scope(): if state.interrupted: break @@ -368,22 +369,32 @@ def process_images(p: StableDiffusionProcessing) -> Processed: x_samples_ddim = p.sd_model.decode_first_stage(samples_ddim) x_samples_ddim = torch.clamp((x_samples_ddim + 1.0) / 2.0, min=0.0, max=1.0) + del samples_ddim + + if shared.cmd_opts.lowvram or shared.cmd_opts.medvram: + lowvram.send_everything_to_cpu() + + devices.torch_gc() + if opts.filter_nsfw: import modules.safety as safety x_samples_ddim = modules.safety.censor_batch(x_samples_ddim) - for i, x_sample in enumerate(x_samples_ddim): + for i, x_sample in enumerate(x_samples_ddim): + with torch.no_grad(), precision_scope("cuda"), ema_scope(): x_sample = 255. * np.moveaxis(x_sample.cpu().numpy(), 0, 2) x_sample = x_sample.astype(np.uint8) - if p.restore_faces: + if p.restore_faces: + with torch.no_grad(), precision_scope("cuda"), ema_scope(): if opts.save and not p.do_not_save_samples and opts.save_images_before_face_restoration: images.save_image(Image.fromarray(x_sample), p.outpath_samples, "", seeds[i], prompts[i], opts.samples_format, info=infotext(n, i), p=p, suffix="-before-face-restoration") - devices.torch_gc() - x_sample = modules.face_restoration.restore_faces(x_sample) + devices.torch_gc() + + with torch.no_grad(), precision_scope("cuda"), ema_scope(): image = Image.fromarray(x_sample) if p.color_corrections is not None and i < len(p.color_corrections): @@ -411,8 +422,13 @@ def process_images(p: StableDiffusionProcessing) -> Processed: infotexts.append(infotext(n, i)) output_images.append(image) - state.nextjob() + del x_samples_ddim + devices.torch_gc() + + state.nextjob() + + with torch.no_grad(), precision_scope("cuda"), ema_scope(): p.color_corrections = None index_of_first_image = 0 @@ -648,4 +664,7 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing): if self.mask is not None: samples = samples * self.nmask + self.init_latent * self.mask + del x + devices.torch_gc() + return samples diff --git a/webui.py b/webui.py index c70a11c7c..b61a318db 100644 --- a/webui.py +++ b/webui.py @@ -22,7 +22,10 @@ import modules.txt2img import modules.img2img import modules.swinir as swinir import modules.sd_models +from torch.nn.functional import silu +import ldm +ldm.modules.diffusionmodules.model.nonlinearity = silu modules.codeformer_model.setup_codeformer() modules.gfpgan_model.setup_gfpgan() From c2d5b29040132c171bc4d77f1f63da972306f22c Mon Sep 17 00:00:00 2001 From: Jairo Correa Date: Thu, 29 Sep 2022 01:14:54 -0300 Subject: [PATCH 004/138] Move silu to sd_hijack --- modules/sd_hijack.py | 12 +++--------- webui.py | 3 --- 2 files changed, 3 insertions(+), 12 deletions(-) diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index bfbd07f9a..4bc58fa2b 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -12,6 +12,7 @@ from ldm.util import default from einops import rearrange import ldm.modules.attention import ldm.modules.diffusionmodules.model +from torch.nn.functional import silu # see https://github.com/basujindal/stable-diffusion/pull/117 for discussion @@ -100,14 +101,6 @@ def split_cross_attention_forward(self, x, context=None, mask=None): return self.to_out(r2) -def nonlinearity_hijack(x): - # swish - t = torch.sigmoid(x) - x *= t - del t - - return x - def cross_attention_attnblock_forward(self, x): h_ = x h_ = self.norm(h_) @@ -245,11 +238,12 @@ class StableDiffusionModelHijack: m.cond_stage_model = FrozenCLIPEmbedderWithCustomWords(m.cond_stage_model, self) self.clip = m.cond_stage_model + ldm.modules.diffusionmodules.model.nonlinearity = silu + if cmd_opts.opt_split_attention_v1: ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward_v1 elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward - ldm.modules.diffusionmodules.model.nonlinearity = nonlinearity_hijack ldm.modules.diffusionmodules.model.AttnBlock.forward = cross_attention_attnblock_forward def flatten(el): diff --git a/webui.py b/webui.py index b61a318db..c70a11c7c 100644 --- a/webui.py +++ b/webui.py @@ -22,10 +22,7 @@ import modules.txt2img import modules.img2img import modules.swinir as swinir import modules.sd_models -from torch.nn.functional import silu -import ldm -ldm.modules.diffusionmodules.model.nonlinearity = silu modules.codeformer_model.setup_codeformer() modules.gfpgan_model.setup_gfpgan() From 98cc6c6e744f454cbbd763033c93a0dceb40c343 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 30 Sep 2022 14:16:26 +0300 Subject: [PATCH 005/138] add embeddings dir --- embeddings/Place Textual Inversion embeddings here.txt | 0 modules/sd_hijack.py | 7 ++++++- 2 files changed, 6 insertions(+), 1 deletion(-) create mode 100644 embeddings/Place Textual Inversion embeddings here.txt diff --git a/embeddings/Place Textual Inversion embeddings here.txt b/embeddings/Place Textual Inversion embeddings here.txt new file mode 100644 index 000000000..e69de29bb diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index 5945b7c23..fa7eaeb89 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -232,7 +232,12 @@ class StableDiffusionModelHijack: for fn in os.listdir(dirname): try: - process_file(os.path.join(dirname, fn), fn) + fullfn = os.path.join(dirname, fn) + + if os.stat(fullfn).st_size == 0: + continue + + process_file(fullfn, fn) except Exception: print(f"Error loading emedding {fn}:", file=sys.stderr) print(traceback.format_exc(), file=sys.stderr) From 980cd1697ae980f57399da2b90462c07d102d935 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 30 Sep 2022 14:23:41 +0300 Subject: [PATCH 006/138] prevent neural network resizing when it is not necessary #1109 --- modules/images.py | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/modules/images.py b/modules/images.py index 2e8305ed8..8f34dcc1f 100644 --- a/modules/images.py +++ b/modules/images.py @@ -213,17 +213,19 @@ def resize_image(resize_mode, im, width, height): if opts.upscaler_for_img2img is None or opts.upscaler_for_img2img == "None" or im.mode == 'L': return im.resize((w, h), resample=LANCZOS) - upscalers = [x for x in shared.sd_upscalers if x.name == opts.upscaler_for_img2img] - assert len(upscalers) > 0, f"could not find upscaler named {opts.upscaler_for_img2img}" - - upscaler = upscalers[0] scale = max(w / im.width, h / im.height) - upscaled = upscaler.scaler.upscale(im, scale, upscaler.data_path) - if upscaled.width != w or upscaled.height != h: - upscaled = im.resize((w, h), resample=LANCZOS) + if scale > 1.0: + upscalers = [x for x in shared.sd_upscalers if x.name == opts.upscaler_for_img2img] + assert len(upscalers) > 0, f"could not find upscaler named {opts.upscaler_for_img2img}" - return upscaled + upscaler = upscalers[0] + im = upscaler.scaler.upscale(im, scale, upscaler.data_path) + + if im.width != w or im.height != h: + im = im.resize((w, h), resample=LANCZOS) + + return im if resize_mode == 0: res = resize(im, width, height) From 41f9449d96488017092d229d65f674c2e8252697 Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 08:33:06 -0500 Subject: [PATCH 007/138] Fix LDSR YAML path --- modules/ldsr_model.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/ldsr_model.py b/modules/ldsr_model.py index 969d1a0d8..877e7e73e 100644 --- a/modules/ldsr_model.py +++ b/modules/ldsr_model.py @@ -24,7 +24,7 @@ class UpscalerLDSR(Upscaler): def load_model(self, path: str): model = load_file_from_url(url=self.model_url, model_dir=self.model_path, file_name="model.pth", progress=True) - yaml = load_file_from_url(url=self.model_url, model_dir=self.model_path, + yaml = load_file_from_url(url=self.yaml_url, model_dir=self.model_path, file_name="project.yaml", progress=True) try: From 64c6b13312ff3a20f48781c4c3780355c4b7b2af Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 08:41:25 -0500 Subject: [PATCH 008/138] Remove LDSR project.yaml if too large. If we accidentally downloaded the wrong file and saved it as project.yaml, this will delete it so it can be re-downloaded. --- modules/ldsr_model.py | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/modules/ldsr_model.py b/modules/ldsr_model.py index 877e7e73e..4d8687c2f 100644 --- a/modules/ldsr_model.py +++ b/modules/ldsr_model.py @@ -22,6 +22,13 @@ class UpscalerLDSR(Upscaler): self.scalers = [scaler_data] def load_model(self, path: str): + # Remove incorrect project.yaml file if too big + yaml_path = os.path.join(self.model_path, "project.yaml") + if os.path.exists(yaml_path): + statinfo = os.stat(yaml_path) + if statinfo.st_size <= 10485760: + print("Removing invalid LDSR YAML file.") + os.remove(yaml_path) model = load_file_from_url(url=self.model_url, model_dir=self.model_path, file_name="model.pth", progress=True) yaml = load_file_from_url(url=self.yaml_url, model_dir=self.model_path, From 8d60645106d7e2daa0da89c5b21d7ffdac61cf9e Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 08:55:04 -0500 Subject: [PATCH 009/138] Fix model paths, ensure we have the right files. Also, clean up logging in the ldsr arch file. --- modules/ldsr_model.py | 9 +++++++-- modules/ldsr_model_arch.py | 3 +-- 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/modules/ldsr_model.py b/modules/ldsr_model.py index 4d8687c2f..7dff0a9c2 100644 --- a/modules/ldsr_model.py +++ b/modules/ldsr_model.py @@ -24,13 +24,18 @@ class UpscalerLDSR(Upscaler): def load_model(self, path: str): # Remove incorrect project.yaml file if too big yaml_path = os.path.join(self.model_path, "project.yaml") + old_model_path = os.path.join(self.model_path, "model.pth") + new_model_path = os.path.join(self.model_path, "model.ckpt") if os.path.exists(yaml_path): statinfo = os.stat(yaml_path) - if statinfo.st_size <= 10485760: + if statinfo.st_size >= 10485760: print("Removing invalid LDSR YAML file.") os.remove(yaml_path) + if os.path.exists(old_model_path): + print("Renaming model from model.pth to model.ckpt") + os.rename(old_model_path, new_model_path) model = load_file_from_url(url=self.model_url, model_dir=self.model_path, - file_name="model.pth", progress=True) + file_name="model.ckpt", progress=True) yaml = load_file_from_url(url=self.yaml_url, model_dir=self.model_path, file_name="project.yaml", progress=True) diff --git a/modules/ldsr_model_arch.py b/modules/ldsr_model_arch.py index 7faac6e18..093a32106 100644 --- a/modules/ldsr_model_arch.py +++ b/modules/ldsr_model_arch.py @@ -100,7 +100,6 @@ class LDSR: # If we can adjust the max upscale size, then the 4 below should be our variable print("Foo") down_sample_rate = target_scale / 4 - print(f"Downsample rate is {down_sample_rate}") wd = width_og * down_sample_rate hd = height_og * down_sample_rate width_downsampled_pre = int(wd) @@ -111,7 +110,7 @@ class LDSR: f'Downsampling from [{width_og}, {height_og}] to [{width_downsampled_pre}, {height_downsampled_pre}]') im_og = im_og.resize((width_downsampled_pre, height_downsampled_pre), Image.LANCZOS) else: - print(f"Down sample rate is 1 from {target_scale} / 4") + print(f"Down sample rate is 1 from {target_scale} / 4 (Not downsampling)") logs = self.run(model["model"], im_og, diffusion_steps, eta) sample = logs["sample"] From 99aa132df7045077a420918d276fcca877fdc9e3 Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 08:56:39 -0500 Subject: [PATCH 010/138] Remove useless print message --- modules/ldsr_model_arch.py | 1 - 1 file changed, 1 deletion(-) diff --git a/modules/ldsr_model_arch.py b/modules/ldsr_model_arch.py index 093a32106..14db50766 100644 --- a/modules/ldsr_model_arch.py +++ b/modules/ldsr_model_arch.py @@ -98,7 +98,6 @@ class LDSR: im_og = image width_og, height_og = im_og.size # If we can adjust the max upscale size, then the 4 below should be our variable - print("Foo") down_sample_rate = target_scale / 4 wd = width_og * down_sample_rate hd = height_og * down_sample_rate From 19eb1467f1acbb45d3b416ad7887ba9ab8a94e75 Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 09:05:00 -0500 Subject: [PATCH 011/138] Fix BSRGAN variable not found. --- modules/bsrgan_model.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/modules/bsrgan_model.py b/modules/bsrgan_model.py index 771415457..339f402bc 100644 --- a/modules/bsrgan_model.py +++ b/modules/bsrgan_model.py @@ -67,9 +67,8 @@ class UpscalerBSRGAN(modules.upscaler.Upscaler): else: filename = path if not os.path.exists(filename) or filename is None: - print("Unable to load %s from %s" % (self.model_dir, filename)) + print("Unable to load %s from %s" % (self.model_path, filename)) return None - print("Loading %s from %s" % (self.model_dir, filename)) model = RRDBNet(in_nc=3, out_nc=3, nf=64, nb=23, gc=32, sf=2) # define network model.load_state_dict(torch.load(filename), strict=True) model.eval() From ca87c09c0ec9f6d3fbfda5d8579907c7017eb629 Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 09:28:28 -0500 Subject: [PATCH 012/138] Fix recursive model loading Ensure we find checkpoints within subdirectories. --- modules/modelloader.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/modules/modelloader.py b/modules/modelloader.py index 1106aeb7f..df29b6ef6 100644 --- a/modules/modelloader.py +++ b/modules/modelloader.py @@ -1,3 +1,4 @@ +import glob import os import shutil import importlib @@ -41,7 +42,7 @@ def load_models(model_path: str, model_url: str = None, command_path: str = None for place in places: if os.path.exists(place): - for file in os.listdir(place): + for file in glob.iglob(place + '**/**', recursive=True): full_path = os.path.join(place, file) if os.path.isdir(full_path): continue @@ -50,6 +51,7 @@ def load_models(model_path: str, model_url: str = None, command_path: str = None if extension not in ext_filter: continue if file not in output: + print(f"FILE: {full_path}") output.append(full_path) if model_url is not None and len(output) == 0: From 5d52231bcb9fc21953b3d3495b577dc0bd82aa6a Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 09:28:48 -0500 Subject: [PATCH 013/138] Logging.... Cleanup. Sorry. --- modules/modelloader.py | 1 - 1 file changed, 1 deletion(-) diff --git a/modules/modelloader.py b/modules/modelloader.py index df29b6ef6..8c862b42f 100644 --- a/modules/modelloader.py +++ b/modules/modelloader.py @@ -51,7 +51,6 @@ def load_models(model_path: str, model_url: str = None, command_path: str = None if extension not in ext_filter: continue if file not in output: - print(f"FILE: {full_path}") output.append(full_path) if model_url is not None and len(output) == 0: From f71d02b85feacf52d03d69d5caca97cd7d90b481 Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 09:29:31 -0500 Subject: [PATCH 014/138] Remove unused LDSR opt --- modules/ldsr_model.py | 1 - 1 file changed, 1 deletion(-) diff --git a/modules/ldsr_model.py b/modules/ldsr_model.py index 7dff0a9c2..1c1070fc6 100644 --- a/modules/ldsr_model.py +++ b/modules/ldsr_model.py @@ -53,5 +53,4 @@ class UpscalerLDSR(Upscaler): print("NO LDSR!") return img ddim_steps = shared.opts.ldsr_steps - pre_scale = shared.opts.ldsr_pre_down return ldsr.super_resolution(img, ddim_steps, self.scale) From 9fc1e49bd2674f8c30dd3545c1c271472ff6d3c2 Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 09:38:57 -0500 Subject: [PATCH 015/138] Set default value for ckpt-dir --- modules/shared.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/shared.py b/modules/shared.py index 8428c7a38..76233ea0f 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -21,7 +21,7 @@ model_path = os.path.join(script_path, 'models') 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=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=None, help="Path to directory with stable diffusion checkpoints") +parser.add_argument("--ckpt-dir", type=str, default=os.path.join(model_path, "Stable-diffusion"), 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=None) parser.add_argument("--no-half", action='store_true', help="do not switch the model to 16-bit floats") From 43c87ef0fcf1771d5511004968e70f804cfd95b8 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 30 Sep 2022 18:07:49 +0300 Subject: [PATCH 016/138] change default inpaint mode to original --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/ui.py b/modules/ui.py index ada9a38e9..249b3eea1 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -599,7 +599,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): mask_mode = gr.Radio(label="Mask mode", show_label=False, choices=["Draw mask", "Upload mask"], type="index", value="Draw mask", elem_id="mask_mode") inpainting_mask_invert = gr.Radio(label='Masking mode', show_label=False, choices=['Inpaint masked', 'Inpaint not masked'], value='Inpaint masked', type="index") - inpainting_fill = gr.Radio(label='Masked content', choices=['fill', 'original', 'latent noise', 'latent nothing'], value='fill', type="index") + inpainting_fill = gr.Radio(label='Masked content', choices=['fill', 'original', 'latent noise', 'latent nothing'], value='original', type="index") with gr.Row(): inpaint_full_res = gr.Checkbox(label='Inpaint at full resolution', value=False) From ac21d308cd0144b8e02e4cfa1fc36722b2d909d5 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 30 Sep 2022 18:55:46 +0300 Subject: [PATCH 017/138] fix for one error with BSRGAN, though it fails to work anyway #1109 --- modules/bsrgan_model.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/modules/bsrgan_model.py b/modules/bsrgan_model.py index 771415457..47346f318 100644 --- a/modules/bsrgan_model.py +++ b/modules/bsrgan_model.py @@ -67,9 +67,8 @@ class UpscalerBSRGAN(modules.upscaler.Upscaler): else: filename = path if not os.path.exists(filename) or filename is None: - print("Unable to load %s from %s" % (self.model_dir, filename)) + print(f"BSRGAN: Unable to load model from {filename}", file=sys.stderr) return None - print("Loading %s from %s" % (self.model_dir, filename)) model = RRDBNet(in_nc=3, out_nc=3, nf=64, nb=23, gc=32, sf=2) # define network model.load_state_dict(torch.load(filename), strict=True) model.eval() From 4794202ebc1f5184e0a15cc11d91ea9a23640f27 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 30 Sep 2022 18:56:10 +0300 Subject: [PATCH 018/138] hello #1109 --- modules/ldsr_model.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/ldsr_model.py b/modules/ldsr_model.py index 969d1a0d8..877e7e73e 100644 --- a/modules/ldsr_model.py +++ b/modules/ldsr_model.py @@ -24,7 +24,7 @@ class UpscalerLDSR(Upscaler): def load_model(self, path: str): model = load_file_from_url(url=self.model_url, model_dir=self.model_path, file_name="model.pth", progress=True) - yaml = load_file_from_url(url=self.model_url, model_dir=self.model_path, + yaml = load_file_from_url(url=self.yaml_url, model_dir=self.model_path, file_name="project.yaml", progress=True) try: From 8f1d412e7bcd279c3c49b0153def0e3d6d941e5a Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 12:51:54 -0500 Subject: [PATCH 019/138] Fix BSRGAN Model loading. --- modules/bsrgan_model.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/modules/bsrgan_model.py b/modules/bsrgan_model.py index 339f402bc..165f77b5b 100644 --- a/modules/bsrgan_model.py +++ b/modules/bsrgan_model.py @@ -69,7 +69,8 @@ class UpscalerBSRGAN(modules.upscaler.Upscaler): if not os.path.exists(filename) or filename is None: print("Unable to load %s from %s" % (self.model_path, filename)) return None - model = RRDBNet(in_nc=3, out_nc=3, nf=64, nb=23, gc=32, sf=2) # define network + print("Loading %s from %s" % (self.model_path, filename)) + model = RRDBNet(in_nc=3, out_nc=3, nf=64, nb=23, gc=32, sf=4) # define network model.load_state_dict(torch.load(filename), strict=True) model.eval() for k, v in model.named_parameters(): From 7ab91d9e1bf3abc11dc0d90e270b199ae8c3c84a Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 13:29:33 -0500 Subject: [PATCH 020/138] Update bsrgan_model.py --- modules/bsrgan_model.py | 1 - 1 file changed, 1 deletion(-) diff --git a/modules/bsrgan_model.py b/modules/bsrgan_model.py index 827a50a54..e62c66577 100644 --- a/modules/bsrgan_model.py +++ b/modules/bsrgan_model.py @@ -69,7 +69,6 @@ class UpscalerBSRGAN(modules.upscaler.Upscaler): if not os.path.exists(filename) or filename is None: print(f"BSRGAN: Unable to load model from {filename}", file=sys.stderr) return None - print("Loading %s from %s" % (self.model_path, filename)) model = RRDBNet(in_nc=3, out_nc=3, nf=64, nb=23, gc=32, sf=4) # define network model.load_state_dict(torch.load(filename), strict=True) model.eval() From 2b03f0bbda1229dff6e7ab6f656b28587eba8308 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 30 Sep 2022 22:16:03 +0300 Subject: [PATCH 021/138] if --ckpt option is specified, load that model --- modules/sd_models.py | 1 + modules/shared.py | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/modules/sd_models.py b/modules/sd_models.py index ab014efb6..2539f14cd 100644 --- a/modules/sd_models.py +++ b/modules/sd_models.py @@ -69,6 +69,7 @@ def list_models(): h = model_hash(cmd_ckpt) title, short_model_name = modeltitle(cmd_ckpt, h) checkpoints_list[title] = CheckpointInfo(cmd_ckpt, title, h, short_model_name) + shared.opts.sd_model_checkpoint = title elif cmd_ckpt is not None and cmd_ckpt != shared.default_sd_model_file: print(f"Checkpoint in --ckpt argument not found (Possible it was moved to {model_path}: {cmd_ckpt}", file=sys.stderr) for filename in model_list: diff --git a/modules/shared.py b/modules/shared.py index 8428c7a38..ac968b2d2 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -20,7 +20,7 @@ default_sd_model_file = sd_model_file model_path = os.path.join(script_path, 'models') 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=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", type=str, default=sd_model_file, help="path to checkpoint of stable diffusion model; if specified, this checkpoint will be added to the list of checkpoints and loaded",) parser.add_argument("--ckpt-dir", type=str, default=None, 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=None) From 3f64e23f81618395428da0a132bf3a6cfd5c42c9 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 30 Sep 2022 22:57:25 +0300 Subject: [PATCH 022/138] fix model checkpoint merger tab failing to save model --- modules/extras.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/modules/extras.py b/modules/extras.py index 1bff58749..6a0d5cb0f 100644 --- a/modules/extras.py +++ b/modules/extras.py @@ -191,9 +191,11 @@ def run_modelmerger(primary_model_name, secondary_model_name, interp_method, int if save_as_half: theta_0[key] = theta_0[key].half() + ckpt_dir = shared.cmd_opts.ckpt_dir or sd_models.model_path + filename = primary_model_info.model_name + '_' + str(round(interp_amount, 2)) + '-' + secondary_model_info.model_name + '_' + str(round((float(1.0) - interp_amount), 2)) + '-' + interp_method.replace(" ", "_") + '-merged.ckpt' filename = filename if custom_name == '' else (custom_name + '.ckpt') - output_modelname = os.path.join(shared.cmd_opts.ckpt_dir, filename) + output_modelname = os.path.join(ckpt_dir, filename) print(f"Saving to {output_modelname}...") torch.save(primary_model, output_modelname) From 89f0acc6bf3726e40bd9a3c634716debe3ece87d Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 30 Sep 2022 23:18:47 +0300 Subject: [PATCH 023/138] remove extra print from BSRGAN --- modules/bsrgan_model_arch.py | 1 - 1 file changed, 1 deletion(-) diff --git a/modules/bsrgan_model_arch.py b/modules/bsrgan_model_arch.py index d72647db0..cb4d1c133 100644 --- a/modules/bsrgan_model_arch.py +++ b/modules/bsrgan_model_arch.py @@ -76,7 +76,6 @@ class RRDBNet(nn.Module): super(RRDBNet, self).__init__() RRDB_block_f = functools.partial(RRDB, nf=nf, gc=gc) self.sf = sf - print([in_nc, out_nc, nf, nb, gc, sf]) self.conv_first = nn.Conv2d(in_nc, nf, 3, 1, 1, bias=True) self.RRDB_trunk = make_layer(RRDB_block_f, nb) From e82ea202997cbcd2ab72891cd075d9ba270eb67d Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 15:26:18 -0500 Subject: [PATCH 024/138] Optimize model loader Child classes only get populated to __subclassess__ when they are imported. We don't actually need to import any of them to webui any more, so clean up webUI imports and make sure loader imports children. Also, fix command line paths not actually being passed to the scalers. --- modules/modelloader.py | 19 ++++++++++++++++--- webui.py | 13 +++---------- 2 files changed, 19 insertions(+), 13 deletions(-) diff --git a/modules/modelloader.py b/modules/modelloader.py index 1106aeb7f..b1721671b 100644 --- a/modules/modelloader.py +++ b/modules/modelloader.py @@ -4,7 +4,6 @@ import importlib from urllib.parse import urlparse from basicsr.utils.download_util import load_file_from_url - from modules import shared from modules.upscaler import Upscaler from modules.paths import script_path, models_path @@ -120,16 +119,30 @@ def move_files(src_path: str, dest_path: str, ext_filter: str = None): def load_upscalers(): + sd = shared.script_path + # We can only do this 'magic' method to dynamically load upscalers if they are referenced, + # so we'll try to import any _model.py files before looking in __subclasses__ + modules_dir = os.path.join(sd, "modules") + for file in os.listdir(modules_dir): + if "_model.py" in file: + model_name = file.replace("_model.py", "") + full_model = f"modules.{model_name}_model" + try: + importlib.import_module(full_model) + except: + pass datas = [] + c_o = vars(shared.cmd_opts) for cls in Upscaler.__subclasses__(): name = cls.__name__ module_name = cls.__module__ module = importlib.import_module(module_name) class_ = getattr(module, name) - cmd_name = f"{name.lower().replace('upscaler', '')}-models-path" + cmd_name = f"{name.lower().replace('upscaler', '')}_models_path" opt_string = None try: - opt_string = shared.opts.__getattr__(cmd_name) + if cmd_name in c_o: + opt_string = c_o[cmd_name] except: pass scaler = class_(opt_string) diff --git a/webui.py b/webui.py index b8cccd546..ebe39a170 100644 --- a/webui.py +++ b/webui.py @@ -1,28 +1,21 @@ import os -import threading - -from modules import devices -from modules.paths import script_path import signal import threading -import modules.paths + import modules.codeformer_model as codeformer -import modules.esrgan_model as esrgan -import modules.bsrgan_model as bsrgan import modules.extras import modules.face_restoration import modules.gfpgan_model as gfpgan import modules.img2img -import modules.ldsr_model as ldsr import modules.lowvram -import modules.realesrgan_model as realesrgan +import modules.paths import modules.scripts import modules.sd_hijack import modules.sd_models import modules.shared as shared -import modules.swinir_model as swinir import modules.txt2img import modules.ui +from modules import devices from modules import modelloader from modules.paths import script_path from modules.shared import cmd_opts From 8deae077004f0332ca607fc3a5d568b1a4705bec Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Fri, 30 Sep 2022 15:28:37 -0500 Subject: [PATCH 025/138] Add ScuNET DeNoiser/Upscaler Q&D Implementation of ScuNET, thanks to our handy model loader. :P https://github.com/cszn/SCUNet --- modules/scunet_model.py | 90 ++++++++++++ modules/scunet_model_arch.py | 265 +++++++++++++++++++++++++++++++++++ modules/shared.py | 1 + 3 files changed, 356 insertions(+) create mode 100644 modules/scunet_model.py create mode 100644 modules/scunet_model_arch.py diff --git a/modules/scunet_model.py b/modules/scunet_model.py new file mode 100644 index 000000000..7987ac145 --- /dev/null +++ b/modules/scunet_model.py @@ -0,0 +1,90 @@ +import os.path +import sys +import traceback + +import PIL.Image +import numpy as np +import torch +from basicsr.utils.download_util import load_file_from_url + +import modules.upscaler +from modules import shared, modelloader +from modules.paths import models_path +from modules.scunet_model_arch import SCUNet as net + + +class UpscalerScuNET(modules.upscaler.Upscaler): + def __init__(self, dirname): + self.name = "ScuNET" + self.model_path = os.path.join(models_path, self.name) + self.model_name = "ScuNET GAN" + self.model_name2 = "ScuNET PSNR" + self.model_url = "https://github.com/cszn/KAIR/releases/download/v1.0/scunet_color_real_gan.pth" + self.model_url2 = "https://github.com/cszn/KAIR/releases/download/v1.0/scunet_color_real_psnr.pth" + self.user_path = dirname + super().__init__() + model_paths = self.find_models(ext_filter=[".pth"]) + scalers = [] + add_model2 = True + for file in model_paths: + if "http" in file: + name = self.model_name + else: + name = modelloader.friendly_name(file) + if name == self.model_name2 or file == self.model_url2: + add_model2 = False + try: + scaler_data = modules.upscaler.UpscalerData(name, file, self, 4) + scalers.append(scaler_data) + except Exception: + print(f"Error loading ScuNET model: {file}", file=sys.stderr) + print(traceback.format_exc(), file=sys.stderr) + if add_model2: + scaler_data2 = modules.upscaler.UpscalerData(self.model_name2, self.model_url2, self) + scalers.append(scaler_data2) + self.scalers = scalers + + def do_upscale(self, img: PIL.Image, selected_file): + torch.cuda.empty_cache() + + model = self.load_model(selected_file) + if model is None: + return img + + device = shared.device + img = np.array(img) + img = img[:, :, ::-1] + img = np.moveaxis(img, 2, 0) / 255 + img = torch.from_numpy(img).float() + img = img.unsqueeze(0).to(shared.device) + + img = img.to(device) + with torch.no_grad(): + output = model(img) + output = output.squeeze().float().cpu().clamp_(0, 1).numpy() + output = 255. * np.moveaxis(output, 0, 2) + output = output.astype(np.uint8) + output = output[:, :, ::-1] + torch.cuda.empty_cache() + return PIL.Image.fromarray(output, 'RGB') + + def load_model(self, path: str): + device = shared.device + if "http" in path: + filename = load_file_from_url(url=self.model_url, model_dir=self.model_path, file_name="%s.pth" % self.name, + progress=True) + else: + filename = path + if not os.path.exists(os.path.join(self.model_path, filename)) or filename is None: + print(f"ScuNET: Unable to load model from {filename}", file=sys.stderr) + return None + + model = net(in_nc=3, config=[4, 4, 4, 4, 4, 4, 4], dim=64) + model.load_state_dict(torch.load(filename), strict=True) + model.eval() + for k, v in model.named_parameters(): + v.requires_grad = False + model = model.to(device) + + return model + diff --git a/modules/scunet_model_arch.py b/modules/scunet_model_arch.py new file mode 100644 index 000000000..972a2639a --- /dev/null +++ b/modules/scunet_model_arch.py @@ -0,0 +1,265 @@ +# -*- coding: utf-8 -*- +import numpy as np +import torch +import torch.nn as nn +from einops import rearrange +from einops.layers.torch import Rearrange +from timm.models.layers import trunc_normal_, DropPath + + +class WMSA(nn.Module): + """ Self-attention module in Swin Transformer + """ + + def __init__(self, input_dim, output_dim, head_dim, window_size, type): + super(WMSA, self).__init__() + self.input_dim = input_dim + self.output_dim = output_dim + self.head_dim = head_dim + self.scale = self.head_dim ** -0.5 + self.n_heads = input_dim // head_dim + self.window_size = window_size + self.type = type + self.embedding_layer = nn.Linear(self.input_dim, 3 * self.input_dim, bias=True) + + self.relative_position_params = nn.Parameter( + torch.zeros((2 * window_size - 1) * (2 * window_size - 1), self.n_heads)) + + self.linear = nn.Linear(self.input_dim, self.output_dim) + + trunc_normal_(self.relative_position_params, std=.02) + self.relative_position_params = torch.nn.Parameter( + self.relative_position_params.view(2 * window_size - 1, 2 * window_size - 1, self.n_heads).transpose(1, + 2).transpose( + 0, 1)) + + def generate_mask(self, h, w, p, shift): + """ generating the mask of SW-MSA + Args: + shift: shift parameters in CyclicShift. + Returns: + attn_mask: should be (1 1 w p p), + """ + # supporting sqaure. + attn_mask = torch.zeros(h, w, p, p, p, p, dtype=torch.bool, device=self.relative_position_params.device) + if self.type == 'W': + return attn_mask + + s = p - shift + attn_mask[-1, :, :s, :, s:, :] = True + attn_mask[-1, :, s:, :, :s, :] = True + attn_mask[:, -1, :, :s, :, s:] = True + attn_mask[:, -1, :, s:, :, :s] = True + attn_mask = rearrange(attn_mask, 'w1 w2 p1 p2 p3 p4 -> 1 1 (w1 w2) (p1 p2) (p3 p4)') + return attn_mask + + def forward(self, x): + """ Forward pass of Window Multi-head Self-attention module. + Args: + x: input tensor with shape of [b h w c]; + attn_mask: attention mask, fill -inf where the value is True; + Returns: + output: tensor shape [b h w c] + """ + if self.type != 'W': x = torch.roll(x, shifts=(-(self.window_size // 2), -(self.window_size // 2)), dims=(1, 2)) + x = rearrange(x, 'b (w1 p1) (w2 p2) c -> b w1 w2 p1 p2 c', p1=self.window_size, p2=self.window_size) + h_windows = x.size(1) + w_windows = x.size(2) + # sqaure validation + # assert h_windows == w_windows + + x = rearrange(x, 'b w1 w2 p1 p2 c -> b (w1 w2) (p1 p2) c', p1=self.window_size, p2=self.window_size) + qkv = self.embedding_layer(x) + q, k, v = rearrange(qkv, 'b nw np (threeh c) -> threeh b nw np c', c=self.head_dim).chunk(3, dim=0) + sim = torch.einsum('hbwpc,hbwqc->hbwpq', q, k) * self.scale + # Adding learnable relative embedding + sim = sim + rearrange(self.relative_embedding(), 'h p q -> h 1 1 p q') + # Using Attn Mask to distinguish different subwindows. + if self.type != 'W': + attn_mask = self.generate_mask(h_windows, w_windows, self.window_size, shift=self.window_size // 2) + sim = sim.masked_fill_(attn_mask, float("-inf")) + + probs = nn.functional.softmax(sim, dim=-1) + output = torch.einsum('hbwij,hbwjc->hbwic', probs, v) + output = rearrange(output, 'h b w p c -> b w p (h c)') + output = self.linear(output) + output = rearrange(output, 'b (w1 w2) (p1 p2) c -> b (w1 p1) (w2 p2) c', w1=h_windows, p1=self.window_size) + + if self.type != 'W': output = torch.roll(output, shifts=(self.window_size // 2, self.window_size // 2), + dims=(1, 2)) + return output + + def relative_embedding(self): + cord = torch.tensor(np.array([[i, j] for i in range(self.window_size) for j in range(self.window_size)])) + relation = cord[:, None, :] - cord[None, :, :] + self.window_size - 1 + # negative is allowed + return self.relative_position_params[:, relation[:, :, 0].long(), relation[:, :, 1].long()] + + +class Block(nn.Module): + def __init__(self, input_dim, output_dim, head_dim, window_size, drop_path, type='W', input_resolution=None): + """ SwinTransformer Block + """ + super(Block, self).__init__() + self.input_dim = input_dim + self.output_dim = output_dim + assert type in ['W', 'SW'] + self.type = type + if input_resolution <= window_size: + self.type = 'W' + + self.ln1 = nn.LayerNorm(input_dim) + self.msa = WMSA(input_dim, input_dim, head_dim, window_size, self.type) + self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() + self.ln2 = nn.LayerNorm(input_dim) + self.mlp = nn.Sequential( + nn.Linear(input_dim, 4 * input_dim), + nn.GELU(), + nn.Linear(4 * input_dim, output_dim), + ) + + def forward(self, x): + x = x + self.drop_path(self.msa(self.ln1(x))) + x = x + self.drop_path(self.mlp(self.ln2(x))) + return x + + +class ConvTransBlock(nn.Module): + def __init__(self, conv_dim, trans_dim, head_dim, window_size, drop_path, type='W', input_resolution=None): + """ SwinTransformer and Conv Block + """ + super(ConvTransBlock, self).__init__() + self.conv_dim = conv_dim + self.trans_dim = trans_dim + self.head_dim = head_dim + self.window_size = window_size + self.drop_path = drop_path + self.type = type + self.input_resolution = input_resolution + + assert self.type in ['W', 'SW'] + if self.input_resolution <= self.window_size: + self.type = 'W' + + self.trans_block = Block(self.trans_dim, self.trans_dim, self.head_dim, self.window_size, self.drop_path, + self.type, self.input_resolution) + self.conv1_1 = nn.Conv2d(self.conv_dim + self.trans_dim, self.conv_dim + self.trans_dim, 1, 1, 0, bias=True) + self.conv1_2 = nn.Conv2d(self.conv_dim + self.trans_dim, self.conv_dim + self.trans_dim, 1, 1, 0, bias=True) + + self.conv_block = nn.Sequential( + nn.Conv2d(self.conv_dim, self.conv_dim, 3, 1, 1, bias=False), + nn.ReLU(True), + nn.Conv2d(self.conv_dim, self.conv_dim, 3, 1, 1, bias=False) + ) + + def forward(self, x): + conv_x, trans_x = torch.split(self.conv1_1(x), (self.conv_dim, self.trans_dim), dim=1) + conv_x = self.conv_block(conv_x) + conv_x + trans_x = Rearrange('b c h w -> b h w c')(trans_x) + trans_x = self.trans_block(trans_x) + trans_x = Rearrange('b h w c -> b c h w')(trans_x) + res = self.conv1_2(torch.cat((conv_x, trans_x), dim=1)) + x = x + res + + return x + + +class SCUNet(nn.Module): + # def __init__(self, in_nc=3, config=[2, 2, 2, 2, 2, 2, 2], dim=64, drop_path_rate=0.0, input_resolution=256): + def __init__(self, in_nc=3, config=None, dim=64, drop_path_rate=0.0, input_resolution=256): + super(SCUNet, self).__init__() + if config is None: + config = [2, 2, 2, 2, 2, 2, 2] + self.config = config + self.dim = dim + self.head_dim = 32 + self.window_size = 8 + + # drop path rate for each layer + dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(config))] + + self.m_head = [nn.Conv2d(in_nc, dim, 3, 1, 1, bias=False)] + + begin = 0 + self.m_down1 = [ConvTransBlock(dim // 2, dim // 2, self.head_dim, self.window_size, dpr[i + begin], + 'W' if not i % 2 else 'SW', input_resolution) + for i in range(config[0])] + \ + [nn.Conv2d(dim, 2 * dim, 2, 2, 0, bias=False)] + + begin += config[0] + self.m_down2 = [ConvTransBlock(dim, dim, self.head_dim, self.window_size, dpr[i + begin], + 'W' if not i % 2 else 'SW', input_resolution // 2) + for i in range(config[1])] + \ + [nn.Conv2d(2 * dim, 4 * dim, 2, 2, 0, bias=False)] + + begin += config[1] + self.m_down3 = [ConvTransBlock(2 * dim, 2 * dim, self.head_dim, self.window_size, dpr[i + begin], + 'W' if not i % 2 else 'SW', input_resolution // 4) + for i in range(config[2])] + \ + [nn.Conv2d(4 * dim, 8 * dim, 2, 2, 0, bias=False)] + + begin += config[2] + self.m_body = [ConvTransBlock(4 * dim, 4 * dim, self.head_dim, self.window_size, dpr[i + begin], + 'W' if not i % 2 else 'SW', input_resolution // 8) + for i in range(config[3])] + + begin += config[3] + self.m_up3 = [nn.ConvTranspose2d(8 * dim, 4 * dim, 2, 2, 0, bias=False), ] + \ + [ConvTransBlock(2 * dim, 2 * dim, self.head_dim, self.window_size, dpr[i + begin], + 'W' if not i % 2 else 'SW', input_resolution // 4) + for i in range(config[4])] + + begin += config[4] + self.m_up2 = [nn.ConvTranspose2d(4 * dim, 2 * dim, 2, 2, 0, bias=False), ] + \ + [ConvTransBlock(dim, dim, self.head_dim, self.window_size, dpr[i + begin], + 'W' if not i % 2 else 'SW', input_resolution // 2) + for i in range(config[5])] + + begin += config[5] + self.m_up1 = [nn.ConvTranspose2d(2 * dim, dim, 2, 2, 0, bias=False), ] + \ + [ConvTransBlock(dim // 2, dim // 2, self.head_dim, self.window_size, dpr[i + begin], + 'W' if not i % 2 else 'SW', input_resolution) + for i in range(config[6])] + + self.m_tail = [nn.Conv2d(dim, in_nc, 3, 1, 1, bias=False)] + + self.m_head = nn.Sequential(*self.m_head) + self.m_down1 = nn.Sequential(*self.m_down1) + self.m_down2 = nn.Sequential(*self.m_down2) + self.m_down3 = nn.Sequential(*self.m_down3) + self.m_body = nn.Sequential(*self.m_body) + self.m_up3 = nn.Sequential(*self.m_up3) + self.m_up2 = nn.Sequential(*self.m_up2) + self.m_up1 = nn.Sequential(*self.m_up1) + self.m_tail = nn.Sequential(*self.m_tail) + # self.apply(self._init_weights) + + def forward(self, x0): + + h, w = x0.size()[-2:] + paddingBottom = int(np.ceil(h / 64) * 64 - h) + paddingRight = int(np.ceil(w / 64) * 64 - w) + x0 = nn.ReplicationPad2d((0, paddingRight, 0, paddingBottom))(x0) + + x1 = self.m_head(x0) + x2 = self.m_down1(x1) + x3 = self.m_down2(x2) + x4 = self.m_down3(x3) + x = self.m_body(x4) + x = self.m_up3(x + x4) + x = self.m_up2(x + x3) + x = self.m_up1(x + x2) + x = self.m_tail(x + x1) + + x = x[..., :h, :w] + + return x + + def _init_weights(self, m): + if isinstance(m, nn.Linear): + trunc_normal_(m.weight, std=.02) + if m.bias is not None: + nn.init.constant_(m.bias, 0) + elif isinstance(m, nn.LayerNorm): + nn.init.constant_(m.bias, 0) + nn.init.constant_(m.weight, 1.0) \ No newline at end of file diff --git a/modules/shared.py b/modules/shared.py index 8428c7a38..a48b995ad 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -40,6 +40,7 @@ parser.add_argument("--gfpgan-models-path", type=str, help="Path to directory wi parser.add_argument("--esrgan-models-path", type=str, help="Path to directory with ESRGAN model file(s).", default=os.path.join(model_path, 'ESRGAN')) parser.add_argument("--bsrgan-models-path", type=str, help="Path to directory with BSRGAN model file(s).", default=os.path.join(model_path, 'BSRGAN')) parser.add_argument("--realesrgan-models-path", type=str, help="Path to directory with RealESRGAN model file(s).", default=os.path.join(model_path, 'RealESRGAN')) +parser.add_argument("--scunet-models-path", type=str, help="Path to directory with ScuNET model file(s).", default=os.path.join(model_path, 'ScuNET')) parser.add_argument("--swinir-models-path", type=str, help="Path to directory with SwinIR model file(s).", default=os.path.join(model_path, 'SwinIR')) parser.add_argument("--ldsr-models-path", type=str, help="Path to directory with LDSR model file(s).", default=os.path.join(model_path, 'LDSR')) parser.add_argument("--opt-split-attention", action='store_true', help="force-enables cross-attention layer optimization. By default, it's on for torch.cuda and off for other torch devices.") From d1db330010de31d48975bb3ec86e9e60d56ae9e9 Mon Sep 17 00:00:00 2001 From: Trung Ngo Date: Fri, 30 Sep 2022 15:31:00 -0500 Subject: [PATCH 026/138] ALT + return will submit prompt when it's in focus --- javascript/ui.js | 10 ++++++++++ modules/ui.py | 2 +- style.css | 2 +- 3 files changed, 12 insertions(+), 2 deletions(-) diff --git a/javascript/ui.js b/javascript/ui.js index 562d2552b..6b52878d9 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -186,10 +186,12 @@ onUiUpdate(function(){ if (!txt2img_textarea) { txt2img_textarea = gradioApp().querySelector("#txt2img_prompt > label > textarea"); txt2img_textarea?.addEventListener("input", () => update_token_counter("txt2img_token_button")); + txt2img_textarea?.addEventListener("keyup", (event) => submit_prompt(event, "txt2img_generate")); } if (!img2img_textarea) { img2img_textarea = gradioApp().querySelector("#img2img_prompt > label > textarea"); img2img_textarea?.addEventListener("input", () => update_token_counter("img2img_token_button")); + txt2img_textarea?.addEventListener("keyup", (event) => submit_prompt(event, "img2img_generate")); } }) @@ -197,6 +199,14 @@ let txt2img_textarea, img2img_textarea = undefined; let wait_time = 800 let token_timeout; +function submit_prompt(event, generate_button_id) { + if (event.altKey && event.keyCode === 13) { + event.preventDefault(); + gradioApp().getElementById(generate_button_id).click(); + return; + } +} + function update_token_counter(button_id) { if (token_timeout) clearTimeout(token_timeout); diff --git a/modules/ui.py b/modules/ui.py index 249b3eea1..15572bb0a 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -380,7 +380,7 @@ def create_toprow(is_img2img): with gr.Column(scale=1): with gr.Row(): interrupt = gr.Button('Interrupt', elem_id=f"{id_part}_interrupt") - submit = gr.Button('Generate', elem_id="generate", variant='primary') + submit = gr.Button('Generate', elem_id=f"{id_part}_generate", variant='primary') interrupt.click( fn=lambda: shared.state.interrupt(), diff --git a/style.css b/style.css index 9709c4eec..79d6bb0dc 100644 --- a/style.css +++ b/style.css @@ -23,7 +23,7 @@ text-align: right; } -#generate{ +#txt2img_generate, #img2img_generate { min-height: 4.5em; } From 6b02966fcd0de851b268cd64d219bc7ba41cd411 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 1 Oct 2022 00:14:44 +0300 Subject: [PATCH 027/138] fix #1405 submitting both tabs --- javascript/ui.js | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/javascript/ui.js b/javascript/ui.js index 6b52878d9..bfe024108 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -191,7 +191,7 @@ onUiUpdate(function(){ if (!img2img_textarea) { img2img_textarea = gradioApp().querySelector("#img2img_prompt > label > textarea"); img2img_textarea?.addEventListener("input", () => update_token_counter("img2img_token_button")); - txt2img_textarea?.addEventListener("keyup", (event) => submit_prompt(event, "img2img_generate")); + img2img_textarea?.addEventListener("keyup", (event) => submit_prompt(event, "img2img_generate")); } }) From 84e97a98c5233119d0f444e0a3a0f6391da23677 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 1 Oct 2022 00:38:48 +0300 Subject: [PATCH 028/138] features updates unused code removed from outpainting mk2 --- README.md | 52 +++++++++++++++++++++++-------------- scripts/outpainting_mk_2.py | 40 +--------------------------- 2 files changed, 33 insertions(+), 59 deletions(-) diff --git a/README.md b/README.md index 219288ac7..5ded94f98 100644 --- a/README.md +++ b/README.md @@ -11,44 +11,56 @@ Check the [custom scripts](https://github.com/AUTOMATIC1111/stable-diffusion-web - One click install and run script (but you still must install python and git) - Outpainting - Inpainting -- Prompt matrix +- Prompt - Stable Diffusion upscale -- Attention -- Loopback -- X/Y plot +- Attention, specify parts of text that the model should pay more attention to + - a man in a ((txuedo)) - will pay more attentinoto tuxedo + - a man in a (txuedo:1.21) - alternative syntax +- Loopback, run img2img procvessing multiple times +- X/Y plot, a way to draw a 2 dimensional plot of images with different parameters - Textual Inversion + - have as many embeddings as you want and use any names you like for them + - use multiple embeddings with different numbers of vectors per token + - works with half precision floating point numbers - Extras tab with: - GFPGAN, neural network that fixes faces - CodeFormer, face restoration tool as an alternative to GFPGAN - RealESRGAN, neural network upscaler - - ESRGAN, neural network with a lot of third party models + - ESRGAN, neural network upscaler with a lot of third party models - SwinIR, neural network upscaler - LDSR, Latent diffusion super resolution upscaling - Resizing aspect ratio options - Sampling method selection - Interrupt processing at any time -- 4GB video card support -- Correct seeds for batches +- 4GB video card support (also reports of 2GB working) +- Correct seeds for batches - Prompt length validation -- Generation parameters added as text to PNG -- Tab to view an existing picture's generation parameters + - get length of prompt in tokensas you type + - get a warning after geenration if some text was truncated +- Generation parameters + - parameters you used to generate images are saved with that image + - in PNG chunks for PNG, in EXIF for JPEG + - can drag the image to PNG info tab to restore generation parameters and automatically copy them into UI + - can be disabled in settings - Settings page -- Running custom code from UI +- Running arbitrary python code from UI (must run with commandline flag to enable) - Mouseover hints for most UI elements - Possible to change defaults/mix/max/step values for UI elements via text config - Random artist button -- Tiling support: UI checkbox to create images that can be tiled like textures +- Tiling support, a checkbox to create images that can be tiled like textures - Progress bar and live image generation preview -- Negative prompt -- Styles -- Variations -- Seed resizing -- CLIP interrogator -- Prompt Editing -- Batch Processing +- Negative prompt, an extra text field that allows you to list what you don't want to see in generated image +- Styles, a way to save part of prompt and easily apply them via dropdown later +- Variations, a way to generate same image but with tiny differences +- Seed resizing, a way to generate same image but at slightly different resolution +- CLIP interrogator, a button that tries to guess prompt from an image +- Prompt Editing, a way to change prompt mid-generation, say to start making a watermelon and switch to anime girl midway +- Batch Processing, process a group of files using img2img - Img2img Alternative -- Highres Fix -- LDSR Upscaling +- Highres Fix, a convenience option to produce high resolution pictures in one click without usual distortions +- Reloading checkpoints on the fly +- Checkpoint Merger, a tab that allows you to merge two checkpoints into one +- [Custom scripts](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Custom-Scripts) with many extensions from community ## Installation and Running Make sure the required [dependencies](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Dependencies) are met and follow the instructions available for both [NVidia](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-NVidia-GPUs) (recommended) and [AMD](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-AMD-GPUs) GPUs. diff --git a/scripts/outpainting_mk_2.py b/scripts/outpainting_mk_2.py index 9719bb8f9..11613ca36 100644 --- a/scripts/outpainting_mk_2.py +++ b/scripts/outpainting_mk_2.py @@ -11,46 +11,8 @@ from modules import images, processing, devices from modules.processing import Processed, process_images from modules.shared import opts, cmd_opts, state -# https://github.com/parlance-zz/g-diffuser-bot -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 - +# this function is taken from https://github.com/parlance-zz/g-diffuser-bot 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): From abdbf1de646f007b6d76cfb3f416fdfaadb57903 Mon Sep 17 00:00:00 2001 From: Liam Date: Thu, 29 Sep 2022 14:40:47 -0400 Subject: [PATCH 029/138] token counters now update when roll artist and style buttons are pressed https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/1194#issuecomment-1261203893 --- javascript/ui.js | 28 ++++++++++++++++++++++------ modules/ui.py | 6 +++++- 2 files changed, 27 insertions(+), 7 deletions(-) diff --git a/javascript/ui.js b/javascript/ui.js index bfe024108..88fd45ae9 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -199,12 +199,21 @@ let txt2img_textarea, img2img_textarea = undefined; let wait_time = 800 let token_timeout; -function submit_prompt(event, generate_button_id) { - if (event.altKey && event.keyCode === 13) { - event.preventDefault(); - gradioApp().getElementById(generate_button_id).click(); - return; - } +function roll_artist_txt2img(prompt_text) { + update_token_counter("txt2img_token_button") + return prompt_text; +} +function roll_artist_img2img(prompt_text) { + update_token_counter("img2img_token_button") + return prompt_text; +} +function update_style_txt2img(prompt_text, negative_prompt, style1, style2) { + update_token_counter("txt2img_token_button") + return [prompt_text, negative_prompt, style1, style2] +} +function update_style_img2img(prompt_text, negative_prompt, style1, style2) { + update_token_counter("img2img_token_button") + return [prompt_text, negative_prompt, style1, style2] } function update_token_counter(button_id) { @@ -212,3 +221,10 @@ function update_token_counter(button_id) { clearTimeout(token_timeout); token_timeout = setTimeout(() => gradioApp().getElementById(button_id)?.click(), wait_time); } +function submit_prompt(event, generate_button_id) { + if (event.altKey && event.keyCode === 13) { + event.preventDefault(); + gradioApp().getElementById(generate_button_id).click(); + return; + } +} \ No newline at end of file diff --git a/modules/ui.py b/modules/ui.py index 15572bb0a..5eea18606 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -539,6 +539,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): roll.click( fn=roll_artist, + _js="roll_artist_txt2img", inputs=[ txt2img_prompt, ], @@ -743,6 +744,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): roll.click( fn=roll_artist, + _js="roll_artist_img2img", inputs=[ img2img_prompt, ], @@ -753,6 +755,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): prompts = [(txt2img_prompt, txt2img_negative_prompt), (img2img_prompt, img2img_negative_prompt)] style_dropdowns = [(txt2img_prompt_style, txt2img_prompt_style2), (img2img_prompt_style, img2img_prompt_style2)] + style_js_funcs = ["update_style_txt2img", "update_style_img2img"] for button, (prompt, negative_prompt) in zip([txt2img_save_style, img2img_save_style], prompts): button.click( @@ -764,9 +767,10 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): outputs=[txt2img_prompt_style, img2img_prompt_style, txt2img_prompt_style2, img2img_prompt_style2], ) - for button, (prompt, negative_prompt), (style1, style2) in zip([txt2img_prompt_style_apply, img2img_prompt_style_apply], prompts, style_dropdowns): + for button, (prompt, negative_prompt), (style1, style2), js_func in zip([txt2img_prompt_style_apply, img2img_prompt_style_apply], prompts, style_dropdowns, style_js_funcs): button.click( fn=apply_styles, + _js=js_func, inputs=[prompt, negative_prompt, style1, style2], outputs=[prompt, negative_prompt, style1, style2], ) From ff8dc1908af088d0ed43fb85baad662733c5ca9c Mon Sep 17 00:00:00 2001 From: Liam Date: Thu, 29 Sep 2022 15:47:06 -0400 Subject: [PATCH 030/138] fixed token counter for prompt editing --- modules/ui.py | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/modules/ui.py b/modules/ui.py index 5eea18606..6bf28562c 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -11,6 +11,7 @@ import time import traceback import platform import subprocess as sp +from functools import reduce import numpy as np import torch @@ -32,6 +33,7 @@ import modules.gfpgan_model import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste +from modules.prompt_parser import get_learned_conditioning_prompt_schedules # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI mimetypes.init() @@ -345,8 +347,11 @@ def connect_reuse_seed(seed: gr.Number, reuse_seed: gr.Button, generation_info: outputs=[seed, dummy_component] ) -def update_token_counter(text): - tokens, token_count, max_length = model_hijack.tokenize(text) +def update_token_counter(text, steps): + prompt_schedules = get_learned_conditioning_prompt_schedules([text], steps) + flat_prompts = reduce(lambda list1, list2: list1+list2, prompt_schedules) + prompts = [prompt_text for step,prompt_text in flat_prompts] + tokens, token_count, max_length = max([model_hijack.tokenize(prompt) for prompt in prompts], key=lambda args: args[1]) style_class = ' class="red"' if (token_count > max_length) else "" return f"{token_count}/{max_length}" @@ -364,8 +369,7 @@ def create_toprow(is_img2img): roll = gr.Button(value=art_symbol, elem_id="roll", visible=len(shared.artist_db.artists) > 0) paste = gr.Button(value=paste_symbol, elem_id="paste") token_counter = gr.HTML(value="", elem_id=f"{id_part}_token_counter") - hidden_button = gr.Button(visible=False, elem_id=f"{id_part}_token_button") - hidden_button.click(fn=update_token_counter, inputs=[prompt], outputs=[token_counter]) + token_button = gr.Button(visible=False, elem_id=f"{id_part}_token_button") with gr.Column(scale=10, elem_id="style_pos_col"): prompt_style = gr.Dropdown(label="Style 1", elem_id=f"{id_part}_style_index", choices=[k for k, v in shared.prompt_styles.styles.items()], value=next(iter(shared.prompt_styles.styles.keys())), visible=len(shared.prompt_styles.styles) > 1) @@ -396,7 +400,7 @@ def create_toprow(is_img2img): prompt_style_apply = gr.Button('Apply style', elem_id="style_apply") save_style = gr.Button('Create style', elem_id="style_create") - return prompt, roll, prompt_style, negative_prompt, prompt_style2, submit, interrogate, prompt_style_apply, save_style, paste + return prompt, roll, prompt_style, negative_prompt, prompt_style2, submit, interrogate, prompt_style_apply, save_style, paste, token_counter, token_button def setup_progressbar(progressbar, preview, id_part): @@ -419,7 +423,7 @@ def setup_progressbar(progressbar, preview, id_part): def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): with gr.Blocks(analytics_enabled=False) as txt2img_interface: - txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, txt2img_prompt_style_apply, txt2img_save_style, paste = create_toprow(is_img2img=False) + txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, txt2img_prompt_style_apply, txt2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=False) dummy_component = gr.Label(visible=False) with gr.Row(elem_id='txt2img_progress_row'): @@ -568,9 +572,10 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): (hr_options, lambda d: gr.Row.update(visible="Denoising strength" in d)), ] modules.generation_parameters_copypaste.connect_paste(paste, txt2img_paste_fields, txt2img_prompt) + token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) with gr.Blocks(analytics_enabled=False) as img2img_interface: - img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_prompt_style_apply, img2img_save_style, paste = create_toprow(is_img2img=True) + img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) with gr.Row(elem_id='img2img_progress_row'): with gr.Column(scale=1): @@ -793,6 +798,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): (denoising_strength, "Denoising strength"), ] modules.generation_parameters_copypaste.connect_paste(paste, img2img_paste_fields, img2img_prompt) + token_button.click(fn=update_token_counter, inputs=[img2img_prompt, steps], outputs=[token_counter]) with gr.Blocks(analytics_enabled=False) as extras_interface: with gr.Row().style(equal_height=False): From 3c6a049fc3c6b54ada3736710a7e86663ea7f3d9 Mon Sep 17 00:00:00 2001 From: Liam Date: Fri, 30 Sep 2022 12:12:44 -0400 Subject: [PATCH 031/138] consolidated token counter functions --- javascript/ui.js | 21 +++++++++------------ modules/ui.py | 6 +++--- 2 files changed, 12 insertions(+), 15 deletions(-) diff --git a/javascript/ui.js b/javascript/ui.js index 88fd45ae9..f94ed081d 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -199,21 +199,18 @@ let txt2img_textarea, img2img_textarea = undefined; let wait_time = 800 let token_timeout; -function roll_artist_txt2img(prompt_text) { +function update_txt2img_tokens(...args) { update_token_counter("txt2img_token_button") - return prompt_text; + if (args.length == 2) + return args[0] + return args; } -function roll_artist_img2img(prompt_text) { + +function update_img2img_tokens(...args) { update_token_counter("img2img_token_button") - return prompt_text; -} -function update_style_txt2img(prompt_text, negative_prompt, style1, style2) { - update_token_counter("txt2img_token_button") - return [prompt_text, negative_prompt, style1, style2] -} -function update_style_img2img(prompt_text, negative_prompt, style1, style2) { - update_token_counter("img2img_token_button") - return [prompt_text, negative_prompt, style1, style2] + if (args.length == 2) + return args[0] + return args; } function update_token_counter(button_id) { diff --git a/modules/ui.py b/modules/ui.py index 6bf28562c..40c089841 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -543,7 +543,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): roll.click( fn=roll_artist, - _js="roll_artist_txt2img", + _js="update_txt2img_tokens", inputs=[ txt2img_prompt, ], @@ -749,7 +749,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): roll.click( fn=roll_artist, - _js="roll_artist_img2img", + _js="update_img2img_tokens", inputs=[ img2img_prompt, ], @@ -760,7 +760,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): prompts = [(txt2img_prompt, txt2img_negative_prompt), (img2img_prompt, img2img_negative_prompt)] style_dropdowns = [(txt2img_prompt_style, txt2img_prompt_style2), (img2img_prompt_style, img2img_prompt_style2)] - style_js_funcs = ["update_style_txt2img", "update_style_img2img"] + style_js_funcs = ["update_txt2img_tokens", "update_img2img_tokens"] for button, (prompt, negative_prompt) in zip([txt2img_save_style, img2img_save_style], prompts): button.click( From bdaa36c84470adbdce3e98c01a69af5e95adfb02 Mon Sep 17 00:00:00 2001 From: brkirch Date: Fri, 30 Sep 2022 23:53:25 -0400 Subject: [PATCH 032/138] When device is MPS, use CPU for GFPGAN instead GFPGAN will not work if the device is MPS, so default to CPU instead. --- modules/devices.py | 2 +- modules/gfpgan_model.py | 6 +++--- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/modules/devices.py b/modules/devices.py index 07bb23397..08bb26d6f 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -34,7 +34,7 @@ errors.run(enable_tf32, "Enabling TF32") device = get_optimal_device() -device_codeformer = cpu if has_mps else device +device_gfpgan = device_codeformer = cpu if device.type == 'mps' else device def randn(seed, shape): diff --git a/modules/gfpgan_model.py b/modules/gfpgan_model.py index bb30d7330..fcd8544a5 100644 --- a/modules/gfpgan_model.py +++ b/modules/gfpgan_model.py @@ -21,7 +21,7 @@ def gfpgann(): global loaded_gfpgan_model global model_path if loaded_gfpgan_model is not None: - loaded_gfpgan_model.gfpgan.to(shared.device) + loaded_gfpgan_model.gfpgan.to(devices.device_gfpgan) return loaded_gfpgan_model if gfpgan_constructor is None: @@ -36,8 +36,8 @@ def gfpgann(): else: print("Unable to load gfpgan model!") return None - model = gfpgan_constructor(model_path=model_file, upscale=1, arch='clean', channel_multiplier=2, bg_upsampler=None) - model.gfpgan.to(shared.device) + model = gfpgan_constructor(model_path=model_file, upscale=1, arch='clean', channel_multiplier=2, bg_upsampler=None, device=devices.device_gfpgan) + model.gfpgan.to(devices.device_gfpgan) loaded_gfpgan_model = model return model From 9de1e56e2dbb405213da9c221e0329d27f411691 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Fri, 30 Sep 2022 01:44:38 +0100 Subject: [PATCH 033/138] add sampler_noise_scheduler_override property --- modules/processing.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/modules/processing.py b/modules/processing.py index 7eeb5191c..1da753a2c 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -79,7 +79,7 @@ class StableDiffusionProcessing: self.paste_to = None self.color_corrections = None self.denoising_strength: float = 0 - + self.sampler_noise_scheduler_override = None self.ddim_discretize = opts.ddim_discretize self.s_churn = opts.s_churn self.s_tmin = opts.s_tmin @@ -130,7 +130,7 @@ class Processed: self.s_tmin = p.s_tmin self.s_tmax = p.s_tmax self.s_noise = p.s_noise - + self.sampler_noise_scheduler_override = p.sampler_noise_scheduler_override self.prompt = self.prompt if type(self.prompt) != list else self.prompt[0] self.negative_prompt = self.negative_prompt if type(self.negative_prompt) != list else self.negative_prompt[0] self.seed = int(self.seed if type(self.seed) != list else self.seed[0]) From bc38c80cfc83d4e2fc09c02dd49355664c05d15c Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Fri, 30 Sep 2022 01:46:06 +0100 Subject: [PATCH 034/138] add sampler_noise_scheduler_override switch --- modules/sd_samplers.py | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index dff89c092..925222148 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -290,7 +290,10 @@ class KDiffusionSampler: def sample_img2img(self, p, x, noise, conditioning, unconditional_conditioning, steps=None): steps, t_enc = setup_img2img_steps(p, steps) - sigmas = self.model_wrap.get_sigmas(steps) + if p.sampler_noise_scheduler_override: + sigmas = p.sampler_noise_scheduler_override(steps) + else: + sigmas = self.model_wrap.get_sigmas(steps) noise = noise * sigmas[steps - t_enc - 1] xi = x + noise @@ -306,7 +309,10 @@ class KDiffusionSampler: def sample(self, p, x, conditioning, unconditional_conditioning, steps=None): steps = steps or p.steps - sigmas = self.model_wrap.get_sigmas(steps) + if p.sampler_noise_scheduler_override: + sigmas = p.sampler_noise_scheduler_override(steps) + else: + sigmas = self.model_wrap.get_sigmas(steps) x = x * sigmas[0] extra_params_kwargs = self.initialize(p) From bd4fc6633f126c4e40448e36115ed79f1c2e746f Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Fri, 30 Sep 2022 02:53:30 +0100 Subject: [PATCH 035/138] add script alternate_sampler_noise_schedules --- scripts/alternate_sampler_noise_schedules.py | 53 ++++++++++++++++++++ 1 file changed, 53 insertions(+) create mode 100644 scripts/alternate_sampler_noise_schedules.py diff --git a/scripts/alternate_sampler_noise_schedules.py b/scripts/alternate_sampler_noise_schedules.py new file mode 100644 index 000000000..343dad411 --- /dev/null +++ b/scripts/alternate_sampler_noise_schedules.py @@ -0,0 +1,53 @@ +import inspect +from modules.processing import Processed, process_images +import gradio as gr +import modules.scripts as scripts +import k_diffusion.sampling +import torch + + +class Script(scripts.Script): + + def title(self): + return "Alternate Sampler Noise Schedules" + + def ui(self, is_img2img): + noise_scheduler = gr.Dropdown(label="Noise Scheduler", choices=['Default','Karras','Exponential', 'Variance Preserving'], value='Default', type="index") + sched_smin = gr.Slider(value=0.1, label="Sigma min", minimum=0.0, maximum=100.0, step=0.5,) + sched_smax = gr.Slider(value=10.0, label="Sigma max", minimum=0.0, maximum=100.0, step=0.5) + sched_rho = gr.Slider(value=7.0, label="Sigma rho (Karras only)", minimum=7.0, maximum=100.0, step=0.5) + sched_beta_d = gr.Slider(value=19.9, label="Beta distribution (VP only)",minimum=0.0, maximum=40.0, step=0.5) + sched_beta_min = gr.Slider(value=0.1, label="Beta min (VP only)", minimum=0.0, maximum=40.0, step=0.1) + sched_eps_s = gr.Slider(value=0.001, label="Epsilon (VP only)", minimum=0.001, maximum=1.0, step=0.001) + + return [noise_scheduler, sched_smin, sched_smax, sched_rho, sched_beta_d, sched_beta_min, sched_eps_s] + + def run(self, p, noise_scheduler, sched_smin, sched_smax, sched_rho, sched_beta_d, sched_beta_min, sched_eps_s): + + noise_scheduler_func_name = ['-','get_sigmas_karras','get_sigmas_exponential','get_sigmas_vp'][noise_scheduler] + + base_params = { + "sigma_min":sched_smin, + "sigma_max":sched_smax, + "rho":sched_rho, + "beta_d":sched_beta_d, + "beta_min":sched_beta_min, + "eps_s":sched_eps_s, + "device":"cuda" if torch.cuda.is_available() else "cpu" + } + + if hasattr(k_diffusion.sampling,noise_scheduler_func_name): + + sigma_func = getattr(k_diffusion.sampling,noise_scheduler_func_name) + sigma_func_kwargs = {} + + for k,v in base_params.items(): + if k in inspect.signature(sigma_func).parameters: + sigma_func_kwargs[k] = v + + def substitute_noise_scheduler(n): + return sigma_func(n,**sigma_func_kwargs) + + p.sampler_noise_scheduler_override = substitute_noise_scheduler + + return process_images(p) \ No newline at end of file From 3f417566b0bda8eab05d247567aebf001c1d1725 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 04:19:32 +0100 Subject: [PATCH 036/138] Delete alternate_sampler_noise_schedules.py --- scripts/alternate_sampler_noise_schedules.py | 53 -------------------- 1 file changed, 53 deletions(-) delete mode 100644 scripts/alternate_sampler_noise_schedules.py diff --git a/scripts/alternate_sampler_noise_schedules.py b/scripts/alternate_sampler_noise_schedules.py deleted file mode 100644 index 343dad411..000000000 --- a/scripts/alternate_sampler_noise_schedules.py +++ /dev/null @@ -1,53 +0,0 @@ -import inspect -from modules.processing import Processed, process_images -import gradio as gr -import modules.scripts as scripts -import k_diffusion.sampling -import torch - - -class Script(scripts.Script): - - def title(self): - return "Alternate Sampler Noise Schedules" - - def ui(self, is_img2img): - noise_scheduler = gr.Dropdown(label="Noise Scheduler", choices=['Default','Karras','Exponential', 'Variance Preserving'], value='Default', type="index") - sched_smin = gr.Slider(value=0.1, label="Sigma min", minimum=0.0, maximum=100.0, step=0.5,) - sched_smax = gr.Slider(value=10.0, label="Sigma max", minimum=0.0, maximum=100.0, step=0.5) - sched_rho = gr.Slider(value=7.0, label="Sigma rho (Karras only)", minimum=7.0, maximum=100.0, step=0.5) - sched_beta_d = gr.Slider(value=19.9, label="Beta distribution (VP only)",minimum=0.0, maximum=40.0, step=0.5) - sched_beta_min = gr.Slider(value=0.1, label="Beta min (VP only)", minimum=0.0, maximum=40.0, step=0.1) - sched_eps_s = gr.Slider(value=0.001, label="Epsilon (VP only)", minimum=0.001, maximum=1.0, step=0.001) - - return [noise_scheduler, sched_smin, sched_smax, sched_rho, sched_beta_d, sched_beta_min, sched_eps_s] - - def run(self, p, noise_scheduler, sched_smin, sched_smax, sched_rho, sched_beta_d, sched_beta_min, sched_eps_s): - - noise_scheduler_func_name = ['-','get_sigmas_karras','get_sigmas_exponential','get_sigmas_vp'][noise_scheduler] - - base_params = { - "sigma_min":sched_smin, - "sigma_max":sched_smax, - "rho":sched_rho, - "beta_d":sched_beta_d, - "beta_min":sched_beta_min, - "eps_s":sched_eps_s, - "device":"cuda" if torch.cuda.is_available() else "cpu" - } - - if hasattr(k_diffusion.sampling,noise_scheduler_func_name): - - sigma_func = getattr(k_diffusion.sampling,noise_scheduler_func_name) - sigma_func_kwargs = {} - - for k,v in base_params.items(): - if k in inspect.signature(sigma_func).parameters: - sigma_func_kwargs[k] = v - - def substitute_noise_scheduler(n): - return sigma_func(n,**sigma_func_kwargs) - - p.sampler_noise_scheduler_override = substitute_noise_scheduler - - return process_images(p) \ No newline at end of file From 4c2478a68a4f11959fe4887d38e0436eac19f97e Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 18:30:53 +0100 Subject: [PATCH 037/138] add script reload method --- modules/scripts.py | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/modules/scripts.py b/modules/scripts.py index 7c3bd5e74..3c14b9e32 100644 --- a/modules/scripts.py +++ b/modules/scripts.py @@ -165,3 +165,12 @@ class ScriptRunner: scripts_txt2img = ScriptRunner() scripts_img2img = ScriptRunner() + +def reload_scripts(basedir): + global scripts_txt2img,scripts_img2img + + scripts_data.clear() + load_scripts(basedir) + + scripts_txt2img = ScriptRunner() + scripts_img2img = ScriptRunner() From 95f35d04ab1636e08f69ca9c0ae2446714870e80 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 18:31:58 +0100 Subject: [PATCH 038/138] Host busy thread, check for reload --- webui.py | 46 +++++++++++++++++++++++++++++++--------------- 1 file changed, 31 insertions(+), 15 deletions(-) diff --git a/webui.py b/webui.py index b8cccd546..4948c394f 100644 --- a/webui.py +++ b/webui.py @@ -86,22 +86,38 @@ def webui(): signal.signal(signal.SIGINT, sigint_handler) - demo = modules.ui.create_ui( - txt2img=wrap_gradio_gpu_call(modules.txt2img.txt2img), - img2img=wrap_gradio_gpu_call(modules.img2img.img2img), - run_extras=wrap_gradio_gpu_call(modules.extras.run_extras), - run_pnginfo=modules.extras.run_pnginfo, - run_modelmerger=modules.extras.run_modelmerger - ) + while 1: - demo.launch( - share=cmd_opts.share, - server_name="0.0.0.0" if cmd_opts.listen else None, - server_port=cmd_opts.port, - debug=cmd_opts.gradio_debug, - auth=[tuple(cred.split(':')) for cred in cmd_opts.gradio_auth.strip('"').split(',')] if cmd_opts.gradio_auth else None, - inbrowser=cmd_opts.autolaunch, - ) + demo = modules.ui.create_ui( + txt2img=wrap_gradio_gpu_call(modules.txt2img.txt2img), + img2img=wrap_gradio_gpu_call(modules.img2img.img2img), + run_extras=wrap_gradio_gpu_call(modules.extras.run_extras), + run_pnginfo=modules.extras.run_pnginfo, + run_modelmerger=modules.extras.run_modelmerger + ) + + + demo.launch( + share=cmd_opts.share, + server_name="0.0.0.0" if cmd_opts.listen else None, + server_port=cmd_opts.port, + debug=cmd_opts.gradio_debug, + auth=[tuple(cred.split(':')) for cred in cmd_opts.gradio_auth.strip('"').split(',')] if cmd_opts.gradio_auth else None, + inbrowser=cmd_opts.autolaunch, + prevent_thread_lock=True + ) + + while 1: + time.sleep(0.5) + if getattr(demo,'do_restart',False): + time.sleep(0.5) + demo.close() + time.sleep(0.5) + break + + print('Reloading Scripts') + modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) + print('Restarting Gradio') if __name__ == "__main__": From 4f8490cd5630823ac44de8b5c5e4325bdbbea7fa Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 18:33:31 +0100 Subject: [PATCH 039/138] add restart button --- modules/ui.py | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/modules/ui.py b/modules/ui.py index 15572bb0a..ec6aaa288 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1002,6 +1002,17 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): _js='function(){}' ) + def request_restart(): + settings_interface.gradio_ref.do_restart = True + + restart_gradio = gr.Button(value='Restart Gradio and Refresh Scripts') + restart_gradio.click( + fn=request_restart, + inputs=[], + outputs=[], + _js='function(){document.body.innerHTML=\'

Reloading

\';setTimeout(function(){location.reload()},2000)}' + ) + if column is not None: column.__exit__() @@ -1026,7 +1037,9 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): css += css_hide_progressbar with gr.Blocks(css=css, analytics_enabled=False, title="Stable Diffusion") as demo: - + + settings_interface.gradio_ref = demo + with gr.Tabs() as tabs: for interface, label, ifid in interfaces: with gr.TabItem(label, id=ifid): From 121ed7d36febe94995774973b5edc1ba2ba84aad Mon Sep 17 00:00:00 2001 From: Alexandre Simard Date: Sat, 1 Oct 2022 14:04:20 -0400 Subject: [PATCH 040/138] Add progress bar for SwinIR in cmd I do not know how to add them to the UI... --- modules/swinir_model.py | 25 ++++++++++++++----------- webui-user.bat | 2 +- 2 files changed, 15 insertions(+), 12 deletions(-) diff --git a/modules/swinir_model.py b/modules/swinir_model.py index 41fda5a7c..9bd454c69 100644 --- a/modules/swinir_model.py +++ b/modules/swinir_model.py @@ -5,6 +5,7 @@ import numpy as np import torch from PIL import Image from basicsr.utils.download_util import load_file_from_url +from tqdm import tqdm from modules import modelloader from modules.paths import models_path @@ -122,18 +123,20 @@ def inference(img, model, tile, tile_overlap, window_size, scale): E = torch.zeros(b, c, h * sf, w * sf, dtype=torch.half, device=device).type_as(img) W = torch.zeros_like(E, dtype=torch.half, device=device) - for h_idx in h_idx_list: - for w_idx in w_idx_list: - in_patch = img[..., h_idx: h_idx + tile, w_idx: w_idx + tile] - out_patch = model(in_patch) - out_patch_mask = torch.ones_like(out_patch) + with tqdm(total=len(h_idx_list) * len(w_idx_list), desc="SwinIR tiles") as pbar: + for h_idx in h_idx_list: + for w_idx in w_idx_list: + in_patch = img[..., h_idx: h_idx + tile, w_idx: w_idx + tile] + out_patch = model(in_patch) + out_patch_mask = torch.ones_like(out_patch) - E[ - ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf - ].add_(out_patch) - W[ - ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf - ].add_(out_patch_mask) + E[ + ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf + ].add_(out_patch) + W[ + ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf + ].add_(out_patch_mask) + pbar.update(1) output = E.div_(W) return output diff --git a/webui-user.bat b/webui-user.bat index e5a257bef..5c7789535 100644 --- a/webui-user.bat +++ b/webui-user.bat @@ -3,6 +3,6 @@ set PYTHON= set GIT= set VENV_DIR= -set COMMANDLINE_ARGS= +set COMMANDLINE_ARGS=--autolaunch call webui.bat From b8a2b0453b62e4e99d0e5c049313402bc79056b5 Mon Sep 17 00:00:00 2001 From: Alexandre Simard Date: Sat, 1 Oct 2022 14:07:20 -0400 Subject: [PATCH 041/138] Set launch options to default --- webui-user.bat | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/webui-user.bat b/webui-user.bat index 5c7789535..e5a257bef 100644 --- a/webui-user.bat +++ b/webui-user.bat @@ -3,6 +3,6 @@ set PYTHON= set GIT= set VENV_DIR= -set COMMANDLINE_ARGS=--autolaunch +set COMMANDLINE_ARGS= call webui.bat From a9044475c06204deb886d2a69467d0d3a9f5c9be Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 21:47:42 +0100 Subject: [PATCH 042/138] add time import --- webui.py | 1 + 1 file changed, 1 insertion(+) diff --git a/webui.py b/webui.py index 4948c394f..e2c4c2baa 100644 --- a/webui.py +++ b/webui.py @@ -1,5 +1,6 @@ import os import threading +import time from modules import devices from modules.paths import script_path From afaa03c5fd05f48ed9c9f15558ea6f0bc4f61628 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 22:43:45 +0100 Subject: [PATCH 043/138] add redefinition guard to gradio_routes_templates_response --- modules/ui.py | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/modules/ui.py b/modules/ui.py index ec6aaa288..fd057916e 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1219,12 +1219,13 @@ for filename in sorted(os.listdir(jsdir)): javascript += f"\n" -def template_response(*args, **kwargs): - res = gradio_routes_templates_response(*args, **kwargs) - res.body = res.body.replace(b'', f'{javascript}'.encode("utf8")) - res.init_headers() - return res +if 'gradio_routes_templates_response' not in globals(): + def template_response(*args, **kwargs): + res = gradio_routes_templates_response(*args, **kwargs) + res.body = res.body.replace(b'', f'{javascript}'.encode("utf8")) + res.init_headers() + return res + gradio_routes_templates_response = gradio.routes.templates.TemplateResponse + gradio.routes.templates.TemplateResponse = template_response -gradio_routes_templates_response = gradio.routes.templates.TemplateResponse -gradio.routes.templates.TemplateResponse = template_response From 30f2e3565840544dd66470c6ef216ec664db6432 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 22:50:03 +0100 Subject: [PATCH 044/138] add importlib.reload --- webui.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/webui.py b/webui.py index e2c4c2baa..ab200045a 100644 --- a/webui.py +++ b/webui.py @@ -1,7 +1,7 @@ import os import threading import time - +import importlib from modules import devices from modules.paths import script_path import signal @@ -116,8 +116,10 @@ def webui(): time.sleep(0.5) break - print('Reloading Scripts') + print('Reloading Custom Scripts') modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) + print('Reloading modules: modules.ui') + importlib.reload(modules.ui) print('Restarting Gradio') From 6048002dade91b82b1ce9fea3c6ff5b5c1f8c990 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 23:10:07 +0100 Subject: [PATCH 045/138] Add scope warning to refresh button --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/ui.py b/modules/ui.py index fd057916e..72846a122 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1005,7 +1005,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): def request_restart(): settings_interface.gradio_ref.do_restart = True - restart_gradio = gr.Button(value='Restart Gradio and Refresh Scripts') + restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') restart_gradio.click( fn=request_restart, inputs=[], From 027c5aae5546ff3650347cb3c2b87df4415ab900 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 23:29:26 +0100 Subject: [PATCH 046/138] update reloading message style --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/ui.py b/modules/ui.py index 72846a122..7b2359c20 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1010,7 +1010,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): fn=request_restart, inputs=[], outputs=[], - _js='function(){document.body.innerHTML=\'

Reloading

\';setTimeout(function(){location.reload()},2000)}' + _js='function(){document.body.innerHTML=\'

Reloading...

\';setTimeout(function(){location.reload()},2000)}' ) if column is not None: From 55b046312c51bb7b2329d3b5b7f1c05956f821bf Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 00:12:49 +0100 Subject: [PATCH 047/138] move JavaScript into ui.js --- javascript/ui.js | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/javascript/ui.js b/javascript/ui.js index bfe024108..e8f289b44 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -212,3 +212,8 @@ function update_token_counter(button_id) { clearTimeout(token_timeout); token_timeout = setTimeout(() => gradioApp().getElementById(button_id)?.click(), wait_time); } + +function restart_reload(){ + document.body.innerHTML='

Reloading...

'; + setTimeout(function(){location.reload()},2000) +} From 0aa354bd5e811e2b41b17a3052cf5d4c8190d533 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 00:13:47 +0100 Subject: [PATCH 048/138] remove styling from python side --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/ui.py b/modules/ui.py index 7b2359c20..cb859ac45 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1010,7 +1010,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): fn=request_restart, inputs=[], outputs=[], - _js='function(){document.body.innerHTML=\'

Reloading...

\';setTimeout(function(){location.reload()},2000)}' + _js='function(){restart_reload()}' ) if column is not None: From cf33268d686986a24f2e04eb615f01ed53bfe308 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 01:18:42 +0100 Subject: [PATCH 049/138] add script body only refresh --- modules/scripts.py | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/modules/scripts.py b/modules/scripts.py index 3c14b9e32..788397f53 100644 --- a/modules/scripts.py +++ b/modules/scripts.py @@ -162,10 +162,33 @@ class ScriptRunner: return processed + def reload_sources(self): + for si,script in list(enumerate(self.scripts)): + with open(script.filename, "r", encoding="utf8") as file: + args_from = script.args_from + args_to = script.args_to + filename = script.filename + text = file.read() + + from types import ModuleType + compiled = compile(text, filename, 'exec') + module = ModuleType(script.filename) + exec(compiled, module.__dict__) + + for key, script_class in module.__dict__.items(): + if type(script_class) == type and issubclass(script_class, Script): + self.scripts[si] = script_class() + self.scripts[si].filename = filename + self.scripts[si].args_from = args_from + self.scripts[si].args_to = args_to scripts_txt2img = ScriptRunner() scripts_img2img = ScriptRunner() +def reload_script_body_only(): + scripts_txt2img.reload_sources() + scripts_img2img.reload_sources() + def reload_scripts(basedir): global scripts_txt2img,scripts_img2img From 07e40ad7f23472fc1c781fe1cc6c1ee403413918 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 01:19:55 +0100 Subject: [PATCH 050/138] add custom script body only refresh option --- modules/ui.py | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/modules/ui.py b/modules/ui.py index cb859ac45..eb7c05852 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1012,6 +1012,17 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): outputs=[], _js='function(){restart_reload()}' ) + + def reload_scripts(): + modules.scripts.reload_script_body_only() + + reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='primary') + reload_script_bodies.click( + fn=reload_scripts, + inputs=[], + outputs=[], + _js='function(){}' + ) if column is not None: column.__exit__() From 2deea867814272f1f089b60e9ba8d587c16b2fb1 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 01:36:30 +0100 Subject: [PATCH 051/138] Put reload buttons in row and add secondary style --- modules/ui.py | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/modules/ui.py b/modules/ui.py index eb7c05852..963a2c611 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1002,27 +1002,30 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): _js='function(){}' ) - def request_restart(): - settings_interface.gradio_ref.do_restart = True + with gr.Row(): + reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') + restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') - restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') - restart_gradio.click( - fn=request_restart, - inputs=[], - outputs=[], - _js='function(){restart_reload()}' - ) def reload_scripts(): modules.scripts.reload_script_body_only() - reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='primary') reload_script_bodies.click( fn=reload_scripts, inputs=[], outputs=[], _js='function(){}' ) + + def request_restart(): + settings_interface.gradio_ref.do_restart = True + + restart_gradio.click( + fn=request_restart, + inputs=[], + outputs=[], + _js='function(){restart_reload()}' + ) if column is not None: column.__exit__() From 3cf1a96006daffedb8ecd0ae142eca4c4da06105 Mon Sep 17 00:00:00 2001 From: RnDMonkey Date: Sat, 1 Oct 2022 21:11:03 -0700 Subject: [PATCH 052/138] added safety for blank directory naming patterns --- modules/images.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/modules/images.py b/modules/images.py index f1aed5d6b..e7894b4cd 100644 --- a/modules/images.py +++ b/modules/images.py @@ -311,7 +311,7 @@ def apply_filename_pattern(x, p, seed, prompt): x = x.replace("[cfg]", str(p.cfg_scale)) x = x.replace("[width]", str(p.width)) x = x.replace("[height]", str(p.height)) - x = x.replace("[styles]", sanitize_filename_part(", ".join([x for x in p.styles if not x == "None"]), replace_spaces=False)) + x = x.replace("[styles]", sanitize_filename_part(", ".join([x for x in p.styles if not x == "None"]) or "No styles", replace_spaces=False)) x = x.replace("[sampler]", sanitize_filename_part(sd_samplers.samplers[p.sampler_index].name, replace_spaces=False)) x = x.replace("[model_hash]", shared.sd_model.sd_model_hash) @@ -374,7 +374,7 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i save_to_dirs = (grid and opts.grid_save_to_dirs) or (not grid and opts.save_to_dirs and not no_prompt) if save_to_dirs: - dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, seed, prompt) + dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, seed, prompt).strip('\\ ') path = os.path.join(path, dirname) os.makedirs(path, exist_ok=True) From 70f526704721a303ae045f6406439dcceee4302e Mon Sep 17 00:00:00 2001 From: RnDMonkey Date: Sat, 1 Oct 2022 21:18:15 -0700 Subject: [PATCH 053/138] use os.path.normpath for better safety checking --- modules/images.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/modules/images.py b/modules/images.py index e7894b4cd..5ef7eb926 100644 --- a/modules/images.py +++ b/modules/images.py @@ -374,8 +374,8 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i save_to_dirs = (grid and opts.grid_save_to_dirs) or (not grid and opts.save_to_dirs and not no_prompt) if save_to_dirs: - dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, seed, prompt).strip('\\ ') - path = os.path.join(path, dirname) + dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, seed, prompt) + path = os.path.normpath(os.path.join(path, dirname)) os.makedirs(path, exist_ok=True) From 32edf1732f27a1fad5133667c22b948adda1b070 Mon Sep 17 00:00:00 2001 From: RnDMonkey Date: Sat, 1 Oct 2022 21:37:14 -0700 Subject: [PATCH 054/138] os.path.normpath wasn't working, reverting to manual strip --- modules/images.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/modules/images.py b/modules/images.py index 5ef7eb926..4998e92cf 100644 --- a/modules/images.py +++ b/modules/images.py @@ -374,8 +374,8 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i save_to_dirs = (grid and opts.grid_save_to_dirs) or (not grid and opts.save_to_dirs and not no_prompt) if save_to_dirs: - dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, seed, prompt) - path = os.path.normpath(os.path.join(path, dirname)) + dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, seed, prompt).strip('\\ /') + path = os.path.join(path, dirname) os.makedirs(path, exist_ok=True) From 820f1dc96b1979d7e92170c161db281ee8bd988b Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 15:03:39 +0300 Subject: [PATCH 055/138] initial support for training textual inversion --- .gitignore | 1 + javascript/progressbar.js | 1 + javascript/textualInversion.js | 8 + modules/devices.py | 3 +- modules/processing.py | 13 +- modules/sd_hijack.py | 324 +++--------------- modules/sd_hijack_optimizations.py | 164 +++++++++ modules/sd_models.py | 4 +- modules/shared.py | 3 +- modules/textual_inversion/dataset.py | 76 ++++ .../textual_inversion/textual_inversion.py | 258 ++++++++++++++ modules/textual_inversion/ui.py | 32 ++ modules/ui.py | 139 +++++++- style.css | 10 +- textual_inversion_templates/style.txt | 19 + .../style_filewords.txt | 19 + textual_inversion_templates/subject.txt | 27 ++ .../subject_filewords.txt | 27 ++ webui.py | 15 +- 19 files changed, 828 insertions(+), 315 deletions(-) create mode 100644 javascript/textualInversion.js create mode 100644 modules/sd_hijack_optimizations.py create mode 100644 modules/textual_inversion/dataset.py create mode 100644 modules/textual_inversion/textual_inversion.py create mode 100644 modules/textual_inversion/ui.py create mode 100644 textual_inversion_templates/style.txt create mode 100644 textual_inversion_templates/style_filewords.txt create mode 100644 textual_inversion_templates/subject.txt create mode 100644 textual_inversion_templates/subject_filewords.txt diff --git a/.gitignore b/.gitignore index 3532dab37..7afc93953 100644 --- a/.gitignore +++ b/.gitignore @@ -25,3 +25,4 @@ __pycache__ /.idea notification.mp3 /SwinIR +/textual_inversion diff --git a/javascript/progressbar.js b/javascript/progressbar.js index 21f25b38d..1e297abbe 100644 --- a/javascript/progressbar.js +++ b/javascript/progressbar.js @@ -30,6 +30,7 @@ function check_progressbar(id_part, id_progressbar, id_progressbar_span, id_inte onUiUpdate(function(){ check_progressbar('txt2img', 'txt2img_progressbar', 'txt2img_progress_span', 'txt2img_interrupt', 'txt2img_preview', 'txt2img_gallery') check_progressbar('img2img', 'img2img_progressbar', 'img2img_progress_span', 'img2img_interrupt', 'img2img_preview', 'img2img_gallery') + check_progressbar('ti', 'ti_progressbar', 'ti_progress_span', 'ti_interrupt', 'ti_preview', 'ti_gallery') }) function requestMoreProgress(id_part, id_progressbar_span, id_interrupt){ diff --git a/javascript/textualInversion.js b/javascript/textualInversion.js new file mode 100644 index 000000000..8061be089 --- /dev/null +++ b/javascript/textualInversion.js @@ -0,0 +1,8 @@ + + +function start_training_textual_inversion(){ + requestProgress('ti') + gradioApp().querySelector('#ti_error').innerHTML='' + + return args_to_array(arguments) +} diff --git a/modules/devices.py b/modules/devices.py index 07bb23397..ff82f2f64 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -32,10 +32,9 @@ def enable_tf32(): errors.run(enable_tf32, "Enabling TF32") - device = get_optimal_device() device_codeformer = cpu if has_mps else device - +dtype = torch.float16 def randn(seed, shape): # Pytorch currently doesn't handle setting randomness correctly when the metal backend is used. diff --git a/modules/processing.py b/modules/processing.py index 7eeb5191c..8223423ab 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -56,7 +56,7 @@ class StableDiffusionProcessing: self.prompt: str = prompt self.prompt_for_display: str = None self.negative_prompt: str = (negative_prompt or "") - self.styles: str = styles + self.styles: list = styles or [] self.seed: int = seed self.subseed: int = subseed self.subseed_strength: float = subseed_strength @@ -271,7 +271,7 @@ def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments, iteration "Variation seed strength": (None if p.subseed_strength == 0 else p.subseed_strength), "Seed resize from": (None if p.seed_resize_from_w == 0 or p.seed_resize_from_h == 0 else f"{p.seed_resize_from_w}x{p.seed_resize_from_h}"), "Denoising strength": getattr(p, 'denoising_strength', None), - "Eta": (None if p.sampler.eta == p.sampler.default_eta else p.sampler.eta), + "Eta": (None if p.sampler is None or p.sampler.eta == p.sampler.default_eta else p.sampler.eta), } generation_params.update(p.extra_generation_params) @@ -295,8 +295,11 @@ def process_images(p: StableDiffusionProcessing) -> Processed: fix_seed(p) - os.makedirs(p.outpath_samples, exist_ok=True) - os.makedirs(p.outpath_grids, exist_ok=True) + if p.outpath_samples is not None: + os.makedirs(p.outpath_samples, exist_ok=True) + + if p.outpath_grids is not None: + os.makedirs(p.outpath_grids, exist_ok=True) modules.sd_hijack.model_hijack.apply_circular(p.tiling) @@ -323,7 +326,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: return create_infotext(p, all_prompts, all_seeds, all_subseeds, comments, iteration, position_in_batch) if os.path.exists(cmd_opts.embeddings_dir): - model_hijack.load_textual_inversion_embeddings(cmd_opts.embeddings_dir, p.sd_model) + model_hijack.embedding_db.load_textual_inversion_embeddings() infotexts = [] output_images = [] diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index fa7eaeb89..fd57e5c54 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -6,244 +6,41 @@ import torch import numpy as np from torch import einsum -from modules import prompt_parser +import modules.textual_inversion.textual_inversion +from modules import prompt_parser, devices, sd_hijack_optimizations, shared from modules.shared import opts, device, cmd_opts -from ldm.util import default -from einops import rearrange import ldm.modules.attention import ldm.modules.diffusionmodules.model - -# see https://github.com/basujindal/stable-diffusion/pull/117 for discussion -def split_cross_attention_forward_v1(self, x, context=None, mask=None): - h = self.heads - - q = self.to_q(x) - context = default(context, x) - k = self.to_k(context) - v = self.to_v(context) - del context, x - - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) - - r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device) - for i in range(0, q.shape[0], 2): - end = i + 2 - s1 = einsum('b i d, b j d -> b i j', q[i:end], k[i:end]) - s1 *= self.scale - - s2 = s1.softmax(dim=-1) - del s1 - - r1[i:end] = einsum('b i j, b j d -> b i d', s2, v[i:end]) - del s2 - - r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) - del r1 - - return self.to_out(r2) +attention_CrossAttention_forward = ldm.modules.attention.CrossAttention.forward +diffusionmodules_model_nonlinearity = ldm.modules.diffusionmodules.model.nonlinearity +diffusionmodules_model_AttnBlock_forward = ldm.modules.diffusionmodules.model.AttnBlock.forward -# taken from https://github.com/Doggettx/stable-diffusion -def split_cross_attention_forward(self, x, context=None, mask=None): - h = self.heads +def apply_optimizations(): + if cmd_opts.opt_split_attention_v1: + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward_v1 + elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward + ldm.modules.diffusionmodules.model.nonlinearity = sd_hijack_optimizations.nonlinearity_hijack + ldm.modules.diffusionmodules.model.AttnBlock.forward = sd_hijack_optimizations.cross_attention_attnblock_forward - q_in = self.to_q(x) - context = default(context, x) - k_in = self.to_k(context) * self.scale - v_in = self.to_v(context) - del context, x - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q_in, k_in, v_in)) - del q_in, k_in, v_in +def undo_optimizations(): + ldm.modules.attention.CrossAttention.forward = attention_CrossAttention_forward + ldm.modules.diffusionmodules.model.nonlinearity = diffusionmodules_model_nonlinearity + ldm.modules.diffusionmodules.model.AttnBlock.forward = diffusionmodules_model_AttnBlock_forward - r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype) - - stats = torch.cuda.memory_stats(q.device) - mem_active = stats['active_bytes.all.current'] - mem_reserved = stats['reserved_bytes.all.current'] - mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) - mem_free_torch = mem_reserved - mem_active - mem_free_total = mem_free_cuda + mem_free_torch - - gb = 1024 ** 3 - tensor_size = q.shape[0] * q.shape[1] * k.shape[1] * q.element_size() - modifier = 3 if q.element_size() == 2 else 2.5 - mem_required = tensor_size * modifier - steps = 1 - - if mem_required > mem_free_total: - steps = 2 ** (math.ceil(math.log(mem_required / mem_free_total, 2))) - # print(f"Expected tensor size:{tensor_size/gb:0.1f}GB, cuda free:{mem_free_cuda/gb:0.1f}GB " - # f"torch free:{mem_free_torch/gb:0.1f} total:{mem_free_total/gb:0.1f} steps:{steps}") - - if steps > 64: - max_res = math.floor(math.sqrt(math.sqrt(mem_free_total / 2.5)) / 8) * 64 - raise RuntimeError(f'Not enough memory, use lower resolution (max approx. {max_res}x{max_res}). ' - f'Need: {mem_required / 64 / gb:0.1f}GB free, Have:{mem_free_total / gb:0.1f}GB free') - - slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] - for i in range(0, q.shape[1], slice_size): - end = i + slice_size - s1 = einsum('b i d, b j d -> b i j', q[:, i:end], k) - - s2 = s1.softmax(dim=-1, dtype=q.dtype) - del s1 - - r1[:, i:end] = einsum('b i j, b j d -> b i d', s2, v) - del s2 - - del q, k, v - - r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) - del r1 - - return self.to_out(r2) - -def nonlinearity_hijack(x): - # swish - t = torch.sigmoid(x) - x *= t - del t - - return x - -def cross_attention_attnblock_forward(self, x): - h_ = x - h_ = self.norm(h_) - q1 = self.q(h_) - k1 = self.k(h_) - v = self.v(h_) - - # compute attention - b, c, h, w = q1.shape - - q2 = q1.reshape(b, c, h*w) - del q1 - - q = q2.permute(0, 2, 1) # b,hw,c - del q2 - - k = k1.reshape(b, c, h*w) # b,c,hw - del k1 - - h_ = torch.zeros_like(k, device=q.device) - - stats = torch.cuda.memory_stats(q.device) - mem_active = stats['active_bytes.all.current'] - mem_reserved = stats['reserved_bytes.all.current'] - mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) - mem_free_torch = mem_reserved - mem_active - mem_free_total = mem_free_cuda + mem_free_torch - - tensor_size = q.shape[0] * q.shape[1] * k.shape[2] * q.element_size() - mem_required = tensor_size * 2.5 - steps = 1 - - if mem_required > mem_free_total: - steps = 2**(math.ceil(math.log(mem_required / mem_free_total, 2))) - - slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] - for i in range(0, q.shape[1], slice_size): - end = i + slice_size - - w1 = torch.bmm(q[:, i:end], k) # b,hw,hw w[b,i,j]=sum_c q[b,i,c]k[b,c,j] - w2 = w1 * (int(c)**(-0.5)) - del w1 - w3 = torch.nn.functional.softmax(w2, dim=2, dtype=q.dtype) - del w2 - - # attend to values - v1 = v.reshape(b, c, h*w) - w4 = w3.permute(0, 2, 1) # b,hw,hw (first hw of k, second of q) - del w3 - - h_[:, :, i:end] = torch.bmm(v1, w4) # b, c,hw (hw of q) h_[b,c,j] = sum_i v[b,c,i] w_[b,i,j] - del v1, w4 - - h2 = h_.reshape(b, c, h, w) - del h_ - - h3 = self.proj_out(h2) - del h2 - - h3 += x - - return h3 class StableDiffusionModelHijack: - ids_lookup = {} - word_embeddings = {} - word_embeddings_checksums = {} fixes = None comments = [] - dir_mtime = None layers = None circular_enabled = False clip = None - def load_textual_inversion_embeddings(self, dirname, model): - mt = os.path.getmtime(dirname) - if self.dir_mtime is not None and mt <= self.dir_mtime: - return - - self.dir_mtime = mt - self.ids_lookup.clear() - self.word_embeddings.clear() - - tokenizer = model.cond_stage_model.tokenizer - - def const_hash(a): - r = 0 - for v in a: - r = (r * 281 ^ int(v) * 997) & 0xFFFFFFFF - return r - - def process_file(path, filename): - name = os.path.splitext(filename)[0] - - data = torch.load(path, map_location="cpu") - - # textual inversion embeddings - if 'string_to_param' in data: - param_dict = data['string_to_param'] - if hasattr(param_dict, '_parameters'): - param_dict = getattr(param_dict, '_parameters') # fix for torch 1.12.1 loading saved file from torch 1.11 - assert len(param_dict) == 1, 'embedding file has multiple terms in it' - emb = next(iter(param_dict.items()))[1] - # diffuser concepts - elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor: - assert len(data.keys()) == 1, 'embedding file has multiple terms in it' - - emb = next(iter(data.values())) - if len(emb.shape) == 1: - emb = emb.unsqueeze(0) - - self.word_embeddings[name] = emb.detach().to(device) - self.word_embeddings_checksums[name] = f'{const_hash(emb.reshape(-1)*100)&0xffff:04x}' - - ids = tokenizer([name], add_special_tokens=False)['input_ids'][0] - - first_id = ids[0] - if first_id not in self.ids_lookup: - self.ids_lookup[first_id] = [] - self.ids_lookup[first_id].append((ids, name)) - - for fn in os.listdir(dirname): - try: - fullfn = os.path.join(dirname, fn) - - if os.stat(fullfn).st_size == 0: - continue - - process_file(fullfn, fn) - except Exception: - print(f"Error loading emedding {fn}:", file=sys.stderr) - print(traceback.format_exc(), file=sys.stderr) - continue - - print(f"Loaded a total of {len(self.word_embeddings)} textual inversion embeddings.") + embedding_db = modules.textual_inversion.textual_inversion.EmbeddingDatabase(cmd_opts.embeddings_dir) def hijack(self, m): model_embeddings = m.cond_stage_model.transformer.text_model.embeddings @@ -253,12 +50,7 @@ class StableDiffusionModelHijack: self.clip = m.cond_stage_model - if cmd_opts.opt_split_attention_v1: - ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward_v1 - elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): - ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward - ldm.modules.diffusionmodules.model.nonlinearity = nonlinearity_hijack - ldm.modules.diffusionmodules.model.AttnBlock.forward = cross_attention_attnblock_forward + apply_optimizations() def flatten(el): flattened = [flatten(children) for children in el.children()] @@ -296,7 +88,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): def __init__(self, wrapped, hijack): super().__init__() self.wrapped = wrapped - self.hijack = hijack + self.hijack: StableDiffusionModelHijack = hijack self.tokenizer = wrapped.tokenizer self.max_length = wrapped.max_length self.token_mults = {} @@ -317,7 +109,6 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): if mult != 1.0: self.token_mults[ident] = mult - def tokenize_line(self, line, used_custom_terms, hijack_comments): id_start = self.wrapped.tokenizer.bos_token_id id_end = self.wrapped.tokenizer.eos_token_id @@ -339,28 +130,19 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - possible_matches = self.hijack.ids_lookup.get(token, None) + embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) - if possible_matches is None: + if embedding is None: remade_tokens.append(token) multipliers.append(weight) + i += 1 else: - found = False - for ids, word in possible_matches: - if tokens[i:i + len(ids)] == ids: - emb_len = int(self.hijack.word_embeddings[word].shape[0]) - fixes.append((len(remade_tokens), word)) - remade_tokens += [0] * emb_len - multipliers += [weight] * emb_len - i += len(ids) - 1 - found = True - used_custom_terms.append((word, self.hijack.word_embeddings_checksums[word])) - break - - if not found: - remade_tokens.append(token) - multipliers.append(weight) - i += 1 + emb_len = int(embedding.vec.shape[0]) + fixes.append((len(remade_tokens), embedding)) + remade_tokens += [0] * emb_len + multipliers += [weight] * emb_len + used_custom_terms.append((embedding.name, embedding.checksum())) + i += emb_len if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} @@ -431,32 +213,23 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - possible_matches = self.hijack.ids_lookup.get(token, None) + embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) mult_change = self.token_mults.get(token) if opts.enable_emphasis else None if mult_change is not None: mult *= mult_change - elif possible_matches is None: + i += 1 + elif embedding is None: remade_tokens.append(token) multipliers.append(mult) + i += 1 else: - found = False - for ids, word in possible_matches: - if tokens[i:i+len(ids)] == ids: - emb_len = int(self.hijack.word_embeddings[word].shape[0]) - fixes.append((len(remade_tokens), word)) - remade_tokens += [0] * emb_len - multipliers += [mult] * emb_len - i += len(ids) - 1 - found = True - used_custom_terms.append((word, self.hijack.word_embeddings_checksums[word])) - break - - if not found: - remade_tokens.append(token) - multipliers.append(mult) - - i += 1 + emb_len = int(embedding.vec.shape[0]) + fixes.append((len(remade_tokens), embedding)) + remade_tokens += [0] * emb_len + multipliers += [mult] * emb_len + used_custom_terms.append((embedding.name, embedding.checksum())) + i += emb_len if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} @@ -464,6 +237,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): overflowing_words = [vocab.get(int(x), "") for x in ovf] overflowing_text = self.wrapped.tokenizer.convert_tokens_to_string(''.join(overflowing_words)) hijack_comments.append(f"Warning: too many input tokens; some ({len(overflowing_words)}) have been truncated:\n{overflowing_text}\n") + token_count = len(remade_tokens) remade_tokens = remade_tokens + [id_end] * (maxlen - 2 - len(remade_tokens)) remade_tokens = [id_start] + remade_tokens[0:maxlen-2] + [id_end] @@ -484,7 +258,6 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): else: batch_multipliers, remade_batch_tokens, used_custom_terms, hijack_comments, hijack_fixes, token_count = self.process_text(text) - self.hijack.fixes = hijack_fixes self.hijack.comments = hijack_comments @@ -517,14 +290,19 @@ class EmbeddingsWithFixes(torch.nn.Module): inputs_embeds = self.wrapped(input_ids) - if batch_fixes is not None: - for fixes, tensor in zip(batch_fixes, inputs_embeds): - for offset, word in fixes: - emb = self.embeddings.word_embeddings[word] - emb_len = min(tensor.shape[0]-offset-1, emb.shape[0]) - tensor[offset+1:offset+1+emb_len] = self.embeddings.word_embeddings[word][0:emb_len] + if batch_fixes is None or len(batch_fixes) == 0 or max([len(x) for x in batch_fixes]) == 0: + return inputs_embeds - return inputs_embeds + vecs = [] + for fixes, tensor in zip(batch_fixes, inputs_embeds): + for offset, embedding in fixes: + emb = embedding.vec + emb_len = min(tensor.shape[0]-offset-1, emb.shape[0]) + tensor = torch.cat([tensor[0:offset+1], emb[0:emb_len], tensor[offset+1+emb_len:]]) + + vecs.append(tensor) + + return torch.stack(vecs) def add_circular_option_to_conv_2d(): diff --git a/modules/sd_hijack_optimizations.py b/modules/sd_hijack_optimizations.py new file mode 100644 index 000000000..9c079e578 --- /dev/null +++ b/modules/sd_hijack_optimizations.py @@ -0,0 +1,164 @@ +import math +import torch +from torch import einsum + +from ldm.util import default +from einops import rearrange + + +# see https://github.com/basujindal/stable-diffusion/pull/117 for discussion +def split_cross_attention_forward_v1(self, x, context=None, mask=None): + h = self.heads + + q = self.to_q(x) + context = default(context, x) + k = self.to_k(context) + v = self.to_v(context) + del context, x + + q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) + + r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device) + for i in range(0, q.shape[0], 2): + end = i + 2 + s1 = einsum('b i d, b j d -> b i j', q[i:end], k[i:end]) + s1 *= self.scale + + s2 = s1.softmax(dim=-1) + del s1 + + r1[i:end] = einsum('b i j, b j d -> b i d', s2, v[i:end]) + del s2 + + r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) + del r1 + + return self.to_out(r2) + + +# taken from https://github.com/Doggettx/stable-diffusion +def split_cross_attention_forward(self, x, context=None, mask=None): + h = self.heads + + q_in = self.to_q(x) + context = default(context, x) + k_in = self.to_k(context) * self.scale + v_in = self.to_v(context) + del context, x + + q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q_in, k_in, v_in)) + del q_in, k_in, v_in + + r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype) + + stats = torch.cuda.memory_stats(q.device) + mem_active = stats['active_bytes.all.current'] + mem_reserved = stats['reserved_bytes.all.current'] + mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) + mem_free_torch = mem_reserved - mem_active + mem_free_total = mem_free_cuda + mem_free_torch + + gb = 1024 ** 3 + tensor_size = q.shape[0] * q.shape[1] * k.shape[1] * q.element_size() + modifier = 3 if q.element_size() == 2 else 2.5 + mem_required = tensor_size * modifier + steps = 1 + + if mem_required > mem_free_total: + steps = 2 ** (math.ceil(math.log(mem_required / mem_free_total, 2))) + # print(f"Expected tensor size:{tensor_size/gb:0.1f}GB, cuda free:{mem_free_cuda/gb:0.1f}GB " + # f"torch free:{mem_free_torch/gb:0.1f} total:{mem_free_total/gb:0.1f} steps:{steps}") + + if steps > 64: + max_res = math.floor(math.sqrt(math.sqrt(mem_free_total / 2.5)) / 8) * 64 + raise RuntimeError(f'Not enough memory, use lower resolution (max approx. {max_res}x{max_res}). ' + f'Need: {mem_required / 64 / gb:0.1f}GB free, Have:{mem_free_total / gb:0.1f}GB free') + + slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] + for i in range(0, q.shape[1], slice_size): + end = i + slice_size + s1 = einsum('b i d, b j d -> b i j', q[:, i:end], k) + + s2 = s1.softmax(dim=-1, dtype=q.dtype) + del s1 + + r1[:, i:end] = einsum('b i j, b j d -> b i d', s2, v) + del s2 + + del q, k, v + + r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) + del r1 + + return self.to_out(r2) + +def nonlinearity_hijack(x): + # swish + t = torch.sigmoid(x) + x *= t + del t + + return x + +def cross_attention_attnblock_forward(self, x): + h_ = x + h_ = self.norm(h_) + q1 = self.q(h_) + k1 = self.k(h_) + v = self.v(h_) + + # compute attention + b, c, h, w = q1.shape + + q2 = q1.reshape(b, c, h*w) + del q1 + + q = q2.permute(0, 2, 1) # b,hw,c + del q2 + + k = k1.reshape(b, c, h*w) # b,c,hw + del k1 + + h_ = torch.zeros_like(k, device=q.device) + + stats = torch.cuda.memory_stats(q.device) + mem_active = stats['active_bytes.all.current'] + mem_reserved = stats['reserved_bytes.all.current'] + mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) + mem_free_torch = mem_reserved - mem_active + mem_free_total = mem_free_cuda + mem_free_torch + + tensor_size = q.shape[0] * q.shape[1] * k.shape[2] * q.element_size() + mem_required = tensor_size * 2.5 + steps = 1 + + if mem_required > mem_free_total: + steps = 2**(math.ceil(math.log(mem_required / mem_free_total, 2))) + + slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] + for i in range(0, q.shape[1], slice_size): + end = i + slice_size + + w1 = torch.bmm(q[:, i:end], k) # b,hw,hw w[b,i,j]=sum_c q[b,i,c]k[b,c,j] + w2 = w1 * (int(c)**(-0.5)) + del w1 + w3 = torch.nn.functional.softmax(w2, dim=2, dtype=q.dtype) + del w2 + + # attend to values + v1 = v.reshape(b, c, h*w) + w4 = w3.permute(0, 2, 1) # b,hw,hw (first hw of k, second of q) + del w3 + + h_[:, :, i:end] = torch.bmm(v1, w4) # b, c,hw (hw of q) h_[b,c,j] = sum_i v[b,c,i] w_[b,i,j] + del v1, w4 + + h2 = h_.reshape(b, c, h, w) + del h_ + + h3 = self.proj_out(h2) + del h2 + + h3 += x + + return h3 diff --git a/modules/sd_models.py b/modules/sd_models.py index 2539f14cd..5b3dbdc79 100644 --- a/modules/sd_models.py +++ b/modules/sd_models.py @@ -8,7 +8,7 @@ from omegaconf import OmegaConf from ldm.util import instantiate_from_config -from modules import shared, modelloader +from modules import shared, modelloader, devices from modules.paths import models_path model_dir = "Stable-diffusion" @@ -134,6 +134,8 @@ def load_model_weights(model, checkpoint_file, sd_model_hash): if not shared.cmd_opts.no_half: model.half() + devices.dtype = torch.float32 if shared.cmd_opts.no_half else torch.float16 + model.sd_model_hash = sd_model_hash model.sd_model_checkpint = checkpoint_file diff --git a/modules/shared.py b/modules/shared.py index ac968b2d2..ac0bc480c 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -78,6 +78,7 @@ class State: current_latent = None current_image = None current_image_sampling_step = 0 + textinfo = None def interrupt(self): self.interrupted = True @@ -88,7 +89,7 @@ class State: self.current_image_sampling_step = 0 def get_job_timestamp(self): - return datetime.datetime.now().strftime("%Y%m%d%H%M%S") + return datetime.datetime.now().strftime("%Y%m%d%H%M%S") # shouldn't this return job_timestamp? state = State() diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py new file mode 100644 index 000000000..7e134a08f --- /dev/null +++ b/modules/textual_inversion/dataset.py @@ -0,0 +1,76 @@ +import os +import numpy as np +import PIL +import torch +from PIL import Image +from torch.utils.data import Dataset +from torchvision import transforms + +import random +import tqdm + + +class PersonalizedBase(Dataset): + def __init__(self, data_root, size=None, repeats=100, flip_p=0.5, placeholder_token="*", width=512, height=512, model=None, device=None, template_file=None): + + self.placeholder_token = placeholder_token + + self.size = size + self.width = width + self.height = height + self.flip = transforms.RandomHorizontalFlip(p=flip_p) + + self.dataset = [] + + with open(template_file, "r") as file: + lines = [x.strip() for x in file.readlines()] + + self.lines = lines + + assert data_root, 'dataset directory not specified' + + self.image_paths = [os.path.join(data_root, file_path) for file_path in os.listdir(data_root)] + print("Preparing dataset...") + for path in tqdm.tqdm(self.image_paths): + image = Image.open(path) + image = image.convert('RGB') + image = image.resize((self.width, self.height), PIL.Image.BICUBIC) + + filename = os.path.basename(path) + filename_tokens = os.path.splitext(filename)[0].replace('_', '-').replace(' ', '-').split('-') + filename_tokens = [token for token in filename_tokens if token.isalpha()] + + npimage = np.array(image).astype(np.uint8) + npimage = (npimage / 127.5 - 1.0).astype(np.float32) + + torchdata = torch.from_numpy(npimage).to(device=device, dtype=torch.float32) + torchdata = torch.moveaxis(torchdata, 2, 0) + + init_latent = model.get_first_stage_encoding(model.encode_first_stage(torchdata.unsqueeze(dim=0))).squeeze() + + self.dataset.append((init_latent, filename_tokens)) + + self.length = len(self.dataset) * repeats + + self.initial_indexes = np.arange(self.length) % len(self.dataset) + self.indexes = None + self.shuffle() + + def shuffle(self): + self.indexes = self.initial_indexes[torch.randperm(self.initial_indexes.shape[0])] + + def __len__(self): + return self.length + + def __getitem__(self, i): + if i % len(self.dataset) == 0: + self.shuffle() + + index = self.indexes[i % len(self.indexes)] + x, filename_tokens = self.dataset[index] + + text = random.choice(self.lines) + text = text.replace("[name]", self.placeholder_token) + text = text.replace("[filewords]", ' '.join(filename_tokens)) + + return x, text diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py new file mode 100644 index 000000000..c0baaace2 --- /dev/null +++ b/modules/textual_inversion/textual_inversion.py @@ -0,0 +1,258 @@ +import os +import sys +import traceback + +import torch +import tqdm +import html +import datetime + +from modules import shared, devices, sd_hijack, processing +import modules.textual_inversion.dataset + + +class Embedding: + def __init__(self, vec, name, step=None): + self.vec = vec + self.name = name + self.step = step + self.cached_checksum = None + + def save(self, filename): + embedding_data = { + "string_to_token": {"*": 265}, + "string_to_param": {"*": self.vec}, + "name": self.name, + "step": self.step, + } + + torch.save(embedding_data, filename) + + def checksum(self): + if self.cached_checksum is not None: + return self.cached_checksum + + def const_hash(a): + r = 0 + for v in a: + r = (r * 281 ^ int(v) * 997) & 0xFFFFFFFF + return r + + self.cached_checksum = f'{const_hash(self.vec.reshape(-1) * 100) & 0xffff:04x}' + return self.cached_checksum + +class EmbeddingDatabase: + def __init__(self, embeddings_dir): + self.ids_lookup = {} + self.word_embeddings = {} + self.dir_mtime = None + self.embeddings_dir = embeddings_dir + + def register_embedding(self, embedding, model): + + self.word_embeddings[embedding.name] = embedding + + ids = model.cond_stage_model.tokenizer([embedding.name], add_special_tokens=False)['input_ids'][0] + + first_id = ids[0] + if first_id not in self.ids_lookup: + self.ids_lookup[first_id] = [] + self.ids_lookup[first_id].append((ids, embedding)) + + return embedding + + def load_textual_inversion_embeddings(self): + mt = os.path.getmtime(self.embeddings_dir) + if self.dir_mtime is not None and mt <= self.dir_mtime: + return + + self.dir_mtime = mt + self.ids_lookup.clear() + self.word_embeddings.clear() + + def process_file(path, filename): + name = os.path.splitext(filename)[0] + + data = torch.load(path, map_location="cpu") + + # textual inversion embeddings + if 'string_to_param' in data: + param_dict = data['string_to_param'] + if hasattr(param_dict, '_parameters'): + param_dict = getattr(param_dict, '_parameters') # fix for torch 1.12.1 loading saved file from torch 1.11 + assert len(param_dict) == 1, 'embedding file has multiple terms in it' + emb = next(iter(param_dict.items()))[1] + # diffuser concepts + elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor: + assert len(data.keys()) == 1, 'embedding file has multiple terms in it' + + emb = next(iter(data.values())) + if len(emb.shape) == 1: + emb = emb.unsqueeze(0) + else: + raise Exception(f"Couldn't identify {filename} as neither textual inversion embedding nor diffuser concept.") + + vec = emb.detach().to(devices.device, dtype=torch.float32) + embedding = Embedding(vec, name) + embedding.step = data.get('step', None) + self.register_embedding(embedding, shared.sd_model) + + for fn in os.listdir(self.embeddings_dir): + try: + fullfn = os.path.join(self.embeddings_dir, fn) + + if os.stat(fullfn).st_size == 0: + continue + + process_file(fullfn, fn) + except Exception: + print(f"Error loading emedding {fn}:", file=sys.stderr) + print(traceback.format_exc(), file=sys.stderr) + continue + + print(f"Loaded a total of {len(self.word_embeddings)} textual inversion embeddings.") + + def find_embedding_at_position(self, tokens, offset): + token = tokens[offset] + possible_matches = self.ids_lookup.get(token, None) + + if possible_matches is None: + return None + + for ids, embedding in possible_matches: + if tokens[offset:offset + len(ids)] == ids: + return embedding + + return None + + + +def create_embedding(name, num_vectors_per_token): + init_text = '*' + + cond_model = shared.sd_model.cond_stage_model + embedding_layer = cond_model.wrapped.transformer.text_model.embeddings + + ids = cond_model.tokenizer(init_text, max_length=num_vectors_per_token, return_tensors="pt", add_special_tokens=False)["input_ids"] + embedded = embedding_layer(ids.to(devices.device)).squeeze(0) + vec = torch.zeros((num_vectors_per_token, embedded.shape[1]), device=devices.device) + + for i in range(num_vectors_per_token): + vec[i] = embedded[i * int(embedded.shape[0]) // num_vectors_per_token] + + fn = os.path.join(shared.cmd_opts.embeddings_dir, f"{name}.pt") + assert not os.path.exists(fn), f"file {fn} already exists" + + embedding = Embedding(vec, name) + embedding.step = 0 + embedding.save(fn) + + return fn + + +def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, create_image_every, save_embedding_every, template_file): + assert embedding_name, 'embedding not selected' + + shared.state.textinfo = "Initializing textual inversion training..." + shared.state.job_count = steps + + filename = os.path.join(shared.cmd_opts.embeddings_dir, f'{embedding_name}.pt') + + log_directory = os.path.join(log_directory, datetime.datetime.now().strftime("%Y-%d-%m"), embedding_name) + + if save_embedding_every > 0: + embedding_dir = os.path.join(log_directory, "embeddings") + os.makedirs(embedding_dir, exist_ok=True) + else: + embedding_dir = None + + if create_image_every > 0: + images_dir = os.path.join(log_directory, "images") + os.makedirs(images_dir, exist_ok=True) + else: + images_dir = None + + cond_model = shared.sd_model.cond_stage_model + + shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." + with torch.autocast("cuda"): + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, size=512, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) + + hijack = sd_hijack.model_hijack + + embedding = hijack.embedding_db.word_embeddings[embedding_name] + embedding.vec.requires_grad = True + + optimizer = torch.optim.AdamW([embedding.vec], lr=learn_rate) + + losses = torch.zeros((32,)) + + last_saved_file = "" + last_saved_image = "" + + ititial_step = embedding.step or 0 + if ititial_step > steps: + return embedding, filename + + pbar = tqdm.tqdm(enumerate(ds), total=steps-ititial_step) + for i, (x, text) in pbar: + embedding.step = i + ititial_step + + if embedding.step > steps: + break + + if shared.state.interrupted: + break + + with torch.autocast("cuda"): + c = cond_model([text]) + loss = shared.sd_model(x.unsqueeze(0), c)[0] + + losses[embedding.step % losses.shape[0]] = loss.item() + + optimizer.zero_grad() + loss.backward() + optimizer.step() + + pbar.set_description(f"loss: {losses.mean():.7f}") + + if embedding.step > 0 and embedding_dir is not None and embedding.step % save_embedding_every == 0: + last_saved_file = os.path.join(embedding_dir, f'{embedding_name}-{embedding.step}.pt') + embedding.save(last_saved_file) + + if embedding.step > 0 and images_dir is not None and embedding.step % create_image_every == 0: + last_saved_image = os.path.join(images_dir, f'{embedding_name}-{embedding.step}.png') + + p = processing.StableDiffusionProcessingTxt2Img( + sd_model=shared.sd_model, + prompt=text, + steps=20, + do_not_save_grid=True, + do_not_save_samples=True, + ) + + processed = processing.process_images(p) + image = processed.images[0] + + shared.state.current_image = image + image.save(last_saved_image) + + last_saved_image += f", prompt: {text}" + + shared.state.job_no = embedding.step + + shared.state.textinfo = f""" +

+Loss: {losses.mean():.7f}
+Step: {embedding.step}
+Last prompt: {html.escape(text)}
+Last saved embedding: {html.escape(last_saved_file)}
+Last saved image: {html.escape(last_saved_image)}
+

+""" + + embedding.cached_checksum = None + embedding.save(filename) + + return embedding, filename + diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py new file mode 100644 index 000000000..ce3677a98 --- /dev/null +++ b/modules/textual_inversion/ui.py @@ -0,0 +1,32 @@ +import html + +import gradio as gr + +import modules.textual_inversion.textual_inversion as ti +from modules import sd_hijack, shared + + +def create_embedding(name, nvpt): + filename = ti.create_embedding(name, nvpt) + + sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() + + return gr.Dropdown.update(choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())), f"Created: {filename}", "" + + +def train_embedding(*args): + + try: + sd_hijack.undo_optimizations() + + embedding, filename = ti.train_embedding(*args) + + res = f""" +Training {'interrupted' if shared.state.interrupted else 'finished'} after {embedding.step} steps. +Embedding saved to {html.escape(filename)} +""" + return res, "" + except Exception: + raise + finally: + sd_hijack.apply_optimizations() diff --git a/modules/ui.py b/modules/ui.py index 15572bb0a..57aef6ff1 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -21,6 +21,7 @@ import gradio as gr import gradio.utils import gradio.routes +from modules import sd_hijack from modules.paths import script_path from modules.shared import opts, cmd_opts import modules.shared as shared @@ -32,6 +33,7 @@ import modules.gfpgan_model import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste +import modules.textual_inversion.ui # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI mimetypes.init() @@ -142,8 +144,8 @@ def save_files(js_data, images, index): return '', '', plaintext_to_html(f"Saved: {filenames[0]}") -def wrap_gradio_call(func): - def f(*args, **kwargs): +def wrap_gradio_call(func, extra_outputs=None): + def f(*args, extra_outputs_array=extra_outputs, **kwargs): run_memmon = opts.memmon_poll_rate > 0 and not shared.mem_mon.disabled if run_memmon: shared.mem_mon.monitor() @@ -159,7 +161,10 @@ def wrap_gradio_call(func): shared.state.job = "" shared.state.job_count = 0 - res = [None, '', f"
{plaintext_to_html(type(e).__name__+': '+str(e))}
"] + if extra_outputs_array is None: + extra_outputs_array = [None, ''] + + res = extra_outputs_array + [f"
{plaintext_to_html(type(e).__name__+': '+str(e))}
"] elapsed = time.perf_counter() - t @@ -179,6 +184,7 @@ def wrap_gradio_call(func): res[-1] += f"

Time taken: {elapsed:.2f}s

{vram_html}
" shared.state.interrupted = False + shared.state.job_count = 0 return tuple(res) @@ -187,7 +193,7 @@ def wrap_gradio_call(func): def check_progress_call(id_part): if shared.state.job_count == 0: - return "", gr_show(False), gr_show(False) + return "", gr_show(False), gr_show(False), gr_show(False) progress = 0 @@ -219,13 +225,19 @@ def check_progress_call(id_part): else: preview_visibility = gr_show(True) - return f"

{progressbar}

", preview_visibility, image + if shared.state.textinfo is not None: + textinfo_result = gr.HTML.update(value=shared.state.textinfo, visible=True) + else: + textinfo_result = gr_show(False) + + return f"

{progressbar}

", preview_visibility, image, textinfo_result def check_progress_call_initial(id_part): shared.state.job_count = -1 shared.state.current_latent = None shared.state.current_image = None + shared.state.textinfo = None return check_progress_call(id_part) @@ -399,13 +411,16 @@ def create_toprow(is_img2img): return prompt, roll, prompt_style, negative_prompt, prompt_style2, submit, interrogate, prompt_style_apply, save_style, paste -def setup_progressbar(progressbar, preview, id_part): +def setup_progressbar(progressbar, preview, id_part, textinfo=None): + if textinfo is None: + textinfo = gr.HTML(visible=False) + check_progress = gr.Button('Check progress', elem_id=f"{id_part}_check_progress", visible=False) check_progress.click( fn=lambda: check_progress_call(id_part), show_progress=False, inputs=[], - outputs=[progressbar, preview, preview], + outputs=[progressbar, preview, preview, textinfo], ) check_progress_initial = gr.Button('Check progress (first)', elem_id=f"{id_part}_check_progress_initial", visible=False) @@ -413,11 +428,14 @@ def setup_progressbar(progressbar, preview, id_part): fn=lambda: check_progress_call_initial(id_part), show_progress=False, inputs=[], - outputs=[progressbar, preview, preview], + outputs=[progressbar, preview, preview, textinfo], ) -def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): +def create_ui(wrap_gradio_gpu_call): + import modules.img2img + import modules.txt2img + with gr.Blocks(analytics_enabled=False) as txt2img_interface: txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, txt2img_prompt_style_apply, txt2img_save_style, paste = create_toprow(is_img2img=False) dummy_component = gr.Label(visible=False) @@ -483,7 +501,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) txt2img_args = dict( - fn=txt2img, + fn=wrap_gradio_gpu_call(modules.txt2img.txt2img), _js="submit", inputs=[ txt2img_prompt, @@ -675,7 +693,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): ) img2img_args = dict( - fn=img2img, + fn=wrap_gradio_gpu_call(modules.img2img.img2img), _js="submit_img2img", inputs=[ dummy_component, @@ -828,7 +846,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): open_extras_folder = gr.Button('Open output directory', elem_id=button_id) submit.click( - fn=run_extras, + fn=wrap_gradio_gpu_call(modules.extras.run_extras), _js="get_extras_tab_index", inputs=[ dummy_component, @@ -878,7 +896,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): pnginfo_send_to_img2img = gr.Button('Send to img2img') image.change( - fn=wrap_gradio_call(run_pnginfo), + fn=wrap_gradio_call(modules.extras.run_pnginfo), inputs=[image], outputs=[html, generation_info, html2], ) @@ -887,7 +905,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): with gr.Row().style(equal_height=False): with gr.Column(variant='panel'): gr.HTML(value="

A merger of the two checkpoints will be generated in your checkpoint directory.

") - + with gr.Row(): primary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_primary_model_name", label="Primary Model Name") secondary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_secondary_model_name", label="Secondary Model Name") @@ -896,10 +914,96 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): interp_method = gr.Radio(choices=["Weighted Sum", "Sigmoid", "Inverse Sigmoid"], value="Weighted Sum", label="Interpolation Method") save_as_half = gr.Checkbox(value=False, label="Safe as float16") modelmerger_merge = gr.Button(elem_id="modelmerger_merge", label="Merge", variant='primary') - + with gr.Column(variant='panel'): submit_result = gr.Textbox(elem_id="modelmerger_result", show_label=False) + sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() + + with gr.Blocks() as textual_inversion_interface: + with gr.Row().style(equal_height=False): + with gr.Column(): + with gr.Group(): + gr.HTML(value="

Create a new embedding

") + + new_embedding_name = gr.Textbox(label="Name") + nvpt = gr.Slider(label="Number of vectors per token", minimum=1, maximum=75, step=1, value=1) + + with gr.Row(): + with gr.Column(scale=3): + gr.HTML(value="") + + with gr.Column(): + create_embedding = gr.Button(value="Create", variant='primary') + + with gr.Group(): + gr.HTML(value="

Train an embedding; must specify a directory with a set of 512x512 images

") + train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) + learn_rate = gr.Number(label='Learning rate', value=5.0e-03) + dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images") + log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") + template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) + steps = gr.Number(label='Max steps', value=100000, precision=0) + create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=1000, precision=0) + save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=1000, precision=0) + + with gr.Row(): + with gr.Column(scale=2): + gr.HTML(value="") + + with gr.Column(): + with gr.Row(): + interrupt_training = gr.Button(value="Interrupt") + train_embedding = gr.Button(value="Train", variant='primary') + + with gr.Column(): + progressbar = gr.HTML(elem_id="ti_progressbar") + ti_output = gr.Text(elem_id="ti_output", value="", show_label=False) + + ti_gallery = gr.Gallery(label='Output', show_label=False, elem_id='ti_gallery').style(grid=4) + ti_preview = gr.Image(elem_id='ti_preview', visible=False) + ti_progress = gr.HTML(elem_id="ti_progress", value="") + ti_outcome = gr.HTML(elem_id="ti_error", value="") + setup_progressbar(progressbar, ti_preview, 'ti', textinfo=ti_progress) + + create_embedding.click( + fn=modules.textual_inversion.ui.create_embedding, + inputs=[ + new_embedding_name, + nvpt, + ], + outputs=[ + train_embedding_name, + ti_output, + ti_outcome, + ] + ) + + train_embedding.click( + fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.train_embedding, extra_outputs=[gr.update()]), + _js="start_training_textual_inversion", + inputs=[ + train_embedding_name, + learn_rate, + dataset_directory, + log_directory, + steps, + create_image_every, + save_embedding_every, + template_file, + ], + outputs=[ + ti_output, + ti_outcome, + ] + ) + + interrupt_training.click( + fn=lambda: shared.state.interrupt(), + inputs=[], + outputs=[], + ) + def create_setting_component(key): def fun(): return opts.data[key] if key in opts.data else opts.data_labels[key].default @@ -1011,6 +1115,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): (extras_interface, "Extras", "extras"), (pnginfo_interface, "PNG Info", "pnginfo"), (modelmerger_interface, "Checkpoint Merger", "modelmerger"), + (textual_inversion_interface, "Textual inversion", "ti"), (settings_interface, "Settings", "settings"), ] @@ -1044,11 +1149,11 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): def modelmerger(*args): try: - results = run_modelmerger(*args) + results = modules.extras.run_modelmerger(*args) except Exception as e: print("Error loading/saving model file:", file=sys.stderr) print(traceback.format_exc(), file=sys.stderr) - modules.sd_models.list_models() #To remove the potentially missing models from the list + modules.sd_models.list_models() # to remove the potentially missing models from the list return ["Error loading/saving model file. It doesn't exist or the name contains illegal characters"] + [gr.Dropdown.update(choices=modules.sd_models.checkpoint_tiles()) for _ in range(3)] return results diff --git a/style.css b/style.css index 79d6bb0dc..39586bf18 100644 --- a/style.css +++ b/style.css @@ -157,7 +157,7 @@ button{ max-width: 10em; } -#txt2img_preview, #img2img_preview{ +#txt2img_preview, #img2img_preview, #ti_preview{ position: absolute; width: 320px; left: 0; @@ -172,18 +172,18 @@ button{ } @media screen and (min-width: 768px) { - #txt2img_preview, #img2img_preview { + #txt2img_preview, #img2img_preview, #ti_preview { position: absolute; } } @media screen and (max-width: 767px) { - #txt2img_preview, #img2img_preview { + #txt2img_preview, #img2img_preview, #ti_preview { position: relative; } } -#txt2img_preview div.left-0.top-0, #img2img_preview div.left-0.top-0{ +#txt2img_preview div.left-0.top-0, #img2img_preview div.left-0.top-0, #ti_preview div.left-0.top-0{ display: none; } @@ -247,7 +247,7 @@ input[type="range"]{ #txt2img_negative_prompt, #img2img_negative_prompt{ } -#txt2img_progressbar, #img2img_progressbar{ +#txt2img_progressbar, #img2img_progressbar, #ti_progressbar{ position: absolute; z-index: 1000; right: 0; diff --git a/textual_inversion_templates/style.txt b/textual_inversion_templates/style.txt new file mode 100644 index 000000000..15af2d6b8 --- /dev/null +++ b/textual_inversion_templates/style.txt @@ -0,0 +1,19 @@ +a painting, art by [name] +a rendering, art by [name] +a cropped painting, art by [name] +the painting, art by [name] +a clean painting, art by [name] +a dirty painting, art by [name] +a dark painting, art by [name] +a picture, art by [name] +a cool painting, art by [name] +a close-up painting, art by [name] +a bright painting, art by [name] +a cropped painting, art by [name] +a good painting, art by [name] +a close-up painting, art by [name] +a rendition, art by [name] +a nice painting, art by [name] +a small painting, art by [name] +a weird painting, art by [name] +a large painting, art by [name] diff --git a/textual_inversion_templates/style_filewords.txt b/textual_inversion_templates/style_filewords.txt new file mode 100644 index 000000000..b3a8159a8 --- /dev/null +++ b/textual_inversion_templates/style_filewords.txt @@ -0,0 +1,19 @@ +a painting of [filewords], art by [name] +a rendering of [filewords], art by [name] +a cropped painting of [filewords], art by [name] +the painting of [filewords], art by [name] +a clean painting of [filewords], art by [name] +a dirty painting of [filewords], art by [name] +a dark painting of [filewords], art by [name] +a picture of [filewords], art by [name] +a cool painting of [filewords], art by [name] +a close-up painting of [filewords], art by [name] +a bright painting of [filewords], art by [name] +a cropped painting of [filewords], art by [name] +a good painting of [filewords], art by [name] +a close-up painting of [filewords], art by [name] +a rendition of [filewords], art by [name] +a nice painting of [filewords], art by [name] +a small painting of [filewords], art by [name] +a weird painting of [filewords], art by [name] +a large painting of [filewords], art by [name] diff --git a/textual_inversion_templates/subject.txt b/textual_inversion_templates/subject.txt new file mode 100644 index 000000000..79f36aa05 --- /dev/null +++ b/textual_inversion_templates/subject.txt @@ -0,0 +1,27 @@ +a photo of a [name] +a rendering of a [name] +a cropped photo of the [name] +the photo of a [name] +a photo of a clean [name] +a photo of a dirty [name] +a dark photo of the [name] +a photo of my [name] +a photo of the cool [name] +a close-up photo of a [name] +a bright photo of the [name] +a cropped photo of a [name] +a photo of the [name] +a good photo of the [name] +a photo of one [name] +a close-up photo of the [name] +a rendition of the [name] +a photo of the clean [name] +a rendition of a [name] +a photo of a nice [name] +a good photo of a [name] +a photo of the nice [name] +a photo of the small [name] +a photo of the weird [name] +a photo of the large [name] +a photo of a cool [name] +a photo of a small [name] diff --git a/textual_inversion_templates/subject_filewords.txt b/textual_inversion_templates/subject_filewords.txt new file mode 100644 index 000000000..008652a6b --- /dev/null +++ b/textual_inversion_templates/subject_filewords.txt @@ -0,0 +1,27 @@ +a photo of a [name], [filewords] +a rendering of a [name], [filewords] +a cropped photo of the [name], [filewords] +the photo of a [name], [filewords] +a photo of a clean [name], [filewords] +a photo of a dirty [name], [filewords] +a dark photo of the [name], [filewords] +a photo of my [name], [filewords] +a photo of the cool [name], [filewords] +a close-up photo of a [name], [filewords] +a bright photo of the [name], [filewords] +a cropped photo of a [name], [filewords] +a photo of the [name], [filewords] +a good photo of the [name], [filewords] +a photo of one [name], [filewords] +a close-up photo of the [name], [filewords] +a rendition of the [name], [filewords] +a photo of the clean [name], [filewords] +a rendition of a [name], [filewords] +a photo of a nice [name], [filewords] +a good photo of a [name], [filewords] +a photo of the nice [name], [filewords] +a photo of the small [name], [filewords] +a photo of the weird [name], [filewords] +a photo of the large [name], [filewords] +a photo of a cool [name], [filewords] +a photo of a small [name], [filewords] diff --git a/webui.py b/webui.py index b8cccd546..19fdcdd4d 100644 --- a/webui.py +++ b/webui.py @@ -12,7 +12,6 @@ import modules.bsrgan_model as bsrgan import modules.extras import modules.face_restoration import modules.gfpgan_model as gfpgan -import modules.img2img import modules.ldsr_model as ldsr import modules.lowvram import modules.realesrgan_model as realesrgan @@ -21,7 +20,6 @@ import modules.sd_hijack import modules.sd_models import modules.shared as shared import modules.swinir_model as swinir -import modules.txt2img import modules.ui from modules import modelloader from modules.paths import script_path @@ -46,7 +44,7 @@ def wrap_queued_call(func): return f -def wrap_gradio_gpu_call(func): +def wrap_gradio_gpu_call(func, extra_outputs=None): def f(*args, **kwargs): devices.torch_gc() @@ -58,6 +56,7 @@ def wrap_gradio_gpu_call(func): shared.state.current_image = None shared.state.current_image_sampling_step = 0 shared.state.interrupted = False + shared.state.textinfo = None with queue_lock: res = func(*args, **kwargs) @@ -69,7 +68,7 @@ def wrap_gradio_gpu_call(func): return res - return modules.ui.wrap_gradio_call(f) + return modules.ui.wrap_gradio_call(f, extra_outputs=extra_outputs) modules.scripts.load_scripts(os.path.join(script_path, "scripts")) @@ -86,13 +85,7 @@ def webui(): signal.signal(signal.SIGINT, sigint_handler) - demo = modules.ui.create_ui( - txt2img=wrap_gradio_gpu_call(modules.txt2img.txt2img), - img2img=wrap_gradio_gpu_call(modules.img2img.img2img), - run_extras=wrap_gradio_gpu_call(modules.extras.run_extras), - run_pnginfo=modules.extras.run_pnginfo, - run_modelmerger=modules.extras.run_modelmerger - ) + demo = modules.ui.create_ui(wrap_gradio_gpu_call=wrap_gradio_gpu_call) demo.launch( share=cmd_opts.share, From 0114057ad672a581bd0b598870b58b674b1a3624 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 15:49:42 +0300 Subject: [PATCH 056/138] fix incorrect use of glob in modelloader for #1410 --- modules/modelloader.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/modelloader.py b/modules/modelloader.py index 8c862b42f..015aeafa3 100644 --- a/modules/modelloader.py +++ b/modules/modelloader.py @@ -43,7 +43,7 @@ def load_models(model_path: str, model_url: str = None, command_path: str = None for place in places: if os.path.exists(place): for file in glob.iglob(place + '**/**', recursive=True): - full_path = os.path.join(place, file) + full_path = file if os.path.isdir(full_path): continue if len(ext_filter) != 0: From 4e72a1aab6d1b3a8d8c09fadc81843a07c05cc18 Mon Sep 17 00:00:00 2001 From: ClashSAN <98228077+ClashSAN@users.noreply.github.com> Date: Sat, 1 Oct 2022 00:15:43 +0000 Subject: [PATCH 057/138] Grammar Fix --- README.md | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/README.md b/README.md index 5ded94f98..15e224e8f 100644 --- a/README.md +++ b/README.md @@ -11,12 +11,12 @@ Check the [custom scripts](https://github.com/AUTOMATIC1111/stable-diffusion-web - One click install and run script (but you still must install python and git) - Outpainting - Inpainting -- Prompt -- Stable Diffusion upscale +- Prompt Matrix +- Stable Diffusion Upscale - Attention, specify parts of text that the model should pay more attention to - - a man in a ((txuedo)) - will pay more attentinoto tuxedo - - a man in a (txuedo:1.21) - alternative syntax -- Loopback, run img2img procvessing multiple times + - a man in a ((tuxedo)) - will pay more attention to tuxedo + - a man in a (tuxedo:1.21) - alternative syntax +- Loopback, run img2img processing multiple times - X/Y plot, a way to draw a 2 dimensional plot of images with different parameters - Textual Inversion - have as many embeddings as you want and use any names you like for them @@ -35,15 +35,15 @@ Check the [custom scripts](https://github.com/AUTOMATIC1111/stable-diffusion-web - 4GB video card support (also reports of 2GB working) - Correct seeds for batches - Prompt length validation - - get length of prompt in tokensas you type - - get a warning after geenration if some text was truncated + - get length of prompt in tokens as you type + - get a warning after generation if some text was truncated - Generation parameters - parameters you used to generate images are saved with that image - in PNG chunks for PNG, in EXIF for JPEG - can drag the image to PNG info tab to restore generation parameters and automatically copy them into UI - can be disabled in settings - Settings page -- Running arbitrary python code from UI (must run with commandline flag to enable) +- Running arbitrary python code from UI (must run with --allow-code to enable) - Mouseover hints for most UI elements - Possible to change defaults/mix/max/step values for UI elements via text config - Random artist button From 0758f6e641b5790ce566a998d43e0ea74a627766 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 17:24:50 +0300 Subject: [PATCH 058/138] fix --ckpt option breaking model selection --- modules/sd_models.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/sd_models.py b/modules/sd_models.py index 5b3dbdc79..9259d69e7 100644 --- a/modules/sd_models.py +++ b/modules/sd_models.py @@ -69,7 +69,7 @@ def list_models(): h = model_hash(cmd_ckpt) title, short_model_name = modeltitle(cmd_ckpt, h) checkpoints_list[title] = CheckpointInfo(cmd_ckpt, title, h, short_model_name) - shared.opts.sd_model_checkpoint = title + shared.opts.data['sd_model_checkpoint'] = title elif cmd_ckpt is not None and cmd_ckpt != shared.default_sd_model_file: print(f"Checkpoint in --ckpt argument not found (Possible it was moved to {model_path}: {cmd_ckpt}", file=sys.stderr) for filename in model_list: From 53a3dc601fb734ce433505b1ca68770919106bad Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 18:21:56 +0300 Subject: [PATCH 059/138] move CLIP out of requirements and into launcher to make it possible to launch the program offline --- launch.py | 4 ++++ requirements.txt | 2 -- requirements_versions.txt | 1 - 3 files changed, 4 insertions(+), 3 deletions(-) diff --git a/launch.py b/launch.py index d2793ed20..57405feab 100644 --- a/launch.py +++ b/launch.py @@ -15,6 +15,7 @@ requirements_file = os.environ.get('REQS_FILE', "requirements_versions.txt") commandline_args = os.environ.get('COMMANDLINE_ARGS', "") gfpgan_package = os.environ.get('GFPGAN_PACKAGE', "git+https://github.com/TencentARC/GFPGAN.git@8d2447a2d918f8eba5a4a01463fd48e45126a379") +clip_package = os.environ.get('CLIP_PACKAGE', "git+https://github.com/openai/CLIP.git@d50d76daa670286dd6cacf3bcd80b5e4823fc8e1") stable_diffusion_commit_hash = os.environ.get('STABLE_DIFFUSION_COMMIT_HASH', "69ae4b35e0a0f6ee1af8bb9a5d0016ccb27e36dc") taming_transformers_commit_hash = os.environ.get('TAMING_TRANSFORMERS_COMMIT_HASH', "24268930bf1dce879235a7fddd0b2355b84d7ea6") @@ -111,6 +112,9 @@ if not skip_torch_cuda_test: if not is_installed("gfpgan"): run_pip(f"install {gfpgan_package}", "gfpgan") +if not is_installed("clip"): + run_pip(f"install {clip_package}", "clip") + os.makedirs(dir_repos, exist_ok=True) git_clone("https://github.com/CompVis/stable-diffusion.git", repo_dir('stable-diffusion'), "Stable Diffusion", stable_diffusion_commit_hash) diff --git a/requirements.txt b/requirements.txt index 7cb9d3293..d4b337fce 100644 --- a/requirements.txt +++ b/requirements.txt @@ -13,14 +13,12 @@ Pillow pytorch_lightning realesrgan scikit-image>=0.19 -git+https://github.com/TencentARC/GFPGAN.git@8d2447a2d918f8eba5a4a01463fd48e45126a379 timm==0.4.12 transformers==4.19.2 torch einops jsonmerge clean-fid -git+https://github.com/openai/CLIP@d50d76daa670286dd6cacf3bcd80b5e4823fc8e1 resize-right torchdiffeq kornia diff --git a/requirements_versions.txt b/requirements_versions.txt index 1e8006e05..8a9acf205 100644 --- a/requirements_versions.txt +++ b/requirements_versions.txt @@ -18,7 +18,6 @@ piexif==1.1.3 einops==0.4.1 jsonmerge==1.8.0 clean-fid==0.1.29 -git+https://github.com/openai/CLIP@d50d76daa670286dd6cacf3bcd80b5e4823fc8e1 resize-right==0.0.2 torchdiffeq==0.2.3 kornia==0.6.7 From 88ec0cf5571883d84abd09196652b3679e359f2e Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 19:40:51 +0300 Subject: [PATCH 060/138] fix for incorrect embedding token length calculation (will break seeds that use embeddings, you're welcome!) add option to input initialization text for embeddings --- modules/sd_hijack.py | 8 ++++---- modules/textual_inversion/textual_inversion.py | 13 +++++-------- modules/textual_inversion/ui.py | 4 ++-- modules/ui.py | 2 ++ 4 files changed, 13 insertions(+), 14 deletions(-) diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index fd57e5c54..3fa062422 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -130,7 +130,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) + embedding, embedding_length_in_tokens = self.hijack.embedding_db.find_embedding_at_position(tokens, i) if embedding is None: remade_tokens.append(token) @@ -142,7 +142,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): remade_tokens += [0] * emb_len multipliers += [weight] * emb_len used_custom_terms.append((embedding.name, embedding.checksum())) - i += emb_len + i += embedding_length_in_tokens if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} @@ -213,7 +213,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) + embedding, embedding_length_in_tokens = self.hijack.embedding_db.find_embedding_at_position(tokens, i) mult_change = self.token_mults.get(token) if opts.enable_emphasis else None if mult_change is not None: @@ -229,7 +229,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): remade_tokens += [0] * emb_len multipliers += [mult] * emb_len used_custom_terms.append((embedding.name, embedding.checksum())) - i += emb_len + i += embedding_length_in_tokens if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index c0baaace2..0c50161db 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -117,24 +117,21 @@ class EmbeddingDatabase: possible_matches = self.ids_lookup.get(token, None) if possible_matches is None: - return None + return None, None for ids, embedding in possible_matches: if tokens[offset:offset + len(ids)] == ids: - return embedding + return embedding, len(ids) - return None + return None, None - -def create_embedding(name, num_vectors_per_token): - init_text = '*' - +def create_embedding(name, num_vectors_per_token, init_text='*'): cond_model = shared.sd_model.cond_stage_model embedding_layer = cond_model.wrapped.transformer.text_model.embeddings ids = cond_model.tokenizer(init_text, max_length=num_vectors_per_token, return_tensors="pt", add_special_tokens=False)["input_ids"] - embedded = embedding_layer(ids.to(devices.device)).squeeze(0) + embedded = embedding_layer.token_embedding.wrapped(ids.to(devices.device)).squeeze(0) vec = torch.zeros((num_vectors_per_token, embedded.shape[1]), device=devices.device) for i in range(num_vectors_per_token): diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py index ce3677a98..66c43ffbe 100644 --- a/modules/textual_inversion/ui.py +++ b/modules/textual_inversion/ui.py @@ -6,8 +6,8 @@ import modules.textual_inversion.textual_inversion as ti from modules import sd_hijack, shared -def create_embedding(name, nvpt): - filename = ti.create_embedding(name, nvpt) +def create_embedding(name, initialization_text, nvpt): + filename = ti.create_embedding(name, nvpt, init_text=initialization_text) sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() diff --git a/modules/ui.py b/modules/ui.py index 3b81a4f74..eca50df0f 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -954,6 +954,7 @@ def create_ui(wrap_gradio_gpu_call): gr.HTML(value="

Create a new embedding

") new_embedding_name = gr.Textbox(label="Name") + initialization_text = gr.Textbox(label="Initialization text", value="*") nvpt = gr.Slider(label="Number of vectors per token", minimum=1, maximum=75, step=1, value=1) with gr.Row(): @@ -997,6 +998,7 @@ def create_ui(wrap_gradio_gpu_call): fn=modules.textual_inversion.ui.create_embedding, inputs=[ new_embedding_name, + initialization_text, nvpt, ], outputs=[ From 71fe7fa49f5eb1a2c89932a9d217ed153c12fc8b Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 19:56:37 +0300 Subject: [PATCH 061/138] fix using aaaa-100 embedding when the prompt has aaaa-10000 and you have both aaaa-100 and aaaa-10000 in the directory with embeddings. --- modules/textual_inversion/textual_inversion.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 0c50161db..9d2241cef 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -57,7 +57,8 @@ class EmbeddingDatabase: first_id = ids[0] if first_id not in self.ids_lookup: self.ids_lookup[first_id] = [] - self.ids_lookup[first_id].append((ids, embedding)) + + self.ids_lookup[first_id] = sorted(self.ids_lookup[first_id] + [(ids, embedding)], key=lambda x: len(x[0]), reverse=True) return embedding From 4ec4af6e0b7addeee5221a03f32d117ccdc875d9 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 20:15:25 +0300 Subject: [PATCH 062/138] add checkpoint info to saved embeddings --- modules/textual_inversion/textual_inversion.py | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 9d2241cef..1183aab76 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -7,7 +7,7 @@ import tqdm import html import datetime -from modules import shared, devices, sd_hijack, processing +from modules import shared, devices, sd_hijack, processing, sd_models import modules.textual_inversion.dataset @@ -17,6 +17,8 @@ class Embedding: self.name = name self.step = step self.cached_checksum = None + self.sd_checkpoint = None + self.sd_checkpoint_name = None def save(self, filename): embedding_data = { @@ -24,6 +26,8 @@ class Embedding: "string_to_param": {"*": self.vec}, "name": self.name, "step": self.step, + "sd_checkpoint": self.sd_checkpoint, + "sd_checkpoint_name": self.sd_checkpoint_name, } torch.save(embedding_data, filename) @@ -41,6 +45,7 @@ class Embedding: self.cached_checksum = f'{const_hash(self.vec.reshape(-1) * 100) & 0xffff:04x}' return self.cached_checksum + class EmbeddingDatabase: def __init__(self, embeddings_dir): self.ids_lookup = {} @@ -96,6 +101,8 @@ class EmbeddingDatabase: vec = emb.detach().to(devices.device, dtype=torch.float32) embedding = Embedding(vec, name) embedding.step = data.get('step', None) + embedding.sd_checkpoint = data.get('hash', None) + embedding.sd_checkpoint_name = data.get('sd_checkpoint_name', None) self.register_embedding(embedding, shared.sd_model) for fn in os.listdir(self.embeddings_dir): @@ -249,6 +256,10 @@ Last saved image: {html.escape(last_saved_image)}

""" + checkpoint = sd_models.select_checkpoint() + + embedding.sd_checkpoint = checkpoint.hash + embedding.sd_checkpoint_name = checkpoint.model_name embedding.cached_checksum = None embedding.save(filename) From 3ff0de2c594b786ef948a89efb1814c59bb42117 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 20:23:40 +0300 Subject: [PATCH 063/138] added --disable-console-progressbars to disable progressbars in console disabled printing prompts to console by default, enabled by --enable-console-prompts --- modules/img2img.py | 4 +++- modules/sd_samplers.py | 8 ++++++-- modules/shared.py | 7 +++++-- modules/txt2img.py | 4 +++- 4 files changed, 17 insertions(+), 6 deletions(-) diff --git a/modules/img2img.py b/modules/img2img.py index 03e934e96..f4455c90f 100644 --- a/modules/img2img.py +++ b/modules/img2img.py @@ -103,7 +103,9 @@ def img2img(mode: int, prompt: str, negative_prompt: str, prompt_style: str, pro inpaint_full_res_padding=inpaint_full_res_padding, inpainting_mask_invert=inpainting_mask_invert, ) - print(f"\nimg2img: {prompt}", file=shared.progress_print_out) + + if shared.cmd_opts.enable_console_prompts: + print(f"\nimg2img: {prompt}", file=shared.progress_print_out) p.extra_generation_params["Mask blur"] = mask_blur diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index 925222148..9316875ab 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -77,7 +77,9 @@ def extended_tdqm(sequence, *args, desc=None, **kwargs): state.sampling_steps = len(sequence) state.sampling_step = 0 - for x in tqdm.tqdm(sequence, *args, desc=state.job, file=shared.progress_print_out, **kwargs): + seq = sequence if cmd_opts.disable_console_progressbars else tqdm.tqdm(sequence, *args, desc=state.job, file=shared.progress_print_out, **kwargs) + + for x in seq: if state.interrupted: break @@ -207,7 +209,9 @@ def extended_trange(sampler, count, *args, **kwargs): state.sampling_steps = count state.sampling_step = 0 - for x in tqdm.trange(count, *args, desc=state.job, file=shared.progress_print_out, **kwargs): + seq = range(count) if cmd_opts.disable_console_progressbars else tqdm.trange(count, *args, desc=state.job, file=shared.progress_print_out, **kwargs) + + for x in seq: if state.interrupted: break diff --git a/modules/shared.py b/modules/shared.py index 5a591dc99..1bf7a6c14 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -58,6 +58,9 @@ parser.add_argument("--opt-channelslast", action='store_true', help="change memo parser.add_argument("--styles-file", type=str, help="filename to use for styles", default=os.path.join(script_path, 'styles.csv')) parser.add_argument("--autolaunch", action='store_true', help="open the webui URL in the system's default browser upon launch", default=False) parser.add_argument("--use-textbox-seed", action='store_true', help="use textbox for seeds in UI (no up/down, but possible to input long seeds)", default=False) +parser.add_argument("--disable-console-progressbars", action='store_true', help="do not output progressbars to console", default=False) +parser.add_argument("--enable-console-prompts", action='store_true', help="print prompts to console when generating with txt2img and img2img", default=False) + cmd_opts = parser.parse_args() device = get_optimal_device() @@ -320,14 +323,14 @@ class TotalTQDM: ) def update(self): - if not opts.multiple_tqdm: + if not opts.multiple_tqdm or cmd_opts.disable_console_progressbars: return if self._tqdm is None: self.reset() self._tqdm.update() def updateTotal(self, new_total): - if not opts.multiple_tqdm: + if not opts.multiple_tqdm or cmd_opts.disable_console_progressbars: return if self._tqdm is None: self.reset() diff --git a/modules/txt2img.py b/modules/txt2img.py index 5368e4d00..d4406c3c0 100644 --- a/modules/txt2img.py +++ b/modules/txt2img.py @@ -34,7 +34,9 @@ def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: denoising_strength=denoising_strength if enable_hr else None, ) - print(f"\ntxt2img: {prompt}", file=shared.progress_print_out) + if cmd_opts.enable_console_prompts: + print(f"\ntxt2img: {prompt}", file=shared.progress_print_out) + processed = modules.scripts.scripts_txt2img.run(p, *args) if processed is None: From 6365a41f5981efa506dfe4e8fa878b43ca2d8d0c Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Sun, 2 Oct 2022 12:58:17 -0500 Subject: [PATCH 064/138] Update esrgan_model.py Use alternate ESRGAN Model download path. --- modules/esrgan_model.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/modules/esrgan_model.py b/modules/esrgan_model.py index ea91abfe8..4aed9283c 100644 --- a/modules/esrgan_model.py +++ b/modules/esrgan_model.py @@ -73,8 +73,8 @@ def fix_model_layers(crt_model, pretrained_net): class UpscalerESRGAN(Upscaler): def __init__(self, dirname): self.name = "ESRGAN" - self.model_url = "https://drive.google.com/u/0/uc?id=1TPrz5QKd8DHHt1k8SRtm6tMiPjz_Qene&export=download" - self.model_name = "ESRGAN 4x" + self.model_url = "https://github.com/cszn/KAIR/releases/download/v1.0/ESRGAN.pth" + self.model_name = "ESRGAN_4x" self.scalers = [] self.user_path = dirname self.model_path = os.path.join(models_path, self.name) From a1cde7e6468f80584030525a1b07cbf0f4ee42eb Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 21:09:10 +0300 Subject: [PATCH 065/138] disabled SD model download after multiple complaints --- modules/sd_models.py | 18 ++++++++---------- modules/textual_inversion/ui.py | 2 +- webui.py | 2 +- 3 files changed, 10 insertions(+), 12 deletions(-) diff --git a/modules/sd_models.py b/modules/sd_models.py index 9259d69e7..9a6b568f0 100644 --- a/modules/sd_models.py +++ b/modules/sd_models.py @@ -13,9 +13,6 @@ from modules.paths import models_path model_dir = "Stable-diffusion" model_path = os.path.abspath(os.path.join(models_path, model_dir)) -model_name = "sd-v1-4.ckpt" -model_url = "https://drive.yerf.org/wl/?id=EBfTrmcCCUAGaQBXVIj5lJmEhjoP1tgl&mode=grid&download=1" -user_dir = None CheckpointInfo = namedtuple("CheckpointInfo", ['filename', 'title', 'hash', 'model_name']) checkpoints_list = {} @@ -30,12 +27,10 @@ except Exception: pass -def setup_model(dirname): - global user_dir - user_dir = dirname +def setup_model(): if not os.path.exists(model_path): os.makedirs(model_path) - checkpoints_list.clear() + list_models() @@ -45,7 +40,7 @@ def checkpoint_tiles(): def list_models(): checkpoints_list.clear() - model_list = modelloader.load_models(model_path=model_path, model_url=model_url, command_path=user_dir, ext_filter=[".ckpt"], download_name=model_name) + model_list = modelloader.load_models(model_path=model_path, command_path=shared.cmd_opts.ckpt_dir, ext_filter=[".ckpt"]) def modeltitle(path, shorthash): abspath = os.path.abspath(path) @@ -106,8 +101,11 @@ def select_checkpoint(): if len(checkpoints_list) == 0: print(f"No checkpoints found. When searching for checkpoints, looked at:", file=sys.stderr) - print(f" - file {os.path.abspath(shared.cmd_opts.ckpt)}", file=sys.stderr) - print(f" - directory {os.path.abspath(shared.cmd_opts.ckpt_dir)}", file=sys.stderr) + if shared.cmd_opts.ckpt is not None: + print(f" - file {os.path.abspath(shared.cmd_opts.ckpt)}", file=sys.stderr) + print(f" - directory {model_path}", file=sys.stderr) + if shared.cmd_opts.ckpt_dir is not None: + print(f" - directory {os.path.abspath(shared.cmd_opts.ckpt_dir)}", file=sys.stderr) print(f"Can't run without a checkpoint. Find and place a .ckpt file into any of those locations. The program will exit.", file=sys.stderr) exit(1) diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py index 66c43ffbe..633037d8e 100644 --- a/modules/textual_inversion/ui.py +++ b/modules/textual_inversion/ui.py @@ -22,7 +22,7 @@ def train_embedding(*args): embedding, filename = ti.train_embedding(*args) res = f""" -Training {'interrupted' if shared.state.interrupted else 'finished'} after {embedding.step} steps. +Training {'interrupted' if shared.state.interrupted else 'finished'} at {embedding.step} steps. Embedding saved to {html.escape(filename)} """ return res, "" diff --git a/webui.py b/webui.py index 424ab9751..dc72ceb8a 100644 --- a/webui.py +++ b/webui.py @@ -23,7 +23,7 @@ from modules.paths import script_path from modules.shared import cmd_opts modelloader.cleanup_models() -modules.sd_models.setup_model(cmd_opts.ckpt_dir) +modules.sd_models.setup_model() codeformer.setup_model(cmd_opts.codeformer_models_path) gfpgan.setup_model(cmd_opts.gfpgan_models_path) shared.face_restorers.append(modules.face_restoration.FaceRestoration()) From 852fd90c0dcda9cb5fbbfdf0c7308ce58034935c Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 21:22:20 +0300 Subject: [PATCH 066/138] emergency fix for disabling SD model download after multiple complaints --- modules/sd_models.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/modules/sd_models.py b/modules/sd_models.py index 9a6b568f0..5f9920647 100644 --- a/modules/sd_models.py +++ b/modules/sd_models.py @@ -45,8 +45,8 @@ def list_models(): def modeltitle(path, shorthash): abspath = os.path.abspath(path) - if user_dir is not None and abspath.startswith(user_dir): - name = abspath.replace(user_dir, '') + if shared.cmd_opts.ckpt_dir is not None and abspath.startswith(shared.cmd_opts.ckpt_dir): + name = abspath.replace(shared.cmd_opts.ckpt_dir, '') elif abspath.startswith(model_path): name = abspath.replace(model_path, '') else: From e808096cf641d868f88465515d70d40fc46125d4 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 19:26:06 +0100 Subject: [PATCH 067/138] correct indent --- modules/scripts.py | 48 ++++++++++++++++++++++++---------------------- modules/ui.py | 23 +++++++++++----------- 2 files changed, 36 insertions(+), 35 deletions(-) diff --git a/modules/scripts.py b/modules/scripts.py index 788397f53..45230f9a1 100644 --- a/modules/scripts.py +++ b/modules/scripts.py @@ -163,37 +163,39 @@ class ScriptRunner: return processed def reload_sources(self): - for si,script in list(enumerate(self.scripts)): - with open(script.filename, "r", encoding="utf8") as file: - args_from = script.args_from - args_to = script.args_to - filename = script.filename - text = file.read() + for si, script in list(enumerate(self.scripts)): + with open(script.filename, "r", encoding="utf8") as file: + args_from = script.args_from + args_to = script.args_to + filename = script.filename + text = file.read() - from types import ModuleType - compiled = compile(text, filename, 'exec') - module = ModuleType(script.filename) - exec(compiled, module.__dict__) + from types import ModuleType - for key, script_class in module.__dict__.items(): - if type(script_class) == type and issubclass(script_class, Script): - self.scripts[si] = script_class() - self.scripts[si].filename = filename - self.scripts[si].args_from = args_from - self.scripts[si].args_to = args_to + compiled = compile(text, filename, 'exec') + module = ModuleType(script.filename) + exec(compiled, module.__dict__) + + for key, script_class in module.__dict__.items(): + if type(script_class) == type and issubclass(script_class, Script): + self.scripts[si] = script_class() + self.scripts[si].filename = filename + self.scripts[si].args_from = args_from + self.scripts[si].args_to = args_to scripts_txt2img = ScriptRunner() scripts_img2img = ScriptRunner() def reload_script_body_only(): - scripts_txt2img.reload_sources() - scripts_img2img.reload_sources() + scripts_txt2img.reload_sources() + scripts_img2img.reload_sources() + def reload_scripts(basedir): - global scripts_txt2img,scripts_img2img + global scripts_txt2img, scripts_img2img - scripts_data.clear() - load_scripts(basedir) + scripts_data.clear() + load_scripts(basedir) - scripts_txt2img = ScriptRunner() - scripts_img2img = ScriptRunner() + scripts_txt2img = ScriptRunner() + scripts_img2img = ScriptRunner() diff --git a/modules/ui.py b/modules/ui.py index 963a2c611..6b30f84ba 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1003,12 +1003,12 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): ) with gr.Row(): - reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') - restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') + reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') + restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') def reload_scripts(): - modules.scripts.reload_script_body_only() + modules.scripts.reload_script_body_only() reload_script_bodies.click( fn=reload_scripts, @@ -1018,7 +1018,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): ) def request_restart(): - settings_interface.gradio_ref.do_restart = True + settings_interface.gradio_ref.do_restart = True restart_gradio.click( fn=request_restart, @@ -1234,12 +1234,11 @@ for filename in sorted(os.listdir(jsdir)): if 'gradio_routes_templates_response' not in globals(): - def template_response(*args, **kwargs): - res = gradio_routes_templates_response(*args, **kwargs) - res.body = res.body.replace(b'', f'{javascript}'.encode("utf8")) - res.init_headers() - return res - - gradio_routes_templates_response = gradio.routes.templates.TemplateResponse - gradio.routes.templates.TemplateResponse = template_response + def template_response(*args, **kwargs): + res = gradio_routes_templates_response(*args, **kwargs) + res.body = res.body.replace(b'', f'{javascript}'.encode("utf8")) + res.init_headers() + return res + gradio_routes_templates_response = gradio.routes.templates.TemplateResponse + gradio.routes.templates.TemplateResponse = template_response From a634c3226fd69486ce96df56f95f3fd63172305c Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 19:26:38 +0100 Subject: [PATCH 068/138] correct indent --- webui.py | 56 ++++++++++++++++++++++++++++---------------------------- 1 file changed, 28 insertions(+), 28 deletions(-) diff --git a/webui.py b/webui.py index ab200045a..140040ca1 100644 --- a/webui.py +++ b/webui.py @@ -89,38 +89,38 @@ def webui(): while 1: - demo = modules.ui.create_ui( - txt2img=wrap_gradio_gpu_call(modules.txt2img.txt2img), - img2img=wrap_gradio_gpu_call(modules.img2img.img2img), - run_extras=wrap_gradio_gpu_call(modules.extras.run_extras), - run_pnginfo=modules.extras.run_pnginfo, - run_modelmerger=modules.extras.run_modelmerger - ) + demo = modules.ui.create_ui( + txt2img=wrap_gradio_gpu_call(modules.txt2img.txt2img), + img2img=wrap_gradio_gpu_call(modules.img2img.img2img), + run_extras=wrap_gradio_gpu_call(modules.extras.run_extras), + run_pnginfo=modules.extras.run_pnginfo, + run_modelmerger=modules.extras.run_modelmerger + ) - demo.launch( - share=cmd_opts.share, - server_name="0.0.0.0" if cmd_opts.listen else None, - server_port=cmd_opts.port, - debug=cmd_opts.gradio_debug, - auth=[tuple(cred.split(':')) for cred in cmd_opts.gradio_auth.strip('"').split(',')] if cmd_opts.gradio_auth else None, - inbrowser=cmd_opts.autolaunch, - prevent_thread_lock=True - ) + demo.launch( + share=cmd_opts.share, + server_name="0.0.0.0" if cmd_opts.listen else None, + server_port=cmd_opts.port, + debug=cmd_opts.gradio_debug, + auth=[tuple(cred.split(':')) for cred in cmd_opts.gradio_auth.strip('"').split(',')] if cmd_opts.gradio_auth else None, + inbrowser=cmd_opts.autolaunch, + prevent_thread_lock=True + ) - while 1: - time.sleep(0.5) - if getattr(demo,'do_restart',False): - time.sleep(0.5) - demo.close() - time.sleep(0.5) - break + while 1: + time.sleep(0.5) + if getattr(demo,'do_restart',False): + time.sleep(0.5) + demo.close() + time.sleep(0.5) + break - print('Reloading Custom Scripts') - modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) - print('Reloading modules: modules.ui') - importlib.reload(modules.ui) - print('Restarting Gradio') + print('Reloading Custom Scripts') + modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) + print('Reloading modules: modules.ui') + importlib.reload(modules.ui) + print('Restarting Gradio') if __name__ == "__main__": From c0389eb3071870240bc158263e5dfb4351ec8eba Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 21:35:29 +0300 Subject: [PATCH 069/138] hello --- webui.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/webui.py b/webui.py index 634956978..47848ba58 100644 --- a/webui.py +++ b/webui.py @@ -103,11 +103,11 @@ def webui(): while 1: time.sleep(0.5) - if getattr(demo,'do_restart',False): - time.sleep(0.5) - demo.close() - time.sleep(0.5) - break + if getattr(demo, 'do_restart', False): + time.sleep(0.5) + demo.close() + time.sleep(0.5) + break print('Reloading Custom Scripts') modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) From 2ef69df9a7c7b6793401f29ced71fb8a781fad4c Mon Sep 17 00:00:00 2001 From: Jocke Date: Sun, 2 Oct 2022 16:10:41 +0200 Subject: [PATCH 070/138] Prevent upscaling when None is selected for SD upscale --- scripts/sd_upscale.py | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/scripts/sd_upscale.py b/scripts/sd_upscale.py index 2653e2d40..cb37ff7e8 100644 --- a/scripts/sd_upscale.py +++ b/scripts/sd_upscale.py @@ -34,7 +34,11 @@ class Script(scripts.Script): seed = p.seed init_img = p.init_images[0] - img = upscaler.scaler.upscale(init_img, 2, upscaler.data_path) + + if(upscaler.name != "None"): + img = upscaler.scaler.upscale(init_img, 2, upscaler.data_path) + else: + img = init_img devices.torch_gc() From 91f327f22bb2feb780c424c74723cc0629dc34a1 Mon Sep 17 00:00:00 2001 From: Lopyter Date: Sun, 2 Oct 2022 18:15:31 +0200 Subject: [PATCH 071/138] make save to dirs optional for imgs saved from ui --- modules/shared.py | 1 + modules/ui.py | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/modules/shared.py b/modules/shared.py index 1bf7a6c14..785e7af6f 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -173,6 +173,7 @@ options_templates.update(options_section(('saving-to-dirs', "Saving to a directo "grid_save_to_dirs": OptionInfo(False, "Save grids to subdirectory"), "directories_filename_pattern": OptionInfo("", "Directory name pattern"), "directories_max_prompt_words": OptionInfo(8, "Max prompt words", gr.Slider, {"minimum": 1, "maximum": 20, "step": 1}), + "use_save_to_dirs_for_ui": OptionInfo(False, "Use \"Save images to a subdirectory\" option for images saved from UI"), })) options_templates.update(options_section(('upscaling', "Upscaling"), { diff --git a/modules/ui.py b/modules/ui.py index 78a15d83a..8912deff4 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -113,7 +113,7 @@ def save_files(js_data, images, index): p = MyObject(data) path = opts.outdir_save - save_to_dirs = opts.save_to_dirs + save_to_dirs = opts.use_save_to_dirs_for_ui if save_to_dirs: dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, p.seed, p.prompt) From c4445225f79f1c57afe52358ff4b205864eb7aac Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 21:50:14 +0300 Subject: [PATCH 072/138] change wording for options --- modules/shared.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/modules/shared.py b/modules/shared.py index 785e7af6f..7246eadc6 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -170,10 +170,10 @@ options_templates.update(options_section(('saving-paths', "Paths for saving"), { options_templates.update(options_section(('saving-to-dirs', "Saving to a directory"), { "save_to_dirs": OptionInfo(False, "Save images to a subdirectory"), - "grid_save_to_dirs": OptionInfo(False, "Save grids to subdirectory"), + "grid_save_to_dirs": OptionInfo(False, "Save grids to a subdirectory"), + "use_save_to_dirs_for_ui": OptionInfo(False, "When using \"Save\" button, save images to a subdirectory"), "directories_filename_pattern": OptionInfo("", "Directory name pattern"), - "directories_max_prompt_words": OptionInfo(8, "Max prompt words", gr.Slider, {"minimum": 1, "maximum": 20, "step": 1}), - "use_save_to_dirs_for_ui": OptionInfo(False, "Use \"Save images to a subdirectory\" option for images saved from UI"), + "directories_max_prompt_words": OptionInfo(8, "Max prompt words for [prompt_words] pattern", gr.Slider, {"minimum": 1, "maximum": 20, "step": 1}), })) options_templates.update(options_section(('upscaling', "Upscaling"), { From c7543d4940da672d970124ae8f2fec9de7bdc1da Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 22:41:21 +0300 Subject: [PATCH 073/138] preprocessing for textual inversion added --- modules/interrogate.py | 1 + modules/textual_inversion/preprocess.py | 75 +++++++++++++++++++ .../textual_inversion/textual_inversion.py | 1 + modules/textual_inversion/ui.py | 14 +++- modules/ui.py | 36 +++++++++ 5 files changed, 124 insertions(+), 3 deletions(-) create mode 100644 modules/textual_inversion/preprocess.py diff --git a/modules/interrogate.py b/modules/interrogate.py index f62a47458..eed87144f 100644 --- a/modules/interrogate.py +++ b/modules/interrogate.py @@ -21,6 +21,7 @@ Category = namedtuple("Category", ["name", "topn", "items"]) re_topn = re.compile(r"\.top(\d+)\.") + class InterrogateModels: blip_model = None clip_model = None diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py new file mode 100644 index 000000000..209e928ff --- /dev/null +++ b/modules/textual_inversion/preprocess.py @@ -0,0 +1,75 @@ +import os +from PIL import Image, ImageOps +import tqdm + +from modules import shared, images + + +def preprocess(process_src, process_dst, process_flip, process_split, process_caption): + size = 512 + src = os.path.abspath(process_src) + dst = os.path.abspath(process_dst) + + assert src != dst, 'same directory specified as source and desitnation' + + os.makedirs(dst, exist_ok=True) + + files = os.listdir(src) + + shared.state.textinfo = "Preprocessing..." + shared.state.job_count = len(files) + + if process_caption: + shared.interrogator.load() + + def save_pic_with_caption(image, index): + if process_caption: + caption = "-" + shared.interrogator.generate_caption(image) + else: + caption = "" + + image.save(os.path.join(dst, f"{index:05}-{subindex[0]}{caption}.png")) + subindex[0] += 1 + + def save_pic(image, index): + save_pic_with_caption(image, index) + + if process_flip: + save_pic_with_caption(ImageOps.mirror(image), index) + + for index, imagefile in enumerate(tqdm.tqdm(files)): + subindex = [0] + filename = os.path.join(src, imagefile) + img = Image.open(filename).convert("RGB") + + if shared.state.interrupted: + break + + ratio = img.height / img.width + is_tall = ratio > 1.35 + is_wide = ratio < 1 / 1.35 + + if process_split and is_tall: + img = img.resize((size, size * img.height // img.width)) + + top = img.crop((0, 0, size, size)) + save_pic(top, index) + + bot = img.crop((0, img.height - size, size, img.height)) + save_pic(bot, index) + elif process_split and is_wide: + img = img.resize((size * img.width // img.height, size)) + + left = img.crop((0, 0, size, size)) + save_pic(left, index) + + right = img.crop((img.width - size, 0, img.width, size)) + save_pic(right, index) + else: + img = images.resize_image(1, img, size, size) + save_pic(img, index) + + shared.state.nextjob() + + if process_caption: + shared.interrogator.send_blip_to_ram() diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 1183aab76..d4e250d87 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -7,6 +7,7 @@ import tqdm import html import datetime + from modules import shared, devices, sd_hijack, processing, sd_models import modules.textual_inversion.dataset diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py index 633037d8e..f19ac5e02 100644 --- a/modules/textual_inversion/ui.py +++ b/modules/textual_inversion/ui.py @@ -2,24 +2,31 @@ import html import gradio as gr -import modules.textual_inversion.textual_inversion as ti +import modules.textual_inversion.textual_inversion +import modules.textual_inversion.preprocess from modules import sd_hijack, shared def create_embedding(name, initialization_text, nvpt): - filename = ti.create_embedding(name, nvpt, init_text=initialization_text) + filename = modules.textual_inversion.textual_inversion.create_embedding(name, nvpt, init_text=initialization_text) sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() return gr.Dropdown.update(choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())), f"Created: {filename}", "" +def preprocess(*args): + modules.textual_inversion.preprocess.preprocess(*args) + + return "Preprocessing finished.", "" + + def train_embedding(*args): try: sd_hijack.undo_optimizations() - embedding, filename = ti.train_embedding(*args) + embedding, filename = modules.textual_inversion.textual_inversion.train_embedding(*args) res = f""" Training {'interrupted' if shared.state.interrupted else 'finished'} at {embedding.step} steps. @@ -30,3 +37,4 @@ Embedding saved to {html.escape(filename)} raise finally: sd_hijack.apply_optimizations() + diff --git a/modules/ui.py b/modules/ui.py index 8912deff4..e7bde53bf 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -961,6 +961,8 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row().style(equal_height=False): with gr.Column(): with gr.Group(): + gr.HTML(value="

See wiki for detailed explanation.

") + gr.HTML(value="

Create a new embedding

") new_embedding_name = gr.Textbox(label="Name") @@ -974,6 +976,24 @@ def create_ui(wrap_gradio_gpu_call): with gr.Column(): create_embedding = gr.Button(value="Create", variant='primary') + with gr.Group(): + gr.HTML(value="

Preprocess images

") + + process_src = gr.Textbox(label='Source directory') + process_dst = gr.Textbox(label='Destination directory') + + with gr.Row(): + process_flip = gr.Checkbox(label='Flip') + process_split = gr.Checkbox(label='Split into two') + process_caption = gr.Checkbox(label='Add caption') + + with gr.Row(): + with gr.Column(scale=3): + gr.HTML(value="") + + with gr.Column(): + run_preprocess = gr.Button(value="Preprocess", variant='primary') + with gr.Group(): gr.HTML(value="

Train an embedding; must specify a directory with a set of 512x512 images

") train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) @@ -1018,6 +1038,22 @@ def create_ui(wrap_gradio_gpu_call): ] ) + run_preprocess.click( + fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), + _js="start_training_textual_inversion", + inputs=[ + process_src, + process_dst, + process_flip, + process_split, + process_caption, + ], + outputs=[ + ti_output, + ti_outcome, + ], + ) + train_embedding.click( fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.train_embedding, extra_outputs=[gr.update()]), _js="start_training_textual_inversion", From 6785331e22d6a488fbf5905fab56d7fec867e038 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 22:59:01 +0300 Subject: [PATCH 074/138] keep textual inversion dataset latents in CPU memory to save a bit of VRAM --- modules/textual_inversion/dataset.py | 2 ++ modules/textual_inversion/textual_inversion.py | 3 +++ modules/ui.py | 4 ++-- 3 files changed, 7 insertions(+), 2 deletions(-) diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py index 7e134a08f..e8394ff65 100644 --- a/modules/textual_inversion/dataset.py +++ b/modules/textual_inversion/dataset.py @@ -8,6 +8,7 @@ from torchvision import transforms import random import tqdm +from modules import devices class PersonalizedBase(Dataset): @@ -47,6 +48,7 @@ class PersonalizedBase(Dataset): torchdata = torch.moveaxis(torchdata, 2, 0) init_latent = model.get_first_stage_encoding(model.encode_first_stage(torchdata.unsqueeze(dim=0))).squeeze() + init_latent = init_latent.to(devices.cpu) self.dataset.append((init_latent, filename_tokens)) diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index d4e250d87..8686f5347 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -212,7 +212,10 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, with torch.autocast("cuda"): c = cond_model([text]) + + x = x.to(devices.device) loss = shared.sd_model(x.unsqueeze(0), c)[0] + del x losses[embedding.step % losses.shape[0]] = loss.item() diff --git a/modules/ui.py b/modules/ui.py index e7bde53bf..d9d02ecef 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1002,8 +1002,8 @@ def create_ui(wrap_gradio_gpu_call): log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) steps = gr.Number(label='Max steps', value=100000, precision=0) - create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=1000, precision=0) - save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=1000, precision=0) + create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) + save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) with gr.Row(): with gr.Column(scale=2): From 166283653cfe7521a422c91e8fb801f3ecb4adc8 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 23:18:13 +0300 Subject: [PATCH 075/138] remove LDSR warning --- modules/paths.py | 1 - 1 file changed, 1 deletion(-) diff --git a/modules/paths.py b/modules/paths.py index ceb804171..606f7d666 100644 --- a/modules/paths.py +++ b/modules/paths.py @@ -20,7 +20,6 @@ path_dirs = [ (os.path.join(sd_path, '../taming-transformers'), 'taming', 'Taming Transformers', []), (os.path.join(sd_path, '../CodeFormer'), 'inference_codeformer.py', 'CodeFormer', []), (os.path.join(sd_path, '../BLIP'), 'models/blip.py', 'BLIP', []), - (os.path.join(sd_path, '../latent-diffusion'), 'LDSR.py', 'LDSR', []), (os.path.join(sd_path, '../k-diffusion'), 'k_diffusion/sampling.py', 'k_diffusion', ["atstart"]), ] From 4c2eccf8e96825333ed400f8a8a2be78141ed8ec Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 23:22:48 +0300 Subject: [PATCH 076/138] credit Rinon Gal --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 15e224e8f..ec3d7532d 100644 --- a/README.md +++ b/README.md @@ -113,6 +113,7 @@ The documentation was moved from this README over to the project's [wiki](https: - LDSR - https://github.com/Hafiidz/latent-diffusion - 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. +- Rinon Gal - Textual Inversion - https://github.com/rinongal/textual_inversion (we're not using his code, but we are using his ideas). - 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 From 138662734c25dab4e73e632b7eaff9ad9c0ce2b4 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Mon, 3 Oct 2022 07:57:59 +0300 Subject: [PATCH 077/138] use dropdown instead of radio for img2img upscaler selection --- modules/shared.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/shared.py b/modules/shared.py index 7246eadc6..2a599e9cf 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -183,7 +183,7 @@ options_templates.update(options_section(('upscaling', "Upscaling"), { "SWIN_tile": OptionInfo(192, "Tile size for all SwinIR.", gr.Slider, {"minimum": 16, "maximum": 512, "step": 16}), "SWIN_tile_overlap": OptionInfo(8, "Tile overlap, in pixels for SwinIR. Low values = visible seam.", gr.Slider, {"minimum": 0, "maximum": 48, "step": 1}), "ldsr_steps": OptionInfo(100, "LDSR processing steps. Lower = faster", gr.Slider, {"minimum": 1, "maximum": 200, "step": 1}), - "upscaler_for_img2img": OptionInfo(None, "Upscaler for img2img", gr.Radio, lambda: {"choices": [x.name for x in sd_upscalers]}), + "upscaler_for_img2img": OptionInfo(None, "Upscaler for img2img", gr.Dropdown, lambda: {"choices": [x.name for x in sd_upscalers]}), })) options_templates.update(options_section(('face-restoration', "Face restoration"), { From e615d4f9d101e2712c7c2d0e3e8feb19cb430c74 Mon Sep 17 00:00:00 2001 From: Hanusz Leszek Date: Sun, 2 Oct 2022 21:08:23 +0200 Subject: [PATCH 078/138] Convert folder icon surrogate pair to valid utf8 --- javascript/hints.js | 2 +- modules/ui.py | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/javascript/hints.js b/javascript/hints.js index 84694eeb3..e72e93381 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -15,7 +15,7 @@ titles = { "\u267b\ufe0f": "Reuse seed from last generation, mostly useful if it was randomed", "\u{1f3a8}": "Add a random artist to the prompt.", "\u2199\ufe0f": "Read generation parameters from prompt into user interface.", - "\uD83D\uDCC2": "Open images output directory", + "\u{1f4c2}": "Open images output directory", "Inpaint a part of image": "Draw a mask over an image, and the script will regenerate the masked area with content according to prompt", "SD upscale": "Upscale image normally, split result into tiles, improve each tile using img2img, merge whole image back", diff --git a/modules/ui.py b/modules/ui.py index d9d02ecef..164321512 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -69,7 +69,7 @@ random_symbol = '\U0001f3b2\ufe0f' # 🎲️ reuse_symbol = '\u267b\ufe0f' # ♻️ art_symbol = '\U0001f3a8' # 🎨 paste_symbol = '\u2199\ufe0f' # ↙ -folder_symbol = '\uD83D\uDCC2' +folder_symbol = '\U0001f4c2' # 📂 def plaintext_to_html(text): text = "

" + "
\n".join([f"{html.escape(x)}" for x in text.split('\n')]) + "

" From 34c638142eaa57f89b86545ba3c72085036398bb Mon Sep 17 00:00:00 2001 From: hentailord85ez <112723046+hentailord85ez@users.noreply.github.com> Date: Fri, 30 Sep 2022 22:38:14 +0100 Subject: [PATCH 079/138] Fixed when eta = 0 Unexpected behavior when using eta = 0 in something like XY, but your default eta was set to something not 0. --- modules/sd_samplers.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index 9316875ab..dbf570d2c 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -127,7 +127,7 @@ class VanillaStableDiffusionSampler: return res def initialize(self, p): - self.eta = p.eta or opts.eta_ddim + self.eta = p.eta if p.eta is not None else opts.eta_ddim for fieldname in ['p_sample_ddim', 'p_sample_plms']: if hasattr(self.sampler, fieldname): From 36ea4ac0f5844e5c8dec124edbdb714ccdd6013c Mon Sep 17 00:00:00 2001 From: RnDMonkey Date: Sun, 2 Oct 2022 22:21:16 -0700 Subject: [PATCH 080/138] moved no-style return outside join function --- modules/images.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/images.py b/modules/images.py index bba55158e..1a046aca6 100644 --- a/modules/images.py +++ b/modules/images.py @@ -315,7 +315,7 @@ def apply_filename_pattern(x, p, seed, prompt): #currently disabled if using the save button, will work otherwise # if enabled it will cause a bug because styles is not included in the save_files data dictionary if hasattr(p, "styles"): - x = x.replace("[styles]", sanitize_filename_part(", ".join([x for x in p.styles if not x == "None"] or "None"), replace_spaces=False)) + x = x.replace("[styles]", sanitize_filename_part(", ".join([x for x in p.styles if not x == "None"]) or "None", replace_spaces=False)) x = x.replace("[sampler]", sanitize_filename_part(sd_samplers.samplers[p.sampler_index].name, replace_spaces=False)) From 6491b09c24ea77f1f69990ea80a216f9ce319589 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Mon, 3 Oct 2022 08:53:52 +0300 Subject: [PATCH 081/138] use existing function for gfpgan --- modules/gfpgan_model.py | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/modules/gfpgan_model.py b/modules/gfpgan_model.py index bb30d7330..dd3fbcab1 100644 --- a/modules/gfpgan_model.py +++ b/modules/gfpgan_model.py @@ -97,11 +97,7 @@ def setup_model(dirname): return "GFPGAN" def restore(self, np_image): - np_image_bgr = np_image[:, :, ::-1] - cropped_faces, restored_faces, gfpgan_output_bgr = gfpgann().enhance(np_image_bgr, has_aligned=False, only_center_face=False, paste_back=True) - np_image = gfpgan_output_bgr[:, :, ::-1] - - return np_image + return gfpgan_fix_faces(np_image) shared.face_restorers.append(FaceRestorerGFPGAN()) except Exception: From 43a74fa595003321200a40bd2431e56c245e75ed Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Mon, 3 Oct 2022 11:48:19 +0300 Subject: [PATCH 082/138] batch processing for img2img with an empty output directory, by request --- modules/img2img.py | 7 +++++-- modules/ui.py | 2 +- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/modules/img2img.py b/modules/img2img.py index f4455c90f..2ff8e2617 100644 --- a/modules/img2img.py +++ b/modules/img2img.py @@ -23,8 +23,10 @@ def process_batch(p, input_dir, output_dir, args): print(f"Will process {len(images)} images, creating {p.n_iter * p.batch_size} new images for each.") + save_normally = output_dir == '' + p.do_not_save_grid = True - p.do_not_save_samples = True + p.do_not_save_samples = not save_normally state.job_count = len(images) * p.n_iter @@ -48,7 +50,8 @@ def process_batch(p, input_dir, output_dir, args): left, right = os.path.splitext(filename) filename = f"{left}-{n}{right}" - processed_image.save(os.path.join(output_dir, filename)) + if not save_normally: + processed_image.save(os.path.join(output_dir, filename)) def img2img(mode: int, prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: str, init_img, init_img_with_mask, init_img_inpaint, init_mask_inpaint, mask_mode, steps: int, sampler_index: int, mask_blur: int, inpainting_fill: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, denoising_strength: float, seed: int, subseed: int, subseed_strength: float, seed_resize_from_h: int, seed_resize_from_w: int, seed_enable_extras: bool, height: int, width: int, resize_mode: int, inpaint_full_res: bool, inpaint_full_res_padding: int, inpainting_mask_invert: int, img2img_batch_input_dir: str, img2img_batch_output_dir: str, *args): diff --git a/modules/ui.py b/modules/ui.py index 164321512..55f7aa953 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -658,7 +658,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.TabItem('Batch img2img', id='batch'): hidden = '
Disabled when launched with --hide-ui-dir-config.' if shared.cmd_opts.hide_ui_dir_config else '' - gr.HTML(f"

Process images in a directory on the same machine where the server is running.{hidden}

") + gr.HTML(f"

Process images in a directory on the same machine where the server is running.
Use an empty output directory to save pictures normally instead of writing to the output directory.{hidden}

") img2img_batch_input_dir = gr.Textbox(label="Input directory", **shared.hide_dirs) img2img_batch_output_dir = gr.Textbox(label="Output directory", **shared.hide_dirs) From 2865ef4b9ab16d56326cc805541bebcf01d099bc Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Mon, 3 Oct 2022 13:10:03 +0300 Subject: [PATCH 083/138] fix broken date in TI --- modules/textual_inversion/textual_inversion.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 8686f5347..cd9f34984 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -164,7 +164,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, filename = os.path.join(shared.cmd_opts.embeddings_dir, f'{embedding_name}.pt') - log_directory = os.path.join(log_directory, datetime.datetime.now().strftime("%Y-%d-%m"), embedding_name) + log_directory = os.path.join(log_directory, datetime.datetime.now().strftime("%Y-%m-%d"), embedding_name) if save_embedding_every > 0: embedding_dir = os.path.join(log_directory, "embeddings") From 5ef0baf5eaec7f21a1666af424405cbee19f3764 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 08:52:11 +0300 Subject: [PATCH 084/138] add support for gelbooru tags in filenames for textual inversion --- modules/textual_inversion/dataset.py | 7 +++++-- modules/textual_inversion/preprocess.py | 4 +++- 2 files changed, 8 insertions(+), 3 deletions(-) diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py index e8394ff65..7c44ea5be 100644 --- a/modules/textual_inversion/dataset.py +++ b/modules/textual_inversion/dataset.py @@ -9,6 +9,9 @@ from torchvision import transforms import random import tqdm from modules import devices +import re + +re_tag = re.compile(r"[a-zA-Z][_\w\d()]+") class PersonalizedBase(Dataset): @@ -38,8 +41,8 @@ class PersonalizedBase(Dataset): image = image.resize((self.width, self.height), PIL.Image.BICUBIC) filename = os.path.basename(path) - filename_tokens = os.path.splitext(filename)[0].replace('_', '-').replace(' ', '-').split('-') - filename_tokens = [token for token in filename_tokens if token.isalpha()] + filename_tokens = os.path.splitext(filename)[0] + filename_tokens = re_tag.findall(filename_tokens) npimage = np.array(image).astype(np.uint8) npimage = (npimage / 127.5 - 1.0).astype(np.float32) diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py index 209e928ff..f545a9937 100644 --- a/modules/textual_inversion/preprocess.py +++ b/modules/textual_inversion/preprocess.py @@ -26,7 +26,9 @@ def preprocess(process_src, process_dst, process_flip, process_split, process_ca if process_caption: caption = "-" + shared.interrogator.generate_caption(image) else: - caption = "" + caption = filename + caption = os.path.splitext(caption)[0] + caption = os.path.basename(caption) image.save(os.path.join(dst, f"{index:05}-{subindex[0]}{caption}.png")) subindex[0] += 1 From 1c5604791da7e57f40880698666b6617a1754c65 Mon Sep 17 00:00:00 2001 From: DoTheSneedful Date: Mon, 3 Oct 2022 22:20:09 -0400 Subject: [PATCH 085/138] Add a prompt order option to XY plot script --- scripts/xy_grid.py | 40 ++++++++++++++++++++++++++++++++++++++-- 1 file changed, 38 insertions(+), 2 deletions(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 146663b0a..044c30e61 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -1,5 +1,6 @@ from collections import namedtuple from copy import copy +from itertools import permutations import random from PIL import Image @@ -28,6 +29,27 @@ def apply_prompt(p, x, xs): p.prompt = p.prompt.replace(xs[0], x) p.negative_prompt = p.negative_prompt.replace(xs[0], x) +def apply_order(p, x, xs): + token_order = [] + + # Initally grab the tokens from the prompt so they can be later be replaced in order of earliest seen in the prompt + for token in x: + token_order.append((p.prompt.find(token), token)) + + token_order.sort(key=lambda t: t[0]) + + search_from_pos = 0 + for idx, token in enumerate(x): + original_pos, old_token = token_order[idx] + + # Get position of the token again as it will likely change as tokens are being replaced + pos = p.prompt.find(old_token) + if original_pos >= 0: + # Avoid trying to replace what was just replaced by searching later in the prompt string + p.prompt = p.prompt[0:search_from_pos] + p.prompt[search_from_pos:].replace(old_token, token, 1) + + search_from_pos = pos + len(token) + samplers_dict = {} for i, sampler in enumerate(modules.sd_samplers.samplers): @@ -60,7 +82,8 @@ def format_value_add_label(p, opt, x): def format_value(p, opt, x): if type(x) == float: x = round(x, 8) - + if type(x) == type(list()): + x = str(x) return x def do_nothing(p, x, xs): @@ -89,6 +112,7 @@ axis_options = [ AxisOption("Sigma max", float, apply_field("s_tmax"), format_value_add_label), AxisOption("Sigma noise", float, apply_field("s_noise"), format_value_add_label), AxisOption("Eta", float, apply_field("eta"), format_value_add_label), + AxisOption("Prompt order", type(list()), apply_order, 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 ] @@ -159,7 +183,11 @@ class Script(scripts.Script): if opt.label == 'Nothing': return [0] - valslist = [x.strip() for x in vals.split(",")] + if opt.type == type(list()): + valslist = [x for x in vals] + else: + valslist = [x.strip() for x in vals.split(",")] + if opt.type == int: valslist_ext = [] @@ -212,9 +240,17 @@ class Script(scripts.Script): return valslist x_opt = axis_options[x_type] + + if x_opt.label == "Prompt order": + x_values = list(permutations([x.strip() for x in x_values.split(",")])) + xs = process_axis(x_opt, x_values) y_opt = axis_options[y_type] + + if y_opt.label == "Prompt order": + y_values = list(permutations([y.strip() for y in y_values.split(",")])) + ys = process_axis(y_opt, y_values) def fix_axis_seeds(axis_opt, axis_list): From 1a6d40db35656083d5bf9d3a3430b45fda4e85eb Mon Sep 17 00:00:00 2001 From: DoTheSneedful Date: Tue, 4 Oct 2022 00:18:15 -0400 Subject: [PATCH 086/138] Fix token ordering in prompt order XY plot --- scripts/xy_grid.py | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 044c30e61..5bcd39217 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -32,24 +32,21 @@ def apply_prompt(p, x, xs): def apply_order(p, x, xs): token_order = [] - # Initally grab the tokens from the prompt so they can be later be replaced in order of earliest seen in the prompt + # Initally grab the tokens from the prompt so they can be be replaced in order of earliest seen for token in x: token_order.append((p.prompt.find(token), token)) token_order.sort(key=lambda t: t[0]) search_from_pos = 0 - for idx, token in enumerate(x): - original_pos, old_token = token_order[idx] - + for idx, (original_pos, old_token) in enumerate(token_order): # Get position of the token again as it will likely change as tokens are being replaced - pos = p.prompt.find(old_token) + pos = search_from_pos + p.prompt[search_from_pos:].find(old_token) if original_pos >= 0: # Avoid trying to replace what was just replaced by searching later in the prompt string - p.prompt = p.prompt[0:search_from_pos] + p.prompt[search_from_pos:].replace(old_token, token, 1) - - search_from_pos = pos + len(token) + p.prompt = p.prompt[0:search_from_pos] + p.prompt[search_from_pos:].replace(old_token, x[idx], 1) + search_from_pos = pos + len(x[idx]) samplers_dict = {} for i, sampler in enumerate(modules.sd_samplers.samplers): From 56371153b545e3a43c3a5f206264019af361f3af Mon Sep 17 00:00:00 2001 From: DoTheSneedful Date: Tue, 4 Oct 2022 01:07:36 -0400 Subject: [PATCH 087/138] XY plot prompt order simplify logic --- scripts/xy_grid.py | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 5bcd39217..7def47f57 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -38,15 +38,21 @@ def apply_order(p, x, xs): token_order.sort(key=lambda t: t[0]) - search_from_pos = 0 - for idx, (original_pos, old_token) in enumerate(token_order): - # Get position of the token again as it will likely change as tokens are being replaced - pos = search_from_pos + p.prompt[search_from_pos:].find(old_token) - if original_pos >= 0: - # Avoid trying to replace what was just replaced by searching later in the prompt string - p.prompt = p.prompt[0:search_from_pos] + p.prompt[search_from_pos:].replace(old_token, x[idx], 1) + prompt_parts = [] - search_from_pos = pos + len(x[idx]) + # Split the prompt up, taking out the tokens + for _, token in token_order: + n = p.prompt.find(token) + prompt_parts.append(p.prompt[0:n]) + p.prompt = p.prompt[n + len(token):] + + # Rebuild the prompt with the tokens in the order we want + prompt_tmp = "" + for idx, part in enumerate(prompt_parts): + prompt_tmp += part + prompt_tmp += x[idx] + p.prompt = prompt_tmp + p.prompt + samplers_dict = {} for i, sampler in enumerate(modules.sd_samplers.samplers): From 556c36b9607e3f4eacdddc85f8e7a78b29476ea7 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 09:18:00 +0300 Subject: [PATCH 088/138] add hint, refactor code for #1607 --- javascript/hints.js | 1 + scripts/xy_grid.py | 35 ++++++++++++++++++----------------- 2 files changed, 19 insertions(+), 17 deletions(-) diff --git a/javascript/hints.js b/javascript/hints.js index e72e93381..8adcd983e 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -47,6 +47,7 @@ titles = { "Custom code": "Run Python code. Advanced user only. Must run program with --allow-code for this to work", "Prompt S/R": "Separate a list of words with commas, and the first word will be used as a keyword: script will search for this word in the prompt, and replace it with others", + "Prompt order": "Separate a list of words with commas, and the script will make a variation of prompt with those words for their every possible order", "Tiling": "Produce an image that can be tiled.", "Tile overlap": "For SD upscale, how much overlap in pixels should there be between tiles. Tiles overlap so that when they are merged back into one picture, there is no clearly visible seam.", diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 7def47f57..1237e754d 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -29,10 +29,11 @@ def apply_prompt(p, x, xs): p.prompt = p.prompt.replace(xs[0], x) p.negative_prompt = p.negative_prompt.replace(xs[0], x) + def apply_order(p, x, xs): token_order = [] - # Initally grab the tokens from the prompt so they can be be replaced in order of earliest seen + # Initally grab the tokens from the prompt, so they can be replaced in order of earliest seen for token in x: token_order.append((p.prompt.find(token), token)) @@ -85,17 +86,26 @@ def format_value_add_label(p, opt, x): def format_value(p, opt, x): if type(x) == float: x = round(x, 8) - if type(x) == type(list()): - x = str(x) return x + +def format_value_join_list(p, opt, x): + return ", ".join(x) + + def do_nothing(p, x, xs): pass + def format_nothing(p, opt, x): return "" +def str_permutations(x): + """dummy function for specifying it in AxisOption's type when you want to get a list of permutations""" + return x + + AxisOption = namedtuple("AxisOption", ["label", "type", "apply", "format_value"]) AxisOptionImg2Img = namedtuple("AxisOptionImg2Img", ["label", "type", "apply", "format_value"]) @@ -108,6 +118,7 @@ axis_options = [ AxisOption("Steps", int, apply_field("steps"), format_value_add_label), AxisOption("CFG Scale", float, apply_field("cfg_scale"), format_value_add_label), AxisOption("Prompt S/R", str, apply_prompt, format_value), + AxisOption("Prompt order", str_permutations, apply_order, format_value_join_list), AxisOption("Sampler", str, apply_sampler, format_value), AxisOption("Checkpoint name", str, apply_checkpoint, format_value), AxisOption("Sigma Churn", float, apply_field("s_churn"), format_value_add_label), @@ -115,7 +126,6 @@ axis_options = [ AxisOption("Sigma max", float, apply_field("s_tmax"), format_value_add_label), AxisOption("Sigma noise", float, apply_field("s_noise"), format_value_add_label), AxisOption("Eta", float, apply_field("eta"), format_value_add_label), - AxisOption("Prompt order", type(list()), apply_order, 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 ] @@ -158,6 +168,7 @@ re_range_float = re.compile(r"\s*([+-]?\s*\d+(?:.\d*)?)\s*-\s*([+-]?\s*\d+(?:.\d re_range_count = re.compile(r"\s*([+-]?\s*\d+)\s*-\s*([+-]?\s*\d+)(?:\s*\[(\d+)\s*\])?\s*") re_range_count_float = re.compile(r"\s*([+-]?\s*\d+(?:.\d*)?)\s*-\s*([+-]?\s*\d+(?:.\d*)?)(?:\s*\[(\d+(?:.\d*)?)\s*\])?\s*") + class Script(scripts.Script): def title(self): return "X/Y plot" @@ -186,11 +197,7 @@ class Script(scripts.Script): if opt.label == 'Nothing': return [0] - if opt.type == type(list()): - valslist = [x for x in vals] - else: - valslist = [x.strip() for x in vals.split(",")] - + valslist = [x.strip() for x in vals.split(",")] if opt.type == int: valslist_ext = [] @@ -237,23 +244,17 @@ class Script(scripts.Script): valslist_ext.append(val) valslist = valslist_ext + elif opt.type == str_permutations: + valslist = list(permutations(valslist)) valslist = [opt.type(x) for x in valslist] return valslist x_opt = axis_options[x_type] - - if x_opt.label == "Prompt order": - x_values = list(permutations([x.strip() for x in x_values.split(",")])) - xs = process_axis(x_opt, x_values) y_opt = axis_options[y_type] - - if y_opt.label == "Prompt order": - y_values = list(permutations([y.strip() for y in y_values.split(",")])) - ys = process_axis(y_opt, y_values) def fix_axis_seeds(axis_opt, axis_list): From eeab7aedf532680a6ae9058ee272450bb07e41eb Mon Sep 17 00:00:00 2001 From: brkirch Date: Tue, 4 Oct 2022 04:24:35 -0400 Subject: [PATCH 089/138] Add --use-cpu command line option Remove MPS detection to use CPU for GFPGAN / CodeFormer and add a --use-cpu command line option. --- modules/devices.py | 5 ++--- modules/esrgan_model.py | 9 ++++----- modules/scunet_model.py | 8 ++++---- modules/shared.py | 9 +++++++-- 4 files changed, 17 insertions(+), 14 deletions(-) diff --git a/modules/devices.py b/modules/devices.py index 5d9c7a076..b5a0cd29e 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -1,8 +1,8 @@ import torch -# has_mps is only available in nightly pytorch (for now), `getattr` for compatibility from modules import errors +# has_mps is only available in nightly pytorch (for now), `getattr` for compatibility has_mps = getattr(torch, 'has_mps', False) cpu = torch.device("cpu") @@ -32,8 +32,7 @@ def enable_tf32(): errors.run(enable_tf32, "Enabling TF32") -device = get_optimal_device() -device_gfpgan = device_codeformer = cpu if device.type == 'mps' else device +device = device_gfpgan = device_esrgan = device_scunet = device_codeformer = get_optimal_device() dtype = torch.float16 def randn(seed, shape): diff --git a/modules/esrgan_model.py b/modules/esrgan_model.py index 4aed9283c..d17e730f9 100644 --- a/modules/esrgan_model.py +++ b/modules/esrgan_model.py @@ -6,8 +6,7 @@ from PIL import Image from basicsr.utils.download_util import load_file_from_url import modules.esrgam_model_arch as arch -from modules import shared, modelloader, images -from modules.devices import has_mps +from modules import shared, modelloader, images, devices from modules.paths import models_path from modules.upscaler import Upscaler, UpscalerData from modules.shared import opts @@ -97,7 +96,7 @@ class UpscalerESRGAN(Upscaler): model = self.load_model(selected_model) if model is None: return img - model.to(shared.device) + model.to(devices.device_esrgan) img = esrgan_upscale(model, img) return img @@ -112,7 +111,7 @@ class UpscalerESRGAN(Upscaler): print("Unable to load %s from %s" % (self.model_path, filename)) return None - pretrained_net = torch.load(filename, map_location='cpu' if has_mps else None) + pretrained_net = torch.load(filename, map_location='cpu' if shared.device.type == 'mps' else None) crt_model = arch.RRDBNet(3, 3, 64, 23, gc=32) pretrained_net = fix_model_layers(crt_model, pretrained_net) @@ -127,7 +126,7 @@ def upscale_without_tiling(model, img): img = img[:, :, ::-1] img = np.moveaxis(img, 2, 0) / 255 img = torch.from_numpy(img).float() - img = img.unsqueeze(0).to(shared.device) + img = img.unsqueeze(0).to(devices.device_esrgan) with torch.no_grad(): output = model(img) output = output.squeeze().float().cpu().clamp_(0, 1).numpy() diff --git a/modules/scunet_model.py b/modules/scunet_model.py index 7987ac145..fb64b7409 100644 --- a/modules/scunet_model.py +++ b/modules/scunet_model.py @@ -8,7 +8,7 @@ import torch from basicsr.utils.download_util import load_file_from_url import modules.upscaler -from modules import shared, modelloader +from modules import devices, modelloader from modules.paths import models_path from modules.scunet_model_arch import SCUNet as net @@ -51,12 +51,12 @@ class UpscalerScuNET(modules.upscaler.Upscaler): if model is None: return img - device = shared.device + device = devices.device_scunet img = np.array(img) img = img[:, :, ::-1] img = np.moveaxis(img, 2, 0) / 255 img = torch.from_numpy(img).float() - img = img.unsqueeze(0).to(shared.device) + img = img.unsqueeze(0).to(device) img = img.to(device) with torch.no_grad(): @@ -69,7 +69,7 @@ class UpscalerScuNET(modules.upscaler.Upscaler): return PIL.Image.fromarray(output, 'RGB') def load_model(self, path: str): - device = shared.device + device = devices.device_scunet if "http" in path: filename = load_file_from_url(url=self.model_url, model_dir=self.model_path, file_name="%s.pth" % self.name, progress=True) diff --git a/modules/shared.py b/modules/shared.py index 2a599e9cf..7899ab8d1 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -12,7 +12,7 @@ import modules.interrogate import modules.memmon import modules.sd_models import modules.styles -from modules.devices import get_optimal_device +import modules.devices as devices from modules.paths import script_path, sd_path sd_model_file = os.path.join(script_path, 'model.ckpt') @@ -46,6 +46,7 @@ parser.add_argument("--ldsr-models-path", type=str, help="Path to directory with parser.add_argument("--opt-split-attention", action='store_true', help="force-enables cross-attention layer optimization. By default, it's on for torch.cuda and off for other torch devices.") parser.add_argument("--disable-opt-split-attention", action='store_true', help="force-disables cross-attention layer optimization") parser.add_argument("--opt-split-attention-v1", action='store_true', help="enable older version of split attention optimization that does not consume all the VRAM it can find") +parser.add_argument("--use-cpu", nargs='+',choices=['SD', 'GFPGAN', 'ESRGAN', 'SCUNet', 'CodeFormer'], help="use CPU for specified modules", default=[]) parser.add_argument("--listen", action='store_true', help="launch gradio with 0.0.0.0 as server name, allowing to respond to network requests") parser.add_argument("--port", type=int, help="launch gradio with given server port, you need root/admin rights for ports < 1024, defaults to 7860 if available", default=None) parser.add_argument("--show-negative-prompt", action='store_true', help="does not do anything", default=False) @@ -63,7 +64,11 @@ parser.add_argument("--enable-console-prompts", action='store_true', help="print cmd_opts = parser.parse_args() -device = get_optimal_device() + +devices.device, devices.device_gfpgan, devices.device_esrgan, devices.device_scunet, devices.device_codeformer = \ +(devices.cpu if x in cmd_opts.use_cpu else devices.get_optimal_device() for x in ['SD', 'GFPGAN', 'ESRGAN', 'SCUNet', 'CodeFormer']) + +device = devices.device batch_cond_uncond = cmd_opts.always_batch_cond_uncond or not (cmd_opts.lowvram or cmd_opts.medvram) parallel_processing_allowed = not cmd_opts.lowvram and not cmd_opts.medvram From 27ddc24fdee1fbe709054a43235ab7f9c51b3e9f Mon Sep 17 00:00:00 2001 From: brkirch Date: Tue, 4 Oct 2022 05:18:17 -0400 Subject: [PATCH 090/138] Add BSRGAN to --add-cpu --- modules/bsrgan_model.py | 6 +++--- modules/devices.py | 2 +- modules/shared.py | 6 +++--- 3 files changed, 7 insertions(+), 7 deletions(-) diff --git a/modules/bsrgan_model.py b/modules/bsrgan_model.py index e62c66577..3bd80791a 100644 --- a/modules/bsrgan_model.py +++ b/modules/bsrgan_model.py @@ -8,7 +8,7 @@ import torch from basicsr.utils.download_util import load_file_from_url import modules.upscaler -from modules import shared, modelloader +from modules import devices, modelloader from modules.bsrgan_model_arch import RRDBNet from modules.paths import models_path @@ -44,13 +44,13 @@ class UpscalerBSRGAN(modules.upscaler.Upscaler): model = self.load_model(selected_file) if model is None: return img - model.to(shared.device) + model.to(devices.device_bsrgan) torch.cuda.empty_cache() img = np.array(img) img = img[:, :, ::-1] img = np.moveaxis(img, 2, 0) / 255 img = torch.from_numpy(img).float() - img = img.unsqueeze(0).to(shared.device) + img = img.unsqueeze(0).to(devices.device_bsrgan) with torch.no_grad(): output = model(img) output = output.squeeze().float().cpu().clamp_(0, 1).numpy() diff --git a/modules/devices.py b/modules/devices.py index b5a0cd29e..b78996322 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -32,7 +32,7 @@ def enable_tf32(): errors.run(enable_tf32, "Enabling TF32") -device = device_gfpgan = device_esrgan = device_scunet = device_codeformer = get_optimal_device() +device = device_gfpgan = device_bsrgan = device_esrgan = device_scunet = device_codeformer = get_optimal_device() dtype = torch.float16 def randn(seed, shape): diff --git a/modules/shared.py b/modules/shared.py index 7899ab8d1..95b98a06e 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -46,7 +46,7 @@ parser.add_argument("--ldsr-models-path", type=str, help="Path to directory with parser.add_argument("--opt-split-attention", action='store_true', help="force-enables cross-attention layer optimization. By default, it's on for torch.cuda and off for other torch devices.") parser.add_argument("--disable-opt-split-attention", action='store_true', help="force-disables cross-attention layer optimization") parser.add_argument("--opt-split-attention-v1", action='store_true', help="enable older version of split attention optimization that does not consume all the VRAM it can find") -parser.add_argument("--use-cpu", nargs='+',choices=['SD', 'GFPGAN', 'ESRGAN', 'SCUNet', 'CodeFormer'], help="use CPU for specified modules", default=[]) +parser.add_argument("--use-cpu", nargs='+',choices=['SD', 'GFPGAN', 'BSRGAN', 'ESRGAN', 'SCUNet', 'CodeFormer'], help="use CPU for specified modules", default=[]) parser.add_argument("--listen", action='store_true', help="launch gradio with 0.0.0.0 as server name, allowing to respond to network requests") parser.add_argument("--port", type=int, help="launch gradio with given server port, you need root/admin rights for ports < 1024, defaults to 7860 if available", default=None) parser.add_argument("--show-negative-prompt", action='store_true', help="does not do anything", default=False) @@ -65,8 +65,8 @@ parser.add_argument("--enable-console-prompts", action='store_true', help="print cmd_opts = parser.parse_args() -devices.device, devices.device_gfpgan, devices.device_esrgan, devices.device_scunet, devices.device_codeformer = \ -(devices.cpu if x in cmd_opts.use_cpu else devices.get_optimal_device() for x in ['SD', 'GFPGAN', 'ESRGAN', 'SCUNet', 'CodeFormer']) +devices.device, devices.device_gfpgan, devices.device_bsrgan, devices.device_esrgan, devices.device_scunet, devices.device_codeformer = \ +(devices.cpu if x in cmd_opts.use_cpu else devices.get_optimal_device() for x in ['SD', 'GFPGAN', 'BSRGAN', 'ESRGAN', 'SCUNet', 'CodeFormer']) device = devices.device From dc9c5a97742e3a34d37da7108642d8adc0dc5858 Mon Sep 17 00:00:00 2001 From: brkirch Date: Tue, 4 Oct 2022 05:22:50 -0400 Subject: [PATCH 091/138] Modify --add-cpu description --- modules/shared.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/shared.py b/modules/shared.py index 95b98a06e..25aff5b0e 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -46,7 +46,7 @@ parser.add_argument("--ldsr-models-path", type=str, help="Path to directory with parser.add_argument("--opt-split-attention", action='store_true', help="force-enables cross-attention layer optimization. By default, it's on for torch.cuda and off for other torch devices.") parser.add_argument("--disable-opt-split-attention", action='store_true', help="force-disables cross-attention layer optimization") parser.add_argument("--opt-split-attention-v1", action='store_true', help="enable older version of split attention optimization that does not consume all the VRAM it can find") -parser.add_argument("--use-cpu", nargs='+',choices=['SD', 'GFPGAN', 'BSRGAN', 'ESRGAN', 'SCUNet', 'CodeFormer'], help="use CPU for specified modules", default=[]) +parser.add_argument("--use-cpu", nargs='+',choices=['SD', 'GFPGAN', 'BSRGAN', 'ESRGAN', 'SCUNet', 'CodeFormer'], help="use CPU as torch device for specified modules", default=[]) parser.add_argument("--listen", action='store_true', help="launch gradio with 0.0.0.0 as server name, allowing to respond to network requests") parser.add_argument("--port", type=int, help="launch gradio with given server port, you need root/admin rights for ports < 1024, defaults to 7860 if available", default=None) parser.add_argument("--show-negative-prompt", action='store_true', help="does not do anything", default=False) From 6c6ae28bf5fd1e8bc3e8f64a3430b6f29f338f77 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 12:32:22 +0300 Subject: [PATCH 092/138] send all three of GFPGAN's and codeformer's models to CPU memory instead of just one for #1283 --- modules/codeformer_model.py | 12 ++++++++++-- modules/devices.py | 10 ++++++++++ modules/gfpgan_model.py | 14 ++++++++++++-- modules/processing.py | 16 +++++++++------- 4 files changed, 41 insertions(+), 11 deletions(-) diff --git a/modules/codeformer_model.py b/modules/codeformer_model.py index a29f38550..e6d9fa4f4 100644 --- a/modules/codeformer_model.py +++ b/modules/codeformer_model.py @@ -69,10 +69,14 @@ def setup_model(dirname): self.net = net self.face_helper = face_helper - self.net.to(devices.device_codeformer) return net, face_helper + def send_model_to(self, device): + self.net.to(device) + self.face_helper.face_det.to(device) + self.face_helper.face_parse.to(device) + def restore(self, np_image, w=None): np_image = np_image[:, :, ::-1] @@ -82,6 +86,8 @@ def setup_model(dirname): if self.net is None or self.face_helper is None: return np_image + self.send_model_to(devices.device_codeformer) + self.face_helper.clean_all() self.face_helper.read_image(np_image) self.face_helper.get_face_landmarks_5(only_center_face=False, resize=640, eye_dist_threshold=5) @@ -113,8 +119,10 @@ def setup_model(dirname): if original_resolution != restored_img.shape[0:2]: restored_img = cv2.resize(restored_img, (0, 0), fx=original_resolution[1]/restored_img.shape[1], fy=original_resolution[0]/restored_img.shape[0], interpolation=cv2.INTER_LINEAR) + self.face_helper.clean_all() + if shared.opts.face_restoration_unload: - self.net.to(devices.cpu) + self.send_model_to(devices.cpu) return restored_img diff --git a/modules/devices.py b/modules/devices.py index ff82f2f64..12aab6652 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -1,3 +1,5 @@ +import contextlib + import torch # has_mps is only available in nightly pytorch (for now), `getattr` for compatibility @@ -57,3 +59,11 @@ def randn_without_seed(shape): return torch.randn(shape, device=device) + +def autocast(): + from modules import shared + + if dtype == torch.float32 or shared.cmd_opts.precision == "full": + return contextlib.nullcontext() + + return torch.autocast("cuda") diff --git a/modules/gfpgan_model.py b/modules/gfpgan_model.py index dd3fbcab1..5586b554b 100644 --- a/modules/gfpgan_model.py +++ b/modules/gfpgan_model.py @@ -37,22 +37,32 @@ def gfpgann(): print("Unable to load gfpgan model!") return None model = gfpgan_constructor(model_path=model_file, upscale=1, arch='clean', channel_multiplier=2, bg_upsampler=None) - model.gfpgan.to(shared.device) loaded_gfpgan_model = model return model +def send_model_to(model, device): + model.gfpgan.to(device) + model.face_helper.face_det.to(device) + model.face_helper.face_parse.to(device) + + def gfpgan_fix_faces(np_image): model = gfpgann() if model is None: return np_image + + send_model_to(model, devices.device) + np_image_bgr = np_image[:, :, ::-1] cropped_faces, restored_faces, gfpgan_output_bgr = model.enhance(np_image_bgr, has_aligned=False, only_center_face=False, paste_back=True) np_image = gfpgan_output_bgr[:, :, ::-1] + model.face_helper.clean_all() + if shared.opts.face_restoration_unload: - model.gfpgan.to(devices.cpu) + send_model_to(model, devices.cpu) return np_image diff --git a/modules/processing.py b/modules/processing.py index 0a4b6198f..9cbecdd83 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -1,4 +1,3 @@ -import contextlib import json import math import os @@ -330,9 +329,8 @@ def process_images(p: StableDiffusionProcessing) -> Processed: infotexts = [] output_images = [] - precision_scope = torch.autocast if cmd_opts.precision == "autocast" else contextlib.nullcontext - ema_scope = (contextlib.nullcontext if cmd_opts.lowvram else p.sd_model.ema_scope) - with torch.no_grad(), precision_scope("cuda"), ema_scope(): + + with torch.no_grad(): p.init(all_prompts, all_seeds, all_subseeds) if state.job_count == -1: @@ -351,8 +349,9 @@ def process_images(p: StableDiffusionProcessing) -> Processed: #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) - c = prompt_parser.get_learned_conditioning(prompts, p.steps) + with devices.autocast(): + uc = prompt_parser.get_learned_conditioning(len(prompts) * [p.negative_prompt], p.steps) + c = prompt_parser.get_learned_conditioning(prompts, p.steps) if len(model_hijack.comments) > 0: for comment in model_hijack.comments: @@ -361,7 +360,9 @@ def process_images(p: StableDiffusionProcessing) -> Processed: if p.n_iter > 1: shared.state.job = f"Batch {n+1} out of {p.n_iter}" - samples_ddim = p.sample(conditioning=c, unconditional_conditioning=uc, seeds=seeds, subseeds=subseeds, subseed_strength=p.subseed_strength) + with devices.autocast(): + samples_ddim = p.sample(conditioning=c, unconditional_conditioning=uc, seeds=seeds, subseeds=subseeds, subseed_strength=p.subseed_strength).to(devices.dtype) + if state.interrupted: # if we are interruped, sample returns just noise @@ -386,6 +387,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: devices.torch_gc() x_sample = modules.face_restoration.restore_faces(x_sample) + devices.torch_gc() image = Image.fromarray(x_sample) From 2f1b61d97987ae0a52a7dfc6bc99c68928bdb594 Mon Sep 17 00:00:00 2001 From: dan Date: Mon, 3 Oct 2022 19:25:36 +0800 Subject: [PATCH 093/138] Allow nested structures inside schedules --- modules/prompt_parser.py | 119 +++++++++++++++++--------------------- requirements.txt | 1 + requirements_versions.txt | 1 + 3 files changed, 55 insertions(+), 66 deletions(-) diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index e811eb9ec..99c8ed99c 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -1,20 +1,11 @@ import re from collections import namedtuple import torch +from lark import Lark, Transformer, Visitor +import functools import modules.shared as shared -re_prompt = re.compile(r''' -(.*?) -\[ - ([^]:]+): - (?:([^]:]*):)? - ([0-9]*\.?[0-9]+) -] -| -(.+) -''', re.X) - # a prompt like this: "fantasy landscape with a [mountain:lake:0.25] and [an oak:a christmas tree:0.75][ in foreground::0.6][ in background:0.25] [shoddy:masterful:0.5]" # will be represented with prompt_schedule like this (assuming steps=100): # [25, 'fantasy landscape with a mountain and an oak in foreground shoddy'] @@ -25,61 +16,57 @@ re_prompt = re.compile(r''' def get_learned_conditioning_prompt_schedules(prompts, steps): - res = [] - cache = {} - - for prompt in prompts: - prompt_schedule: list[list[str | int]] = [[steps, ""]] - - cached = cache.get(prompt, None) - if cached is not None: - res.append(cached) - continue - - for m in re_prompt.finditer(prompt): - plaintext = m.group(1) if m.group(5) is None else m.group(5) - concept_from = m.group(2) - concept_to = m.group(3) - if concept_to is None: - concept_to = concept_from - concept_from = "" - swap_position = float(m.group(4)) if m.group(4) is not None else None - - if swap_position is not None: - if swap_position < 1: - swap_position = swap_position * steps - swap_position = int(min(swap_position, steps)) - - swap_index = None - found_exact_index = False - for i in range(len(prompt_schedule)): - end_step = prompt_schedule[i][0] - prompt_schedule[i][1] += plaintext - - if swap_position is not None and swap_index is None: - if swap_position == end_step: - swap_index = i - found_exact_index = True - - if swap_position < end_step: - swap_index = i - - if swap_index is not None: - if not found_exact_index: - prompt_schedule.insert(swap_index, [swap_position, prompt_schedule[swap_index][1]]) - - for i in range(len(prompt_schedule)): - end_step = prompt_schedule[i][0] - must_replace = swap_position < end_step - - prompt_schedule[i][1] += concept_to if must_replace else concept_from - - res.append(prompt_schedule) - cache[prompt] = prompt_schedule - #for t in prompt_schedule: - # print(t) - - return res + grammar = r""" + start: prompt + prompt: (emphasized | scheduled | weighted | plain)* + !emphasized: "(" prompt ")" + | "(" prompt ":" prompt ")" + | "[" prompt "]" + scheduled: "[" (prompt ":")? prompt ":" NUMBER "]" + !weighted: "{" weighted_item ("|" weighted_item)* "}" + !weighted_item: prompt (":" prompt)? + plain: /([^\\\[\](){}:|]|\\.)+/ + %import common.SIGNED_NUMBER -> NUMBER + """ + parser = Lark(grammar, parser='lalr') + def collect_steps(steps, tree): + l = [steps] + class CollectSteps(Visitor): + def scheduled(self, tree): + tree.children[-1] = float(tree.children[-1]) + if tree.children[-1] < 1: + tree.children[-1] *= steps + tree.children[-1] = min(steps, int(tree.children[-1])) + l.append(tree.children[-1]) + CollectSteps().visit(tree) + return sorted(set(l)) + def at_step(step, tree): + class AtStep(Transformer): + def scheduled(self, args): + if len(args) == 2: + before, after, when = (), *args + else: + before, after, when = args + yield before if step <= when else after + def start(self, args): + def flatten(x): + if type(x) == str: + yield x + else: + for gen in x: + yield from flatten(gen) + return ''.join(flatten(args[0])) + def plain(self, args): + yield args[0].value + def __default__(self, data, children, meta): + for child in children: + yield from child + return AtStep().transform(tree) + @functools.cache + def get_schedule(prompt): + tree = parser.parse(prompt) + return [[t, at_step(t, tree)] for t in collect_steps(steps, tree)] + return [get_schedule(prompt) for prompt in prompts] ScheduledPromptConditioning = namedtuple("ScheduledPromptConditioning", ["end_at_step", "cond"]) diff --git a/requirements.txt b/requirements.txt index d4b337fce..631fe616a 100644 --- a/requirements.txt +++ b/requirements.txt @@ -22,3 +22,4 @@ clean-fid resize-right torchdiffeq kornia +lark diff --git a/requirements_versions.txt b/requirements_versions.txt index 8a9acf205..fdff26878 100644 --- a/requirements_versions.txt +++ b/requirements_versions.txt @@ -21,3 +21,4 @@ clean-fid==0.1.29 resize-right==0.0.2 torchdiffeq==0.2.3 kornia==0.6.7 +lark==1.1.2 From 61652461242951966e5b4cee83ce359cefa91c17 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 14:23:22 +0300 Subject: [PATCH 094/138] support interrupting after the previous change --- modules/processing.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/modules/processing.py b/modules/processing.py index 9cbecdd83..6f5599c7d 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -361,7 +361,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: shared.state.job = f"Batch {n+1} out of {p.n_iter}" with devices.autocast(): - samples_ddim = p.sample(conditioning=c, unconditional_conditioning=uc, seeds=seeds, subseeds=subseeds, subseed_strength=p.subseed_strength).to(devices.dtype) + samples_ddim = p.sample(conditioning=c, unconditional_conditioning=uc, seeds=seeds, subseeds=subseeds, subseed_strength=p.subseed_strength) if state.interrupted: @@ -369,6 +369,8 @@ def process_images(p: StableDiffusionProcessing) -> Processed: # use the image collected previously in sampler loop samples_ddim = shared.state.current_latent + samples_ddim = samples_ddim.to(devices.dtype) + x_samples_ddim = p.sd_model.decode_first_stage(samples_ddim) x_samples_ddim = torch.clamp((x_samples_ddim + 1.0) / 2.0, min=0.0, max=1.0) From d5bba20a58f43a9f984bb67b4e17f48661f6b818 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 14:35:12 +0300 Subject: [PATCH 095/138] ignore errors in parse for purposes of token counting for #1564 --- modules/ui.py | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/modules/ui.py b/modules/ui.py index 55f7aa953..20dc8c379 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -386,14 +386,22 @@ def connect_reuse_seed(seed: gr.Number, reuse_seed: gr.Button, generation_info: outputs=[seed, dummy_component] ) + def update_token_counter(text, steps): - prompt_schedules = get_learned_conditioning_prompt_schedules([text], steps) + try: + prompt_schedules = get_learned_conditioning_prompt_schedules([text], steps) + except Exception: + # a parsing error can happen here during typing, and we don't want to bother the user with + # messages related to it in console + prompt_schedules = [[[steps, text]]] + flat_prompts = reduce(lambda list1, list2: list1+list2, prompt_schedules) - prompts = [prompt_text for step,prompt_text in flat_prompts] + prompts = [prompt_text for step, prompt_text in flat_prompts] tokens, token_count, max_length = max([model_hijack.tokenize(prompt) for prompt in prompts], key=lambda args: args[1]) style_class = ' class="red"' if (token_count > max_length) else "" return f"{token_count}/{max_length}" + def create_toprow(is_img2img): id_part = "img2img" if is_img2img else "txt2img" From accd00d6b8258c12b5168918a4c546b02357924a Mon Sep 17 00:00:00 2001 From: Justin Riddiough Date: Tue, 4 Oct 2022 01:14:28 -0500 Subject: [PATCH 096/138] Explain how to use second progress bar in pycharm --- modules/shared.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/shared.py b/modules/shared.py index 25aff5b0e..11bdf01a7 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -200,7 +200,7 @@ options_templates.update(options_section(('face-restoration', "Face restoration" options_templates.update(options_section(('system', "System"), { "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}), "samples_log_stdout": OptionInfo(False, "Always print all generation info to standard output"), - "multiple_tqdm": OptionInfo(True, "Add a second progress bar to the console that shows progress for an entire job. Broken in PyCharm console."), + "multiple_tqdm": OptionInfo(True, "Add a second progress bar to the console that shows progress for an entire job. In PyCharm select 'emulate terminal in console output'."), })) options_templates.update(options_section(('sd', "Stable Diffusion"), { From ea6b0d98a64290a0305e27126ea59ce1da7959a2 Mon Sep 17 00:00:00 2001 From: Justin Riddiough Date: Tue, 4 Oct 2022 06:38:45 -0500 Subject: [PATCH 097/138] Remove pycharm note, fix typo --- modules/shared.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/modules/shared.py b/modules/shared.py index 11bdf01a7..a7d13b2d4 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -200,7 +200,7 @@ options_templates.update(options_section(('face-restoration', "Face restoration" options_templates.update(options_section(('system', "System"), { "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}), "samples_log_stdout": OptionInfo(False, "Always print all generation info to standard output"), - "multiple_tqdm": OptionInfo(True, "Add a second progress bar to the console that shows progress for an entire job. In PyCharm select 'emulate terminal in console output'."), + "multiple_tqdm": OptionInfo(True, "Add a second progress bar to the console that shows progress for an entire job."), })) options_templates.update(options_section(('sd', "Stable Diffusion"), { @@ -209,7 +209,7 @@ options_templates.update(options_section(('sd', "Stable Diffusion"), { "save_images_before_color_correction": OptionInfo(False, "Save a copy of image before applying color correction to img2img results"), "img2img_fix_steps": OptionInfo(False, "With img2img, do exactly the amount of steps the slider specifies (normally you'd do less with less denoising)."), "enable_quantization": OptionInfo(False, "Enable quantization in K samplers for sharper and cleaner results. This may change existing seeds. Requires restart to apply."), - "enable_emphasis": OptionInfo(True, "Eemphasis: use (text) to make model pay more attention to text and [text] to make it pay less attention"), + "enable_emphasis": OptionInfo(True, "Emphasis: use (text) to make model pay more attention to text and [text] to make it pay less attention"), "use_old_emphasis_implementation": OptionInfo(False, "Use old emphasis implementation. Can be useful to reproduce old seeds."), "enable_batch_seeds": OptionInfo(True, "Make K-diffusion samplers produce same images in a batch as when making a single image"), "filter_nsfw": OptionInfo(False, "Filter NSFW content"), From eec1b39bd54711ca31e43022d2d6ac8c6d7281da Mon Sep 17 00:00:00 2001 From: Milly Date: Tue, 4 Oct 2022 20:16:52 +0900 Subject: [PATCH 098/138] Apply prompt pattern last --- modules/images.py | 43 ++++++++++++++++++++++--------------------- 1 file changed, 22 insertions(+), 21 deletions(-) diff --git a/modules/images.py b/modules/images.py index bba55158e..5b56c7e37 100644 --- a/modules/images.py +++ b/modules/images.py @@ -287,32 +287,13 @@ def apply_filename_pattern(x, p, seed, prompt): if seed is not None: x = x.replace("[seed]", str(seed)) - if prompt is not None: - x = x.replace("[prompt]", sanitize_filename_part(prompt)) - if "[prompt_no_styles]" in x: - prompt_no_style = prompt - for style in shared.prompt_styles.get_style_prompts(p.styles): - if len(style) > 0: - style_parts = [y for y in style.split("{prompt}")] - for part in style_parts: - prompt_no_style = prompt_no_style.replace(part, "").replace(", ,", ",").strip().strip(',') - prompt_no_style = prompt_no_style.replace(style, "").strip().strip(',').strip() - x = x.replace("[prompt_no_styles]", sanitize_filename_part(prompt_no_style, replace_spaces=False)) - - x = x.replace("[prompt_spaces]", sanitize_filename_part(prompt, replace_spaces=False)) - if "[prompt_words]" in x: - words = [x for x in re_nonletters.split(prompt or "") if len(x) > 0] - if len(words) == 0: - words = ["empty"] - x = x.replace("[prompt_words]", sanitize_filename_part(" ".join(words[0:max_prompt_words]), replace_spaces=False)) - if p is not None: x = x.replace("[steps]", str(p.steps)) x = x.replace("[cfg]", str(p.cfg_scale)) x = x.replace("[width]", str(p.width)) x = x.replace("[height]", str(p.height)) - - #currently disabled if using the save button, will work otherwise + + #currently disabled if using the save button, will work otherwise # if enabled it will cause a bug because styles is not included in the save_files data dictionary if hasattr(p, "styles"): x = x.replace("[styles]", sanitize_filename_part(", ".join([x for x in p.styles if not x == "None"] or "None"), replace_spaces=False)) @@ -324,6 +305,26 @@ def apply_filename_pattern(x, p, seed, prompt): x = x.replace("[datetime]", datetime.datetime.now().strftime("%Y%m%d%H%M%S")) x = x.replace("[job_timestamp]", shared.state.job_timestamp) + # Apply [prompt] at last. Because it may contain any replacement word.^M + if prompt is not None: + x = x.replace("[prompt]", sanitize_filename_part(prompt)) + if "[prompt_no_styles]" in x: + prompt_no_style = prompt + for style in shared.prompt_styles.get_style_prompts(p.styles): + if len(style) > 0: + style_parts = [y for y in style.split("{prompt}")] + for part in style_parts: + prompt_no_style = prompt_no_style.replace(part, "").replace(", ,", ",").strip().strip(',') + prompt_no_style = prompt_no_style.replace(style, "").strip().strip(',').strip() + x = x.replace("[prompt_no_styles]", sanitize_filename_part(prompt_no_style, replace_spaces=False)) + + x = x.replace("[prompt_spaces]", sanitize_filename_part(prompt, replace_spaces=False)) + if "[prompt_words]" in x: + words = [x for x in re_nonletters.split(prompt or "") if len(x) > 0] + if len(words) == 0: + words = ["empty"] + x = x.replace("[prompt_words]", sanitize_filename_part(" ".join(words[0:max_prompt_words]), replace_spaces=False)) + if cmd_opts.hide_ui_dir_config: x = re.sub(r'^[\\/]+|\.{2,}[\\/]+|[\\/]+\.{2,}', '', x) From 52cef36f6ba169a8e606ecdcaed73d47378f0e8e Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 16:54:31 +0300 Subject: [PATCH 099/138] emergency fix for img2img --- modules/processing.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/modules/processing.py b/modules/processing.py index 6f5599c7d..e9c453942 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -331,7 +331,8 @@ def process_images(p: StableDiffusionProcessing) -> Processed: output_images = [] with torch.no_grad(): - p.init(all_prompts, all_seeds, all_subseeds) + with devices.autocast(): + p.init(all_prompts, all_seeds, all_subseeds) if state.job_count == -1: state.job_count = p.n_iter From 957e29a8e9cb8ca069799ec69263e188c89ed6a6 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 17:23:48 +0300 Subject: [PATCH 100/138] option to not show images in web ui --- modules/img2img.py | 3 +++ modules/shared.py | 1 + modules/txt2img.py | 3 +++ 3 files changed, 7 insertions(+) diff --git a/modules/img2img.py b/modules/img2img.py index 2ff8e2617..da212d72b 100644 --- a/modules/img2img.py +++ b/modules/img2img.py @@ -129,4 +129,7 @@ def img2img(mode: int, prompt: str, negative_prompt: str, prompt_style: str, pro if opts.samples_log_stdout: print(generation_info_js) + if opts.do_not_show_images: + processed.images = [] + return processed.images, generation_info_js, plaintext_to_html(processed.info) diff --git a/modules/shared.py b/modules/shared.py index a7d13b2d4..ff4e5fa39 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -229,6 +229,7 @@ options_templates.update(options_section(('ui', "User interface"), { "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}), "return_grid": OptionInfo(True, "Show grid in results for web"), + "do_not_show_images": OptionInfo(False, "Do not show any images in results for web"), "add_model_hash_to_info": OptionInfo(True, "Add model hash to generation information"), "font": OptionInfo("", "Font for image grids that have text"), "js_modal_lightbox": OptionInfo(True, "Enable full page image viewer"), diff --git a/modules/txt2img.py b/modules/txt2img.py index d4406c3c0..e985242b3 100644 --- a/modules/txt2img.py +++ b/modules/txt2img.py @@ -48,5 +48,8 @@ def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: if opts.samples_log_stdout: print(generation_info_js) + if opts.do_not_show_images: + processed.images = [] + return processed.images, generation_info_js, plaintext_to_html(processed.info) From e1b128d8e46bddb9c0b2fd3ee0eefd57e0527ee0 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 17:36:39 +0300 Subject: [PATCH 101/138] do not touch p.seed/p.subseed during processing #1181 --- modules/processing.py | 26 +++++++++++++++++--------- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/modules/processing.py b/modules/processing.py index e9c453942..8180c63d8 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -248,9 +248,16 @@ def create_random_tensors(shape, seeds, subseeds=None, subseed_strength=0.0, see return x +def get_fixed_seed(seed): + if seed is None or seed == '' or seed == -1: + return int(random.randrange(4294967294)) + + return seed + + def fix_seed(p): - p.seed = int(random.randrange(4294967294)) if p.seed is None or p.seed == '' or p.seed == -1 else p.seed - p.subseed = int(random.randrange(4294967294)) if p.subseed is None or p.subseed == '' or p.subseed == -1 else p.subseed + p.seed = get_fixed_seed(p.seed) + p.subseed = get_fixed_seed(p.subseed) def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments, iteration=0, position_in_batch=0): @@ -292,7 +299,8 @@ def process_images(p: StableDiffusionProcessing) -> Processed: devices.torch_gc() - fix_seed(p) + seed = get_fixed_seed(p.seed) + subseed = get_fixed_seed(p.subseed) if p.outpath_samples is not None: os.makedirs(p.outpath_samples, exist_ok=True) @@ -311,15 +319,15 @@ def process_images(p: StableDiffusionProcessing) -> Processed: else: all_prompts = p.batch_size * p.n_iter * [p.prompt] - if type(p.seed) == list: - all_seeds = p.seed + if type(seed) == list: + all_seeds = seed else: - all_seeds = [int(p.seed) + (x if p.subseed_strength == 0 else 0) for x in range(len(all_prompts))] + all_seeds = [int(seed) + (x if p.subseed_strength == 0 else 0) for x in range(len(all_prompts))] - if type(p.subseed) == list: - all_subseeds = p.subseed + if type(subseed) == list: + all_subseeds = subseed else: - all_subseeds = [int(p.subseed) + x for x in range(len(all_prompts))] + all_subseeds = [int(subseed) + x for x in range(len(all_prompts))] def infotext(iteration=0, position_in_batch=0): return create_infotext(p, all_prompts, all_seeds, all_subseeds, comments, iteration, position_in_batch) From 1eb588cbf19924333b88beaa1ac0041904966640 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 18:02:01 +0300 Subject: [PATCH 102/138] remove functools.cache as some people are having issues with it --- modules/prompt_parser.py | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index 99c8ed99c..5d58c4ed9 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -29,6 +29,7 @@ def get_learned_conditioning_prompt_schedules(prompts, steps): %import common.SIGNED_NUMBER -> NUMBER """ parser = Lark(grammar, parser='lalr') + def collect_steps(steps, tree): l = [steps] class CollectSteps(Visitor): @@ -40,6 +41,7 @@ def get_learned_conditioning_prompt_schedules(prompts, steps): l.append(tree.children[-1]) CollectSteps().visit(tree) return sorted(set(l)) + def at_step(step, tree): class AtStep(Transformer): def scheduled(self, args): @@ -62,11 +64,13 @@ def get_learned_conditioning_prompt_schedules(prompts, steps): for child in children: yield from child return AtStep().transform(tree) - @functools.cache + def get_schedule(prompt): tree = parser.parse(prompt) return [[t, at_step(t, tree)] for t in collect_steps(steps, tree)] - return [get_schedule(prompt) for prompt in prompts] + + promptdict = {prompt: get_schedule(prompt) for prompt in set(prompts)} + return [promptdict[prompt] for prompt in prompts] ScheduledPromptConditioning = namedtuple("ScheduledPromptConditioning", ["end_at_step", "cond"]) From 90e911fd546e76f879b38a764473569911a0f845 Mon Sep 17 00:00:00 2001 From: Rae Fu Date: Tue, 4 Oct 2022 09:49:51 -0600 Subject: [PATCH 103/138] prompt_parser: allow spaces in schedules, add test, log/ignore errors Only build the parser once (at import time) instead of for each step. doctest is run by simply executing modules/prompt_parser.py --- modules/processing.py | 10 +-- modules/prompt_parser.py | 139 +++++++++++++++++++++++++-------------- 2 files changed, 95 insertions(+), 54 deletions(-) diff --git a/modules/processing.py b/modules/processing.py index 8180c63d8..bb94033b1 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -84,7 +84,7 @@ class StableDiffusionProcessing: self.s_tmin = opts.s_tmin self.s_tmax = float('inf') # not representable as a standard ui option self.s_noise = opts.s_noise - + if not seed_enable_extras: self.subseed = -1 self.subseed_strength = 0 @@ -296,7 +296,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: assert(len(p.prompt) > 0) else: assert p.prompt is not None - + devices.torch_gc() seed = get_fixed_seed(p.seed) @@ -359,8 +359,8 @@ def process_images(p: StableDiffusionProcessing) -> Processed: #uc = p.sd_model.get_learned_conditioning(len(prompts) * [p.negative_prompt]) #c = p.sd_model.get_learned_conditioning(prompts) with devices.autocast(): - uc = prompt_parser.get_learned_conditioning(len(prompts) * [p.negative_prompt], p.steps) - c = prompt_parser.get_learned_conditioning(prompts, p.steps) + uc = prompt_parser.get_learned_conditioning(shared.sd_model, len(prompts) * [p.negative_prompt], p.steps) + c = prompt_parser.get_learned_conditioning(shared.sd_model, prompts, p.steps) if len(model_hijack.comments) > 0: for comment in model_hijack.comments: @@ -527,7 +527,7 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): # GC now before running the next img2img to prevent running out of memory x = None devices.torch_gc() - + samples = self.sampler.sample_img2img(self, samples, noise, conditioning, unconditional_conditioning, steps=self.steps) return samples diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index 5d58c4ed9..a3b124219 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -1,10 +1,7 @@ import re from collections import namedtuple -import torch -from lark import Lark, Transformer, Visitor -import functools -import modules.shared as shared +import lark # a prompt like this: "fantasy landscape with a [mountain:lake:0.25] and [an oak:a christmas tree:0.75][ in foreground::0.6][ in background:0.25] [shoddy:masterful:0.5]" # will be represented with prompt_schedule like this (assuming steps=100): @@ -14,25 +11,48 @@ import modules.shared as shared # [75, 'fantasy landscape with a lake and an oak in background masterful'] # [100, 'fantasy landscape with a lake and a christmas tree in background masterful'] +schedule_parser = lark.Lark(r""" +!start: (prompt | /[][():]/+)* +prompt: (emphasized | scheduled | plain | WHITESPACE)* +!emphasized: "(" prompt ")" + | "(" prompt ":" prompt ")" + | "[" prompt "]" +scheduled: "[" [prompt ":"] prompt ":" [WHITESPACE] NUMBER "]" +WHITESPACE: /\s+/ +plain: /([^\\\[\]():]|\\.)+/ +%import common.SIGNED_NUMBER -> NUMBER +""") def get_learned_conditioning_prompt_schedules(prompts, steps): - grammar = r""" - start: prompt - prompt: (emphasized | scheduled | weighted | plain)* - !emphasized: "(" prompt ")" - | "(" prompt ":" prompt ")" - | "[" prompt "]" - scheduled: "[" (prompt ":")? prompt ":" NUMBER "]" - !weighted: "{" weighted_item ("|" weighted_item)* "}" - !weighted_item: prompt (":" prompt)? - plain: /([^\\\[\](){}:|]|\\.)+/ - %import common.SIGNED_NUMBER -> NUMBER """ - parser = Lark(grammar, parser='lalr') + >>> g = lambda p: get_learned_conditioning_prompt_schedules([p], 10)[0] + >>> g("test") + [[10, 'test']] + >>> g("a [b:3]") + [[3, 'a '], [10, 'a b']] + >>> g("a [b: 3]") + [[3, 'a '], [10, 'a b']] + >>> g("a [[[b]]:2]") + [[2, 'a '], [10, 'a [[b]]']] + >>> g("[(a:2):3]") + [[3, ''], [10, '(a:2)']] + >>> g("a [b : c : 1] d") + [[1, 'a b d'], [10, 'a c d']] + >>> g("a[b:[c:d:2]:1]e") + [[1, 'abe'], [2, 'ace'], [10, 'ade']] + >>> g("a [unbalanced") + [[10, 'a [unbalanced']] + >>> g("a [b:.5] c") + [[5, 'a c'], [10, 'a b c']] + >>> g("a [{b|d{:.5] c") # not handling this right now + [[5, 'a c'], [10, 'a {b|d{ c']] + >>> g("((a][:b:c [d:3]") + [[3, '((a][:b:c '], [10, '((a][:b:c d']] + """ def collect_steps(steps, tree): l = [steps] - class CollectSteps(Visitor): + class CollectSteps(lark.Visitor): def scheduled(self, tree): tree.children[-1] = float(tree.children[-1]) if tree.children[-1] < 1: @@ -43,13 +63,10 @@ def get_learned_conditioning_prompt_schedules(prompts, steps): return sorted(set(l)) def at_step(step, tree): - class AtStep(Transformer): + class AtStep(lark.Transformer): def scheduled(self, args): - if len(args) == 2: - before, after, when = (), *args - else: - before, after, when = args - yield before if step <= when else after + before, after, _, when = args + yield before or () if step <= when else after def start(self, args): def flatten(x): if type(x) == str: @@ -57,16 +74,22 @@ def get_learned_conditioning_prompt_schedules(prompts, steps): else: for gen in x: yield from flatten(gen) - return ''.join(flatten(args[0])) + return ''.join(flatten(args)) def plain(self, args): yield args[0].value def __default__(self, data, children, meta): for child in children: yield from child return AtStep().transform(tree) - + def get_schedule(prompt): - tree = parser.parse(prompt) + try: + tree = schedule_parser.parse(prompt) + except lark.exceptions.LarkError as e: + if 0: + import traceback + traceback.print_exc() + return [[steps, prompt]] return [[t, at_step(t, tree)] for t in collect_steps(steps, tree)] promptdict = {prompt: get_schedule(prompt) for prompt in set(prompts)} @@ -77,8 +100,7 @@ ScheduledPromptConditioning = namedtuple("ScheduledPromptConditioning", ["end_at ScheduledPromptBatch = namedtuple("ScheduledPromptBatch", ["shape", "schedules"]) -def get_learned_conditioning(prompts, steps): - +def get_learned_conditioning(model, prompts, steps): res = [] prompt_schedules = get_learned_conditioning_prompt_schedules(prompts, steps) @@ -92,7 +114,7 @@ def get_learned_conditioning(prompts, steps): continue texts = [x[1] for x in prompt_schedule] - conds = shared.sd_model.get_learned_conditioning(texts) + conds = model.get_learned_conditioning(texts) cond_schedule = [] for i, (end_at_step, text) in enumerate(prompt_schedule): @@ -105,12 +127,13 @@ def get_learned_conditioning(prompts, steps): def reconstruct_cond_batch(c: ScheduledPromptBatch, current_step): - res = torch.zeros(c.shape, device=shared.device, dtype=next(shared.sd_model.parameters()).dtype) + param = c.schedules[0][0].cond + res = torch.zeros(c.shape, device=param.device, dtype=param.dtype) for i, cond_schedule in enumerate(c.schedules): target_index = 0 - for curret_index, (end_at, cond) in enumerate(cond_schedule): + for current, (end_at, cond) in enumerate(cond_schedule): if current_step <= end_at: - target_index = curret_index + target_index = current break res[i] = cond_schedule[target_index].cond @@ -148,23 +171,26 @@ def parse_prompt_attention(text): \\ - literal character '\' anything else - just text - Example: - - 'a (((house:1.3)) [on] a (hill:0.5), sun, (((sky))).' - - produces: - - [ - ['a ', 1.0], - ['house', 1.5730000000000004], - [' ', 1.1], - ['on', 1.0], - [' a ', 1.1], - ['hill', 0.55], - [', sun, ', 1.1], - ['sky', 1.4641000000000006], - ['.', 1.1] - ] + >>> parse_prompt_attention('normal text') + [['normal text', 1.0]] + >>> parse_prompt_attention('an (important) word') + [['an ', 1.0], ['important', 1.1], [' word', 1.0]] + >>> parse_prompt_attention('(unbalanced') + [['unbalanced', 1.1]] + >>> parse_prompt_attention('\(literal\]') + [['(literal]', 1.0]] + >>> parse_prompt_attention('(unnecessary)(parens)') + [['unnecessaryparens', 1.1]] + >>> parse_prompt_attention('a (((house:1.3)) [on] a (hill:0.5), sun, (((sky))).') + [['a ', 1.0], + ['house', 1.5730000000000004], + [' ', 1.1], + ['on', 1.0], + [' a ', 1.1], + ['hill', 0.55], + [', sun, ', 1.1], + ['sky', 1.4641000000000006], + ['.', 1.1]] """ res = [] @@ -206,4 +232,19 @@ def parse_prompt_attention(text): if len(res) == 0: res = [["", 1.0]] + # merge runs of identical weights + i = 0 + while i + 1 < len(res): + if res[i][1] == res[i + 1][1]: + res[i][0] += res[i + 1][0] + res.pop(i + 1) + else: + i += 1 + return res + +if __name__ == "__main__": + import doctest + doctest.testmod(optionflags=doctest.NORMALIZE_WHITESPACE) +else: + import torch # doctest faster From b32852ef037251eb3d846af76e2965594e1ac7a5 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 20:49:54 +0300 Subject: [PATCH 104/138] add editor to img2img --- modules/shared.py | 1 + modules/ui.py | 2 +- style.css | 4 ++++ 3 files changed, 6 insertions(+), 1 deletion(-) diff --git a/modules/shared.py b/modules/shared.py index ff4e5fa39..e52c9b1d1 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -55,6 +55,7 @@ parser.add_argument("--hide-ui-dir-config", action='store_true', help="hide dire parser.add_argument("--ui-settings-file", type=str, help="filename to use for ui settings", default=os.path.join(script_path, 'config.json')) parser.add_argument("--gradio-debug", action='store_true', help="launch gradio with --debug option") parser.add_argument("--gradio-auth", type=str, help='set gradio authentication like "username:password"; or comma-delimit multiple like "u1:p1,u2:p2,u3:p3"', default=None) +parser.add_argument("--gradio-img2img-tool", type=str, help='gradio image uploader tool: can be either editor for ctopping, or color-sketch for drawing', choices=["color-sketch", "editor"], default="color-sketch") parser.add_argument("--opt-channelslast", action='store_true', help="change memory type for stable diffusion to channels last") parser.add_argument("--styles-file", type=str, help="filename to use for styles", default=os.path.join(script_path, 'styles.csv')) parser.add_argument("--autolaunch", action='store_true', help="open the webui URL in the system's default browser upon launch", default=False) diff --git a/modules/ui.py b/modules/ui.py index 20dc8c379..6cd6761b8 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -644,7 +644,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Tabs(elem_id="mode_img2img") as tabs_img2img_mode: with gr.TabItem('img2img', id='img2img'): - init_img = gr.Image(label="Image for img2img", show_label=False, source="upload", interactive=True, type="pil") + init_img = gr.Image(label="Image for img2img", elem_id="img2img_image", show_label=False, source="upload", interactive=True, type="pil", tool=cmd_opts.gradio_img2img_tool) with gr.TabItem('Inpaint', id='inpaint'): init_img_with_mask = gr.Image(label="Image for inpainting with mask", show_label=False, elem_id="img2maskimg", source="upload", interactive=True, type="pil", tool="sketch", image_mode="RGBA") diff --git a/style.css b/style.css index 39586bf18..e8f4cb752 100644 --- a/style.css +++ b/style.css @@ -403,3 +403,7 @@ input[type="range"]{ .red { color: red; } + +#img2img_image div.h-60{ + height: 480px; +} \ No newline at end of file From ef40e4cd4d383a3405e03f1da3f5b5a1820a8f53 Mon Sep 17 00:00:00 2001 From: xpscyho Date: Tue, 4 Oct 2022 15:12:38 -0400 Subject: [PATCH 105/138] Display time taken in mins, secs when relevant Fixes #1656 --- modules/ui.py | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/modules/ui.py b/modules/ui.py index 6cd6761b8..de6342a48 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -196,6 +196,11 @@ def wrap_gradio_call(func, extra_outputs=None): res = extra_outputs_array + [f"
{plaintext_to_html(type(e).__name__+': '+str(e))}
"] elapsed = time.perf_counter() - t + elapsed_m = int(elapsed // 60) + elapsed_s = elapsed % 60 + elapsed_text = f"{elapsed_s:.2f}s" + if (elapsed_m > 0): + elapsed_text = f"{elapsed_m}m "+elapsed_text if run_memmon: mem_stats = {k: -(v//-(1024*1024)) for k, v in shared.mem_mon.stop().items()} @@ -210,7 +215,7 @@ def wrap_gradio_call(func, extra_outputs=None): vram_html = '' # last item is always HTML - res[-1] += f"

Time taken: {elapsed:.2f}s

{vram_html}
" + res[-1] += f"

Time taken: {elapsed_text}

{vram_html}
" shared.state.interrupted = False shared.state.job_count = 0 From 82380d9ac18614c87bebba1b4cfd4b147cc76a18 Mon Sep 17 00:00:00 2001 From: Jairo Correa Date: Tue, 4 Oct 2022 22:28:50 -0300 Subject: [PATCH 106/138] Removing parts no longer needed to fix vram --- modules/devices.py | 3 +-- modules/processing.py | 21 ++++++++------------- 2 files changed, 9 insertions(+), 15 deletions(-) diff --git a/modules/devices.py b/modules/devices.py index 6db4e57c9..0158b11fc 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -1,7 +1,6 @@ import contextlib import torch -import gc from modules import errors @@ -20,8 +19,8 @@ def get_optimal_device(): return cpu + def torch_gc(): - gc.collect() if torch.cuda.is_available(): torch.cuda.empty_cache() torch.cuda.ipc_collect() diff --git a/modules/processing.py b/modules/processing.py index e7f9c85e1..f666ba811 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -345,8 +345,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: if state.job_count == -1: state.job_count = p.n_iter - for n in range(p.n_iter): - with torch.no_grad(), precision_scope("cuda"), ema_scope(): + for n in range(p.n_iter): if state.interrupted: break @@ -395,22 +394,19 @@ def process_images(p: StableDiffusionProcessing) -> Processed: import modules.safety as safety x_samples_ddim = modules.safety.censor_batch(x_samples_ddim) - for i, x_sample in enumerate(x_samples_ddim): - with torch.no_grad(), precision_scope("cuda"), ema_scope(): + for i, x_sample in enumerate(x_samples_ddim): x_sample = 255. * np.moveaxis(x_sample.cpu().numpy(), 0, 2) x_sample = x_sample.astype(np.uint8) - if p.restore_faces: - with torch.no_grad(), precision_scope("cuda"), ema_scope(): + if p.restore_faces: if opts.save and not p.do_not_save_samples and opts.save_images_before_face_restoration: images.save_image(Image.fromarray(x_sample), p.outpath_samples, "", seeds[i], prompts[i], opts.samples_format, info=infotext(n, i), p=p, suffix="-before-face-restoration") + devices.torch_gc() + x_sample = modules.face_restoration.restore_faces(x_sample) devices.torch_gc() - devices.torch_gc() - - with torch.no_grad(), precision_scope("cuda"), ema_scope(): image = Image.fromarray(x_sample) if p.color_corrections is not None and i < len(p.color_corrections): @@ -438,13 +434,12 @@ def process_images(p: StableDiffusionProcessing) -> Processed: infotexts.append(infotext(n, i)) output_images.append(image) - del x_samples_ddim + del x_samples_ddim - devices.torch_gc() + devices.torch_gc() - state.nextjob() + state.nextjob() - with torch.no_grad(), precision_scope("cuda"), ema_scope(): p.color_corrections = None index_of_first_image = 0 From bbdbbd36eda870cf0bd49fdf28476c78919a123e Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Wed, 5 Oct 2022 04:43:05 +0100 Subject: [PATCH 107/138] shared.state.interrupt when restart is requested --- modules/ui.py | 1 + 1 file changed, 1 insertion(+) diff --git a/modules/ui.py b/modules/ui.py index de6342a48..523ab25b3 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1210,6 +1210,7 @@ def create_ui(wrap_gradio_gpu_call): ) def request_restart(): + shared.state.interrupt() settings_interface.gradio_ref.do_restart = True restart_gradio.click( From 67d011b02eddc20202b654dfea56528de3d5edf7 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Wed, 5 Oct 2022 04:44:22 +0100 Subject: [PATCH 108/138] Show generation progress in window title --- javascript/progressbar.js | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/javascript/progressbar.js b/javascript/progressbar.js index 1e297abbe..3e3220c3f 100644 --- a/javascript/progressbar.js +++ b/javascript/progressbar.js @@ -4,6 +4,21 @@ global_progressbars = {} function check_progressbar(id_part, id_progressbar, id_progressbar_span, id_interrupt, id_preview, id_gallery){ var progressbar = gradioApp().getElementById(id_progressbar) var interrupt = gradioApp().getElementById(id_interrupt) + + if(progressbar && progressbar.offsetParent){ + if(progressbar.innerText){ + let newtitle = 'Stable Diffusion - ' + progressbar.innerText + if(document.title != newtitle){ + document.title = newtitle; + } + }else{ + let newtitle = 'Stable Diffusion' + if(document.title != newtitle){ + document.title = newtitle; + } + } + } + if(progressbar!= null && progressbar != global_progressbars[id_progressbar]){ global_progressbars[id_progressbar] = progressbar From c26732fbee2a57e621ac22bf70decf7496daa4cd Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 5 Oct 2022 23:16:27 +0300 Subject: [PATCH 109/138] added support for AND from https://energy-based-model.github.io/Compositional-Visual-Generation-with-Composable-Diffusion-Models/ --- modules/processing.py | 2 +- modules/prompt_parser.py | 114 ++++++++++++++++++++++++++++++++++++--- modules/sd_samplers.py | 35 ++++++++---- modules/ui.py | 6 ++- 4 files changed, 138 insertions(+), 19 deletions(-) diff --git a/modules/processing.py b/modules/processing.py index bb94033b1..d8c6b8d57 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -360,7 +360,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: #c = p.sd_model.get_learned_conditioning(prompts) with devices.autocast(): uc = prompt_parser.get_learned_conditioning(shared.sd_model, len(prompts) * [p.negative_prompt], p.steps) - c = prompt_parser.get_learned_conditioning(shared.sd_model, prompts, p.steps) + c = prompt_parser.get_multicond_learned_conditioning(shared.sd_model, prompts, p.steps) if len(model_hijack.comments) > 0: for comment in model_hijack.comments: diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index a3b124219..f7420daf9 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -97,10 +97,26 @@ def get_learned_conditioning_prompt_schedules(prompts, steps): ScheduledPromptConditioning = namedtuple("ScheduledPromptConditioning", ["end_at_step", "cond"]) -ScheduledPromptBatch = namedtuple("ScheduledPromptBatch", ["shape", "schedules"]) def get_learned_conditioning(model, prompts, steps): + """converts a list of prompts into a list of prompt schedules - each schedule is a list of ScheduledPromptConditioning, specifying the comdition (cond), + and the sampling step at which this condition is to be replaced by the next one. + + Input: + (model, ['a red crown', 'a [blue:green:5] jeweled crown'], 20) + + Output: + [ + [ + ScheduledPromptConditioning(end_at_step=20, cond=tensor([[-0.3886, 0.0229, -0.0523, ..., -0.4901, -0.3066, 0.0674], ..., [ 0.3317, -0.5102, -0.4066, ..., 0.4119, -0.7647, -1.0160]], device='cuda:0')) + ], + [ + ScheduledPromptConditioning(end_at_step=5, cond=tensor([[-0.3886, 0.0229, -0.0522, ..., -0.4901, -0.3067, 0.0673], ..., [-0.0192, 0.3867, -0.4644, ..., 0.1135, -0.3696, -0.4625]], device='cuda:0')), + ScheduledPromptConditioning(end_at_step=20, cond=tensor([[-0.3886, 0.0229, -0.0522, ..., -0.4901, -0.3067, 0.0673], ..., [-0.7352, -0.4356, -0.7888, ..., 0.6994, -0.4312, -1.2593]], device='cuda:0')) + ] + ] + """ res = [] prompt_schedules = get_learned_conditioning_prompt_schedules(prompts, steps) @@ -123,13 +139,75 @@ def get_learned_conditioning(model, prompts, steps): cache[prompt] = cond_schedule res.append(cond_schedule) - return ScheduledPromptBatch((len(prompts),) + res[0][0].cond.shape, res) + return res -def reconstruct_cond_batch(c: ScheduledPromptBatch, current_step): - param = c.schedules[0][0].cond - res = torch.zeros(c.shape, device=param.device, dtype=param.dtype) - for i, cond_schedule in enumerate(c.schedules): +re_AND = re.compile(r"\bAND\b") +re_weight = re.compile(r"^(.*?)(?:\s*:\s*([-+]?\s*(?:\d+|\d*\.\d+)?))?\s*$") + + +def get_multicond_prompt_list(prompts): + res_indexes = [] + + prompt_flat_list = [] + prompt_indexes = {} + + for prompt in prompts: + subprompts = re_AND.split(prompt) + + indexes = [] + for subprompt in subprompts: + text, weight = re_weight.search(subprompt).groups() + + weight = float(weight) if weight is not None else 1.0 + + index = prompt_indexes.get(text, None) + if index is None: + index = len(prompt_flat_list) + prompt_flat_list.append(text) + prompt_indexes[text] = index + + indexes.append((index, weight)) + + res_indexes.append(indexes) + + return res_indexes, prompt_flat_list, prompt_indexes + + +class ComposableScheduledPromptConditioning: + def __init__(self, schedules, weight=1.0): + self.schedules: list[ScheduledPromptConditioning] = schedules + self.weight: float = weight + + +class MulticondLearnedConditioning: + def __init__(self, shape, batch): + self.shape: tuple = shape # the shape field is needed to send this object to DDIM/PLMS + self.batch: list[list[ComposableScheduledPromptConditioning]] = batch + + +def get_multicond_learned_conditioning(model, prompts, steps) -> MulticondLearnedConditioning: + """same as get_learned_conditioning, but returns a list of ScheduledPromptConditioning along with the weight objects for each prompt. + For each prompt, the list is obtained by splitting the prompt using the AND separator. + + https://energy-based-model.github.io/Compositional-Visual-Generation-with-Composable-Diffusion-Models/ + """ + + res_indexes, prompt_flat_list, prompt_indexes = get_multicond_prompt_list(prompts) + + learned_conditioning = get_learned_conditioning(model, prompt_flat_list, steps) + + res = [] + for indexes in res_indexes: + res.append([ComposableScheduledPromptConditioning(learned_conditioning[i], weight) for i, weight in indexes]) + + return MulticondLearnedConditioning(shape=(len(prompts),), batch=res) + + +def reconstruct_cond_batch(c: list[list[ScheduledPromptConditioning]], current_step): + param = c[0][0].cond + res = torch.zeros((len(c),) + param.shape, device=param.device, dtype=param.dtype) + for i, cond_schedule in enumerate(c): target_index = 0 for current, (end_at, cond) in enumerate(cond_schedule): if current_step <= end_at: @@ -140,6 +218,30 @@ def reconstruct_cond_batch(c: ScheduledPromptBatch, current_step): return res +def reconstruct_multicond_batch(c: MulticondLearnedConditioning, current_step): + param = c.batch[0][0].schedules[0].cond + + tensors = [] + conds_list = [] + + for batch_no, composable_prompts in enumerate(c.batch): + conds_for_batch = [] + + for cond_index, composable_prompt in enumerate(composable_prompts): + target_index = 0 + for current, (end_at, cond) in enumerate(composable_prompt.schedules): + if current_step <= end_at: + target_index = current + break + + conds_for_batch.append((len(tensors), composable_prompt.weight)) + tensors.append(composable_prompt.schedules[target_index].cond) + + conds_list.append(conds_for_batch) + + return conds_list, torch.stack(tensors).to(device=param.device, dtype=param.dtype) + + re_attention = re.compile(r""" \\\(| \\\)| diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index dbf570d2c..d27c547b3 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -109,9 +109,12 @@ class VanillaStableDiffusionSampler: return 0 def p_sample_ddim_hook(self, x_dec, cond, ts, unconditional_conditioning, *args, **kwargs): - cond = prompt_parser.reconstruct_cond_batch(cond, self.step) + conds_list, tensor = prompt_parser.reconstruct_multicond_batch(cond, self.step) unconditional_conditioning = prompt_parser.reconstruct_cond_batch(unconditional_conditioning, self.step) + assert all([len(conds) == 1 for conds in conds_list]), 'composition via AND is not supported for DDIM/PLMS samplers' + cond = tensor + if self.mask is not None: img_orig = self.sampler.model.q_sample(self.init_latent, ts) x_dec = img_orig * self.mask + self.nmask * x_dec @@ -183,19 +186,31 @@ class CFGDenoiser(torch.nn.Module): self.step = 0 def forward(self, x, sigma, uncond, cond, cond_scale): - cond = prompt_parser.reconstruct_cond_batch(cond, self.step) + conds_list, tensor = prompt_parser.reconstruct_multicond_batch(cond, self.step) uncond = prompt_parser.reconstruct_cond_batch(uncond, self.step) + batch_size = len(conds_list) + repeats = [len(conds_list[i]) for i in range(batch_size)] + + x_in = torch.cat([torch.stack([x[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [x]) + sigma_in = torch.cat([torch.stack([sigma[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [sigma]) + cond_in = torch.cat([tensor, uncond]) + if shared.batch_cond_uncond: - x_in = torch.cat([x] * 2) - sigma_in = torch.cat([sigma] * 2) - cond_in = torch.cat([uncond, cond]) - uncond, cond = self.inner_model(x_in, sigma_in, cond=cond_in).chunk(2) - denoised = uncond + (cond - uncond) * cond_scale + x_out = self.inner_model(x_in, sigma_in, cond=cond_in) else: - uncond = self.inner_model(x, sigma, cond=uncond) - cond = self.inner_model(x, sigma, cond=cond) - denoised = uncond + (cond - uncond) * cond_scale + x_out = torch.zeros_like(x_in) + for batch_offset in range(0, x_out.shape[0], batch_size): + a = batch_offset + b = a + batch_size + x_out[a:b] = self.inner_model(x_in[a:b], sigma_in[a:b], cond=cond_in[a:b]) + + denoised_uncond = x_out[-batch_size:] + denoised = torch.clone(denoised_uncond) + + for i, conds in enumerate(conds_list): + for cond_index, weight in conds: + denoised[i] += (x_out[cond_index] - denoised_uncond[i]) * (weight * cond_scale) if self.mask is not None: denoised = self.init_latent * self.mask + self.nmask * denoised diff --git a/modules/ui.py b/modules/ui.py index 523ab25b3..9620350fc 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -34,7 +34,7 @@ import modules.gfpgan_model import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste -from modules.prompt_parser import get_learned_conditioning_prompt_schedules +from modules import prompt_parser from modules.images import apply_filename_pattern, get_next_sequence_number import modules.textual_inversion.ui @@ -394,7 +394,9 @@ def connect_reuse_seed(seed: gr.Number, reuse_seed: gr.Button, generation_info: def update_token_counter(text, steps): try: - prompt_schedules = get_learned_conditioning_prompt_schedules([text], steps) + _, prompt_flat_list, _ = prompt_parser.get_multicond_prompt_list([text]) + prompt_schedules = prompt_parser.get_learned_conditioning_prompt_schedules(prompt_flat_list, steps) + except Exception: # a parsing error can happen here during typing, and we don't want to bother the user with # messages related to it in console From f8e41a96bb30a04dd5e294c7e1178c1c3b09d481 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 5 Oct 2022 23:52:05 +0300 Subject: [PATCH 110/138] fix various float parsing errors --- modules/prompt_parser.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index f7420daf9..800b12c75 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -143,8 +143,7 @@ def get_learned_conditioning(model, prompts, steps): re_AND = re.compile(r"\bAND\b") -re_weight = re.compile(r"^(.*?)(?:\s*:\s*([-+]?\s*(?:\d+|\d*\.\d+)?))?\s*$") - +re_weight = re.compile(r"^(.*?)(?:\s*:\s*([-+]?(?:\d+\.?|\d*\.\d+)))?\s*$") def get_multicond_prompt_list(prompts): res_indexes = [] From 20f8ec877a99ce2ebf193cb1e2e773cfc77b7c41 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 00:09:32 +0300 Subject: [PATCH 111/138] remove type annotations in new code because presumably they don't work in 3.7 --- modules/prompt_parser.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index 800b12c75..ee4c5d02d 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -175,14 +175,14 @@ def get_multicond_prompt_list(prompts): class ComposableScheduledPromptConditioning: def __init__(self, schedules, weight=1.0): - self.schedules: list[ScheduledPromptConditioning] = schedules + self.schedules = schedules # : list[ScheduledPromptConditioning] self.weight: float = weight class MulticondLearnedConditioning: def __init__(self, shape, batch): self.shape: tuple = shape # the shape field is needed to send this object to DDIM/PLMS - self.batch: list[list[ComposableScheduledPromptConditioning]] = batch + self.batch = batch # : list[list[ComposableScheduledPromptConditioning]] def get_multicond_learned_conditioning(model, prompts, steps) -> MulticondLearnedConditioning: @@ -203,7 +203,7 @@ def get_multicond_learned_conditioning(model, prompts, steps) -> MulticondLearne return MulticondLearnedConditioning(shape=(len(prompts),), batch=res) -def reconstruct_cond_batch(c: list[list[ScheduledPromptConditioning]], current_step): +def reconstruct_cond_batch(c, current_step): # c: list[list[ScheduledPromptConditioning]] param = c[0][0].cond res = torch.zeros((len(c),) + param.shape, device=param.device, dtype=param.dtype) for i, cond_schedule in enumerate(c): From 34c358d10d52817f7a889ae4c52096ee654f3fe6 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Wed, 5 Oct 2022 22:11:30 +0100 Subject: [PATCH 112/138] use typing.list in prompt_parser.py for wider python version support --- modules/prompt_parser.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index 800b12c75..fdfa21ae6 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -1,6 +1,6 @@ import re from collections import namedtuple - +from typing import List import lark # a prompt like this: "fantasy landscape with a [mountain:lake:0.25] and [an oak:a christmas tree:0.75][ in foreground::0.6][ in background:0.25] [shoddy:masterful:0.5]" @@ -175,14 +175,14 @@ def get_multicond_prompt_list(prompts): class ComposableScheduledPromptConditioning: def __init__(self, schedules, weight=1.0): - self.schedules: list[ScheduledPromptConditioning] = schedules + self.schedules: List[ScheduledPromptConditioning] = schedules self.weight: float = weight class MulticondLearnedConditioning: def __init__(self, shape, batch): self.shape: tuple = shape # the shape field is needed to send this object to DDIM/PLMS - self.batch: list[list[ComposableScheduledPromptConditioning]] = batch + self.batch: List[List[ComposableScheduledPromptConditioning]] = batch def get_multicond_learned_conditioning(model, prompts, steps) -> MulticondLearnedConditioning: @@ -203,7 +203,7 @@ def get_multicond_learned_conditioning(model, prompts, steps) -> MulticondLearne return MulticondLearnedConditioning(shape=(len(prompts),), batch=res) -def reconstruct_cond_batch(c: list[list[ScheduledPromptConditioning]], current_step): +def reconstruct_cond_batch(c: List[List[ScheduledPromptConditioning]], current_step): param = c[0][0].cond res = torch.zeros((len(c),) + param.shape, device=param.device, dtype=param.dtype) for i, cond_schedule in enumerate(c): From 55400c981b7c1389482057a35ed6ea11f08da194 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Thu, 6 Oct 2022 03:11:15 +0100 Subject: [PATCH 113/138] Set gradio-img2img-tool default to 'editor' --- modules/shared.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/shared.py b/modules/shared.py index e52c9b1d1..bab0fe6ee 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -55,7 +55,7 @@ parser.add_argument("--hide-ui-dir-config", action='store_true', help="hide dire parser.add_argument("--ui-settings-file", type=str, help="filename to use for ui settings", default=os.path.join(script_path, 'config.json')) parser.add_argument("--gradio-debug", action='store_true', help="launch gradio with --debug option") parser.add_argument("--gradio-auth", type=str, help='set gradio authentication like "username:password"; or comma-delimit multiple like "u1:p1,u2:p2,u3:p3"', default=None) -parser.add_argument("--gradio-img2img-tool", type=str, help='gradio image uploader tool: can be either editor for ctopping, or color-sketch for drawing', choices=["color-sketch", "editor"], default="color-sketch") +parser.add_argument("--gradio-img2img-tool", type=str, help='gradio image uploader tool: can be either editor for ctopping, or color-sketch for drawing', choices=["color-sketch", "editor"], default="editor") parser.add_argument("--opt-channelslast", action='store_true', help="change memory type for stable diffusion to channels last") parser.add_argument("--styles-file", type=str, help="filename to use for styles", default=os.path.join(script_path, 'styles.csv')) parser.add_argument("--autolaunch", action='store_true', help="open the webui URL in the system's default browser upon launch", default=False) From 2499fb4e1910d31ff12c24110f161b20641b8835 Mon Sep 17 00:00:00 2001 From: Raphael Stoeckli Date: Wed, 5 Oct 2022 21:57:18 +0200 Subject: [PATCH 114/138] Add sanitizer for captions in Textual inversion --- modules/textual_inversion/preprocess.py | 28 +++++++++++++++++++++++++ 1 file changed, 28 insertions(+) diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py index f545a9937..4f3df4bd9 100644 --- a/modules/textual_inversion/preprocess.py +++ b/modules/textual_inversion/preprocess.py @@ -1,5 +1,8 @@ +from cmath import log import os from PIL import Image, ImageOps +import platform +import sys import tqdm from modules import shared, images @@ -25,6 +28,7 @@ def preprocess(process_src, process_dst, process_flip, process_split, process_ca def save_pic_with_caption(image, index): if process_caption: caption = "-" + shared.interrogator.generate_caption(image) + caption = sanitize_caption(os.path.join(dst, f"{index:05}-{subindex[0]}"), caption, ".png") else: caption = filename caption = os.path.splitext(caption)[0] @@ -75,3 +79,27 @@ def preprocess(process_src, process_dst, process_flip, process_split, process_ca if process_caption: shared.interrogator.send_blip_to_ram() + +def sanitize_caption(base_path, original_caption, suffix): + operating_system = platform.system().lower() + if (operating_system == "windows"): + invalid_path_characters = "\\/:*?\"<>|" + max_path_length = 259 + else: + invalid_path_characters = "/" #linux/macos + max_path_length = 1023 + caption = original_caption + for invalid_character in invalid_path_characters: + caption = caption.replace(invalid_character, "") + fixed_path_length = len(base_path) + len(suffix) + if fixed_path_length + len(caption) <= max_path_length: + return caption + caption_tokens = caption.split() + new_caption = "" + for token in caption_tokens: + last_caption = new_caption + new_caption = new_caption + token + " " + if (len(new_caption) + fixed_path_length - 1 > max_path_length): + break + print(f"\nPath will be too long. Truncated caption: {original_caption}\nto: {last_caption}", file=sys.stderr) + return last_caption.strip() From 4288e53fc2ea25fa49715bf5b7f14603553c9e38 Mon Sep 17 00:00:00 2001 From: Raphael Stoeckli Date: Wed, 5 Oct 2022 23:11:32 +0200 Subject: [PATCH 115/138] removed unused import, fixed typo --- modules/textual_inversion/preprocess.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py index 4f3df4bd9..f1c002a2b 100644 --- a/modules/textual_inversion/preprocess.py +++ b/modules/textual_inversion/preprocess.py @@ -1,4 +1,3 @@ -from cmath import log import os from PIL import Image, ImageOps import platform @@ -13,7 +12,7 @@ def preprocess(process_src, process_dst, process_flip, process_split, process_ca src = os.path.abspath(process_src) dst = os.path.abspath(process_dst) - assert src != dst, 'same directory specified as source and desitnation' + assert src != dst, 'same directory specified as source and destination' os.makedirs(dst, exist_ok=True) From a93c3ffbfd264ed6b5d989922352300c9d3efbe4 Mon Sep 17 00:00:00 2001 From: Jocke Date: Wed, 5 Oct 2022 16:31:48 +0200 Subject: [PATCH 116/138] Outpainting mk2, prevent generation of a completely random image every time even when global seed is static --- scripts/outpainting_mk_2.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/scripts/outpainting_mk_2.py b/scripts/outpainting_mk_2.py index 11613ca36..a6468e09a 100644 --- a/scripts/outpainting_mk_2.py +++ b/scripts/outpainting_mk_2.py @@ -85,8 +85,11 @@ def get_matched_noise(_np_src_image, np_mask_rgb, noise_q=1, color_variation=0.0 src_dist = np.absolute(src_fft) src_phase = src_fft / src_dist + # create a generator with a static seed to make outpainting deterministic / only follow global seed + rng = np.random.default_rng(0) + 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_rgb = rng.random((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): From 6e7057b31b9762a9720282c7da486e4f264dee28 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 12:08:06 +0300 Subject: [PATCH 117/138] support for downloading new commit hash for git repos --- launch.py | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/launch.py b/launch.py index 57405feab..2f91f586c 100644 --- a/launch.py +++ b/launch.py @@ -86,6 +86,15 @@ def git_clone(url, dir, name, commithash=None): # TODO clone into temporary dir and move if successful if os.path.exists(dir): + if commithash is None: + return + + current_hash = run(f'"{git}" -C {dir} rev-parse HEAD', None, "Couldn't determine {name}'s hash: {commithash}").strip() + if current_hash == commithash: + return + + run(f'"{git}" -C {dir} fetch', f"Fetching updates for {name}...", f"Couldn't fetch {name}") + run(f'"{git}" -C {dir} checkout {commithash}', f"Checking out commint for {name} with hash: {commithash}...", f"Couldn't checkout commit {commithash} for {name}") return run(f'"{git}" clone "{url}" "{dir}"', f"Cloning {name} into {dir}...", f"Couldn't clone {name}") From 5f24b7bcf4a074fbdec757617fcd1bc82e76551b Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 12:08:48 +0300 Subject: [PATCH 118/138] option to let users select which samplers they want to hide --- modules/processing.py | 13 ++++++------- modules/sd_samplers.py | 19 +++++++++++++++++-- modules/shared.py | 15 +++++++++------ webui.py | 4 +++- 4 files changed, 35 insertions(+), 16 deletions(-) diff --git a/modules/processing.py b/modules/processing.py index d8c6b8d57..e01c8b3f6 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -11,9 +11,8 @@ import cv2 from skimage import exposure import modules.sd_hijack -from modules import devices, prompt_parser, masking +from modules import devices, prompt_parser, masking, sd_samplers from modules.sd_hijack import model_hijack -from modules.sd_samplers import samplers, samplers_for_img2img from modules.shared import opts, cmd_opts, state import modules.shared as shared import modules.face_restoration @@ -110,7 +109,7 @@ class Processed: self.width = p.width self.height = p.height self.sampler_index = p.sampler_index - self.sampler = samplers[p.sampler_index].name + self.sampler = sd_samplers.samplers[p.sampler_index].name self.cfg_scale = p.cfg_scale self.steps = p.steps self.batch_size = p.batch_size @@ -265,7 +264,7 @@ def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments, iteration generation_params = { "Steps": p.steps, - "Sampler": samplers[p.sampler_index].name, + "Sampler": sd_samplers.samplers[p.sampler_index].name, "CFG scale": p.cfg_scale, "Seed": all_seeds[index], "Face restoration": (opts.face_restoration_model if p.restore_faces else None), @@ -478,7 +477,7 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): self.firstphase_height_truncated = int(scale * self.height) def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength): - self.sampler = samplers[self.sampler_index].constructor(self.sd_model) + self.sampler = sd_samplers.samplers[self.sampler_index].constructor(self.sd_model) if not self.enable_hr: x = create_random_tensors([opt_C, self.height // opt_f, self.width // opt_f], seeds=seeds, subseeds=subseeds, subseed_strength=self.subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self) @@ -521,7 +520,7 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): shared.state.nextjob() - self.sampler = samplers[self.sampler_index].constructor(self.sd_model) + self.sampler = sd_samplers.samplers[self.sampler_index].constructor(self.sd_model) noise = create_random_tensors(samples.shape[1:], seeds=seeds, subseeds=subseeds, subseed_strength=subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self) # GC now before running the next img2img to prevent running out of memory @@ -556,7 +555,7 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing): self.nmask = None def init(self, all_prompts, all_seeds, all_subseeds): - self.sampler = samplers_for_img2img[self.sampler_index].constructor(self.sd_model) + self.sampler = sd_samplers.samplers_for_img2img[self.sampler_index].constructor(self.sd_model) crop_region = None if self.image_mask is not None: diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index d27c547b3..2e1f77153 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -32,12 +32,27 @@ samplers_data_k_diffusion = [ if hasattr(k_diffusion.sampling, funcname) ] -samplers = [ +all_samplers = [ *samplers_data_k_diffusion, SamplerData('DDIM', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.ddim.DDIMSampler, model), []), SamplerData('PLMS', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.plms.PLMSSampler, model), []), ] -samplers_for_img2img = [x for x in samplers if x.name not in ['PLMS', 'DPM fast', 'DPM adaptive']] + +samplers = [] +samplers_for_img2img = [] + + +def set_samplers(): + global samplers, samplers_for_img2img + + hidden = set(opts.hide_samplers) + hidden_img2img = set(opts.hide_samplers + ['PLMS', 'DPM fast', 'DPM adaptive']) + + samplers = [x for x in all_samplers if x.name not in hidden] + samplers_for_img2img = [x for x in all_samplers if x.name not in hidden_img2img] + + +set_samplers() sampler_extra_params = { 'sample_euler': ['s_churn', 's_tmin', 's_tmax', 's_noise'], diff --git a/modules/shared.py b/modules/shared.py index bab0fe6ee..ca2e4c742 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -13,6 +13,7 @@ import modules.memmon import modules.sd_models import modules.styles import modules.devices as devices +from modules import sd_samplers from modules.paths import script_path, sd_path sd_model_file = os.path.join(script_path, 'model.ckpt') @@ -238,14 +239,16 @@ options_templates.update(options_section(('ui', "User interface"), { })) options_templates.update(options_section(('sampler-params', "Sampler parameters"), { - "eta_ddim": OptionInfo(0.0, "eta (noise multiplier) for DDIM", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), - "eta_ancestral": OptionInfo(1.0, "eta (noise multiplier) for ancestral samplers", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), - "ddim_discretize": OptionInfo('uniform', "img2img DDIM discretize", gr.Radio, {"choices": ['uniform', 'quad']}), - 's_churn': OptionInfo(0.0, "sigma churn", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), - 's_tmin': OptionInfo(0.0, "sigma tmin", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), - 's_noise': OptionInfo(1.0, "sigma noise", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), + "hide_samplers": OptionInfo([], "Hide samplers in user interface (requires restart)", gr.CheckboxGroup, lambda: {"choices": [x.name for x in sd_samplers.all_samplers]}), + "eta_ddim": OptionInfo(0.0, "eta (noise multiplier) for DDIM", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), + "eta_ancestral": OptionInfo(1.0, "eta (noise multiplier) for ancestral samplers", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), + "ddim_discretize": OptionInfo('uniform', "img2img DDIM discretize", gr.Radio, {"choices": ['uniform', 'quad']}), + 's_churn': OptionInfo(0.0, "sigma churn", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), + 's_tmin': OptionInfo(0.0, "sigma tmin", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), + 's_noise': OptionInfo(1.0, "sigma noise", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), })) + class Options: data = None data_labels = options_templates diff --git a/webui.py b/webui.py index 47848ba58..9ef124274 100644 --- a/webui.py +++ b/webui.py @@ -2,7 +2,7 @@ import os import threading import time import importlib -from modules import devices +from modules import devices, sd_samplers from modules.paths import script_path import signal import threading @@ -109,6 +109,8 @@ def webui(): time.sleep(0.5) break + sd_samplers.set_samplers() + print('Reloading Custom Scripts') modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) print('Reloading modules: modules.ui') From 2d3ea42a2d1e909bbccdb6b49561b187c60a9402 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 13:21:12 +0300 Subject: [PATCH 119/138] workaround for a mysterious bug where prompt weights can't be matched --- modules/prompt_parser.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index a7a6aa314..f00256f28 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -156,7 +156,9 @@ def get_multicond_prompt_list(prompts): indexes = [] for subprompt in subprompts: - text, weight = re_weight.search(subprompt).groups() + match = re_weight.search(subprompt) + + text, weight = match.groups() if match is not None else (subprompt, 1.0) weight = float(weight) if weight is not None else 1.0 From 2a532804957e47bc36c67c8f5b104dcfa8e8f3f0 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 13:21:32 +0300 Subject: [PATCH 120/138] reorder imports to fix the bug with k-diffusion on some version --- webui.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/webui.py b/webui.py index 9ef124274..480360fe0 100644 --- a/webui.py +++ b/webui.py @@ -2,11 +2,12 @@ import os import threading import time import importlib -from modules import devices, sd_samplers -from modules.paths import script_path import signal import threading +from modules.paths import script_path + +from modules import devices, sd_samplers import modules.codeformer_model as codeformer import modules.extras import modules.face_restoration From c30c06db207a580d76544fd10fc1e03cd58ce85e Mon Sep 17 00:00:00 2001 From: C43H66N12O12S2 <36072735+C43H66N12O12S2@users.noreply.github.com> Date: Mon, 3 Oct 2022 12:48:16 +0300 Subject: [PATCH 121/138] update k-diffusion --- launch.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/launch.py b/launch.py index 2f91f586c..c2713c645 100644 --- a/launch.py +++ b/launch.py @@ -19,7 +19,7 @@ clip_package = os.environ.get('CLIP_PACKAGE', "git+https://github.com/openai/CLI stable_diffusion_commit_hash = os.environ.get('STABLE_DIFFUSION_COMMIT_HASH', "69ae4b35e0a0f6ee1af8bb9a5d0016ccb27e36dc") taming_transformers_commit_hash = os.environ.get('TAMING_TRANSFORMERS_COMMIT_HASH', "24268930bf1dce879235a7fddd0b2355b84d7ea6") -k_diffusion_commit_hash = os.environ.get('K_DIFFUSION_COMMIT_HASH', "a7ec1974d4ccb394c2dca275f42cd97490618924") +k_diffusion_commit_hash = os.environ.get('K_DIFFUSION_COMMIT_HASH', "567e11f7062ba20ae32b5a8cd07fb0fc4b9410cf") codeformer_commit_hash = os.environ.get('CODEFORMER_COMMIT_HASH', "c5b4593074ba6214284d6acd5f1719b6c5d739af") blip_commit_hash = os.environ.get('BLIP_COMMIT_HASH', "48211a1594f1321b00f14c9f7a5b4813144b2fb9") From c1a068ed0acc788774afc1541ca69342fd1d94ad Mon Sep 17 00:00:00 2001 From: C43H66N12O12S2 <36072735+C43H66N12O12S2@users.noreply.github.com> Date: Mon, 3 Oct 2022 12:49:17 +0300 Subject: [PATCH 122/138] Create alternate_sampler_noise_schedules.py --- scripts/alternate_sampler_noise_schedules.py | 53 ++++++++++++++++++++ 1 file changed, 53 insertions(+) create mode 100644 scripts/alternate_sampler_noise_schedules.py diff --git a/scripts/alternate_sampler_noise_schedules.py b/scripts/alternate_sampler_noise_schedules.py new file mode 100644 index 000000000..4f3ed8fb1 --- /dev/null +++ b/scripts/alternate_sampler_noise_schedules.py @@ -0,0 +1,53 @@ +import inspect +from modules.processing import Processed, process_images +import gradio as gr +import modules.scripts as scripts +import k_diffusion.sampling +import torch + + +class Script(scripts.Script): + + def title(self): + return "Alternate Sampler Noise Schedules" + + def ui(self, is_img2img): + noise_scheduler = gr.Dropdown(label="Noise Scheduler", choices=['Default','Karras','Exponential', 'Variance Preserving'], value='Default', type="index") + sched_smin = gr.Slider(value=0.1, label="Sigma min", minimum=0.0, maximum=100.0, step=0.5,) + sched_smax = gr.Slider(value=10.0, label="Sigma max", minimum=0.0, maximum=100.0, step=0.5) + sched_rho = gr.Slider(value=7.0, label="Sigma rho (Karras only)", minimum=7.0, maximum=100.0, step=0.5) + sched_beta_d = gr.Slider(value=19.9, label="Beta distribution (VP only)",minimum=0.0, maximum=40.0, step=0.5) + sched_beta_min = gr.Slider(value=0.1, label="Beta min (VP only)", minimum=0.0, maximum=40.0, step=0.1) + sched_eps_s = gr.Slider(value=0.001, label="Epsilon (VP only)", minimum=0.001, maximum=1.0, step=0.001) + + return [noise_scheduler, sched_smin, sched_smax, sched_rho, sched_beta_d, sched_beta_min, sched_eps_s] + + def run(self, p, noise_scheduler, sched_smin, sched_smax, sched_rho, sched_beta_d, sched_beta_min, sched_eps_s): + + noise_scheduler_func_name = ['-','get_sigmas_karras','get_sigmas_exponential','get_sigmas_vp'][noise_scheduler] + + base_params = { + "sigma_min":sched_smin, + "sigma_max":sched_smax, + "rho":sched_rho, + "beta_d":sched_beta_d, + "beta_min":sched_beta_min, + "eps_s":sched_eps_s, + "device":"cuda" if torch.cuda.is_available() else "cpu" + } + + if hasattr(k_diffusion.sampling,noise_scheduler_func_name): + + sigma_func = getattr(k_diffusion.sampling,noise_scheduler_func_name) + sigma_func_kwargs = {} + + for k,v in base_params.items(): + if k in inspect.signature(sigma_func).parameters: + sigma_func_kwargs[k] = v + + def substitute_noise_scheduler(n): + return sigma_func(n,**sigma_func_kwargs) + + p.sampler_noise_scheduler_override = substitute_noise_scheduler + + return process_images(p) From 71901b3d3bea1d035bf4a7229d19356b4b062151 Mon Sep 17 00:00:00 2001 From: C43H66N12O12S2 <36072735+C43H66N12O12S2@users.noreply.github.com> Date: Wed, 5 Oct 2022 14:30:57 +0300 Subject: [PATCH 123/138] add karras scheduling variants --- modules/sd_samplers.py | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index 2e1f77153..8d6eb7620 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -26,6 +26,17 @@ samplers_k_diffusion = [ ('DPM adaptive', 'sample_dpm_adaptive', ['k_dpm_ad']), ] +if opts.show_karras_scheduler_variants: + k_diffusion.sampling.sample_dpm_2_ka = k_diffusion.sampling.sample_dpm_2 + k_diffusion.sampling.sample_dpm_2_ancestral_ka = k_diffusion.sampling.sample_dpm_2_ancestral + k_diffusion.sampling.sample_lms_ka = k_diffusion.sampling.sample_lms + samplers_k_diffusion_ka = [ + ('LMS K Scheduling', 'sample_lms_ka', ['k_lms_ka']), + ('DPM2 K Scheduling', 'sample_dpm_2_ka', ['k_dpm_2_ka']), + ('DPM2 a K Scheduling', 'sample_dpm_2_ancestral_ka', ['k_dpm_2_a_ka']), + ] + samplers_k_diffusion.extend(samplers_k_diffusion_ka) + samplers_data_k_diffusion = [ SamplerData(label, lambda model, funcname=funcname: KDiffusionSampler(funcname, model), aliases) for label, funcname, aliases in samplers_k_diffusion @@ -345,6 +356,8 @@ class KDiffusionSampler: if p.sampler_noise_scheduler_override: sigmas = p.sampler_noise_scheduler_override(steps) + elif self.funcname.endswith('ka'): + sigmas = k_diffusion.sampling.get_sigmas_karras(n=steps, sigma_min=0.1, sigma_max=10, device=shared.device) else: sigmas = self.model_wrap.get_sigmas(steps) x = x * sigmas[0] From 3ddf80a9db8793188e2fe9488233d2b272cceb33 Mon Sep 17 00:00:00 2001 From: C43H66N12O12S2 <36072735+C43H66N12O12S2@users.noreply.github.com> Date: Wed, 5 Oct 2022 14:31:51 +0300 Subject: [PATCH 124/138] add variant setting --- modules/shared.py | 1 + 1 file changed, 1 insertion(+) diff --git a/modules/shared.py b/modules/shared.py index ca2e4c742..9e4860a28 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -236,6 +236,7 @@ options_templates.update(options_section(('ui', "User interface"), { "font": OptionInfo("", "Font for image grids that have text"), "js_modal_lightbox": OptionInfo(True, "Enable full page image viewer"), "js_modal_lightbox_initialy_zoomed": OptionInfo(True, "Show images zoomed in by default in full page image viewer"), + "show_karras_scheduler_variants": OptionInfo(True, "Show Karras scheduling variants for select samplers. Try these variants if your K sampled images suffer from excessive noise."), })) options_templates.update(options_section(('sampler-params', "Sampler parameters"), { From a971e4a767118ec41ec0f129770122babfb16a16 Mon Sep 17 00:00:00 2001 From: C43H66N12O12S2 <36072735+C43H66N12O12S2@users.noreply.github.com> Date: Thu, 6 Oct 2022 13:34:42 +0300 Subject: [PATCH 125/138] update k-diff once again --- launch.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/launch.py b/launch.py index c2713c645..9fe0fd675 100644 --- a/launch.py +++ b/launch.py @@ -19,7 +19,7 @@ clip_package = os.environ.get('CLIP_PACKAGE', "git+https://github.com/openai/CLI stable_diffusion_commit_hash = os.environ.get('STABLE_DIFFUSION_COMMIT_HASH', "69ae4b35e0a0f6ee1af8bb9a5d0016ccb27e36dc") taming_transformers_commit_hash = os.environ.get('TAMING_TRANSFORMERS_COMMIT_HASH', "24268930bf1dce879235a7fddd0b2355b84d7ea6") -k_diffusion_commit_hash = os.environ.get('K_DIFFUSION_COMMIT_HASH', "567e11f7062ba20ae32b5a8cd07fb0fc4b9410cf") +k_diffusion_commit_hash = os.environ.get('K_DIFFUSION_COMMIT_HASH', "f4e99857772fc3a126ba886aadf795a332774878") codeformer_commit_hash = os.environ.get('CODEFORMER_COMMIT_HASH', "c5b4593074ba6214284d6acd5f1719b6c5d739af") blip_commit_hash = os.environ.get('BLIP_COMMIT_HASH', "48211a1594f1321b00f14c9f7a5b4813144b2fb9") From 5993df24a1026225cb8af89237547c1d9101ce69 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 14:12:52 +0300 Subject: [PATCH 126/138] integrate the new samplers PR --- modules/processing.py | 7 ++- modules/sd_samplers.py | 59 ++++++++++---------- modules/shared.py | 1 - scripts/alternate_sampler_noise_schedules.py | 53 ------------------ scripts/img2imgalt.py | 3 +- 5 files changed, 36 insertions(+), 87 deletions(-) delete mode 100644 scripts/alternate_sampler_noise_schedules.py diff --git a/modules/processing.py b/modules/processing.py index e01c8b3f6..e567956ce 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -477,7 +477,7 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): self.firstphase_height_truncated = int(scale * self.height) def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength): - self.sampler = sd_samplers.samplers[self.sampler_index].constructor(self.sd_model) + self.sampler = sd_samplers.create_sampler_with_index(sd_samplers.samplers, self.sampler_index, self.sd_model) if not self.enable_hr: x = create_random_tensors([opt_C, self.height // opt_f, self.width // opt_f], seeds=seeds, subseeds=subseeds, subseed_strength=self.subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self) @@ -520,7 +520,8 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): shared.state.nextjob() - self.sampler = sd_samplers.samplers[self.sampler_index].constructor(self.sd_model) + self.sampler = sd_samplers.create_sampler_with_index(sd_samplers.samplers, self.sampler_index, self.sd_model) + noise = create_random_tensors(samples.shape[1:], seeds=seeds, subseeds=subseeds, subseed_strength=subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self) # GC now before running the next img2img to prevent running out of memory @@ -555,7 +556,7 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing): self.nmask = None def init(self, all_prompts, all_seeds, all_subseeds): - self.sampler = sd_samplers.samplers_for_img2img[self.sampler_index].constructor(self.sd_model) + self.sampler = sd_samplers.create_sampler_with_index(sd_samplers.samplers_for_img2img, self.sampler_index, self.sd_model) crop_region = None if self.image_mask is not None: diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index 8d6eb7620..497df9430 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -13,46 +13,46 @@ from modules.shared import opts, cmd_opts, state import modules.shared as shared -SamplerData = namedtuple('SamplerData', ['name', 'constructor', 'aliases']) +SamplerData = namedtuple('SamplerData', ['name', 'constructor', 'aliases', 'options']) samplers_k_diffusion = [ - ('Euler a', 'sample_euler_ancestral', ['k_euler_a']), - ('Euler', 'sample_euler', ['k_euler']), - ('LMS', 'sample_lms', ['k_lms']), - ('Heun', 'sample_heun', ['k_heun']), - ('DPM2', 'sample_dpm_2', ['k_dpm_2']), - ('DPM2 a', 'sample_dpm_2_ancestral', ['k_dpm_2_a']), - ('DPM fast', 'sample_dpm_fast', ['k_dpm_fast']), - ('DPM adaptive', 'sample_dpm_adaptive', ['k_dpm_ad']), + ('Euler a', 'sample_euler_ancestral', ['k_euler_a'], {}), + ('Euler', 'sample_euler', ['k_euler'], {}), + ('LMS', 'sample_lms', ['k_lms'], {}), + ('Heun', 'sample_heun', ['k_heun'], {}), + ('DPM2', 'sample_dpm_2', ['k_dpm_2'], {}), + ('DPM2 a', 'sample_dpm_2_ancestral', ['k_dpm_2_a'], {}), + ('DPM fast', 'sample_dpm_fast', ['k_dpm_fast'], {}), + ('DPM adaptive', 'sample_dpm_adaptive', ['k_dpm_ad'], {}), + ('LMS Karras', 'sample_lms', ['k_lms_ka'], {'scheduler': 'karras'}), + ('DPM2 Karras', 'sample_dpm_2', ['k_dpm_2_ka'], {'scheduler': 'karras'}), + ('DPM2 a Karras', 'sample_dpm_2_ancestral', ['k_dpm_2_a_ka'], {'scheduler': 'karras'}), ] -if opts.show_karras_scheduler_variants: - k_diffusion.sampling.sample_dpm_2_ka = k_diffusion.sampling.sample_dpm_2 - k_diffusion.sampling.sample_dpm_2_ancestral_ka = k_diffusion.sampling.sample_dpm_2_ancestral - k_diffusion.sampling.sample_lms_ka = k_diffusion.sampling.sample_lms - samplers_k_diffusion_ka = [ - ('LMS K Scheduling', 'sample_lms_ka', ['k_lms_ka']), - ('DPM2 K Scheduling', 'sample_dpm_2_ka', ['k_dpm_2_ka']), - ('DPM2 a K Scheduling', 'sample_dpm_2_ancestral_ka', ['k_dpm_2_a_ka']), - ] - samplers_k_diffusion.extend(samplers_k_diffusion_ka) - samplers_data_k_diffusion = [ - SamplerData(label, lambda model, funcname=funcname: KDiffusionSampler(funcname, model), aliases) - for label, funcname, aliases in samplers_k_diffusion + SamplerData(label, lambda model, funcname=funcname: KDiffusionSampler(funcname, model), aliases, options) + for label, funcname, aliases, options in samplers_k_diffusion if hasattr(k_diffusion.sampling, funcname) ] all_samplers = [ *samplers_data_k_diffusion, - SamplerData('DDIM', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.ddim.DDIMSampler, model), []), - SamplerData('PLMS', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.plms.PLMSSampler, model), []), + SamplerData('DDIM', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.ddim.DDIMSampler, model), [], {}), + SamplerData('PLMS', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.plms.PLMSSampler, model), [], {}), ] samplers = [] samplers_for_img2img = [] +def create_sampler_with_index(list_of_configs, index, model): + config = list_of_configs[index] + sampler = config.constructor(model) + sampler.config = config + + return sampler + + def set_samplers(): global samplers, samplers_for_img2img @@ -130,6 +130,7 @@ class VanillaStableDiffusionSampler: self.step = 0 self.eta = None self.default_eta = 0.0 + self.config = None def number_of_needed_noises(self, p): return 0 @@ -291,6 +292,7 @@ class KDiffusionSampler: self.stop_at = None self.eta = None self.default_eta = 1.0 + self.config = None def callback_state(self, d): store_latent(d["denoised"]) @@ -355,11 +357,12 @@ class KDiffusionSampler: steps = steps or p.steps if p.sampler_noise_scheduler_override: - sigmas = p.sampler_noise_scheduler_override(steps) - elif self.funcname.endswith('ka'): - sigmas = k_diffusion.sampling.get_sigmas_karras(n=steps, sigma_min=0.1, sigma_max=10, device=shared.device) + sigmas = p.sampler_noise_scheduler_override(steps) + elif self.config is not None and self.config.options.get('scheduler', None) == 'karras': + sigmas = k_diffusion.sampling.get_sigmas_karras(n=steps, sigma_min=0.1, sigma_max=10, device=shared.device) else: - sigmas = self.model_wrap.get_sigmas(steps) + sigmas = self.model_wrap.get_sigmas(steps) + x = x * sigmas[0] extra_params_kwargs = self.initialize(p) diff --git a/modules/shared.py b/modules/shared.py index 9e4860a28..ca2e4c742 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -236,7 +236,6 @@ options_templates.update(options_section(('ui', "User interface"), { "font": OptionInfo("", "Font for image grids that have text"), "js_modal_lightbox": OptionInfo(True, "Enable full page image viewer"), "js_modal_lightbox_initialy_zoomed": OptionInfo(True, "Show images zoomed in by default in full page image viewer"), - "show_karras_scheduler_variants": OptionInfo(True, "Show Karras scheduling variants for select samplers. Try these variants if your K sampled images suffer from excessive noise."), })) options_templates.update(options_section(('sampler-params', "Sampler parameters"), { diff --git a/scripts/alternate_sampler_noise_schedules.py b/scripts/alternate_sampler_noise_schedules.py deleted file mode 100644 index 4f3ed8fb1..000000000 --- a/scripts/alternate_sampler_noise_schedules.py +++ /dev/null @@ -1,53 +0,0 @@ -import inspect -from modules.processing import Processed, process_images -import gradio as gr -import modules.scripts as scripts -import k_diffusion.sampling -import torch - - -class Script(scripts.Script): - - def title(self): - return "Alternate Sampler Noise Schedules" - - def ui(self, is_img2img): - noise_scheduler = gr.Dropdown(label="Noise Scheduler", choices=['Default','Karras','Exponential', 'Variance Preserving'], value='Default', type="index") - sched_smin = gr.Slider(value=0.1, label="Sigma min", minimum=0.0, maximum=100.0, step=0.5,) - sched_smax = gr.Slider(value=10.0, label="Sigma max", minimum=0.0, maximum=100.0, step=0.5) - sched_rho = gr.Slider(value=7.0, label="Sigma rho (Karras only)", minimum=7.0, maximum=100.0, step=0.5) - sched_beta_d = gr.Slider(value=19.9, label="Beta distribution (VP only)",minimum=0.0, maximum=40.0, step=0.5) - sched_beta_min = gr.Slider(value=0.1, label="Beta min (VP only)", minimum=0.0, maximum=40.0, step=0.1) - sched_eps_s = gr.Slider(value=0.001, label="Epsilon (VP only)", minimum=0.001, maximum=1.0, step=0.001) - - return [noise_scheduler, sched_smin, sched_smax, sched_rho, sched_beta_d, sched_beta_min, sched_eps_s] - - def run(self, p, noise_scheduler, sched_smin, sched_smax, sched_rho, sched_beta_d, sched_beta_min, sched_eps_s): - - noise_scheduler_func_name = ['-','get_sigmas_karras','get_sigmas_exponential','get_sigmas_vp'][noise_scheduler] - - base_params = { - "sigma_min":sched_smin, - "sigma_max":sched_smax, - "rho":sched_rho, - "beta_d":sched_beta_d, - "beta_min":sched_beta_min, - "eps_s":sched_eps_s, - "device":"cuda" if torch.cuda.is_available() else "cpu" - } - - if hasattr(k_diffusion.sampling,noise_scheduler_func_name): - - sigma_func = getattr(k_diffusion.sampling,noise_scheduler_func_name) - sigma_func_kwargs = {} - - for k,v in base_params.items(): - if k in inspect.signature(sigma_func).parameters: - sigma_func_kwargs[k] = v - - def substitute_noise_scheduler(n): - return sigma_func(n,**sigma_func_kwargs) - - p.sampler_noise_scheduler_override = substitute_noise_scheduler - - return process_images(p) diff --git a/scripts/img2imgalt.py b/scripts/img2imgalt.py index 0ef137f7d..f9894cb01 100644 --- a/scripts/img2imgalt.py +++ b/scripts/img2imgalt.py @@ -8,7 +8,6 @@ import gradio as gr from modules import processing, shared, sd_samplers, prompt_parser from modules.processing import Processed -from modules.sd_samplers import samplers from modules.shared import opts, cmd_opts, state import torch @@ -159,7 +158,7 @@ class Script(scripts.Script): combined_noise = ((1 - randomness) * rec_noise + randomness * rand_noise) / ((randomness**2 + (1-randomness)**2) ** 0.5) - sampler = samplers[p.sampler_index].constructor(p.sd_model) + sampler = sd_samplers.create_sampler_with_index(sd_samplers.samplers, p.sampler_index, p.sd_model) sigmas = sampler.model_wrap.get_sigmas(p.steps) From f5490674a8fd84162b4e80c045e675633afb9ee7 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 17:41:49 +0300 Subject: [PATCH 127/138] fix bad output for error when updating a git repo --- launch.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/launch.py b/launch.py index 9fe0fd675..75edb66a9 100644 --- a/launch.py +++ b/launch.py @@ -89,7 +89,7 @@ def git_clone(url, dir, name, commithash=None): if commithash is None: return - current_hash = run(f'"{git}" -C {dir} rev-parse HEAD', None, "Couldn't determine {name}'s hash: {commithash}").strip() + current_hash = run(f'"{git}" -C {dir} rev-parse HEAD', None, f"Couldn't determine {name}'s hash: {commithash}").strip() if current_hash == commithash: return From be71115b1a1201d04f0e2a11e718fb31cbd26474 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Thu, 6 Oct 2022 01:09:44 +0100 Subject: [PATCH 128/138] Update shared.py --- modules/shared.py | 1 + 1 file changed, 1 insertion(+) diff --git a/modules/shared.py b/modules/shared.py index ca2e4c742..9f7c6efe5 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -236,6 +236,7 @@ options_templates.update(options_section(('ui', "User interface"), { "font": OptionInfo("", "Font for image grids that have text"), "js_modal_lightbox": OptionInfo(True, "Enable full page image viewer"), "js_modal_lightbox_initialy_zoomed": OptionInfo(True, "Show images zoomed in by default in full page image viewer"), + "show_progress_in_title": OptionInfo(False, "Show generation progress in window title."), })) options_templates.update(options_section(('sampler-params', "Sampler parameters"), { From c06298d1d003aa034007978ee7508af636c18124 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Thu, 6 Oct 2022 01:10:38 +0100 Subject: [PATCH 129/138] add check for progress in title setting --- javascript/progressbar.js | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/javascript/progressbar.js b/javascript/progressbar.js index 3e3220c3f..f9e9290e2 100644 --- a/javascript/progressbar.js +++ b/javascript/progressbar.js @@ -5,7 +5,7 @@ function check_progressbar(id_part, id_progressbar, id_progressbar_span, id_inte var progressbar = gradioApp().getElementById(id_progressbar) var interrupt = gradioApp().getElementById(id_interrupt) - if(progressbar && progressbar.offsetParent){ + if(opts.show_progress_in_title && progressbar && progressbar.offsetParent){ if(progressbar.innerText){ let newtitle = 'Stable Diffusion - ' + progressbar.innerText if(document.title != newtitle){ From fec71e4de24b65b0f205a3c071b71651bbcb0dfc Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Thu, 6 Oct 2022 01:35:07 +0100 Subject: [PATCH 130/138] Default window title progress updates on --- modules/shared.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/modules/shared.py b/modules/shared.py index 9f7c6efe5..5c16f0257 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -236,7 +236,7 @@ options_templates.update(options_section(('ui', "User interface"), { "font": OptionInfo("", "Font for image grids that have text"), "js_modal_lightbox": OptionInfo(True, "Enable full page image viewer"), "js_modal_lightbox_initialy_zoomed": OptionInfo(True, "Show images zoomed in by default in full page image viewer"), - "show_progress_in_title": OptionInfo(False, "Show generation progress in window title."), + "show_progress_in_title": OptionInfo(True, "Show generation progress in window title."), })) options_templates.update(options_section(('sampler-params', "Sampler parameters"), { From 5d0e6ab8567bda2ee8f5ed31f332ca07c1b84b98 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Thu, 6 Oct 2022 04:04:50 +0100 Subject: [PATCH 131/138] Allow escaping of commas in xy_grid --- scripts/xy_grid.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 1237e754d..210829a79 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -168,6 +168,7 @@ re_range_float = re.compile(r"\s*([+-]?\s*\d+(?:.\d*)?)\s*-\s*([+-]?\s*\d+(?:.\d re_range_count = re.compile(r"\s*([+-]?\s*\d+)\s*-\s*([+-]?\s*\d+)(?:\s*\[(\d+)\s*\])?\s*") re_range_count_float = re.compile(r"\s*([+-]?\s*\d+(?:.\d*)?)\s*-\s*([+-]?\s*\d+(?:.\d*)?)(?:\s*\[(\d+(?:.\d*)?)\s*\])?\s*") +re_non_escaped_comma = re.compile(r"(? Date: Thu, 6 Oct 2022 11:55:21 +0100 Subject: [PATCH 132/138] use csv.reader --- scripts/xy_grid.py | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 210829a79..1a625898f 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -1,8 +1,9 @@ from collections import namedtuple from copy import copy -from itertools import permutations +from itertools import permutations, chain import random - +import csv +from io import StringIO from PIL import Image import numpy as np @@ -168,8 +169,6 @@ re_range_float = re.compile(r"\s*([+-]?\s*\d+(?:.\d*)?)\s*-\s*([+-]?\s*\d+(?:.\d re_range_count = re.compile(r"\s*([+-]?\s*\d+)\s*-\s*([+-]?\s*\d+)(?:\s*\[(\d+)\s*\])?\s*") re_range_count_float = re.compile(r"\s*([+-]?\s*\d+(?:.\d*)?)\s*-\s*([+-]?\s*\d+(?:.\d*)?)(?:\s*\[(\d+(?:.\d*)?)\s*\])?\s*") -re_non_escaped_comma = re.compile(r"(? Date: Thu, 6 Oct 2022 12:32:17 +0100 Subject: [PATCH 133/138] strip() split comma delimited lines --- scripts/xy_grid.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 1a625898f..ec27e58bc 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -197,7 +197,7 @@ class Script(scripts.Script): if opt.label == 'Nothing': return [0] - valslist = list(chain.from_iterable(csv.reader(StringIO(s)))) + valslist = list(map(str.strip,chain.from_iterable(csv.reader(StringIO(s))))) if opt.type == int: valslist_ext = [] From 82eb8ea452b1e63535c58d15ec6db2ad2342faa8 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Thu, 6 Oct 2022 15:22:51 +0100 Subject: [PATCH 134/138] Update xy_grid.py split vals not 's' from tests --- scripts/xy_grid.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index ec27e58bc..210c7b6e9 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -197,7 +197,7 @@ class Script(scripts.Script): if opt.label == 'Nothing': return [0] - valslist = list(map(str.strip,chain.from_iterable(csv.reader(StringIO(s))))) + valslist = list(map(str.strip,chain.from_iterable(csv.reader(StringIO(vals))))) if opt.type == int: valslist_ext = [] From 0bb458f0ca06a7be27cf1a1003c536d1f06a5bd3 Mon Sep 17 00:00:00 2001 From: Milly Date: Wed, 5 Oct 2022 01:19:50 +0900 Subject: [PATCH 135/138] Removed duplicate image saving codes Use `modules.images.save_image()` instead. --- modules/images.py | 7 ++++--- modules/ui.py | 46 ++++++++++------------------------------------ 2 files changed, 14 insertions(+), 39 deletions(-) diff --git a/modules/images.py b/modules/images.py index c2fadab99..810f1446e 100644 --- a/modules/images.py +++ b/modules/images.py @@ -353,7 +353,7 @@ def get_next_sequence_number(path, basename): return result + 1 -def save_image(image, path, basename, seed=None, prompt=None, extension='png', info=None, short_filename=False, no_prompt=False, grid=False, pnginfo_section_name='parameters', p=None, existing_info=None, forced_filename=None, suffix=""): +def save_image(image, path, basename, seed=None, prompt=None, extension='png', info=None, short_filename=False, no_prompt=False, grid=False, pnginfo_section_name='parameters', p=None, existing_info=None, forced_filename=None, suffix="", save_to_dirs=None): if short_filename or prompt is None or seed is None: file_decoration = "" elif opts.save_to_dirs: @@ -377,7 +377,8 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i else: pnginfo = None - save_to_dirs = (grid and opts.grid_save_to_dirs) or (not grid and opts.save_to_dirs and not no_prompt) + if save_to_dirs is None: + save_to_dirs = (grid and opts.grid_save_to_dirs) or (not grid and opts.save_to_dirs and not no_prompt) if save_to_dirs: dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, seed, prompt).strip('\\ /') @@ -431,4 +432,4 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i with open(f"{fullfn_without_extension}.txt", "w", encoding="utf8") as file: file.write(info + "\n") - + return fullfn diff --git a/modules/ui.py b/modules/ui.py index 9620350fc..4f18126fb 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -35,7 +35,7 @@ import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste from modules import prompt_parser -from modules.images import apply_filename_pattern, get_next_sequence_number +from modules.images import save_image import modules.textual_inversion.ui # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI @@ -114,20 +114,13 @@ def save_files(js_data, images, index): p = MyObject(data) path = opts.outdir_save save_to_dirs = opts.use_save_to_dirs_for_ui - - if save_to_dirs: - dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, p.seed, p.prompt) - path = os.path.join(opts.outdir_save, dirname) - - os.makedirs(path, exist_ok=True) - + extension: str = opts.samples_format + start_index = 0 if index > -1 and opts.save_selected_only and (index >= data["index_of_first_image"]): # ensures we are looking at a specific non-grid picture, and we have save_selected_only images = [images[index]] - infotexts = [data["infotexts"][index]] - else: - infotexts = data["infotexts"] + start_index = index with open(os.path.join(opts.outdir_save, "log.csv"), "a", encoding="utf8", newline='') as file: at_start = file.tell() == 0 @@ -135,37 +128,18 @@ def save_files(js_data, images, index): if at_start: writer.writerow(["prompt", "seed", "width", "height", "sampler", "cfgs", "steps", "filename", "negative_prompt"]) - file_decoration = opts.samples_filename_pattern or "[seed]-[prompt_spaces]" - if file_decoration != "": - file_decoration = "-" + file_decoration.lower() - file_decoration = apply_filename_pattern(file_decoration, p, p.seed, p.prompt) - truncated = (file_decoration[:240] + '..') if len(file_decoration) > 240 else file_decoration - filename_base = truncated - extension = opts.samples_format.lower() - - basecount = get_next_sequence_number(path, "") - for i, filedata in enumerate(images): - file_number = f"{basecount+i:05}" - filename = file_number + filename_base + f".{extension}" - filepath = os.path.join(path, filename) - - + for image_index, filedata in enumerate(images, start_index): if filedata.startswith("data:image/png;base64,"): filedata = filedata[len("data:image/png;base64,"):] image = Image.open(io.BytesIO(base64.decodebytes(filedata.encode('utf-8')))) - if opts.enable_pnginfo and extension == 'png': - pnginfo = PngImagePlugin.PngInfo() - pnginfo.add_text('parameters', infotexts[i]) - image.save(filepath, pnginfo=pnginfo) - else: - image.save(filepath, quality=opts.jpeg_quality) - if opts.enable_pnginfo and extension in ("jpg", "jpeg", "webp"): - piexif.insert(piexif.dump({"Exif": { - piexif.ExifIFD.UserComment: piexif.helper.UserComment.dump(infotexts[i], encoding="unicode") - }}), filepath) + is_grid = image_index < p.index_of_first_image + i = 0 if is_grid else (image_index - p.index_of_first_image) + fullfn = save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) + + filename = os.path.relpath(fullfn, path) filenames.append(filename) writer.writerow([data["prompt"], data["seed"], data["width"], data["height"], data["sampler"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]]) From 1069ec49a35d04c1e85c92534e92a2d6aa59cb75 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 20:16:21 +0300 Subject: [PATCH 136/138] revert back to using list comprehension rather than list and map --- scripts/xy_grid.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 210c7b6e9..6344e612f 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -197,7 +197,7 @@ class Script(scripts.Script): if opt.label == 'Nothing': return [0] - valslist = list(map(str.strip,chain.from_iterable(csv.reader(StringIO(vals))))) + valslist = [x.strip() for x in chain.from_iterable(csv.reader(StringIO(vals)))] if opt.type == int: valslist_ext = [] From dbc8a4d35129b08eab30776bbbaf3a2e7ac10a6c Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 6 Oct 2022 20:27:50 +0300 Subject: [PATCH 137/138] add generation parameters to images shown in web ui --- modules/processing.py | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/modules/processing.py b/modules/processing.py index de818d5b9..8faf90956 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -430,7 +430,9 @@ def process_images(p: StableDiffusionProcessing) -> Processed: if opts.samples_save and not p.do_not_save_samples: images.save_image(image, p.outpath_samples, "", seeds[i], prompts[i], opts.samples_format, info=infotext(n, i), p=p) - infotexts.append(infotext(n, i)) + text = infotext(n, i) + infotexts.append(text) + image.info["parameters"] = text output_images.append(image) del x_samples_ddim @@ -447,7 +449,9 @@ def process_images(p: StableDiffusionProcessing) -> Processed: grid = images.image_grid(output_images, p.batch_size) if opts.return_grid: - infotexts.insert(0, infotext()) + text = infotext() + infotexts.insert(0, text) + grid.info["parameters"] = text output_images.insert(0, grid) index_of_first_image = 1 From cf7c784fcc0c84a8a4edd8d3aca4dda4c7025c43 Mon Sep 17 00:00:00 2001 From: Milly Date: Fri, 7 Oct 2022 00:19:52 +0900 Subject: [PATCH 138/138] Removed duplicate defined models_path Use `modules.paths.models_path` instead `modules.shared.model_path`. --- modules/shared.py | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/modules/shared.py b/modules/shared.py index 5c16f0257..25bb6e6c9 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -14,11 +14,10 @@ import modules.sd_models import modules.styles import modules.devices as devices from modules import sd_samplers -from modules.paths import script_path, sd_path +from modules.paths import models_path, script_path, sd_path sd_model_file = os.path.join(script_path, 'model.ckpt') default_sd_model_file = sd_model_file -model_path = os.path.join(script_path, 'models') 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=sd_model_file, help="path to checkpoint of stable diffusion model; if specified, this checkpoint will be added to the list of checkpoints and loaded",) @@ -36,14 +35,14 @@ parser.add_argument("--always-batch-cond-uncond", action='store_true', help="dis parser.add_argument("--unload-gfpgan", action='store_true', help="does not do anything.") parser.add_argument("--precision", type=str, help="evaluate at this precision", choices=["full", "autocast"], default="autocast") parser.add_argument("--share", action='store_true', help="use share=True for gradio and make the UI accessible through their site (doesn't work for me but you might have better luck)") -parser.add_argument("--codeformer-models-path", type=str, help="Path to directory with codeformer model file(s).", default=os.path.join(model_path, 'Codeformer')) -parser.add_argument("--gfpgan-models-path", type=str, help="Path to directory with GFPGAN model file(s).", default=os.path.join(model_path, 'GFPGAN')) -parser.add_argument("--esrgan-models-path", type=str, help="Path to directory with ESRGAN model file(s).", default=os.path.join(model_path, 'ESRGAN')) -parser.add_argument("--bsrgan-models-path", type=str, help="Path to directory with BSRGAN model file(s).", default=os.path.join(model_path, 'BSRGAN')) -parser.add_argument("--realesrgan-models-path", type=str, help="Path to directory with RealESRGAN model file(s).", default=os.path.join(model_path, 'RealESRGAN')) -parser.add_argument("--scunet-models-path", type=str, help="Path to directory with ScuNET model file(s).", default=os.path.join(model_path, 'ScuNET')) -parser.add_argument("--swinir-models-path", type=str, help="Path to directory with SwinIR model file(s).", default=os.path.join(model_path, 'SwinIR')) -parser.add_argument("--ldsr-models-path", type=str, help="Path to directory with LDSR model file(s).", default=os.path.join(model_path, 'LDSR')) +parser.add_argument("--codeformer-models-path", type=str, help="Path to directory with codeformer model file(s).", default=os.path.join(models_path, 'Codeformer')) +parser.add_argument("--gfpgan-models-path", type=str, help="Path to directory with GFPGAN model file(s).", default=os.path.join(models_path, 'GFPGAN')) +parser.add_argument("--esrgan-models-path", type=str, help="Path to directory with ESRGAN model file(s).", default=os.path.join(models_path, 'ESRGAN')) +parser.add_argument("--bsrgan-models-path", type=str, help="Path to directory with BSRGAN model file(s).", default=os.path.join(models_path, 'BSRGAN')) +parser.add_argument("--realesrgan-models-path", type=str, help="Path to directory with RealESRGAN model file(s).", default=os.path.join(models_path, 'RealESRGAN')) +parser.add_argument("--scunet-models-path", type=str, help="Path to directory with ScuNET model file(s).", default=os.path.join(models_path, 'ScuNET')) +parser.add_argument("--swinir-models-path", type=str, help="Path to directory with SwinIR model file(s).", default=os.path.join(models_path, 'SwinIR')) +parser.add_argument("--ldsr-models-path", type=str, help="Path to directory with LDSR model file(s).", default=os.path.join(models_path, 'LDSR')) parser.add_argument("--opt-split-attention", action='store_true', help="force-enables cross-attention layer optimization. By default, it's on for torch.cuda and off for other torch devices.") parser.add_argument("--disable-opt-split-attention", action='store_true', help="force-disables cross-attention layer optimization") parser.add_argument("--opt-split-attention-v1", action='store_true', help="enable older version of split attention optimization that does not consume all the VRAM it can find")