processing.py

from __future__ import annotations
from modules.shared import opts
import json
import logging
import math
import os
import sys
import hashlib
from dataclasses import dataclass, field

import torch
import numpy as np
from PIL import Image, ImageOps
import random
import cv2
from skimage import exposure
from typing import Any

import modules.sd_hijack
from modules import devices, prompt_parser, masking, sd_samplers, lowvram, infotext_utils, extra_networks, sd_vae_approx, scripts, sd_samplers_common, sd_unet, errors, rng, profiling
from modules.rng import slerp # noqa: F401
from modules.sd_hijack import model_hijack
from modules.sd_samplers_common import images_tensor_to_samples, decode_first_stage, approximation_indexes
from modules.shared import opts, cmd_opts, state
import modules.processing_scripts.comments as comments_parser
import modules.shared as shared
import modules.paths as paths
import modules.face_restoration
import modules.images as images
import modules.styles
import modules.sd_models as sd_models
import modules.sd_vae as sd_vae
from ldm.data.util import AddMiDaS
from ldm.models.diffusion.ddpm import LatentDepth2ImageDiffusion
from ldm_patched.contrib.external_model_advanced import rescale_zero_terminal_snr_sigmas

from einops import repeat, rearrange
from blendmodes.blend import blendLayers, BlendType
from modules.sd_models import apply_token_merging
from modules_forge.forge_util import apply_circular_forge


# some of those options should not be changed at all because they would break the model, so I removed them from options.
opt_C = 4
opt_f = 8


def setup_color_correction(image):
    logging.info("Calibrating color correction.")
    correction_target = cv2.cvtColor(np.asarray(image.copy()), cv2.COLOR_RGB2LAB)
    return correction_target


def apply_color_correction(correction, original_image):
    logging.info("Applying color correction.")
    image = Image.fromarray(cv2.cvtColor(exposure.match_histograms(
        cv2.cvtColor(
            np.asarray(original_image),
            cv2.COLOR_RGB2LAB
        ),
        correction,
        channel_axis=2
    ), cv2.COLOR_LAB2RGB).astype("uint8"))

    image = blendLayers(image, original_image, BlendType.LUMINOSITY)

    return image.convert('RGB')


def uncrop(image, dest_size, paste_loc):
    x, y, w, h = paste_loc
    base_image = Image.new('RGBA', dest_size)
    image = images.resize_image(1, image, w, h)
    base_image.paste(image, (x, y))
    image = base_image

    return image


def apply_overlay(image, paste_loc, overlay):
    if overlay is None:
        return image, image.copy()

    if paste_loc is not None:
        image = uncrop(image, (overlay.width, overlay.height), paste_loc)

    original_denoised_image = image.copy()

    image = image.convert('RGBA')
    image.alpha_composite(overlay)
    image = image.convert('RGB')

    return image, original_denoised_image

def create_binary_mask(image, round=True):
    if image.mode == 'RGBA' and image.getextrema()[-1] != (255, 255):
        if round:
            image = image.split()[-1].convert("L").point(lambda x: 255 if x > 128 else 0)
        else:
            image = image.split()[-1].convert("L")
    else:
        image = image.convert('L')
    return image

def txt2img_image_conditioning(sd_model, x, width, height):
    if sd_model.model.conditioning_key in {'hybrid', 'concat'}: # Inpainting models

        # The "masked-image" in this case will just be all 0.5 since the entire image is masked.
        image_conditioning = torch.ones(x.shape[0], 3, height, width, device=x.device) * 0.5
        image_conditioning = images_tensor_to_samples(image_conditioning, approximation_indexes.get(opts.sd_vae_encode_method))

        # Add the fake full 1s mask to the first dimension.
        image_conditioning = torch.nn.functional.pad(image_conditioning, (0, 0, 0, 0, 1, 0), value=1.0)
        image_conditioning = image_conditioning.to(x.dtype)

        return image_conditioning

    elif sd_model.model.conditioning_key == "crossattn-adm": # UnCLIP models

        return x.new_zeros(x.shape[0], 2*sd_model.noise_augmentor.time_embed.dim, dtype=x.dtype, device=x.device)

    else:
        if sd_model.is_sdxl_inpaint:
            # The "masked-image" in this case will just be all 0.5 since the entire image is masked.
            image_conditioning = torch.ones(x.shape[0], 3, height, width, device=x.device) * 0.5
            image_conditioning = images_tensor_to_samples(image_conditioning,
                                                            approximation_indexes.get(opts.sd_vae_encode_method))

            # Add the fake full 1s mask to the first dimension.
            image_conditioning = torch.nn.functional.pad(image_conditioning, (0, 0, 0, 0, 1, 0), value=1.0)
            image_conditioning = image_conditioning.to(x.dtype)

            return image_conditioning

        # Dummy zero conditioning if we're not using inpainting or unclip models.
        # Still takes up a bit of memory, but no encoder call.
        # Pretty sure we can just make this a 1x1 image since its not going to be used besides its batch size.
        return x.new_zeros(x.shape[0], 5, 1, 1, dtype=x.dtype, device=x.device)


@dataclass(repr=False)
class StableDiffusionProcessing:
    sd_model: object = None
    outpath_samples: str = None
    outpath_grids: str = None
    prompt: str = ""
    prompt_for_display: str = None
    negative_prompt: str = ""
    styles: list[str] = None
    seed: int = -1
    subseed: int = -1
    subseed_strength: float = 0
    seed_resize_from_h: int = -1
    seed_resize_from_w: int = -1
    seed_enable_extras: bool = True
    sampler_name: str = None
    scheduler: str = None
    batch_size: int = 1
    n_iter: int = 1
    steps: int = 50
    cfg_scale: float = 7.0
    width: int = 512
    height: int = 512
    restore_faces: bool = None
    tiling: bool = None
    do_not_save_samples: bool = False
    do_not_save_grid: bool = False
    extra_generation_params: dict[str, Any] = None
    overlay_images: list = None
    eta: float = None
    do_not_reload_embeddings: bool = False
    denoising_strength: float = None
    ddim_discretize: str = None
    s_min_uncond: float = None
    s_churn: float = None
    s_tmax: float = None
    s_tmin: float = None
    s_noise: float = None
    override_settings: dict[str, Any] = None
    override_settings_restore_afterwards: bool = True
    sampler_index: int = None
    refiner_checkpoint: str = None
    refiner_switch_at: float = None
    token_merging_ratio = 0
    token_merging_ratio_hr = 0
    disable_extra_networks: bool = False
    firstpass_image: Image = None

    scripts_value: scripts.ScriptRunner = field(default=None, init=False)
    script_args_value: list = field(default=None, init=False)
    scripts_setup_complete: bool = field(default=False, init=False)

    cached_uc = [None, None, None]
    cached_c = [None, None, None]

    comments: dict = None
    sampler: sd_samplers_common.Sampler | None = field(default=None, init=False)
    is_using_inpainting_conditioning: bool = field(default=False, init=False)
    paste_to: tuple | None = field(default=None, init=False)

    is_hr_pass: bool = field(default=False, init=False)

    c: tuple = field(default=None, init=False)
    uc: tuple = field(default=None, init=False)

    rng: rng.ImageRNG | None = field(default=None, init=False)
    step_multiplier: int = field(default=1, init=False)
    color_corrections: list = field(default=None, init=False)

    all_prompts: list = field(default=None, init=False)
    all_negative_prompts: list = field(default=None, init=False)
    all_seeds: list = field(default=None, init=False)
    all_subseeds: list = field(default=None, init=False)
    iteration: int = field(default=0, init=False)
    main_prompt: str = field(default=None, init=False)
    main_negative_prompt: str = field(default=None, init=False)

    prompts: list = field(default=None, init=False)
    negative_prompts: list = field(default=None, init=False)
    seeds: list = field(default=None, init=False)
    subseeds: list = field(default=None, init=False)
    extra_network_data: dict = field(default=None, init=False)

    user: str = field(default=None, init=False)

    sd_model_name: str = field(default=None, init=False)
    sd_model_hash: str = field(default=None, init=False)
    sd_vae_name: str = field(default=None, init=False)
    sd_vae_hash: str = field(default=None, init=False)

    is_api: bool = field(default=False, init=False)

    latents_after_sampling = []
    pixels_after_sampling = []

    def __post_init__(self):
        if self.sampler_index is not None:
            print("sampler_index argument for StableDiffusionProcessing does not do anything; use sampler_name", file=sys.stderr)

        self.comments = {}

        if self.styles is None:
            self.styles = []

        self.sampler_noise_scheduler_override = None

        self.extra_generation_params = self.extra_generation_params or {}
        self.override_settings = self.override_settings or {}
        self.script_args = self.script_args or {}

        self.refiner_checkpoint_info = None

        if not self.seed_enable_extras:
            self.subseed = -1
            self.subseed_strength = 0
            self.seed_resize_from_h = 0
            self.seed_resize_from_w = 0

        self.cached_uc = StableDiffusionProcessing.cached_uc
        self.cached_c = StableDiffusionProcessing.cached_c

        self.extra_result_images = []
        self.latents_after_sampling = []
        self.pixels_after_sampling = []
        self.modified_noise = None

    if opts.sd_sampling == "A1111":
        def fill_fields_from_opts(self):
            self.s_min_uncond = self.s_min_uncond if self.s_min_uncond is not None else opts.s_min_uncond
            self.s_churn = self.s_churn if self.s_churn is not None else opts.s_churn
            self.s_tmin = self.s_tmin if self.s_tmin is not None else opts.s_tmin
            self.s_tmax = (self.s_tmax if self.s_tmax is not None else opts.s_tmax) or float('inf')
            self.s_noise = self.s_noise if self.s_noise is not None else opts.s_noise
    else:
        pass

    @property
    def sd_model(self):
        return shared.sd_model

    @sd_model.setter
    def sd_model(self, value):
        pass

    @property
    def scripts(self):
        return self.scripts_value

    @scripts.setter
    def scripts(self, value):
        self.scripts_value = value

        if self.scripts_value and self.script_args_value and not self.scripts_setup_complete:
            self.setup_scripts()

    @property
    def script_args(self):
        return self.script_args_value

    @script_args.setter
    def script_args(self, value):
        self.script_args_value = value

        if self.scripts_value and self.script_args_value and not self.scripts_setup_complete:
            self.setup_scripts()

    def setup_scripts(self):
        self.scripts_setup_complete = True

        self.scripts.setup_scrips(self, is_ui=not self.is_api)

    def comment(self, text):
        self.comments[text] = 1

    def txt2img_image_conditioning(self, x, width=None, height=None):
        self.is_using_inpainting_conditioning = self.sd_model.model.conditioning_key in {'hybrid', 'concat'}

        return txt2img_image_conditioning(self.sd_model, x, width or self.width, height or self.height)

    def depth2img_image_conditioning(self, source_image):
        # Use the AddMiDaS helper to Format our source image to suit the MiDaS model
        transformer = AddMiDaS(model_type="dpt_hybrid")
        transformed = transformer({"jpg": rearrange(source_image[0], "c h w -> h w c")})
        midas_in = torch.from_numpy(transformed["midas_in"][None, ...]).to(device=shared.device)
        midas_in = repeat(midas_in, "1 ... -> n ...", n=self.batch_size)

        conditioning_image = images_tensor_to_samples(source_image*0.5+0.5, approximation_indexes.get(opts.sd_vae_encode_method))
        conditioning = torch.nn.functional.interpolate(
            self.sd_model.depth_model(midas_in),
            size=conditioning_image.shape[2:],
            mode="bicubic",
            align_corners=False,
        )

        (depth_min, depth_max) = torch.aminmax(conditioning)
        conditioning = 2. * (conditioning - depth_min) / (depth_max - depth_min) - 1.
        return conditioning

    def edit_image_conditioning(self, source_image):
        conditioning_image = shared.sd_model.encode_first_stage(source_image).mode()

        return conditioning_image

    def unclip_image_conditioning(self, source_image):
        c_adm = self.sd_model.embedder(source_image)
        if self.sd_model.noise_augmentor is not None:
            noise_level = 0 # TODO: Allow other noise levels?
            c_adm, noise_level_emb = self.sd_model.noise_augmentor(c_adm, noise_level=repeat(torch.tensor([noise_level]).to(c_adm.device), '1 -> b', b=c_adm.shape[0]))
            c_adm = torch.cat((c_adm, noise_level_emb), 1)
        return c_adm

    def inpainting_image_conditioning(self, source_image, latent_image, image_mask=None, round_image_mask=True):
        self.is_using_inpainting_conditioning = True

        # Handle the different mask inputs
        if image_mask is not None:
            if torch.is_tensor(image_mask):
                conditioning_mask = image_mask
            else:
                conditioning_mask = np.array(image_mask.convert("L"))
                conditioning_mask = conditioning_mask.astype(np.float32) / 255.0
                conditioning_mask = torch.from_numpy(conditioning_mask[None, None])

                if round_image_mask:
                    # Caller is requesting a discretized mask as input, so we round to either 1.0 or 0.0
                    conditioning_mask = torch.round(conditioning_mask)

        else:
            conditioning_mask = source_image.new_ones(1, 1, *source_image.shape[-2:])

        # Create another latent image, this time with a masked version of the original input.
        # Smoothly interpolate between the masked and unmasked latent conditioning image using a parameter.
        conditioning_mask = conditioning_mask.to(device=source_image.device, dtype=source_image.dtype)
        conditioning_image = torch.lerp(
            source_image,
            source_image * (1.0 - conditioning_mask),
            getattr(self, "inpainting_mask_weight", shared.opts.inpainting_mask_weight)
        )

        # Encode the new masked image using first stage of network.
        conditioning_image = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(conditioning_image))

        # Create the concatenated conditioning tensor to be fed to `c_concat`
        conditioning_mask = torch.nn.functional.interpolate(conditioning_mask, size=latent_image.shape[-2:])
        conditioning_mask = conditioning_mask.expand(conditioning_image.shape[0], -1, -1, -1)
        image_conditioning = torch.cat([conditioning_mask, conditioning_image], dim=1)
        image_conditioning = image_conditioning.to(shared.device).type(self.sd_model.dtype)

        return image_conditioning

    def img2img_image_conditioning(self, source_image, latent_image, image_mask=None, round_image_mask=True):
        source_image = devices.cond_cast_float(source_image)

        # HACK: Using introspection as the Depth2Image model doesn't appear to uniquely
        # identify itself with a field common to all models. The conditioning_key is also hybrid.
        if isinstance(self.sd_model, LatentDepth2ImageDiffusion):
            return self.depth2img_image_conditioning(source_image)

        if self.sd_model.cond_stage_key == "edit":
            return self.edit_image_conditioning(source_image)

        if self.sampler.conditioning_key in {'hybrid', 'concat'}:
            return self.inpainting_image_conditioning(source_image, latent_image, image_mask=image_mask, round_image_mask=round_image_mask)

        if self.sampler.conditioning_key == "crossattn-adm":
            return self.unclip_image_conditioning(source_image)

        if self.sampler.model_wrap.inner_model.is_sdxl_inpaint:
            return self.inpainting_image_conditioning(source_image, latent_image, image_mask=image_mask)

        # Dummy zero conditioning if we're not using inpainting or depth model.
        return latent_image.new_zeros(latent_image.shape[0], 5, 1, 1)

    def init(self, all_prompts, all_seeds, all_subseeds):
        pass

    def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts):
        raise NotImplementedError()

    def close(self):
        self.sampler = None
        self.c = None
        self.uc = None
        if not opts.persistent_cond_cache:
            StableDiffusionProcessing.cached_c = [None, None]
            StableDiffusionProcessing.cached_uc = [None, None]

    def get_token_merging_ratio(self, for_hr=False):
        if for_hr:
            return self.token_merging_ratio_hr or opts.token_merging_ratio_hr or self.token_merging_ratio or opts.token_merging_ratio

        return self.token_merging_ratio or opts.token_merging_ratio

    def setup_prompts(self):
        if isinstance(self.prompt,list):
            self.all_prompts = self.prompt
        elif isinstance(self.negative_prompt, list):
            self.all_prompts = [self.prompt] * len(self.negative_prompt)
        else:
            self.all_prompts = self.batch_size * self.n_iter * [self.prompt]

        if isinstance(self.negative_prompt, list):
            self.all_negative_prompts = self.negative_prompt
        else:
            self.all_negative_prompts = [self.negative_prompt] * len(self.all_prompts)

        if len(self.all_prompts) != len(self.all_negative_prompts):
            raise RuntimeError(f"Received a different number of prompts ({len(self.all_prompts)}) and negative prompts ({len(self.all_negative_prompts)})")

        self.all_prompts = [shared.prompt_styles.apply_styles_to_prompt(x, self.styles) for x in self.all_prompts]
        self.all_negative_prompts = [shared.prompt_styles.apply_negative_styles_to_prompt(x, self.styles) for x in self.all_negative_prompts]

        self.main_prompt = self.all_prompts[0]
        self.main_negative_prompt = self.all_negative_prompts[0]

    def cached_params(self, required_prompts, steps, extra_network_data, hires_steps=None, use_old_scheduling=False):
        """Returns parameters that invalidate the cond cache if changed"""

        return (
            required_prompts,
            steps,
            hires_steps,
            use_old_scheduling,
            opts.CLIP_stop_at_last_layers,
            shared.sd_model.sd_checkpoint_info,
            extra_network_data,
            opts.sdxl_crop_left,
            opts.sdxl_crop_top,
            self.width,
            self.height,
            opts.fp8_storage,
            opts.cache_fp16_weight,
            opts.emphasis,
        )

    def get_conds_with_caching(self, function, required_prompts, steps, caches, extra_network_data, hires_steps=None):
        """
        Returns the result of calling function(shared.sd_model, required_prompts, steps)
        using a cache to store the result if the same arguments have been used before.

        cache is an array containing two elements. The first element is a tuple
        representing the previously used arguments, or None if no arguments
        have been used before. The second element is where the previously
        computed result is stored.

        caches is a list with items described above.
        """

        if shared.opts.use_old_scheduling:
            old_schedules = prompt_parser.get_learned_conditioning_prompt_schedules(required_prompts, steps, hires_steps, False)
            new_schedules = prompt_parser.get_learned_conditioning_prompt_schedules(required_prompts, steps, hires_steps, True)
            if old_schedules != new_schedules:
                self.extra_generation_params["Old prompt editing timelines"] = True

        cached_params = self.cached_params(required_prompts, steps, extra_network_data, hires_steps, shared.opts.use_old_scheduling)

        for cache in caches:
            if cache[0] is not None and cached_params == cache[0]:
                if len(cache) > 2:
                    modules.sd_hijack.model_hijack.extra_generation_params.update(cache[2])
                return cache[1]

        cache = caches[0]

        with devices.autocast():
            cache[1] = function(shared.sd_model, required_prompts, steps, hires_steps, shared.opts.use_old_scheduling)
            if len(cache) > 2:
                cache[2] = modules.sd_hijack.model_hijack.extra_generation_params

        cache[0] = cached_params
        return cache[1]

    def setup_conds(self):
        prompts = prompt_parser.SdConditioning(self.prompts, width=self.width, height=self.height)
        negative_prompts = prompt_parser.SdConditioning(self.negative_prompts, width=self.width, height=self.height, is_negative_prompt=True)

        sampler_config = sd_samplers.find_sampler_config(self.sampler_name)
        total_steps = sampler_config.total_steps(self.steps) if sampler_config else self.steps
        self.step_multiplier = total_steps // self.steps
        self.firstpass_steps = total_steps

        self.uc = self.get_conds_with_caching(prompt_parser.get_learned_conditioning, negative_prompts, total_steps, [self.cached_uc], self.extra_network_data)
        self.c = self.get_conds_with_caching(prompt_parser.get_multicond_learned_conditioning, prompts, total_steps, [self.cached_c], self.extra_network_data)

    def get_conds(self):
        return self.c, self.uc

    def parse_extra_network_prompts(self):
        self.prompts, self.extra_network_data = extra_networks.parse_prompts(self.prompts)

    def save_samples(self) -> bool:
        """Returns whether generated images need to be written to disk"""
        return opts.samples_save and not self.do_not_save_samples and (opts.save_incomplete_images or not state.interrupted and not state.skipped)


class Processed:
    def __init__(self, p: StableDiffusionProcessing, images_list, seed=-1, info="", subseed=None, all_prompts=None, all_negative_prompts=None, all_seeds=None, all_subseeds=None, index_of_first_image=0, infotexts=None, comments="", extra_images_list=[]):
        self.images = images_list
        self.extra_images = extra_images_list
        self.prompt = p.prompt
        self.negative_prompt = p.negative_prompt
        self.seed = seed
        self.subseed = subseed
        self.subseed_strength = p.subseed_strength
        self.info = info
        self.comments = "".join(f"{comment}\n" for comment in p.comments)
        self.width = p.width
        self.height = p.height
        self.sampler_name = p.sampler_name
        self.cfg_scale = p.cfg_scale
        self.image_cfg_scale = getattr(p, 'image_cfg_scale', None)
        self.steps = p.steps
        self.batch_size = p.batch_size
        self.restore_faces = p.restore_faces
        self.face_restoration_model = opts.face_restoration_model if p.restore_faces else None
        self.sd_model_name = p.sd_model_name
        self.sd_model_hash = p.sd_model_hash
        self.sd_vae_name = p.sd_vae_name
        self.sd_vae_hash = p.sd_vae_hash
        self.seed_resize_from_w = p.seed_resize_from_w
        self.seed_resize_from_h = p.seed_resize_from_h
        self.denoising_strength = getattr(p, 'denoising_strength', None)
        self.extra_generation_params = p.extra_generation_params
        self.index_of_first_image = index_of_first_image
        self.styles = p.styles
        self.job_timestamp = state.job_timestamp
        self.clip_skip = opts.CLIP_stop_at_last_layers
        self.token_merging_ratio = p.token_merging_ratio
        self.token_merging_ratio_hr = p.token_merging_ratio_hr

        self.eta = p.eta
        self.ddim_discretize = p.ddim_discretize
        self.s_churn = p.s_churn
        self.s_tmin = p.s_tmin
        self.s_tmax = p.s_tmax
        self.s_noise = p.s_noise
        self.s_min_uncond = p.s_min_uncond
        self.sampler_noise_scheduler_override = p.sampler_noise_scheduler_override
        self.prompt = self.prompt if not isinstance(self.prompt, list) else self.prompt[0]
        self.negative_prompt = self.negative_prompt if not isinstance(self.negative_prompt, list) else self.negative_prompt[0]
        self.seed = int(self.seed if not isinstance(self.seed, list) else self.seed[0]) if self.seed is not None else -1
        self.subseed = int(self.subseed if not isinstance(self.subseed, list) else self.subseed[0]) if self.subseed is not None else -1
        self.is_using_inpainting_conditioning = p.is_using_inpainting_conditioning

        self.all_prompts = all_prompts or p.all_prompts or [self.prompt]
        self.all_negative_prompts = all_negative_prompts or p.all_negative_prompts or [self.negative_prompt]
        self.all_seeds = all_seeds or p.all_seeds or [self.seed]
        self.all_subseeds = all_subseeds or p.all_subseeds or [self.subseed]
        self.infotexts = infotexts or [info] * len(images_list)
        self.version = program_version()

    def js(self):
        obj = {
            "prompt": self.all_prompts[0],
            "all_prompts": self.all_prompts,
            "negative_prompt": self.all_negative_prompts[0],
            "all_negative_prompts": self.all_negative_prompts,
            "seed": self.seed,
            "all_seeds": self.all_seeds,
            "subseed": self.subseed,
            "all_subseeds": self.all_subseeds,
            "subseed_strength": self.subseed_strength,
            "width": self.width,
            "height": self.height,
            "sampler_name": self.sampler_name,
            "cfg_scale": self.cfg_scale,
            "steps": self.steps,
            "batch_size": self.batch_size,
            "restore_faces": self.restore_faces,
            "face_restoration_model": self.face_restoration_model,
            "sd_model_name": self.sd_model_name,
            "sd_model_hash": self.sd_model_hash,
            "sd_vae_name": self.sd_vae_name,
            "sd_vae_hash": self.sd_vae_hash,
            "seed_resize_from_w": self.seed_resize_from_w,
            "seed_resize_from_h": self.seed_resize_from_h,
            "denoising_strength": self.denoising_strength,
            "extra_generation_params": self.extra_generation_params,
            "index_of_first_image": self.index_of_first_image,
            "infotexts": self.infotexts,
            "styles": self.styles,
            "job_timestamp": self.job_timestamp,
            "clip_skip": self.clip_skip,
            "is_using_inpainting_conditioning": self.is_using_inpainting_conditioning,
            "version": self.version,
        }

        return json.dumps(obj, default=lambda o: None)

    def infotext(self, p: StableDiffusionProcessing, index):
        return create_infotext(p, self.all_prompts, self.all_seeds, self.all_subseeds, comments=[], position_in_batch=index % self.batch_size, iteration=index // self.batch_size)

    def get_token_merging_ratio(self, for_hr=False):
        return self.token_merging_ratio_hr if for_hr else self.token_merging_ratio


def create_random_tensors(shape, seeds, subseeds=None, subseed_strength=0.0, seed_resize_from_h=0, seed_resize_from_w=0, p=None):
    g = rng.ImageRNG(shape, seeds, subseeds=subseeds, subseed_strength=subseed_strength, seed_resize_from_h=seed_resize_from_h, seed_resize_from_w=seed_resize_from_w)
    return g.next()


class DecodedSamples(list):
    already_decoded = True


def decode_latent_batch(model, batch, target_device=None, check_for_nans=False):
    samples = DecodedSamples()
    samples_pytorch = decode_first_stage(model, batch).to(target_device)

    for x in samples_pytorch:
        samples.append(x)

    return samples


def get_fixed_seed(seed):
    if seed == '' or seed is None:
        seed = -1
    elif isinstance(seed, str):
        try:
            seed = int(seed)
        except Exception:
            seed = -1

    if seed == -1:
        return int(random.randrange(4294967294))

    return seed


def fix_seed(p):
    p.seed = get_fixed_seed(p.seed)
    p.subseed = get_fixed_seed(p.subseed)


def program_version():
    import launch

    res = launch.git_tag()
    if res == "<none>":
        res = None

    return res


def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments=None, iteration=0, position_in_batch=0, use_main_prompt=False, index=None, all_negative_prompts=None):
    """
    this function is used to generate the infotext that is stored in the generated images, it's contains the parameters that are required to generate the imagee
    Args:
        p: StableDiffusionProcessing
        all_prompts: list[str]
        all_seeds: list[int]
        all_subseeds: list[int]
        comments: list[str]
        iteration: int
        position_in_batch: int
        use_main_prompt: bool
        index: int
        all_negative_prompts: list[str]

    Returns: str

    Extra generation params
    p.extra_generation_params dictionary allows for additional parameters to be added to the infotext
    this can be use by the base webui or extensions.
    To add a new entry, add a new key value pair, the dictionary key will be used as the key of the parameter in the infotext
    the value generation_params can be defined as:
        - str | None
        - List[str|None]
        - callable func(**kwargs) -> str | None

    When defined as a string, it will be used as without extra processing; this is this most common use case.

    Defining as a list allows for parameter that changes across images in the job, for example, the 'Seed' parameter.
    The list should have the same length as the total number of images in the entire job.

    Defining as a callable function allows parameter cannot be generated earlier or when extra logic is required.
    For example 'Hires prompt', due to reasons the hr_prompt might be changed by process in the pipeline or extensions
    and may vary across different images, defining as a static string or list would not work.

    The function takes locals() as **kwargs, as such will have access to variables like 'p' and 'index'.
    the base signature of the function should be:
        func(**kwargs) -> str | None
    optionally it can have additional arguments that will be used in the function:
        func(p, index, **kwargs) -> str | None
    note: for better future compatibility even though this function will have access to all variables in the locals(),
        it is recommended to only use the arguments present in the function signature of create_infotext.
    For actual implementation examples, see StableDiffusionProcessingTxt2Img.init > get_hr_prompt.
    """

    if use_main_prompt:
        index = 0
    elif index is None:
        index = position_in_batch + iteration * p.batch_size

    if all_negative_prompts is None:
        all_negative_prompts = p.all_negative_prompts

    clip_skip = getattr(p, 'clip_skip', opts.CLIP_stop_at_last_layers)
    enable_hr = getattr(p, 'enable_hr', False)
    token_merging_ratio = p.get_token_merging_ratio()
    token_merging_ratio_hr = p.get_token_merging_ratio(for_hr=True)

    prompt_text = p.main_prompt if use_main_prompt else all_prompts[index]
    negative_prompt = p.main_negative_prompt if use_main_prompt else all_negative_prompts[index]

    uses_ensd = opts.eta_noise_seed_delta != 0
    if uses_ensd:
        uses_ensd = sd_samplers_common.is_sampler_using_eta_noise_seed_delta(p)

    generation_params = {
        "Steps": p.steps,
        "Sampler": p.sampler_name,
        "Schedule type": p.scheduler,
        "CFG scale": p.cfg_scale,
        "Image CFG scale": getattr(p, 'image_cfg_scale', None),
        "Seed": p.all_seeds[0] if use_main_prompt else all_seeds[index],
        "Face restoration": opts.face_restoration_model if p.restore_faces else None,
        "Size": f"{p.width}x{p.height}",
        "Model hash": p.sd_model_hash if opts.add_model_hash_to_info else None,
        "Model": p.sd_model_name if opts.add_model_name_to_info else None,
        "FP8 weight": opts.fp8_storage if devices.fp8 else None,
        "Cache FP16 weight for LoRA": opts.cache_fp16_weight if devices.fp8 else None,
        "VAE hash": p.sd_vae_hash if opts.add_vae_hash_to_info else None,
        "VAE": p.sd_vae_name if opts.add_vae_name_to_info else None,
        "Variation seed": (None if p.subseed_strength == 0 else (p.all_subseeds[0] if use_main_prompt 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": p.extra_generation_params.get("Denoising strength"),
        "Conditional mask weight": getattr(p, "inpainting_mask_weight", shared.opts.inpainting_mask_weight) if p.is_using_inpainting_conditioning else None,
        "Clip skip": None if clip_skip <= 1 else clip_skip,
        "ENSD": opts.eta_noise_seed_delta if uses_ensd else None,
        "Token merging ratio": None if token_merging_ratio == 0 else token_merging_ratio,
        "Token merging ratio hr": None if not enable_hr or token_merging_ratio_hr == 0 else token_merging_ratio_hr,
        "Init image hash": getattr(p, 'init_img_hash', None),
        "RNG": opts.randn_source if opts.randn_source != "GPU" else None,
        "Tiling": "True" if p.tiling else None,
        **p.extra_generation_params,
        "Version": program_version() if opts.add_version_to_infotext else None,
        "User": p.user if opts.add_user_name_to_info else None,
    }

    for key, value in generation_params.items():
        try:
            if isinstance(value, list):
                generation_params[key] = value[index]
            elif callable(value):
                generation_params[key] = value(**locals())
        except Exception:
            errors.report(f'Error creating infotext for key "{key}"', exc_info=True)
            generation_params[key] = None

    generation_params_text = ", ".join([k if k == v else f'{k}: {infotext_utils.quote(v)}' for k, v in generation_params.items() if v is not None])

    negative_prompt_text = f"\nNegative prompt: {negative_prompt}" if negative_prompt else ""

    return f"{prompt_text}{negative_prompt_text}\n{generation_params_text}".strip()


def process_images(p: StableDiffusionProcessing) -> Processed:
    if p.scripts is not None:
        p.scripts.before_process(p)

    stored_opts = {k: opts.data[k] if k in opts.data else opts.get_default(k) for k in p.override_settings.keys() if k in opts.data}

    try:
        # if no checkpoint override or the override checkpoint can't be found, remove override entry and load opts checkpoint
        # and if after running refiner, the refiner model is not unloaded - webui swaps back to main model here, if model over is present it will be reloaded afterwards
        if sd_models.checkpoint_aliases.get(p.override_settings.get('sd_model_checkpoint')) is None:
            p.override_settings.pop('sd_model_checkpoint', None)
            sd_models.reload_model_weights()

        for k, v in p.override_settings.items():
            opts.set(k, v, is_api=True, run_callbacks=False)

            if k == 'sd_model_checkpoint':
                sd_models.reload_model_weights()

            if k == 'sd_vae':
                sd_vae.reload_vae_weights()

        # backwards compatibility, fix sampler and scheduler if invalid
        sd_samplers.fix_p_invalid_sampler_and_scheduler(p)

        with profiling.Profiler():
            res = process_images_inner(p)

    finally:
        # restore opts to original state
        if p.override_settings_restore_afterwards:
            for k, v in stored_opts.items():
                setattr(opts, k, v)

                if k == 'sd_vae':
                    sd_vae.reload_vae_weights()

    return res


def process_images_inner(p: StableDiffusionProcessing) -> Processed:
    """this is the main loop that both txt2img and img2img use; it calls func_init once inside all the scopes and func_sample once per batch"""

    if isinstance(p.prompt, list):
        assert(len(p.prompt) > 0)
    else:
        assert p.prompt is not None

    devices.torch_gc()

    seed = get_fixed_seed(p.seed)
    subseed = get_fixed_seed(p.subseed)

    if p.restore_faces is None:
        p.restore_faces = opts.face_restoration

    if p.tiling is None:
        p.tiling = opts.tiling

    if p.refiner_checkpoint not in (None, "", "None", "none"):
        p.refiner_checkpoint_info = sd_models.get_closet_checkpoint_match(p.refiner_checkpoint)
        if p.refiner_checkpoint_info is None:
            raise Exception(f'Could not find checkpoint with name {p.refiner_checkpoint}')

    if hasattr(shared.sd_model, 'fix_dimensions'):
        p.width, p.height = shared.sd_model.fix_dimensions(p.width, p.height)

    p.sd_model_name = shared.sd_model.sd_checkpoint_info.name_for_extra
    p.sd_model_hash = shared.sd_model.sd_model_hash
    p.sd_vae_name = sd_vae.get_loaded_vae_name()
    p.sd_vae_hash = sd_vae.get_loaded_vae_hash()

    apply_circular_forge(p.sd_model, p.tiling)
    modules.sd_hijack.model_hijack.clear_comments()

    if opts.sd_sampling == "A1111":
        p.fill_fields_from_opts()
    else:
        pass
    p.setup_prompts()

    if isinstance(seed, list):
        p.all_seeds = seed
    else:
        p.all_seeds = [int(seed) + (x if p.subseed_strength == 0 else 0) for x in range(len(p.all_prompts))]

    if isinstance(subseed, list):
        p.all_subseeds = subseed
    else:
        p.all_subseeds = [int(subseed) + x for x in range(len(p.all_prompts))]

    if os.path.exists(cmd_opts.embeddings_dir) and not p.do_not_reload_embeddings:
        model_hijack.embedding_db.load_textual_inversion_embeddings()

    if p.scripts is not None:
        p.scripts.process(p)

    infotexts = []
    output_images = []
    with torch.inference_mode():
        with devices.autocast():
            p.init(p.all_prompts, p.all_seeds, p.all_subseeds)

            # for OSX, loading the model during sampling changes the generated picture, so it is loaded here
            if shared.opts.live_previews_enable and opts.show_progress_type == "Approx NN":
                sd_vae_approx.model()

            sd_unet.apply_unet()

        if state.job_count == -1:
            state.job_count = p.n_iter

        for n in range(p.n_iter):
            p.iteration = n

            if state.skipped:
                state.skipped = False

            if state.interrupted or state.stopping_generation:
                break

            sd_models.reload_model_weights()  # model can be changed for example by refiner

            p.sd_model.forge_objects = p.sd_model.forge_objects_original.shallow_copy()
            p.prompts = p.all_prompts[n * p.batch_size:(n + 1) * p.batch_size]
            p.negative_prompts = p.all_negative_prompts[n * p.batch_size:(n + 1) * p.batch_size]
            p.seeds = p.all_seeds[n * p.batch_size:(n + 1) * p.batch_size]
            p.subseeds = p.all_subseeds[n * p.batch_size:(n + 1) * p.batch_size]

            latent_channels = getattr(shared.sd_model, 'latent_channels', opt_C)
            p.rng = rng.ImageRNG((latent_channels, p.height // opt_f, p.width // opt_f), p.seeds, subseeds=p.subseeds, subseed_strength=p.subseed_strength, seed_resize_from_h=p.seed_resize_from_h, seed_resize_from_w=p.seed_resize_from_w)

            if p.scripts is not None:
                p.scripts.before_process_batch(p, batch_number=n, prompts=p.prompts, seeds=p.seeds, subseeds=p.subseeds)

            if len(p.prompts) == 0:
                break

            p.parse_extra_network_prompts()

            if not p.disable_extra_networks:
                extra_networks.activate(p, p.extra_network_data)

            p.sd_model.forge_objects = p.sd_model.forge_objects_after_applying_lora.shallow_copy()

            if p.scripts is not None:
                p.scripts.process_batch(p, batch_number=n, prompts=p.prompts, seeds=p.seeds, subseeds=p.subseeds)

            p.setup_conds()

            p.extra_generation_params.update(model_hijack.extra_generation_params)

            # params.txt should be saved after scripts.process_batch, since the
            # infotext could be modified by that callback
            # Example: a wildcard processed by process_batch sets an extra model
            # strength, which is saved as "Model Strength: 1.0" in the infotext
            if n == 0 and not cmd_opts.no_prompt_history:
                with open(os.path.join(paths.data_path, "params.txt"), "w", encoding="utf8") as file:
                    processed = Processed(p, [])
                    file.write(processed.infotext(p, 0))

            for comment in model_hijack.comments:
                p.comment(comment)

            if p.n_iter > 1:
                shared.state.job = f"Batch {n+1} out of {p.n_iter}"

            advanced_model_sampling_script = next((x for x in p.scripts.alwayson_scripts if x.name == 'advanced model sampling for reforge'), None)
            force_apply_ztsnr = (
                advanced_model_sampling_script is not None
                and p.script_args[advanced_model_sampling_script.args_from:advanced_model_sampling_script.args_to][0]
                and p.script_args[advanced_model_sampling_script.args_from:advanced_model_sampling_script.args_to][3]
            )
            if (
                (
                    opts.sd_noise_schedule == "Zero Terminal SNR"
                    or (hasattr(p.sd_model, 'ztsnr') and p.sd_model.ztsnr)
                    or force_apply_ztsnr
                )
                and p is not None
            ):
                p.sd_model.sigmas_original = p.sd_model.forge_objects.unet.model.model_sampling.sigmas
                p.sd_model.alphas_cumprod_original = p.sd_model.alphas_cumprod
                if not getattr(opts, 'use_old_clip_g_load_and_ztsnr_application', False):
                    sd_models.apply_alpha_schedule_override(p.sd_model, p, force_apply=force_apply_ztsnr)
                p.sd_model.forge_objects.unet.model.model_sampling.set_sigmas(rescale_zero_terminal_snr_sigmas(p.sd_model.forge_objects.unet.model.model_sampling.sigmas).to(p.sd_model.forge_objects.unet.model.device))

            samples_ddim = p.sample(conditioning=p.c, unconditional_conditioning=p.uc, seeds=p.seeds, subseeds=p.subseeds, subseed_strength=p.subseed_strength, prompts=p.prompts)

            if samples_ddim is not None:
                for x_sample in samples_ddim:
                    p.latents_after_sampling.append(x_sample)
            else:
                print("Warning: samples_ddim is None. Skipping latent processing.")

            if hasattr(p.sd_model, 'sigmas_original'):
                p.sd_model.forge_objects.unet.model.model_sampling.set_sigmas(p.sd_model.sigmas_original)
            if hasattr(p.sd_model, 'alphas_cumprod_original'):
                p.sd_model.alphas_cumprod = p.sd_model.alphas_cumprod_original

            if p.scripts is not None: