From 873371f1e5c26457f87871c576737664fe58b5e0 Mon Sep 17 00:00:00 2001 From: Vladimir Mandic Date: Tue, 30 May 2023 14:13:29 -0400 Subject: [PATCH] merge stage one --- .eslintignore | 4 + .eslintrc.json | 89 +++++++++++++++++++ CHANGELOG.md | 16 ++-- extensions-builtin/LDSR/ldsr_model_arch.py | 15 ++-- extensions-builtin/LDSR/preload.py | 6 ++ extensions-builtin/LDSR/scripts/ldsr_model.py | 7 +- .../LDSR/sd_hijack_autoencoder.py | 28 +++--- extensions-builtin/LDSR/sd_hijack_ddpm_v1.py | 66 +++++++------- .../Lora/extra_networks_lora.py | 18 ++++ extensions-builtin/Lora/lora.py | 81 +++++++++++++---- .../Lora/scripts/lora_script.py | 43 ++++++++- .../Lora/ui_extra_networks_lora.py | 8 +- extensions-builtin/ScuNET/preload.py | 6 ++ .../ScuNET/scripts/scunet_model.py | 19 ++-- .../ScuNET/scunet_model_arch.py | 11 ++- extensions-builtin/SwinIR/preload.py | 6 ++ .../SwinIR/scripts/swinir_model.py | 33 ++++--- .../SwinIR/swinir_model_arch.py | 6 +- .../SwinIR/swinir_model_arch_v2.py | 58 ++++++------ .../javascript/prompt-bracket-checker.js | 52 +++++------ html/licenses.html | 26 ++++++ installer.py | 2 +- javascript/.eslintrc.json | 22 ----- javascript/hints.js | 7 ++ modules/hashes.py | 43 ++++++--- modules/sd_samplers.py | 7 +- modules/sd_samplers_common.py | 31 +++++-- modules/sd_samplers_compvis.py | 11 ++- modules/sd_samplers_kdiffusion.py | 79 ++++++++-------- modules/sd_vae_taesd.py | 88 ++++++++++++++++++ modules/shared.py | 26 +++++- 31 files changed, 647 insertions(+), 267 deletions(-) create mode 100644 .eslintignore create mode 100644 .eslintrc.json create mode 100644 extensions-builtin/LDSR/preload.py create mode 100644 extensions-builtin/ScuNET/preload.py create mode 100644 extensions-builtin/SwinIR/preload.py delete mode 100644 javascript/.eslintrc.json create mode 100644 modules/sd_vae_taesd.py diff --git a/.eslintignore b/.eslintignore new file mode 100644 index 000000000..c098feaa7 --- /dev/null +++ b/.eslintignore @@ -0,0 +1,4 @@ +extensions +extensions-disabled +repositories +venv diff --git a/.eslintrc.json b/.eslintrc.json new file mode 100644 index 000000000..2958fbc79 --- /dev/null +++ b/.eslintrc.json @@ -0,0 +1,89 @@ +{ + "env": { + "browser": true, + "commonjs": false, + "node": false, + "jquery": false, + "es2020": true + }, + "parserOptions": { "ecmaVersion": 2020 }, + "plugins": [], + "extends": ["eslint:recommended", "airbnb-base"], + "rules": { + "arrow-spacing": "error", + "block-spacing": "error", + "brace-style": "error", + "comma-dangle": ["error", "only-multiline"], + "comma-spacing": "error", + "comma-style": ["error", "last"], + "curly": ["error", "multi-line", "consistent"], + "eol-last": "error", + "func-call-spacing": "error", + "function-call-argument-newline": ["error", "consistent"], + "function-paren-newline": ["error", "consistent"], + "indent": ["error", 4], + "key-spacing": "error", + "keyword-spacing": "error", + "linebreak-style": ["error", "unix"], + "no-extra-semi": "error", + "no-mixed-spaces-and-tabs": "error", + "no-multi-spaces": "error", + "no-redeclare": ["error", { "builtinGlobals": false }], + "no-trailing-spaces": "error", + "no-unused-vars": "off", + "no-whitespace-before-property": "error", + "object-curly-newline": ["error", { "consistent": true, "multiline": true }], + "object-curly-spacing": ["error", "never"], + "operator-linebreak": ["error", "after"], + "quote-props": ["error", "consistent-as-needed"], + "semi": ["error", "always"], + "semi-spacing": "error", + "semi-style": ["error", "last"], + "space-before-blocks": "error", + "space-before-function-paren": ["error", "never"], + "space-in-parens": ["error", "never"], + "space-infix-ops": "error", + "space-unary-ops": "error", + "switch-colon-spacing": "error", + "template-curly-spacing": ["error", "never"], + "unicode-bom": "error" + }, + "globals": { + //script.js + "gradioApp": "readonly", + "onUiLoaded": "readonly", + "onUiUpdate": "readonly", + "onOptionsChanged": "readonly", + "uiCurrentTab": "writable", + "uiElementIsVisible": "readonly", + "uiElementInSight": "readonly", + "executeCallbacks": "readonly", + //ui.js + "opts": "writable", + "all_gallery_buttons": "readonly", + "selected_gallery_button": "readonly", + "selected_gallery_index": "readonly", + "switch_to_txt2img": "readonly", + "switch_to_img2img_tab": "readonly", + "switch_to_img2img": "readonly", + "switch_to_sketch": "readonly", + "switch_to_inpaint": "readonly", + "witch_to_inpaint_sketch": "readonly", + "switch_to_extras": "readonly", + "get_tab_index": "readonly", + "create_submit_args": "readonly", + "restart_reload": "readonly", + "updateInput": "readonly", + //extraNetworks.js + "requestGet": "readonly", + "popup": "readonly", + // from python + "localization": "readonly", + // progrssbar.js + "randomId": "readonly", + "requestProgress": "readonly", + // imageviewer.js + "modalPrevImage": "readonly", + "modalNextImage": "readonly" + } +} diff --git a/CHANGELOG.md b/CHANGELOG.md index 7684d64e4..67e92961b 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,14 +1,18 @@ # Change Log for SD.Next -## Update for 05/28/2023 +## Update for 05/30/2023 -- settings search option +- new live preview mode: taesd + i really like this one, so its enabled as default for new installs +- settings search feature +- new sampler: sde++ 2m sde - fully common save/zip/delete (new) options in all tabs which (again) meant as rework of process image tab -- system info live gpu memory and load graphs for nvidia gpus -- better controlnet interface -- minor style changes -- add new k-diffusion sampler +- system info tab: live gpu utilization/memory graphs for nvidia gpus +- updated controlnet interface +- minor style changes +- updated lora, swinir, scunet and ldsr code from upstream +- start of merge from a1111 v1.3 ## Update for 05/26/2023 diff --git a/extensions-builtin/LDSR/ldsr_model_arch.py b/extensions-builtin/LDSR/ldsr_model_arch.py index c776fc720..7f450086f 100644 --- a/extensions-builtin/LDSR/ldsr_model_arch.py +++ b/extensions-builtin/LDSR/ldsr_model_arch.py @@ -40,7 +40,7 @@ class LDSR: model = model.to(shared.device) if half_attention: model = model.half() - if shared.opts.opt_channelslast: + if shared.cmd_opts.opt_channelslast: model = model.to(memory_format=torch.channels_last) sd_hijack.model_hijack.hijack(model) # apply optimization @@ -88,7 +88,7 @@ class LDSR: x_t = None logs = None - for n in range(n_runs): + for _ in range(n_runs): if custom_shape is not None: x_t = torch.randn(1, custom_shape[1], custom_shape[2], custom_shape[3]).to(model.device) x_t = repeat(x_t, '1 c h w -> b c h w', b=custom_shape[0]) @@ -110,7 +110,6 @@ class LDSR: diffusion_steps = int(steps) eta = 1.0 - down_sample_method = 'Lanczos' gc.collect() if torch.cuda.is_available: @@ -131,11 +130,11 @@ class LDSR: 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 (Not downsampling)") - + # pad width and height to multiples of 64, pads with the edge values of image to avoid artifacts pad_w, pad_h = np.max(((2, 2), np.ceil(np.array(im_og.size) / 64).astype(int)), axis=0) * 64 - im_og.size im_padded = Image.fromarray(np.pad(np.array(im_og), ((0, pad_h), (0, pad_w), (0, 0)), mode='edge')) - + logs = self.run(model["model"], im_padded, diffusion_steps, eta) sample = logs["sample"] @@ -158,7 +157,7 @@ class LDSR: def get_cond(selected_path): - example = dict() + example = {} up_f = 4 c = selected_path.convert('RGB') c = torch.unsqueeze(torchvision.transforms.ToTensor()(c), 0) @@ -196,7 +195,7 @@ def convsample_ddim(model, cond, steps, shape, eta=1.0, callback=None, normals_s @torch.no_grad() def make_convolutional_sample(batch, model, custom_steps=None, eta=1.0, quantize_x0=False, custom_shape=None, temperature=1., noise_dropout=0., corrector=None, corrector_kwargs=None, x_T=None, ddim_use_x0_pred=False): - log = dict() + log = {} z, c, x, xrec, xc = model.get_input(batch, model.first_stage_key, return_first_stage_outputs=True, @@ -244,7 +243,7 @@ def make_convolutional_sample(batch, model, custom_steps=None, eta=1.0, quantize x_sample_noquant = model.decode_first_stage(sample, force_not_quantize=True) log["sample_noquant"] = x_sample_noquant log["sample_diff"] = torch.abs(x_sample_noquant - x_sample) - except: + except Exception: pass log["sample"] = x_sample diff --git a/extensions-builtin/LDSR/preload.py b/extensions-builtin/LDSR/preload.py new file mode 100644 index 000000000..cfd478d54 --- /dev/null +++ b/extensions-builtin/LDSR/preload.py @@ -0,0 +1,6 @@ +import os +from modules import paths + + +def preload(parser): + parser.add_argument("--ldsr-models-path", type=str, help="Path to directory with LDSR model file(s).", default=os.path.join(paths.models_path, 'LDSR')) diff --git a/extensions-builtin/LDSR/scripts/ldsr_model.py b/extensions-builtin/LDSR/scripts/ldsr_model.py index da19cff12..c4da79f31 100644 --- a/extensions-builtin/LDSR/scripts/ldsr_model.py +++ b/extensions-builtin/LDSR/scripts/ldsr_model.py @@ -7,7 +7,8 @@ from basicsr.utils.download_util import load_file_from_url from modules.upscaler import Upscaler, UpscalerData from ldsr_model_arch import LDSR from modules import shared, script_callbacks -import sd_hijack_autoencoder, sd_hijack_ddpm_v1 +import sd_hijack_autoencoder # noqa: F401 +import sd_hijack_ddpm_v1 # noqa: F401 class UpscalerLDSR(Upscaler): @@ -44,9 +45,9 @@ class UpscalerLDSR(Upscaler): if local_safetensors_path is not None and os.path.exists(local_safetensors_path): model = local_safetensors_path else: - model = local_ckpt_path if local_ckpt_path is not None else load_file_from_url(url=self.model_url, model_dir=self.model_path, file_name="model.ckpt", progress=True) + model = local_ckpt_path if local_ckpt_path is not None else load_file_from_url(url=self.model_url, model_dir=self.model_download_path, file_name="model.ckpt", progress=True) - yaml = local_yaml_path if local_yaml_path is not None else load_file_from_url(url=self.yaml_url, model_dir=self.model_path, file_name="project.yaml", progress=True) + yaml = local_yaml_path if local_yaml_path is not None else load_file_from_url(url=self.yaml_url, model_dir=self.model_download_path, file_name="project.yaml", progress=True) try: return LDSR(model, yaml) diff --git a/extensions-builtin/LDSR/sd_hijack_autoencoder.py b/extensions-builtin/LDSR/sd_hijack_autoencoder.py index 8e03c7f89..81c5101b7 100644 --- a/extensions-builtin/LDSR/sd_hijack_autoencoder.py +++ b/extensions-builtin/LDSR/sd_hijack_autoencoder.py @@ -1,16 +1,21 @@ # The content of this file comes from the ldm/models/autoencoder.py file of the compvis/stable-diffusion repo # The VQModel & VQModelInterface were subsequently removed from ldm/models/autoencoder.py when we moved to the stability-ai/stablediffusion repo # As the LDSR upscaler relies on VQModel & VQModelInterface, the hijack aims to put them back into the ldm.models.autoencoder - +import numpy as np import torch import pytorch_lightning as pl import torch.nn.functional as F from contextlib import contextmanager + +from torch.optim.lr_scheduler import LambdaLR + +from ldm.modules.ema import LitEma from taming.modules.vqvae.quantize import VectorQuantizer2 as VectorQuantizer from ldm.modules.diffusionmodules.model import Encoder, Decoder from ldm.util import instantiate_from_config import ldm.models.autoencoder +from packaging import version class VQModel(pl.LightningModule): def __init__(self, @@ -19,7 +24,7 @@ class VQModel(pl.LightningModule): n_embed, embed_dim, ckpt_path=None, - ignore_keys=[], + ignore_keys=None, image_key="image", colorize_nlabels=None, monitor=None, @@ -57,7 +62,7 @@ class VQModel(pl.LightningModule): print(f"Keeping EMAs of {len(list(self.model_ema.buffers()))}.") if ckpt_path is not None: - self.init_from_ckpt(ckpt_path, ignore_keys=ignore_keys) + self.init_from_ckpt(ckpt_path, ignore_keys=ignore_keys or []) self.scheduler_config = scheduler_config self.lr_g_factor = lr_g_factor @@ -76,11 +81,11 @@ class VQModel(pl.LightningModule): if context is not None: print(f"{context}: Restored training weights") - def init_from_ckpt(self, path, ignore_keys=list()): + def init_from_ckpt(self, path, ignore_keys=None): sd = torch.load(path, map_location="cpu")["state_dict"] keys = list(sd.keys()) for k in keys: - for ik in ignore_keys: + for ik in ignore_keys or []: if k.startswith(ik): print("Deleting key {} from state_dict.".format(k)) del sd[k] @@ -165,7 +170,7 @@ class VQModel(pl.LightningModule): def validation_step(self, batch, batch_idx): log_dict = self._validation_step(batch, batch_idx) with self.ema_scope(): - log_dict_ema = self._validation_step(batch, batch_idx, suffix="_ema") + self._validation_step(batch, batch_idx, suffix="_ema") return log_dict def _validation_step(self, batch, batch_idx, suffix=""): @@ -232,7 +237,7 @@ class VQModel(pl.LightningModule): return self.decoder.conv_out.weight def log_images(self, batch, only_inputs=False, plot_ema=False, **kwargs): - log = dict() + log = {} x = self.get_input(batch, self.image_key) x = x.to(self.device) if only_inputs: @@ -249,7 +254,8 @@ class VQModel(pl.LightningModule): if plot_ema: with self.ema_scope(): xrec_ema, _ = self(x) - if x.shape[1] > 3: xrec_ema = self.to_rgb(xrec_ema) + if x.shape[1] > 3: + xrec_ema = self.to_rgb(xrec_ema) log["reconstructions_ema"] = xrec_ema return log @@ -264,7 +270,7 @@ class VQModel(pl.LightningModule): class VQModelInterface(VQModel): def __init__(self, embed_dim, *args, **kwargs): - super().__init__(embed_dim=embed_dim, *args, **kwargs) + super().__init__(*args, embed_dim=embed_dim, **kwargs) self.embed_dim = embed_dim def encode(self, x): @@ -282,5 +288,5 @@ class VQModelInterface(VQModel): dec = self.decoder(quant) return dec -setattr(ldm.models.autoencoder, "VQModel", VQModel) -setattr(ldm.models.autoencoder, "VQModelInterface", VQModelInterface) +ldm.models.autoencoder.VQModel = VQModel +ldm.models.autoencoder.VQModelInterface = VQModelInterface diff --git a/extensions-builtin/LDSR/sd_hijack_ddpm_v1.py b/extensions-builtin/LDSR/sd_hijack_ddpm_v1.py index 5c0488e5f..631a08ef0 100644 --- a/extensions-builtin/LDSR/sd_hijack_ddpm_v1.py +++ b/extensions-builtin/LDSR/sd_hijack_ddpm_v1.py @@ -48,7 +48,7 @@ class DDPMV1(pl.LightningModule): beta_schedule="linear", loss_type="l2", ckpt_path=None, - ignore_keys=[], + ignore_keys=None, load_only_unet=False, monitor="val/loss", use_ema=True, @@ -100,7 +100,7 @@ class DDPMV1(pl.LightningModule): if monitor is not None: self.monitor = monitor if ckpt_path is not None: - self.init_from_ckpt(ckpt_path, ignore_keys=ignore_keys, only_model=load_only_unet) + self.init_from_ckpt(ckpt_path, ignore_keys=ignore_keys or [], only_model=load_only_unet) self.register_schedule(given_betas=given_betas, beta_schedule=beta_schedule, timesteps=timesteps, linear_start=linear_start, linear_end=linear_end, cosine_s=cosine_s) @@ -182,13 +182,13 @@ class DDPMV1(pl.LightningModule): if context is not None: print(f"{context}: Restored training weights") - def init_from_ckpt(self, path, ignore_keys=list(), only_model=False): + def init_from_ckpt(self, path, ignore_keys=None, only_model=False): sd = torch.load(path, map_location="cpu") if "state_dict" in list(sd.keys()): sd = sd["state_dict"] keys = list(sd.keys()) for k in keys: - for ik in ignore_keys: + for ik in ignore_keys or []: if k.startswith(ik): print("Deleting key {} from state_dict.".format(k)) del sd[k] @@ -375,7 +375,7 @@ class DDPMV1(pl.LightningModule): @torch.no_grad() def log_images(self, batch, N=8, n_row=2, sample=True, return_keys=None, **kwargs): - log = dict() + log = {} x = self.get_input(batch, self.first_stage_key) N = min(x.shape[0], N) n_row = min(x.shape[0], n_row) @@ -383,7 +383,7 @@ class DDPMV1(pl.LightningModule): log["inputs"] = x # get diffusion row - diffusion_row = list() + diffusion_row = [] x_start = x[:n_row] for t in range(self.num_timesteps): @@ -444,13 +444,13 @@ class LatentDiffusionV1(DDPMV1): conditioning_key = None ckpt_path = kwargs.pop("ckpt_path", None) ignore_keys = kwargs.pop("ignore_keys", []) - super().__init__(conditioning_key=conditioning_key, *args, **kwargs) + super().__init__(*args, conditioning_key=conditioning_key, **kwargs) self.concat_mode = concat_mode self.cond_stage_trainable = cond_stage_trainable self.cond_stage_key = cond_stage_key try: self.num_downs = len(first_stage_config.params.ddconfig.ch_mult) - 1 - except: + except Exception: self.num_downs = 0 if not scale_by_std: self.scale_factor = scale_factor @@ -460,7 +460,7 @@ class LatentDiffusionV1(DDPMV1): self.instantiate_cond_stage(cond_stage_config) self.cond_stage_forward = cond_stage_forward self.clip_denoised = False - self.bbox_tokenizer = None + self.bbox_tokenizer = None self.restarted_from_ckpt = False if ckpt_path is not None: @@ -792,7 +792,7 @@ class LatentDiffusionV1(DDPMV1): z = z.view((z.shape[0], -1, ks[0], ks[1], z.shape[-1])) # (bn, nc, ks[0], ks[1], L ) # 2. apply model loop over last dim - if isinstance(self.first_stage_model, VQModelInterface): + if isinstance(self.first_stage_model, VQModelInterface): output_list = [self.first_stage_model.decode(z[:, :, :, :, i], force_not_quantize=predict_cids or force_not_quantize) for i in range(z.shape[-1])] @@ -877,16 +877,6 @@ class LatentDiffusionV1(DDPMV1): c = self.q_sample(x_start=c, t=tc, noise=torch.randn_like(c.float())) return self.p_losses(x, c, t, *args, **kwargs) - def _rescale_annotations(self, bboxes, crop_coordinates): # TODO: move to dataset - def rescale_bbox(bbox): - x0 = clamp((bbox[0] - crop_coordinates[0]) / crop_coordinates[2]) - y0 = clamp((bbox[1] - crop_coordinates[1]) / crop_coordinates[3]) - w = min(bbox[2] / crop_coordinates[2], 1 - x0) - h = min(bbox[3] / crop_coordinates[3], 1 - y0) - return x0, y0, w, h - - return [rescale_bbox(b) for b in bboxes] - def apply_model(self, x_noisy, t, cond, return_ids=False): if isinstance(cond, dict): @@ -900,7 +890,7 @@ class LatentDiffusionV1(DDPMV1): if hasattr(self, "split_input_params"): assert len(cond) == 1 # todo can only deal with one conditioning atm - assert not return_ids + assert not return_ids ks = self.split_input_params["ks"] # eg. (128, 128) stride = self.split_input_params["stride"] # eg. (64, 64) @@ -1126,7 +1116,7 @@ class LatentDiffusionV1(DDPMV1): if cond is not None: if isinstance(cond, dict): cond = {key: cond[key][:batch_size] if not isinstance(cond[key], list) else - list(map(lambda x: x[:batch_size], cond[key])) for key in cond} + [x[:batch_size] for x in cond[key]] for key in cond} else: cond = [c[:batch_size] for c in cond] if isinstance(cond, list) else cond[:batch_size] @@ -1157,8 +1147,10 @@ class LatentDiffusionV1(DDPMV1): if i % log_every_t == 0 or i == timesteps - 1: intermediates.append(x0_partial) - if callback: callback(i) - if img_callback: img_callback(img, i) + if callback: + callback(i) + if img_callback: + img_callback(img, i) return img, intermediates @torch.no_grad() @@ -1205,8 +1197,10 @@ class LatentDiffusionV1(DDPMV1): if i % log_every_t == 0 or i == timesteps - 1: intermediates.append(img) - if callback: callback(i) - if img_callback: img_callback(img, i) + if callback: + callback(i) + if img_callback: + img_callback(img, i) if return_intermediates: return img, intermediates @@ -1221,7 +1215,7 @@ class LatentDiffusionV1(DDPMV1): if cond is not None: if isinstance(cond, dict): cond = {key: cond[key][:batch_size] if not isinstance(cond[key], list) else - list(map(lambda x: x[:batch_size], cond[key])) for key in cond} + [x[:batch_size] for x in cond[key]] for key in cond} else: cond = [c[:batch_size] for c in cond] if isinstance(cond, list) else cond[:batch_size] return self.p_sample_loop(cond, @@ -1253,7 +1247,7 @@ class LatentDiffusionV1(DDPMV1): use_ddim = ddim_steps is not None - log = dict() + log = {} z, c, x, xrec, xc = self.get_input(batch, self.first_stage_key, return_first_stage_outputs=True, force_c_encode=True, @@ -1280,7 +1274,7 @@ class LatentDiffusionV1(DDPMV1): if plot_diffusion_rows: # get diffusion row - diffusion_row = list() + diffusion_row = [] z_start = z[:n_row] for t in range(self.num_timesteps): if t % self.log_every_t == 0 or t == self.num_timesteps - 1: @@ -1322,7 +1316,7 @@ class LatentDiffusionV1(DDPMV1): if inpaint: # make a simple center square - b, h, w = z.shape[0], z.shape[2], z.shape[3] + h, w = z.shape[2], z.shape[3] mask = torch.ones(N, h, w).to(self.device) # zeros will be filled in mask[:, h // 4:3 * h // 4, w // 4:3 * w // 4] = 0. @@ -1424,10 +1418,10 @@ class Layout2ImgDiffusionV1(LatentDiffusionV1): # TODO: move all layout-specific hacks to this class def __init__(self, cond_stage_key, *args, **kwargs): assert cond_stage_key == 'coordinates_bbox', 'Layout2ImgDiffusion only for cond_stage_key="coordinates_bbox"' - super().__init__(cond_stage_key=cond_stage_key, *args, **kwargs) + super().__init__(*args, cond_stage_key=cond_stage_key, **kwargs) def log_images(self, batch, N=8, *args, **kwargs): - logs = super().log_images(batch=batch, N=N, *args, **kwargs) + logs = super().log_images(*args, batch=batch, N=N, **kwargs) key = 'train' if self.training else 'validation' dset = self.trainer.datamodule.datasets[key] @@ -1443,7 +1437,7 @@ class Layout2ImgDiffusionV1(LatentDiffusionV1): logs['bbox_image'] = cond_img return logs -setattr(ldm.models.diffusion.ddpm, "DDPMV1", DDPMV1) -setattr(ldm.models.diffusion.ddpm, "LatentDiffusionV1", LatentDiffusionV1) -setattr(ldm.models.diffusion.ddpm, "DiffusionWrapperV1", DiffusionWrapperV1) -setattr(ldm.models.diffusion.ddpm, "Layout2ImgDiffusionV1", Layout2ImgDiffusionV1) +ldm.models.diffusion.ddpm.DDPMV1 = DDPMV1 +ldm.models.diffusion.ddpm.LatentDiffusionV1 = LatentDiffusionV1 +ldm.models.diffusion.ddpm.DiffusionWrapperV1 = DiffusionWrapperV1 +ldm.models.diffusion.ddpm.Layout2ImgDiffusionV1 = Layout2ImgDiffusionV1 diff --git a/extensions-builtin/Lora/extra_networks_lora.py b/extensions-builtin/Lora/extra_networks_lora.py index ccb249ac7..b5fea4d2e 100644 --- a/extensions-builtin/Lora/extra_networks_lora.py +++ b/extensions-builtin/Lora/extra_networks_lora.py @@ -23,5 +23,23 @@ class ExtraNetworkLora(extra_networks.ExtraNetwork): lora.load_loras(names, multipliers) + if shared.opts.lora_add_hashes_to_infotext: + lora_hashes = [] + for item in lora.loaded_loras: + shorthash = item.lora_on_disk.shorthash + if not shorthash: + continue + + alias = item.mentioned_name + if not alias: + continue + + alias = alias.replace(":", "").replace(",", "") + + lora_hashes.append(f"{alias}: {shorthash}") + + if lora_hashes: + p.extra_generation_params["Lora hashes"] = ", ".join(lora_hashes) + def deactivate(self, p): pass diff --git a/extensions-builtin/Lora/lora.py b/extensions-builtin/Lora/lora.py index 5a12f1836..eec147122 100644 --- a/extensions-builtin/Lora/lora.py +++ b/extensions-builtin/Lora/lora.py @@ -1,8 +1,9 @@ import os import re -from typing import Union import torch -from modules import shared, devices, sd_models, errors, scripts +from typing import Union + +from modules import shared, devices, sd_models, errors, scripts, sd_hijack, hashes metadata_tags_order = {"ss_sd_model_name": 1, "ss_resolution": 2, "ss_clip_skip": 3, "ss_num_train_images": 10, "ss_tag_frequency": 20} @@ -75,9 +76,9 @@ class LoraOnDisk: self.name = name self.filename = filename self.metadata = {} + self.is_safetensors = os.path.splitext(filename)[1].lower() == ".safetensors" - _, ext = os.path.splitext(filename) - if ext.lower() == ".safetensors": + if self.is_safetensors: try: self.metadata = sd_models.read_metadata_from_safetensors(filename) except Exception as e: @@ -93,14 +94,43 @@ class LoraOnDisk: self.ssmd_cover_images = self.metadata.pop('ssmd_cover_images', None) # those are cover images and they are too big to display in UI as text self.alias = self.metadata.get('ss_output_name', self.name) + self.hash = None + self.shorthash = None + self.set_hash( + self.metadata.get('sshs_model_hash') or + hashes.sha256_from_cache(self.filename, "lora/" + self.name, use_addnet_hash=self.is_safetensors) or + '' + ) + + def set_hash(self, v): + self.hash = v + self.shorthash = self.hash[0:12] + + if self.shorthash: + available_lora_hash_lookup[self.shorthash] = self + + def read_hash(self): + if not self.hash: + self.set_hash(hashes.sha256(self.filename, "lora/" + self.name, use_addnet_hash=self.is_safetensors) or '') + + def get_alias(self): + if shared.opts.lora_preferred_name == "Filename" or self.alias.lower() in forbidden_lora_aliases: + return self.name + else: + return self.alias + class LoraModule: - def __init__(self, name): + def __init__(self, name, lora_on_disk: LoraOnDisk): self.name = name + self.lora_on_disk = lora_on_disk self.multiplier = 1.0 self.modules = {} self.mtime = None + self.mentioned_name = None + """the text that was used to add lora to prompt - can be either name or an alias""" + class LoraUpDownModule: def __init__(self): @@ -125,11 +155,11 @@ def assign_lora_names_to_compvis_modules(sd_model): sd_model.lora_layer_mapping = lora_layer_mapping -def load_lora(name, filename): - lora = LoraModule(name) - lora.mtime = os.path.getmtime(filename) +def load_lora(name, lora_on_disk): + lora = LoraModule(name, lora_on_disk) + lora.mtime = os.path.getmtime(lora_on_disk.filename) - sd = sd_models.read_state_dict(filename) + sd = sd_models.read_state_dict(lora_on_disk.filename) # this should not be needed but is here as an emergency fix for an unknown error people are experiencing in 1.2.0 if not hasattr(shared.sd_model, 'lora_layer_mapping'): @@ -175,6 +205,7 @@ def load_lora(name, filename): else: print(f'Lora layer {key_diffusers} matched a layer with unsupported type: {type(sd_module).__name__}') continue + raise AssertionError(f"Lora layer {key_diffusers} matched a layer with unsupported type: {type(sd_module).__name__}") with torch.no_grad(): module.weight.copy_(weight) @@ -186,10 +217,10 @@ def load_lora(name, filename): elif lora_key == "lora_down.weight": lora_module.down = module else: - assert False, f'Bad Lora layer name: {key_diffusers} - must end in lora_up.weight, lora_down.weight or alpha' + raise AssertionError(f"Bad Lora layer name: {key_diffusers} - must end in lora_up.weight, lora_down.weight or alpha") if len(keys_failed_to_match) > 0: - print(f"Failed to match keys when loading Lora {filename}: {len(keys_failed_to_match)}") + print(f"Failed to match keys when loading Lora {lora_on_disk.filename}: {keys_failed_to_match}") return lora @@ -204,30 +235,41 @@ def load_loras(names, multipliers=None): loaded_loras.clear() loras_on_disk = [available_lora_aliases.get(name, None) for name in names] - if any([x is None for x in loras_on_disk]): + if any(x is None for x in loras_on_disk): list_available_loras() loras_on_disk = [available_lora_aliases.get(name, None) for name in names] + failed_to_load_loras = [] + for i, name in enumerate(names): lora = already_loaded.get(name, None) lora_on_disk = loras_on_disk[i] + if lora_on_disk is not None: if lora is None or os.path.getmtime(lora_on_disk.filename) > lora.mtime: try: - lora = load_lora(name, lora_on_disk.filename) + lora = load_lora(name, lora_on_disk) except Exception as e: errors.display(e, f"loading Lora {lora_on_disk.filename}") continue + lora.mentioned_name = name + + lora_on_disk.read_hash() + if lora is None: + failed_to_load_loras.append(name) print(f"Couldn't find Lora with name {name}") continue lora.multiplier = multipliers[i] if multipliers else 1.0 loaded_loras.append(lora) + if len(failed_to_load_loras) > 0: + sd_hijack.model_hijack.comments.append("Failed to find Loras: " + ", ".join(failed_to_load_loras)) + def lora_calc_updown(lora, module, target): with torch.no_grad(): @@ -311,7 +353,7 @@ def lora_apply_weights(self: Union[torch.nn.Conv2d, torch.nn.Linear, torch.nn.Mu print(f'failed to calculate lora weights for layer {lora_layer_name}') - setattr(self, "lora_current_names", wanted_names) + self.lora_current_names = wanted_names def lora_forward(module, input, original_forward): @@ -345,8 +387,8 @@ def lora_forward(module, input, original_forward): def lora_reset_cached_weight(self: Union[torch.nn.Conv2d, torch.nn.Linear]): - setattr(self, "lora_current_names", ()) - setattr(self, "lora_weights_backup", None) + self.lora_current_names = () + self.lora_weights_backup = None def lora_Linear_forward(self, input): @@ -395,7 +437,8 @@ def list_available_loras(): available_loras.clear() available_lora_aliases.clear() forbidden_lora_aliases.clear() - forbidden_lora_aliases.update({"none": 1}) + available_lora_hash_lookup.clear() + forbidden_lora_aliases.update({"none": 1, "Addams": 1}) os.makedirs(shared.cmd_opts.lora_dir, exist_ok=True) @@ -425,7 +468,7 @@ def infotext_pasted(infotext, params): added = [] - for k, v in params.items(): + for k in params: if not k.startswith("AddNet Model "): continue @@ -449,8 +492,10 @@ def infotext_pasted(infotext, params): if added: params["Prompt"] += "\n" + "".join(added) + available_loras = {} available_lora_aliases = {} +available_lora_hash_lookup = {} forbidden_lora_aliases = {} loaded_loras = [] diff --git a/extensions-builtin/Lora/scripts/lora_script.py b/extensions-builtin/Lora/scripts/lora_script.py index 1f0677938..e650f469f 100644 --- a/extensions-builtin/Lora/scripts/lora_script.py +++ b/extensions-builtin/Lora/scripts/lora_script.py @@ -1,3 +1,5 @@ +import re + import torch import gradio as gr from fastapi import FastAPI @@ -20,6 +22,7 @@ def before_ui(): ui_extra_networks.register_page(ui_extra_networks_lora.ExtraNetworksPageLora()) extra_networks.register_extra_network(extra_networks_lora.ExtraNetworkLora()) + if not hasattr(torch.nn, 'Linear_forward_before_lora'): torch.nn.Linear_forward_before_lora = torch.nn.Linear.forward @@ -52,8 +55,13 @@ script_callbacks.on_infotext_pasted(lora.infotext_pasted) shared.options_templates.update(shared.options_section(('extra_networks', "Extra Networks"), { - "sd_lora": shared.OptionInfo("None", "Add Lora to prompt", gr.Dropdown, lambda: {"choices": ["None"] + [x for x in lora.available_loras]}, refresh=lora.list_available_loras), - "lora_preferred_name": shared.OptionInfo("Alias from file", "When adding to prompt, refer to lora by", gr.Radio, {"choices": ["Alias from file", "Filename"]}), + "sd_lora": shared.OptionInfo("None", "Add Lora to prompt", gr.Dropdown, lambda: {"choices": ["None", *lora.available_loras]}, refresh=lora.list_available_loras), + "lora_preferred_name": shared.OptionInfo("Alias from file", "When adding to prompt, refer to Lora by", gr.Radio, {"choices": ["Alias from file", "Filename"]}), + "lora_add_hashes_to_infotext": shared.OptionInfo(True, "Add Lora hashes to infotext"), +})) + + +shared.options_templates.update(shared.options_section(('compatibility', "Compatibility"), { "lora_functional": shared.OptionInfo(False, "Lora: use old method that takes longer when you have multiple Loras active and produces same results as kohya-ss/sd-webui-additional-networks extension"), })) @@ -72,6 +80,37 @@ def api_loras(_: gr.Blocks, app: FastAPI): async def get_loras(): return [create_lora_json(obj) for obj in lora.available_loras.values()] + @app.post("/sdapi/v1/refresh-loras") + async def refresh_loras(): + return lora.list_available_loras() + script_callbacks.on_app_started(api_loras) +re_lora = re.compile(" b w1 w2 p1 p2 c', p1=self.window_size, p2=self.window_size) h_windows = x.size(1) w_windows = x.size(2) @@ -85,8 +87,9 @@ class WMSA(nn.Module): 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)) + 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): @@ -262,4 +265,4 @@ class SCUNet(nn.Module): 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 + nn.init.constant_(m.weight, 1.0) diff --git a/extensions-builtin/SwinIR/preload.py b/extensions-builtin/SwinIR/preload.py new file mode 100644 index 000000000..e912c6402 --- /dev/null +++ b/extensions-builtin/SwinIR/preload.py @@ -0,0 +1,6 @@ +import os +from modules import paths + + +def preload(parser): + parser.add_argument("--swinir-models-path", type=str, help="Path to directory with SwinIR model file(s).", default=os.path.join(paths.models_path, 'SwinIR')) diff --git a/extensions-builtin/SwinIR/scripts/swinir_model.py b/extensions-builtin/SwinIR/scripts/swinir_model.py index 619f52e6d..1c7bf325e 100644 --- a/extensions-builtin/SwinIR/scripts/swinir_model.py +++ b/extensions-builtin/SwinIR/scripts/swinir_model.py @@ -1,10 +1,10 @@ import os + 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 rich import progress from modules import modelloader, devices, script_callbacks, shared from modules.shared import opts, state @@ -44,31 +44,31 @@ class UpscalerSwinIR(Upscaler): img = upscale(img, model) try: torch.cuda.empty_cache() - except: + except Exception: pass return img def load_model(self, path, scale=4): if "http" in path: dl_name = "%s%s" % (self.model_name.replace(" ", "_"), ".pth") - filename = load_file_from_url(url=path, model_dir=self.model_path, file_name=dl_name, progress=True) + filename = load_file_from_url(url=path, model_dir=self.model_download_path, file_name=dl_name, progress=True) else: filename = path if filename is None or not os.path.exists(filename): return None if filename.endswith(".v2.pth"): model = net2( - upscale=scale, - in_chans=3, - img_size=64, - window_size=8, - img_range=1.0, - depths=[6, 6, 6, 6, 6, 6], - embed_dim=180, - num_heads=[6, 6, 6, 6, 6, 6], - mlp_ratio=2, - upsampler="nearest+conv", - resi_connection="1conv", + upscale=scale, + in_chans=3, + img_size=64, + window_size=8, + img_range=1.0, + depths=[6, 6, 6, 6, 6, 6], + embed_dim=180, + num_heads=[6, 6, 6, 6, 6, 6], + mlp_ratio=2, + upsampler="nearest+conv", + resi_connection="1conv", ) params = None else: @@ -87,9 +87,8 @@ class UpscalerSwinIR(Upscaler): ) params = "params_ema" - with progress.open(filename, 'rb', description=f'Loading weights: [cyan]{filename}', auto_refresh=True) as f: - pretrained_model = torch.load(f) - if params is not None and params in pretrained_model: + pretrained_model = torch.load(filename) + if params is not None: model.load_state_dict(pretrained_model[params], strict=True) else: model.load_state_dict(pretrained_model, strict=True) diff --git a/extensions-builtin/SwinIR/swinir_model_arch.py b/extensions-builtin/SwinIR/swinir_model_arch.py index 863f42db6..93b932747 100644 --- a/extensions-builtin/SwinIR/swinir_model_arch.py +++ b/extensions-builtin/SwinIR/swinir_model_arch.py @@ -644,7 +644,7 @@ class SwinIR(nn.Module): """ def __init__(self, img_size=64, patch_size=1, in_chans=3, - embed_dim=96, depths=[6, 6, 6, 6], num_heads=[6, 6, 6, 6], + embed_dim=96, depths=(6, 6, 6, 6), num_heads=(6, 6, 6, 6), window_size=7, mlp_ratio=4., qkv_bias=True, qk_scale=None, drop_rate=0., attn_drop_rate=0., drop_path_rate=0.1, norm_layer=nn.LayerNorm, ape=False, patch_norm=True, @@ -805,7 +805,7 @@ class SwinIR(nn.Module): def forward(self, x): H, W = x.shape[2:] x = self.check_image_size(x) - + self.mean = self.mean.type_as(x) x = (x - self.mean) * self.img_range @@ -844,7 +844,7 @@ class SwinIR(nn.Module): H, W = self.patches_resolution flops += H * W * 3 * self.embed_dim * 9 flops += self.patch_embed.flops() - for i, layer in enumerate(self.layers): + for layer in self.layers: flops += layer.flops() flops += H * W * 3 * self.embed_dim * self.embed_dim flops += self.upsample.flops() diff --git a/extensions-builtin/SwinIR/swinir_model_arch_v2.py b/extensions-builtin/SwinIR/swinir_model_arch_v2.py index 0e28ae6ee..dad22cca2 100644 --- a/extensions-builtin/SwinIR/swinir_model_arch_v2.py +++ b/extensions-builtin/SwinIR/swinir_model_arch_v2.py @@ -74,7 +74,7 @@ class WindowAttention(nn.Module): """ def __init__(self, dim, window_size, num_heads, qkv_bias=True, attn_drop=0., proj_drop=0., - pretrained_window_size=[0, 0]): + pretrained_window_size=(0, 0)): super().__init__() self.dim = dim @@ -241,7 +241,7 @@ class SwinTransformerBlock(nn.Module): attn_mask = None self.register_buffer("attn_mask", attn_mask) - + def calculate_mask(self, x_size): # calculate attention mask for SW-MSA H, W = x_size @@ -263,7 +263,7 @@ class SwinTransformerBlock(nn.Module): attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2) attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0)) - return attn_mask + return attn_mask def forward(self, x, x_size): H, W = x_size @@ -288,7 +288,7 @@ class SwinTransformerBlock(nn.Module): attn_windows = self.attn(x_windows, mask=self.attn_mask) # nW*B, window_size*window_size, C else: attn_windows = self.attn(x_windows, mask=self.calculate_mask(x_size).to(x.device)) - + # merge windows attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C) shifted_x = window_reverse(attn_windows, self.window_size, H, W) # B H' W' C @@ -369,7 +369,7 @@ class PatchMerging(nn.Module): H, W = self.input_resolution flops = (H // 2) * (W // 2) * 4 * self.dim * 2 * self.dim flops += H * W * self.dim // 2 - return flops + return flops class BasicLayer(nn.Module): """ A basic Swin Transformer layer for one stage. @@ -447,7 +447,7 @@ class BasicLayer(nn.Module): nn.init.constant_(blk.norm1.weight, 0) nn.init.constant_(blk.norm2.bias, 0) nn.init.constant_(blk.norm2.weight, 0) - + class PatchEmbed(nn.Module): r""" Image to Patch Embedding Args: @@ -492,7 +492,7 @@ class PatchEmbed(nn.Module): flops = Ho * Wo * self.embed_dim * self.in_chans * (self.patch_size[0] * self.patch_size[1]) if self.norm is not None: flops += Ho * Wo * self.embed_dim - return flops + return flops class RSTB(nn.Module): """Residual Swin Transformer Block (RSTB). @@ -531,7 +531,7 @@ class RSTB(nn.Module): num_heads=num_heads, window_size=window_size, mlp_ratio=mlp_ratio, - qkv_bias=qkv_bias, + qkv_bias=qkv_bias, drop=drop, attn_drop=attn_drop, drop_path=drop_path, norm_layer=norm_layer, @@ -622,7 +622,7 @@ class Upsample(nn.Sequential): else: raise ValueError(f'scale {scale} is not supported. ' 'Supported scales: 2^n and 3.') super(Upsample, self).__init__(*m) - + class Upsample_hf(nn.Sequential): """Upsample module. @@ -642,7 +642,7 @@ class Upsample_hf(nn.Sequential): m.append(nn.PixelShuffle(3)) else: raise ValueError(f'scale {scale} is not supported. ' 'Supported scales: 2^n and 3.') - super(Upsample_hf, self).__init__(*m) + super(Upsample_hf, self).__init__(*m) class UpsampleOneStep(nn.Sequential): @@ -667,8 +667,8 @@ class UpsampleOneStep(nn.Sequential): H, W = self.input_resolution flops = H * W * self.num_feat * 3 * 9 return flops - - + + class Swin2SR(nn.Module): r""" Swin2SR @@ -698,8 +698,8 @@ class Swin2SR(nn.Module): """ def __init__(self, img_size=64, patch_size=1, in_chans=3, - embed_dim=96, depths=[6, 6, 6, 6], num_heads=[6, 6, 6, 6], - window_size=7, mlp_ratio=4., qkv_bias=True, + embed_dim=96, depths=(6, 6, 6, 6), num_heads=(6, 6, 6, 6), + window_size=7, mlp_ratio=4., qkv_bias=True, drop_rate=0., attn_drop_rate=0., drop_path_rate=0.1, norm_layer=nn.LayerNorm, ape=False, patch_norm=True, use_checkpoint=False, upscale=2, img_range=1., upsampler='', resi_connection='1conv', @@ -764,7 +764,7 @@ class Swin2SR(nn.Module): num_heads=num_heads[i_layer], window_size=window_size, mlp_ratio=self.mlp_ratio, - qkv_bias=qkv_bias, + qkv_bias=qkv_bias, drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])], # no impact on SR results norm_layer=norm_layer, @@ -776,7 +776,7 @@ class Swin2SR(nn.Module): ) self.layers.append(layer) - + if self.upsampler == 'pixelshuffle_hf': self.layers_hf = nn.ModuleList() for i_layer in range(self.num_layers): @@ -787,7 +787,7 @@ class Swin2SR(nn.Module): num_heads=num_heads[i_layer], window_size=window_size, mlp_ratio=self.mlp_ratio, - qkv_bias=qkv_bias, + qkv_bias=qkv_bias, drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])], # no impact on SR results norm_layer=norm_layer, @@ -799,7 +799,7 @@ class Swin2SR(nn.Module): ) self.layers_hf.append(layer) - + self.norm = norm_layer(self.num_features) # build the last conv layer in deep feature extraction @@ -829,10 +829,10 @@ class Swin2SR(nn.Module): self.conv_aux = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1) self.conv_after_aux = nn.Sequential( nn.Conv2d(3, num_feat, 3, 1, 1), - nn.LeakyReLU(inplace=True)) + nn.LeakyReLU(inplace=True)) self.upsample = Upsample(upscale, num_feat) self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1) - + elif self.upsampler == 'pixelshuffle_hf': self.conv_before_upsample = nn.Sequential(nn.Conv2d(embed_dim, num_feat, 3, 1, 1), nn.LeakyReLU(inplace=True)) @@ -846,7 +846,7 @@ class Swin2SR(nn.Module): nn.Conv2d(embed_dim, num_feat, 3, 1, 1), nn.LeakyReLU(inplace=True)) self.conv_last_hf = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1) - + elif self.upsampler == 'pixelshuffledirect': # for lightweight SR (to save parameters) self.upsample = UpsampleOneStep(upscale, embed_dim, num_out_ch, @@ -905,7 +905,7 @@ class Swin2SR(nn.Module): x = self.patch_unembed(x, x_size) return x - + def forward_features_hf(self, x): x_size = (x.shape[2], x.shape[3]) x = self.patch_embed(x) @@ -919,7 +919,7 @@ class Swin2SR(nn.Module): x = self.norm(x) # B L C x = self.patch_unembed(x, x_size) - return x + return x def forward(self, x): H, W = x.shape[2:] @@ -951,7 +951,7 @@ class Swin2SR(nn.Module): x = self.conv_after_body(self.forward_features(x)) + x x_before = self.conv_before_upsample(x) x_out = self.conv_last(self.upsample(x_before)) - + x_hf = self.conv_first_hf(x_before) x_hf = self.conv_after_body_hf(self.forward_features_hf(x_hf)) + x_hf x_hf = self.conv_before_upsample_hf(x_hf) @@ -977,15 +977,15 @@ class Swin2SR(nn.Module): x_first = self.conv_first(x) res = self.conv_after_body(self.forward_features(x_first)) + x_first x = x + self.conv_last(res) - + x = x / self.img_range + self.mean if self.upsampler == "pixelshuffle_aux": return x[:, :, :H*self.upscale, :W*self.upscale], aux - + elif self.upsampler == "pixelshuffle_hf": x_out = x_out / self.img_range + self.mean return x_out[:, :, :H*self.upscale, :W*self.upscale], x[:, :, :H*self.upscale, :W*self.upscale], x_hf[:, :, :H*self.upscale, :W*self.upscale] - + else: return x[:, :, :H*self.upscale, :W*self.upscale] @@ -994,7 +994,7 @@ class Swin2SR(nn.Module): H, W = self.patches_resolution flops += H * W * 3 * self.embed_dim * 9 flops += self.patch_embed.flops() - for i, layer in enumerate(self.layers): + for layer in self.layers: flops += layer.flops() flops += H * W * 3 * self.embed_dim * self.embed_dim flops += self.upsample.flops() @@ -1014,4 +1014,4 @@ if __name__ == '__main__': x = torch.randn((1, 3, height, width)) x = model(x) - print(x.shape) \ No newline at end of file + print(x.shape) diff --git a/extensions-builtin/prompt-bracket-checker/javascript/prompt-bracket-checker.js b/extensions-builtin/prompt-bracket-checker/javascript/prompt-bracket-checker.js index 5c7a836a2..114cf94cc 100644 --- a/extensions-builtin/prompt-bracket-checker/javascript/prompt-bracket-checker.js +++ b/extensions-builtin/prompt-bracket-checker/javascript/prompt-bracket-checker.js @@ -4,39 +4,39 @@ // If there's a mismatch, the keyword counter turns red and if you hover on it, a tooltip tells you what's wrong. function checkBrackets(textArea, counterElt) { - var counts = {}; - (textArea.value.match(/[(){}\[\]]/g) || []).forEach(bracket => { - counts[bracket] = (counts[bracket] || 0) + 1; - }); - var errors = []; + var counts = {}; + (textArea.value.match(/[(){}[\]]/g) || []).forEach(bracket => { + counts[bracket] = (counts[bracket] || 0) + 1; + }); + var errors = []; - function checkPair(open, close, kind) { - if (counts[open] !== counts[close]) { - errors.push( - `${open}...${close} - Detected ${counts[open] || 0} opening and ${counts[close] || 0} closing ${kind}.` - ); + function checkPair(open, close, kind) { + if (counts[open] !== counts[close]) { + errors.push( + `${open}...${close} - Detected ${counts[open] || 0} opening and ${counts[close] || 0} closing ${kind}.` + ); + } } - } - checkPair('(', ')', 'round brackets'); - checkPair('[', ']', 'square brackets'); - checkPair('{', '}', 'curly brackets'); - counterElt.title = errors.join('\n'); - counterElt.classList.toggle('error', errors.length !== 0); + checkPair('(', ')', 'round brackets'); + checkPair('[', ']', 'square brackets'); + checkPair('{', '}', 'curly brackets'); + counterElt.title = errors.join('\n'); + counterElt.classList.toggle('error', errors.length !== 0); } function setupBracketChecking(id_prompt, id_counter) { - var textarea = gradioApp().querySelector("#" + id_prompt + " > label > textarea"); - var counter = gradioApp().getElementById(id_counter) + var textarea = gradioApp().querySelector("#" + id_prompt + " > label > textarea"); + var counter = gradioApp().getElementById(id_counter); - if (textarea && counter) { - textarea.addEventListener("input", () => checkBrackets(textarea, counter)); - } + if (textarea && counter) { + textarea.addEventListener("input", () => checkBrackets(textarea, counter)); + } } -onUiLoaded(function () { - setupBracketChecking('txt2img_prompt', 'txt2img_token_counter'); - setupBracketChecking('txt2img_neg_prompt', 'txt2img_negative_token_counter'); - setupBracketChecking('img2img_prompt', 'img2img_token_counter'); - setupBracketChecking('img2img_neg_prompt', 'img2img_negative_token_counter'); +onUiLoaded(function() { + setupBracketChecking('txt2img_prompt', 'txt2img_token_counter'); + setupBracketChecking('txt2img_neg_prompt', 'txt2img_negative_token_counter'); + setupBracketChecking('img2img_prompt', 'img2img_token_counter'); + setupBracketChecking('img2img_neg_prompt', 'img2img_negative_token_counter'); }); diff --git a/html/licenses.html b/html/licenses.html index bc995aa07..ef6f2c0a4 100644 --- a/html/licenses.html +++ b/html/licenses.html @@ -661,4 +661,30 @@ AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. + + +

TAESD

+Tiny AutoEncoder for Stable Diffusion option for live previews +
+MIT License
+
+Copyright (c) 2023 Ollin Boer Bohan
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
\ No newline at end of file diff --git a/installer.py b/installer.py index d5b464156..120169bc0 100644 --- a/installer.py +++ b/installer.py @@ -268,7 +268,7 @@ def check_torch(): else: machine = platform.machine() if sys.platform == 'darwin': - torch_command = os.environ.get('TORCH_COMMAND', 'torch==2.0.0 torchvision==0.15.1') + torch_command = os.environ.get('TORCH_COMMAND', 'torch==2.0.1 torchvision==0.15.2') elif allow_directml and args.use_directml and ('arm' not in machine and 'aarch' not in machine): log.info('Using DirectML Backend') torch_command = os.environ.get('TORCH_COMMAND', 'torch-directml') diff --git a/javascript/.eslintrc.json b/javascript/.eslintrc.json deleted file mode 100644 index 5fcb88998..000000000 --- a/javascript/.eslintrc.json +++ /dev/null @@ -1,22 +0,0 @@ -{ - "globals": {}, - "env": { - "browser": true, - "commonjs": false, - "node": false, - "jquery": false, - "es2020": true - }, - "parserOptions": { "ecmaVersion": 2020 }, - "plugins": [], - "extends": ["eslint:recommended", "airbnb-base"], - "rules": { - "max-len": [1, 220, 3], - "camelcase":"off", - "no-unused-vars":"off", - "no-plusplus":"off", - "no-param-reassign":"off", - "no-restricted-syntax":"off", - "no-mixed-operators":"off" - } -} diff --git a/javascript/hints.js b/javascript/hints.js index 61da3e357..e0fc63634 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -22,6 +22,13 @@ titles = { '\u{1f4cb}': 'Apply selected styles to current prompt', '\u{1f4d2}': 'Paste available values into the field', '\u{1f3b4}': 'Show/hide extra networks', + '\u{1F4D8}': 'Read generation parameters from prompt or last generation if prompt is empty into user interface.', + '\u{1F6C5}': 'Save style', + '\u{1F9F3}': 'Apply selected styles to current prompt', + '\u{1F6AE}': 'Clear prompt', + '\u{1F310}': 'Show/hide extra networks', + + '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/hashes.py b/modules/hashes.py index f36291362..fb7ce62fa 100644 --- a/modules/hashes.py +++ b/modules/hashes.py @@ -17,19 +17,19 @@ def dump_cache(): def cache(subsection): - global cache_data # pylint: disable=global-statement + global cache_data + if cache_data is None: with filelock.FileLock(f"{cache_filename}.lock"): if not os.path.isfile(cache_filename): cache_data = {} else: - try: - with open(cache_filename, "r", encoding="utf8") as file: - cache_data = json.load(file) - except: - cache_data = None + with open(cache_filename, "r", encoding="utf8") as file: + cache_data = json.load(file) + s = cache_data.get(subsection, {}) cache_data[subsection] = s + return s @@ -42,8 +42,8 @@ def calculate_sha256(filename): return hash_sha256.hexdigest() -def sha256_from_cache(filename, title): - hashes = cache("hashes") +def sha256_from_cache(filename, title, use_addnet_hash=False): + hashes = cache("hashes-addnet") if use_addnet_hash else cache("hashes") ondisk_mtime = os.path.getmtime(filename) if title not in hashes: return None @@ -54,17 +54,36 @@ def sha256_from_cache(filename, title): return cached_sha256 -def sha256(filename, title): - hashes = cache("hashes") - sha256_value = sha256_from_cache(filename, title) +def sha256(filename, title, use_addnet_hash=False): + hashes = cache("hashes-addnet") if use_addnet_hash else cache("hashes") + sha256_value = sha256_from_cache(filename, title, use_addnet_hash) if sha256_value is not None: return sha256_value if shared.cmd_opts.no_hashing: return None - sha256_value = calculate_sha256(filename) + if use_addnet_hash: + with progress.open(filename, 'rb', description=f'Calculating model hash: [cyan]{filename}', auto_refresh=True) as f: + sha256_value = addnet_hash_safetensors(f) + else: + sha256_value = calculate_sha256(filename) hashes[title] = { "mtime": os.path.getmtime(filename), "sha256": sha256_value, } dump_cache() return sha256_value + + +def addnet_hash_safetensors(b): + """kohya-ss hash for safetensors from https://github.com/kohya-ss/sd-scripts/blob/main/library/train_util.py""" + hash_sha256 = hashlib.sha256() + blksize = 1024 * 1024 + b.seek(0) + header = b.read(8) + n = int.from_bytes(header, "little") + offset = n + 8 + b.seek(offset) + for chunk in iter(lambda: b.read(blksize), b""): + hash_sha256.update(chunk) + return hash_sha256.hexdigest() + diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index 8ac3f46ef..a16bc5354 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -17,11 +17,16 @@ samplers_for_img2img = all_samplers samplers_map = {} -def create_sampler(name, model): +def find_sampler_config(name): if name is not None: config = all_samplers_map.get(name, None) else: config = all_samplers[0] + return config + + +def create_sampler(name, model): + config = find_sampler_config(name) assert config is not None, f'bad sampler name: {name}' if backend == Backend.ORIGINAL: sampler = config.constructor(model) diff --git a/modules/sd_samplers_common.py b/modules/sd_samplers_common.py index 888f9a30e..1eacedca1 100644 --- a/modules/sd_samplers_common.py +++ b/modules/sd_samplers_common.py @@ -2,7 +2,7 @@ from collections import namedtuple import numpy as np import torch from PIL import Image -from modules import devices, processing, images, sd_vae_approx +from modules import devices, processing, images, sd_vae_approx, sd_samplers, sd_vae_taesd from modules.shared import opts, state import modules.shared as shared @@ -22,19 +22,22 @@ def setup_img2img_steps(p, steps=None): return steps, t_enc -approximation_indexes = {"Full": 0, "Approx NN": 1, "Approx cheap": 2} +approximation_indexes = {"Full": 0, "Approx NN": 1, "Approx cheap": 2, "TAESD": 3} def single_sample_to_image(sample, approximation=None): if approximation is None: approximation = approximation_indexes.get(opts.show_progress_type, 0) - if approximation == 2: - x_sample = sd_vae_approx.cheap_approximation(sample) - elif approximation == 1: - x_sample = sd_vae_approx.model()(sample.to(devices.device, devices.dtype).unsqueeze(0))[0].detach() + if approximation == 1: + x_sample = sd_vae_approx.model()(sample.to(devices.device, devices.dtype).unsqueeze(0))[0].detach() * 0.5 + 0.5 + elif approximation == 2: + x_sample = sd_vae_approx.cheap_approximation(sample) * 0.5 + 0.5 + elif approximation == 3: + x_sample = sample * 1.5 + x_sample = sd_vae_taesd.model()(x_sample.to(devices.device, devices.dtype).unsqueeze(0))[0].detach() else: - x_sample = processing.decode_first_stage(shared.sd_model, sample.unsqueeze(0))[0] - x_sample = torch.clamp((x_sample + 1.0) / 2.0, min=0.0, max=1.0) + x_sample = processing.decode_first_stage(shared.sd_model, sample.unsqueeze(0))[0] * 0.5 + 0.5 + x_sample = torch.clamp(x_sample, min=0.0, max=1.0) x_sample = 255. * np.moveaxis(x_sample.cpu().numpy(), 0, 2) x_sample = x_sample.astype(np.uint8) return Image.fromarray(x_sample) @@ -55,6 +58,18 @@ def store_latent(decoded): if not shared.parallel_processing_allowed: shared.state.assign_current_image(sample_to_image(decoded)) +def is_sampler_using_eta_noise_seed_delta(p): + """returns whether sampler from config will use eta noise seed delta for image creation""" + sampler_config = sd_samplers.find_sampler_config(p.sampler_name) + eta = p.eta + if eta is None and p.sampler is not None: + eta = p.sampler.eta + if eta is None and sampler_config is not None: + eta = 0 if sampler_config.options.get("default_eta_is_0", False) else 1.0 + if eta == 0: + return False + return sampler_config.options.get("uses_ensd", False) + class InterruptedException(BaseException): pass diff --git a/modules/sd_samplers_compvis.py b/modules/sd_samplers_compvis.py index 5f5544479..e07eb7d80 100644 --- a/modules/sd_samplers_compvis.py +++ b/modules/sd_samplers_compvis.py @@ -11,7 +11,7 @@ import modules.models.diffusion.uni_pc samplers_data_compvis = [ - sd_samplers_common.SamplerData('DDIM', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.ddim.DDIMSampler, model), [], {}), + sd_samplers_common.SamplerData('DDIM', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.ddim.DDIMSampler, model), [], {"default_eta_is_0": True, "uses_ensd": True}), sd_samplers_common.SamplerData('PLMS', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.plms.PLMSSampler, model), [], {}), sd_samplers_common.SamplerData('UniPC', lambda model: VanillaStableDiffusionSampler(modules.models.diffusion.uni_pc.UniPCSampler, model), [], {}), ] @@ -55,7 +55,7 @@ class VanillaStableDiffusionSampler: def p_sample_ddim_hook(self, x_dec, cond, ts, unconditional_conditioning, *args, **kwargs): x_dec, ts, cond, unconditional_conditioning = self.before_sample(x_dec, ts, cond, unconditional_conditioning) - res = self.orig_p_sample_ddim(x_dec, cond, ts, unconditional_conditioning=unconditional_conditioning, *args, **kwargs) + res = self.orig_p_sample_ddim(x_dec, cond, ts, *args, unconditional_conditioning=unconditional_conditioning, **kwargs) x_dec, ts, cond, unconditional_conditioning, res = self.after_sample(x_dec, ts, cond, unconditional_conditioning, res) @@ -83,7 +83,7 @@ class VanillaStableDiffusionSampler: 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' + assert all(len(conds) == 1 for conds in conds_list), 'composition via AND is not supported for DDIM/PLMS samplers' cond = tensor # for DDIM, shapes must match, we can't just process cond and uncond independently; @@ -132,7 +132,10 @@ class VanillaStableDiffusionSampler: self.update_step(x) def initialize(self, p): - self.eta = p.eta if p.eta is not None else shared.opts.eta_ddim + if self.is_ddim: + self.eta = p.eta if p.eta is not None else shared.opts.eta_ddim + else: + self.eta = 0.0 if self.eta != 0.0: p.extra_generation_params["Eta DDIM"] = self.eta diff --git a/modules/sd_samplers_kdiffusion.py b/modules/sd_samplers_kdiffusion.py index fbaa26873..b138a662c 100644 --- a/modules/sd_samplers_kdiffusion.py +++ b/modules/sd_samplers_kdiffusion.py @@ -11,24 +11,25 @@ from modules.script_callbacks import CFGDenoisedParams, cfg_denoised_callback from modules.script_callbacks import AfterCFGCallbackParams, cfg_after_cfg_callback samplers_k_diffusion = [ - ('Euler a', 'sample_euler_ancestral', ['k_euler_a', 'k_euler_ancestral'], {}), + ('Euler a', 'sample_euler_ancestral', ['k_euler_a', 'k_euler_ancestral'], {"uses_ensd": True}), ('Euler', 'sample_euler', ['k_euler'], {}), ('LMS', 'sample_lms', ['k_lms'], {}), - ('Heun', 'sample_heun', ['k_heun'], {}), + ('Heun', 'sample_heun', ['k_heun'], {"second_order": True}), ('DPM2', 'sample_dpm_2', ['k_dpm_2'], {'discard_next_to_last_sigma': True}), - ('DPM2 a', 'sample_dpm_2_ancestral', ['k_dpm_2_a'], {'discard_next_to_last_sigma': True}), - ('DPM++ 2S a', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a'], {}), + ('DPM2 a', 'sample_dpm_2_ancestral', ['k_dpm_2_a'], {'discard_next_to_last_sigma': True, "uses_ensd": True}), + ('DPM++ 2S a', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a'], {"uses_ensd": True, "second_order": True}), ('DPM++ 2M', 'sample_dpmpp_2m', ['k_dpmpp_2m'], {}), - ('DPM++ SDE', 'sample_dpmpp_sde', ['k_dpmpp_sde'], {}), - ('DPM fast', 'sample_dpm_fast', ['k_dpm_fast'], {}), - ('DPM adaptive', 'sample_dpm_adaptive', ['k_dpm_ad'], {}), + ('DPM++ SDE', 'sample_dpmpp_sde', ['k_dpmpp_sde'], {"second_order": True, "brownian_noise": True}), + ('DPM++ 2M SDE', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_ka'], {"brownian_noise": True, 'discard_next_to_last_sigma': True}), + ('DPM fast', 'sample_dpm_fast', ['k_dpm_fast'], {"uses_ensd": True}), + ('DPM adaptive', 'sample_dpm_adaptive', ['k_dpm_ad'], {"uses_ensd": True}), ('LMS Karras', 'sample_lms', ['k_lms_ka'], {'scheduler': 'karras'}), - ('DPM2 Karras', 'sample_dpm_2', ['k_dpm_2_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True}), - ('DPM2 a Karras', 'sample_dpm_2_ancestral', ['k_dpm_2_a_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True}), - ('DPM++ 2S a Karras', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a_ka'], {'scheduler': 'karras'}), + ('DPM2 Karras', 'sample_dpm_2', ['k_dpm_2_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True, "uses_ensd": True, "second_order": True}), + ('DPM2 a Karras', 'sample_dpm_2_ancestral', ['k_dpm_2_a_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True, "uses_ensd": True, "second_order": True}), + ('DPM++ 2S a Karras', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a_ka'], {'scheduler': 'karras', "uses_ensd": True, "second_order": True}), ('DPM++ 2M Karras', 'sample_dpmpp_2m', ['k_dpmpp_2m_ka'], {'scheduler': 'karras'}), - ('DPM++ SDE Karras', 'sample_dpmpp_sde', ['k_dpmpp_sde_ka'], {'scheduler': 'karras'}), - ('DPM++ 2M SDE', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde'], {}), + ('DPM++ SDE Karras', 'sample_dpmpp_sde', ['k_dpmpp_sde_ka'], {'scheduler': 'karras', "second_order": True, "brownian_noise": True}), + ('DPM++ 2M SDE Karras', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_ka'], {'scheduler': 'karras', "brownian_noise": True, 'discard_next_to_last_sigma': True}), ] samplers_data_k_diffusion = [ @@ -83,22 +84,22 @@ class CFGDenoiser(torch.nn.Module): # at self.image_cfg_scale == 1.0 produced results for edit model are the same as with normal sampling, # so is_edit_model is set to False to support AND composition. - is_edit_model = (shared.sd_model is not None) and hasattr(shared.sd_model, 'cond_stage_key') and (shared.sd_model.cond_stage_key == "edit") and (self.image_cfg_scale is not None) and (self.image_cfg_scale != 1.0) + is_edit_model = shared.sd_model.cond_stage_key == "edit" and self.image_cfg_scale is not None and self.image_cfg_scale != 1.0 conds_list, tensor = prompt_parser.reconstruct_multicond_batch(cond, self.step) uncond = prompt_parser.reconstruct_cond_batch(uncond, self.step) - assert not is_edit_model or all([len(conds) == 1 for conds in conds_list]), "AND is not supported for InstructPix2Pix checkpoint (unless using Image CFG scale = 1.0)" + assert not is_edit_model or all(len(conds) == 1 for conds in conds_list), "AND is not supported for InstructPix2Pix checkpoint (unless using Image CFG scale = 1.0)" batch_size = len(conds_list) repeats = [len(conds_list[i]) for i in range(batch_size)] if shared.sd_model.model.conditioning_key == "crossattn-adm": image_uncond = torch.zeros_like(image_cond) - make_condition_dict = lambda c_crossattn, c_adm: {"c_crossattn": c_crossattn, "c_adm": c_adm} # pylint: disable=unnecessary-lambda-assignment + make_condition_dict = lambda c_crossattn, c_adm: {"c_crossattn": c_crossattn, "c_adm": c_adm} # pylint: disable=C3001 else: image_uncond = image_cond - make_condition_dict = lambda c_crossattn, c_concat: {"c_crossattn": c_crossattn, "c_concat": [c_concat]} # pylint: disable=unnecessary-lambda-assignment + make_condition_dict = lambda c_crossattn, c_concat: {"c_crossattn": c_crossattn, "c_concat": [c_concat]} # pylint: disable=C3001 if not is_edit_model: x_in = torch.cat([torch.stack([x[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [x]) @@ -168,8 +169,6 @@ class CFGDenoiser(torch.nn.Module): devices.test_for_nans(x_out, "unet") if opts.live_preview_content == "Prompt": - p_step = len(x_out) // batch_size - 1 - p_step = p_step if p_step > 1 else 1 sd_samplers_common.store_latent(torch.cat([x_out[i:i+1] for i in denoised_image_indexes])) elif opts.live_preview_content == "Negative prompt": sd_samplers_common.store_latent(x_out[-uncond.shape[0]:]) @@ -186,11 +185,9 @@ class CFGDenoiser(torch.nn.Module): after_cfg_callback_params = AfterCFGCallbackParams(denoised, state.sampling_step, state.sampling_steps) cfg_after_cfg_callback(after_cfg_callback_params) - if after_cfg_callback_params.output_altered: - denoised = after_cfg_callback_params.x + denoised = after_cfg_callback_params.x self.step += 1 - return denoised @@ -215,7 +212,7 @@ class TorchHijack: if noise.shape == x.shape: return noise - if x.device.type == 'mps': + if opts.randn_source == "CPU" or x.device.type == 'mps': return torch.randn_like(x, device=devices.cpu).to(x.device) else: return torch.randn_like(x) @@ -233,9 +230,10 @@ class KDiffusionSampler: self.sampler_noises = None self.stop_at = None self.eta = None - self.config = None + self.config = None # set by the function calling the constructor self.last_latent = None self.s_min_uncond = None + self.conditioning_key = sd_model.model.conditioning_key def callback_state(self, d): @@ -249,6 +247,7 @@ class KDiffusionSampler: raise sd_samplers_common.InterruptedException state.sampling_step = step + shared.total_tqdm.update() def launch_sampling(self, steps, func): state.sampling_steps = steps @@ -296,7 +295,7 @@ class KDiffusionSampler: if p.sampler_noise_scheduler_override: sigmas = p.sampler_noise_scheduler_override(steps) elif self.config is not None and self.config.options.get('scheduler', None) == 'karras': - sigma_min, sigma_max = (self.model_wrap.sigmas[0].item(), self.model_wrap.sigmas[-1].item()) + sigma_min, sigma_max = (0.1, 10) if opts.use_old_karras_scheduler_sigmas else (self.model_wrap.sigmas[0].item(), self.model_wrap.sigmas[-1].item()) sigmas = k_diffusion.sampling.get_sigmas_karras(n=steps, sigma_min=sigma_min, sigma_max=sigma_max, device=shared.device) else: @@ -308,16 +307,12 @@ class KDiffusionSampler: return sigmas def create_noise_sampler(self, x, sigmas, p): + """For DPM++ SDE: manually create noise sampler to enable deterministic results across different batch sizes""" + if shared.opts.no_dpmpp_sde_batch_determinism: + return None + from k_diffusion.sampling import BrownianTreeNoiseSampler - - positive_sigmas = sigmas[sigmas > 0] - - if positive_sigmas.numel() > 0: - sigma_min = positive_sigmas.min(dim=0)[0] - else: - sigma_min = 0 - - sigma_max = sigmas.max() + sigma_min, sigma_max = sigmas[sigmas > 0].min(), sigmas.max() current_iter_seeds = p.all_seeds[p.iteration * p.batch_size:(p.iteration + 1) * p.batch_size] return BrownianTreeNoiseSampler(x, sigma_min, sigma_max, seed=current_iter_seeds) @@ -344,16 +339,16 @@ class KDiffusionSampler: if 'sigmas' in parameters: extra_params_kwargs['sigmas'] = sigma_sched - if self.funcname == 'sample_dpmpp_sde': + if self.config.options.get('brownian_noise', False): noise_sampler = self.create_noise_sampler(x, sigmas, p) extra_params_kwargs['noise_sampler'] = noise_sampler self.model_wrap_cfg.init_latent = x self.last_latent = x - extra_args={ - 'cond': conditioning, - 'image_cond': image_conditioning, - 'uncond': unconditional_conditioning, + extra_args = { + 'cond': conditioning, + 'image_cond': image_conditioning, + 'uncond': unconditional_conditioning, 'cond_scale': p.cfg_scale, 's_min_uncond': self.s_min_uncond } @@ -380,15 +375,15 @@ class KDiffusionSampler: else: extra_params_kwargs['sigmas'] = sigmas - if self.funcname == 'sample_dpmpp_sde': + if self.config.options.get('brownian_noise', False): noise_sampler = self.create_noise_sampler(x, sigmas, p) extra_params_kwargs['noise_sampler'] = noise_sampler self.last_latent = x samples = self.launch_sampling(steps, lambda: self.func(self.model_wrap_cfg, x, extra_args={ - 'cond': conditioning, - 'image_cond': image_conditioning, - 'uncond': unconditional_conditioning, + 'cond': conditioning, + 'image_cond': image_conditioning, + 'uncond': unconditional_conditioning, 'cond_scale': p.cfg_scale, 's_min_uncond': self.s_min_uncond }, disable=False, callback=self.callback_state, **extra_params_kwargs)) diff --git a/modules/sd_vae_taesd.py b/modules/sd_vae_taesd.py new file mode 100644 index 000000000..74ad13926 --- /dev/null +++ b/modules/sd_vae_taesd.py @@ -0,0 +1,88 @@ +""" +Tiny AutoEncoder for Stable Diffusion +(DNN for encoding / decoding SD's latent space) + +https://github.com/madebyollin/taesd +""" +import os +import torch +import torch.nn as nn + +from modules import devices, paths_internal + +sd_vae_taesd = None + + +def conv(n_in, n_out, **kwargs): + return nn.Conv2d(n_in, n_out, 3, padding=1, **kwargs) + + +class Clamp(nn.Module): + @staticmethod + def forward(x): + return torch.tanh(x / 3) * 3 + + +class Block(nn.Module): + def __init__(self, n_in, n_out): + super().__init__() + self.conv = nn.Sequential(conv(n_in, n_out), nn.ReLU(), conv(n_out, n_out), nn.ReLU(), conv(n_out, n_out)) + self.skip = nn.Conv2d(n_in, n_out, 1, bias=False) if n_in != n_out else nn.Identity() + self.fuse = nn.ReLU() + + def forward(self, x): + return self.fuse(self.conv(x) + self.skip(x)) + + +def decoder(): + return nn.Sequential( + Clamp(), conv(4, 64), nn.ReLU(), + Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False), + Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False), + Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False), + Block(64, 64), conv(64, 3), + ) + + +class TAESD(nn.Module): # pylint: disable=abstract-method + latent_magnitude = 3 + latent_shift = 0.5 + + def __init__(self, decoder_path="taesd_decoder.pth"): + """Initialize pretrained TAESD on the given device from the given checkpoints.""" + super().__init__() + self.decoder = decoder() + self.decoder.load_state_dict( + torch.load(decoder_path, map_location='cpu' if devices.device.type != 'cuda' else None)) + + @staticmethod + def unscale_latents(x): + """[0, 1] -> raw latents""" + return x.sub(TAESD.latent_shift).mul(2 * TAESD.latent_magnitude) + + +def download_model(model_path): + model_url = 'https://github.com/madebyollin/taesd/raw/main/taesd_decoder.pth' + + if not os.path.exists(model_path): + os.makedirs(os.path.dirname(model_path), exist_ok=True) + + print(f'Downloading TAESD decoder to: {model_path}') + torch.hub.download_url_to_file(model_url, model_path) + + +def model(): + global sd_vae_taesd # pylint: disable=global-statement + + if sd_vae_taesd is None: + model_path = os.path.join(paths_internal.models_path, "VAE-taesd", "taesd_decoder.pth") + download_model(model_path) + + if os.path.exists(model_path): + sd_vae_taesd = TAESD(model_path) + sd_vae_taesd.eval() + sd_vae_taesd.to(devices.device, devices.dtype) + else: + raise FileNotFoundError('TAESD model not found') + + return sd_vae_taesd.decoder diff --git a/modules/shared.py b/modules/shared.py index a1fce4156..df2638f70 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -177,7 +177,7 @@ state.server_start = time.time() class OptionInfo: - def __init__(self, default=None, label="", component=None, component_args=None, onchange=None, section=None, refresh=None): + def __init__(self, default=None, label="", component=None, component_args=None, onchange=None, section=None, refresh=None, comment_before='', comment_after=''): self.default = default self.label = label self.component = component @@ -185,10 +185,28 @@ class OptionInfo: self.onchange = onchange self.section = section self.refresh = refresh + self.comment_before = comment_before # HTML text that will be added after label in UI + self.comment_after = comment_after # HTML text that will be added before label in UI + + def link(self, label, uri): + self.comment_before += f"[{label}]" + return self + + def js(self, label, js_func): + self.comment_before += f"[{label}]" + return self + + def info(self, info): + self.comment_after += f"({info})" + return self + + def needs_restart(self): + self.comment_after += " (requires restart)" + return self def options_section(section_identifier, options_dict): - for _k, v in options_dict.items(): + for v in options_dict.values(): v.section = section_identifier return options_dict @@ -450,8 +468,8 @@ options_templates.update(options_section(('ui', "Live previews"), { "show_progress_grid": OptionInfo(True, "Show previews of all images generated in a batch as a grid"), "notification_audio_enable": OptionInfo(False, "Play a sound when images are finished generating"), "notification_audio_path": OptionInfo("html/notification.mp3","Path to notification sound", component_args=hide_dirs), - "show_progress_every_n_steps": OptionInfo(1, "Show new live preview image every N sampling steps. Set to -1 to show after completion of batch.", gr.Slider, {"minimum": -1, "maximum": 32, "step": 1}), - "show_progress_type": OptionInfo("Approx NN", "Image creation progress preview mode", gr.Radio, {"choices": ["Full", "Approx NN", "Approx cheap"]}), + "show_progress_every_n_steps": OptionInfo(1, "Live preview display period", gr.Slider, {"minimum": -1, "maximum": 32, "step": 1}).info("in sampling steps - show new live preview image every N sampling steps; -1 = only show after completion of batch"), + "show_progress_type": OptionInfo("TAESD", "Live preview method", gr.Radio, {"choices": ["Full", "Approx NN", "Approx cheap", "TAESD"]}).info("Full = slow but pretty; Approx NN and TAESD = fast but low quality; Approx cheap = super fast but terrible otherwise"), "live_preview_content": OptionInfo("Combined", "Live preview subject", gr.Radio, {"choices": ["Combined", "Prompt", "Negative prompt"]}), "live_preview_refresh_period": OptionInfo(250, "Progressbar/preview update period, in milliseconds") }))