Merge pull request AUTOMATIC1111#4 from Klace/img2img_integration

Img2img integration

Author: Klace

aes_scores.json

@@ -76,7 +76,7 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args):
                 processed_image.save(os.path.join(output_dir, filename))
 
 
-def img2img(id_task: str, mode: int, prompt: str, negative_prompt: str, prompt_styles, init_img, sketch, init_img_with_mask, inpaint_color_sketch, inpaint_color_sketch_orig, init_img_inpaint, init_mask_inpaint, steps: int, sampler_index: int, mask_blur: int, mask_alpha: float, 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, img2img_batch_inpaint_mask_dir: str, override_settings_texts, *args):
+def img2img(id_task: str, mode: int, prompt: str, negative_prompt: str, prompt_styles, init_img, sketch, init_img_with_mask, inpaint_color_sketch, inpaint_color_sketch_orig, init_img_inpaint, init_mask_inpaint, steps: int, sampler_index: int, mask_blur: int, mask_alpha: float, inpainting_fill: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, image_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, img2img_batch_inpaint_mask_dir: str, override_settings_texts, *args):
     override_settings = create_override_settings_dict(override_settings_texts)
 
     is_batch = mode == 5
@@ -132,6 +132,7 @@ def img2img(id_task: str, mode: int, prompt: str, negative_prompt: str, prompt_s
         n_iter=n_iter,
         steps=steps,
         cfg_scale=cfg_scale,
+        image_cfg_scale=image_cfg_scale,
         width=width,
         height=height,
         restore_faces=restore_faces,

exif_data.json

@@ -11,6 +11,8 @@
 import cv2
 from skimage import exposure
 from typing import Any, Dict, List, Optional
+from torch import autocast
+
 
 import modules.sd_hijack
 from modules import devices, prompt_parser, masking, sd_samplers, lowvram, generation_parameters_copypaste, script_callbacks, extra_networks, sd_vae_approx, scripts
@@ -186,7 +188,11 @@ def depth2img_image_conditioning(self, source_image):
         return conditioning
 
     def edit_image_conditioning(self, source_image):
-        conditioning_image = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(source_image))
+        #source_image = 2 * torch.tensor(np.array(source_image)).float() / 255 - 1
+        #source_image = rearrange(source_image, "h w c -> 1 c h w").to(shared.device)
+        #source_image = rearrange(source_image, "h w c -> 1 c h w").to(shared.device)
+        #conditioning_image = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(source_image))
+        conditioning_image = self.sd_model.encode_first_stage(source_image).mode()
 
         return conditioning_image
 
@@ -450,11 +456,14 @@ def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments=None, iter
         "Size": f"{p.width}x{p.height}",
         "Model hash": getattr(p, 'sd_model_hash', None if not opts.add_model_hash_to_info or not shared.sd_model.sd_model_hash else shared.sd_model.sd_model_hash),
         "Model": (None if not opts.add_model_name_to_info or not shared.sd_model.sd_checkpoint_info.model_name else shared.sd_model.sd_checkpoint_info.model_name.replace(',', '').replace(':', '')),
+        "Batch size": (None if p.batch_size < 2 else p.batch_size),
+        "Batch pos": (None if p.batch_size < 2 else position_in_batch),
         "Variation seed": (None if p.subseed_strength == 0 else all_subseeds[index]),
         "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),
         "Conditional mask weight": getattr(p, "inpainting_mask_weight", shared.opts.inpainting_mask_weight) if p.is_using_inpainting_conditioning else None,
+        "Eta": (None),
         "Clip skip": None if clip_skip <= 1 else clip_skip,
         "ENSD": None if opts.eta_noise_seed_delta == 0 else opts.eta_noise_seed_delta,
     }
@@ -622,15 +631,17 @@ def get_conds_with_caching(function, required_prompts, steps, cache):
             if p.n_iter > 1:
                 shared.state.job = f"Batch {n+1} out of {p.n_iter}"
 
+            print(f"c = {c} and uc = {uc}")
             with devices.without_autocast() if devices.unet_needs_upcast else devices.autocast():
                 samples_ddim = p.sample(conditioning=c, unconditional_conditioning=uc, seeds=seeds, subseeds=subseeds, subseed_strength=p.subseed_strength, prompts=prompts)
 
             x_samples_ddim = [decode_first_stage(p.sd_model, samples_ddim[i:i+1].to(dtype=devices.dtype_vae))[0].cpu() for i in range(samples_ddim.size(0))]
-            for x in x_samples_ddim:
-                devices.test_for_nans(x, "vae")
+            #for x in x_samples_ddim:
+            #    devices.test_for_nans(x, "vae")
 
             x_samples_ddim = torch.stack(x_samples_ddim).float()
             x_samples_ddim = torch.clamp((x_samples_ddim + 1.0) / 2.0, min=0.0, max=1.0)
+            #x_samples_ddim = 255.0 * rearrange(x_samples_ddim, "1 c h w -> h w c")
 
             del samples_ddim
 
@@ -645,7 +656,7 @@ def get_conds_with_caching(function, required_prompts, steps, cache):
             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)
-
+                #x_sample = 255.0 * rearrange(x_sample, "1 c h w -> h w c")
                 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")
@@ -868,8 +879,8 @@ def save_intermediate(image, index):
                 save_intermediate(image, i)
 
                 image = images.resize_image(0, image, target_width, target_height, upscaler_name=self.hr_upscaler)
-                image = np.array(image).astype(np.float32) / 255.0
-                image = np.moveaxis(image, 2, 0)
+                image = np.array(image).astype(np.float32) / 255.0 - 1
+                #image = np.moveaxis(image, 2, 0)
                 batch_images.append(image)
 
             decoded_samples = torch.from_numpy(np.array(batch_images))
@@ -901,7 +912,7 @@ def save_intermediate(image, index):
 class StableDiffusionProcessingImg2Img(StableDiffusionProcessing):
     sampler = None
 
-    def __init__(self, init_images: list = None, resize_mode: int = 0, denoising_strength: float = 0.75, mask: Any = None, mask_blur: int = 4, inpainting_fill: int = 0, inpaint_full_res: bool = True, inpaint_full_res_padding: int = 0, inpainting_mask_invert: int = 0, initial_noise_multiplier: float = None, **kwargs):
+    def __init__(self, init_images: list = None, resize_mode: int = 0, denoising_strength: float = 0.75, mask: Any = None, mask_blur: int = 4, inpainting_fill: int = 0, inpaint_full_res: bool = True, inpaint_full_res_padding: int = 0, inpainting_mask_invert: int = 0, image_cfg_scale: float = 7.5, initial_noise_multiplier: float = None, **kwargs):
         super().__init__(**kwargs)
 
         self.init_images = init_images
@@ -916,6 +927,7 @@ def __init__(self, init_images: list = None, resize_mode: int = 0, denoising_str
         self.inpaint_full_res = inpaint_full_res
         self.inpaint_full_res_padding = inpaint_full_res_padding
         self.inpainting_mask_invert = inpainting_mask_invert
+        self.image_cfg_scale=image_cfg_scale
         self.initial_noise_multiplier = opts.initial_noise_multiplier if initial_noise_multiplier is None else initial_noise_multiplier
         self.mask = None
         self.nmask = None
@@ -983,9 +995,16 @@ def init(self, all_prompts, all_seeds, all_subseeds):
 
             if add_color_corrections:
                 self.color_corrections.append(setup_color_correction(image))
-
-            image = np.array(image).astype(np.float32) / 255.0
-            image = np.moveaxis(image, 2, 0)
+            width, height = image.size
+            factor = self.width / max(width, height)
+            factor = math.ceil(min(width, height) * factor / 64) * 64 / min(width, height)
+            width = int((width * factor) // 64) * 64
+            height = int((height * factor) // 64) * 64
+            image = ImageOps.fit(image, (width, height), method=Image.Resampling.LANCZOS)   
+                  
+            #image = 2 * torch.tensor(np.array(image)).float() / 255 - 1
+            #image = np.array(image).astype(np.float32) / 255.0
+            #image = np.moveaxis(image, 2, 0)
 
             imgs.append(image)
 
@@ -1002,10 +1021,22 @@ def init(self, all_prompts, all_seeds, all_subseeds):
             batch_images = np.array(imgs)
         else:
             raise RuntimeError(f"bad number of images passed: {len(imgs)}; expecting {self.batch_size} or less")
-
-        image = torch.from_numpy(batch_images)
-        image = 2. * image - 1.
-        image = image.to(shared.device)
+        
+        #image = torch.from_numpy(batch_images)
+        #width, height = image.size
+        #factor = 512 / max(width, height)
+        ###factor = math.ceil(min(width, height) * factor / 64) * 64 / min(width, height)
+        #width = int((width * factor) // 64) * 64
+        #height = int((height * factor) // 64) * 64
+        #image = ImageOps.fit(image, (width, height), method=Image.Resampling.LANCZOS)
+        ##image = 2. * image - 1.
+        #image = rearrange(image, "h w c -> 1 c h w")
+        #image = image.to(shared.device)
+        #image = torch.from_numpy(batch_images)
+        #image = 2. * image - 1.
+        image = 2 * torch.tensor(np.array(image)).float() / 255 - 1
+        image = rearrange(image, "h w c -> 1 c h w").to(shared.device)
+        #image = image.to(shared.device)
 
         self.init_latent = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(image))
 
@@ -1032,6 +1063,7 @@ def init(self, all_prompts, all_seeds, all_subseeds):
         self.image_conditioning = self.img2img_image_conditioning(image, self.init_latent, image_mask)
 
     def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts):
+        
         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)
 
         if self.initial_noise_multiplier != 1.0:

modules/img2img.py

@@ -3,6 +3,7 @@
 import torch
 from PIL import Image
 import torchsde._brownian.brownian_interval
+from einops import rearrange
 from modules import devices, processing, images, sd_vae_approx
 
 from modules.shared import opts, state
@@ -38,8 +39,10 @@ def single_sample_to_image(sample, approximation=None):
         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 = 255.0 * rearrange(x_sample, "1 c h w -> h w c")
     x_sample = 255. * np.moveaxis(x_sample.cpu().numpy(), 0, 2)
     x_sample = x_sample.astype(np.uint8)
+    
     return Image.fromarray(x_sample)
 
 

modules/processing.py

@@ -1,6 +1,7 @@
 from collections import deque
 import torch
 import inspect
+import einops
 import k_diffusion.sampling
 from modules import prompt_parser, devices, sd_samplers_common
 
@@ -57,17 +58,17 @@ def __init__(self, model):
         self.init_latent = None
         self.step = 0
 
-    def combine_denoised(self, x_out, conds_list, uncond, cond_scale):
+    def combine_denoised(self, x_out, conds_list, uncond, cond_scale, image_scale):
         denoised_uncond = x_out[-uncond.shape[0]:]
         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)
-
+                denoised[i] += cond_scale * (x_out[cond_index] - denoised_uncond[i]) + image_scale * (denoised_uncond[i] - x_out[cond_index])
+    
         return denoised
 
-    def forward(self, x, sigma, uncond, cond, cond_scale, image_cond):
+    def forward(self, x, sigma, uncond, cond, cond_scale, image_cond, image_scale):
         if state.interrupted or state.skipped:
             raise sd_samplers_common.InterruptedException
 
@@ -76,10 +77,9 @@ def forward(self, x, sigma, uncond, cond, cond_scale, image_cond):
 
         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])
-        image_cond_in = torch.cat([torch.stack([image_cond[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [image_cond])
-        sigma_in = torch.cat([torch.stack([sigma[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [sigma])
+        x_in = einops.repeat(x, "1 ... -> n ...", n=3)
+        sigma_in = einops.repeat(sigma, "1 ... -> n ...", n=3)
+        image_cond_in = torch.cat([torch.stack([image_cond[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [image_cond] + [image_cond])
 
         denoiser_params = CFGDenoiserParams(x_in, image_cond_in, sigma_in, state.sampling_step, state.sampling_steps)
         cfg_denoiser_callback(denoiser_params)
@@ -88,7 +88,7 @@ def forward(self, x, sigma, uncond, cond, cond_scale, image_cond):
         sigma_in = denoiser_params.sigma
 
         if tensor.shape[1] == uncond.shape[1]:
-            cond_in = torch.cat([tensor, uncond])
+            cond_in = torch.cat([tensor, uncond, uncond])
 
             if shared.batch_cond_uncond:
                 x_out = self.inner_model(x_in, sigma_in, cond={"c_crossattn": [cond_in], "c_concat": [image_cond_in]})
@@ -115,7 +115,7 @@ def forward(self, x, sigma, uncond, cond, cond_scale, image_cond):
         elif opts.live_preview_content == "Negative prompt":
             sd_samplers_common.store_latent(x_out[-uncond.shape[0]:])
 
-        denoised = self.combine_denoised(x_out, conds_list, uncond, cond_scale)
+        denoised = self.combine_denoised(x_out, conds_list, uncond, cond_scale, image_scale)
 
         if self.mask is not None:
             denoised = self.init_latent * self.mask + self.nmask * denoised
@@ -124,6 +124,32 @@ def forward(self, x, sigma, uncond, cond, cond_scale, image_cond):
 
         return denoised
 
+class CFGDenoiserIp2p(torch.nn.Module):
+    def __init__(self, model):
+        super().__init__()
+        self.inner_model = model
+        self.mask = None
+        self.nmask = None
+        self.init_latent = None
+        self.step = 0
+
+    def forward(self, z, sigma, uncond, cond, cond_scale, image_cond):
+        if state.interrupted or state.skipped:
+            raise sd_samplers_common.InterruptedException
+        cfg_z = einops.repeat(z, "1 ... -> n ...", n=3)
+        cfg_sigma = einops.repeat(sigma, "1 ... -> n ...", n=3)
+        image_cond_in = einops.repeat(image_cond, "1 ... -> n ...", n=3)
+        
+        conds_list, tensor = prompt_parser.reconstruct_multicond_batch(cond, self.step)
+        uncond = prompt_parser.reconstruct_cond_batch(uncond, self.step)
+        cond_in = torch.cat([tensor, uncond])
+        cfg_cond = {
+            "c_crossattn": [cond_in],
+            "c_concat": [image_cond_in],
+        }
+        out_cond, out_img_cond, out_uncond = self.inner_model(cfg_z, cfg_sigma, cond=cfg_cond).chunk(3)
+        return out_uncond + cond_scale * (out_cond - out_img_cond) + 1.5 * (out_img_cond - out_uncond)
+
 
 class TorchHijack:
     def __init__(self, sampler_noises):
@@ -265,7 +291,8 @@ def sample_img2img(self, p, x, noise, conditioning, unconditional_conditioning,
             'cond': conditioning, 
             'image_cond': image_conditioning, 
             'uncond': unconditional_conditioning, 
-            'cond_scale': p.cfg_scale
+            'cond_scale': p.cfg_scale,
+            'image_scale': p.image_cfg_scale
         }, disable=False, callback=self.callback_state, **extra_params_kwargs))
 
         return samples

modules/sd_samplers_common.py

@@ -765,7 +765,8 @@ def copy_image(img):
 
                     elif category == "cfg":
                         with FormGroup():
-                            cfg_scale = gr.Slider(minimum=1.0, maximum=30.0, step=0.5, label='CFG Scale', value=7.0, elem_id="img2img_cfg_scale")
+                            cfg_scale = gr.Slider(minimum=1.0, maximum=30.0, step=0.5, label='CFG Scale', value=7.5, elem_id="img2img_cfg_scale")
+                            image_cfg_scale = gr.Slider(minimum=1.0, maximum=30.0, step=0.5, label='Image CFG Scale', value=1.5, elem_id="img2img_cfg_scale")
                             denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.75, elem_id="img2img_denoising_strength")
 
                     elif category == "seed":
@@ -861,6 +862,7 @@ def select_img2img_tab(tab):
                     batch_count,
                     batch_size,
                     cfg_scale,
+                    image_cfg_scale,
                     denoising_strength,
                     seed,
                     subseed, subseed_strength, seed_resize_from_h, seed_resize_from_w, seed_checkbox,