Merge pull request #999 from cmdr2/beta

Beta

CHANGES.md

@@ -1,5 +1,77 @@
 # What's new?
 
+## v2.5
+### Major Changes
+- **Nearly twice as fast** - significantly faster speed of image generation. We're now pretty close to automatic1111's speed. Code contributions are welcome to make our project even faster: https://github.com/easydiffusion/sdkit/#is-it-fast
+- **Mac M1/M2 support** - Experimental support for Mac M1/M2. Thanks @michaelgallacher, @JeLuf and vishae.
+- **Full support for Stable Diffusion 2.1 (including CPU)** - supports loading v1.4 or v2.0 or v2.1 models seamlessly. No need to enable "Test SD2", and no need to add `sd2_` to your SD 2.0 model file names. Works on CPU as well.
+- **Memory optimized Stable Diffusion 2.1** - you can now use Stable Diffusion 2.1 models, with the same low VRAM optimizations that we've always had for SD 1.4. Please note, the SD 2.0 and 2.1 models require more GPU and System RAM, as compared to the SD 1.4 and 1.5 models.
+- **11 new samplers!** - explore the new samplers, some of which can generate great images in less than 10 inference steps! We've added the Karras and UniPC samplers. Thanks @Schorny for the UniPC samplers.
+- **Model Merging** - You can now merge two models (`.ckpt` or `.safetensors`) and output `.ckpt` or `.safetensors` models, optionally in `fp16` precision. Details: https://github.com/cmdr2/stable-diffusion-ui/wiki/Model-Merging . Thanks @JeLuf.
+- **Fast loading/unloading of VAEs** - No longer needs to reload the entire Stable Diffusion model, each time you change the VAE
+- **Database of known models** - automatically picks the right configuration for known models. E.g. we automatically detect and apply "v" parameterization (required for some SD 2.0 models), and "fp32" attention precision (required for some SD 2.1 models).
+- **Color correction for img2img** - an option to preserve the color profile (histogram) of the initial image. This is especially useful if you're getting red-tinted images after inpainting/masking.
+- **Three GPU Memory Usage Settings** - `High` (fastest, maximum VRAM usage), `Balanced` (default - almost as fast, significantly lower VRAM usage), `Low` (slowest, very low VRAM usage). The `Low` setting is applied automatically for GPUs with less than 4 GB of VRAM.
+- **Find models in sub-folders** - This allows you to organize your models into sub-folders inside `models/stable-diffusion`, instead of keeping them all in a single folder. Thanks @patriceac and @ogmaresca.
+- **Custom Modifier Categories** - Ability to create custom modifiers with thumbnails, and custom categories (and hierarchy of categories). Details: https://github.com/cmdr2/stable-diffusion-ui/wiki/Custom-Modifiers . Thanks @ogmaresca.
+- **Embed metadata, or save as TXT/JSON** - You can now embed the metadata directly into the images, or save them as text or json files (choose in the Settings tab). Thanks @patriceac.
+- **Major rewrite of the code** - Most of the codebase has been reorganized and rewritten, to make it more manageable and easier for new developers to contribute features. We've separated our core engine into a new project called `sdkit`, which allows anyone to easily integrate Stable Diffusion (and related modules like GFPGAN etc) into their programming projects (via a simple `pip install sdkit`): https://github.com/easydiffusion/sdkit/
+- **Name change** - Last, and probably the least, the UI is now called "Easy Diffusion". It indicates the focus of this project - an easy way for people to play with Stable Diffusion.
+
+Our focus continues to remain on an easy installation experience, and an easy user-interface. While still remaining pretty powerful, in terms of features and speed.
+
+### Detailed changelog
+* 2.5.24 - 11 Mar 2023 - Button to load an image mask from a file.
+* 2.5.24 - 10 Mar 2023 - Logo change. Image credit: @lazlo_vii.
+* 2.5.23 - 8 Mar 2023 - Experimental support for Mac M1/M2. Thanks @michaelgallacher, @JeLuf and vishae!
+* 2.5.23 - 8 Mar 2023 - Ability to create custom modifiers with thumbnails, and custom categories (and hierarchy of categories). More details - https://github.com/cmdr2/stable-diffusion-ui/wiki/Custom-Modifiers . Thanks @ogmaresca.
+* 2.5.22 - 28 Feb 2023 - Minor styling changes to UI buttons, and the models dropdown.
+* 2.5.22 - 28 Feb 2023 - Lots of UI-related bug fixes. Thanks @patriceac.
+* 2.5.21 - 22 Feb 2023 - An option to control the size of the image thumbnails. You can use the `Display options` in the top-right corner to change this. Thanks @JeLuf.
+* 2.5.20 - 20 Feb 2023 - Support saving images in WEBP format (which consumes less disk space, with similar quality). Thanks @ogmaresca.
+* 2.5.20 - 18 Feb 2023 - A setting to block NSFW images from being generated. You can enable this setting in the Settings tab.
+* 2.5.19 - 17 Feb 2023 - Initial support for server-side plugins. Currently supports overriding the `get_cond_and_uncond()` function.
+* 2.5.18 - 17 Feb 2023 - 5 new samplers! UniPC samplers, some of which produce images in less than 15 steps. Thanks @Schorny.
+* 2.5.16 - 13 Feb 2023 - Searchable dropdown for models. This is useful if you have a LOT of models. You can type part of the model name, to auto-search through your models. Thanks @patriceac for the feature, and @AssassinJN for help in UI tweaks!
+* 2.5.16 - 13 Feb 2023 - Lots of fixes and improvements to the installer. First round of changes to add Mac support. Thanks @JeLuf.
+* 2.5.16 - 13 Feb 2023 - UI bug fixes for the inpainter editor. Thanks @patriceac.
+* 2.5.16 - 13 Feb 2023 - Fix broken task reorder. Thanks @JeLuf.
+* 2.5.16 - 13 Feb 2023 - Remove a task if all the images inside it have been removed. Thanks @AssassinJN.
+* 2.5.16 - 10 Feb 2023 - Embed metadata into the JPG/PNG images, if selected in the "Settings" tab (under "Metadata format"). Thanks @patriceac.
+* 2.5.16 - 10 Feb 2023 - Sort models alphabetically in the models dropdown. Thanks @ogmaresca.
+* 2.5.16 - 10 Feb 2023 - Support multiple GFPGAN models. Download new GFPGAN models into the `models/gfpgan` folder, and refresh the UI to use it. Thanks @JeLuf.
+* 2.5.16 - 10 Feb 2023 - Allow a server to enforce a fixed directory path to save images. This is useful if the server is exposed to a lot of users. This can be set in the `config.json` file as `force_save_path: "/path/to/fixed/save/dir"`. E.g. `force_save_path: "D:/user_images"`. Thanks @JeLuf.
+* 2.5.16 - 10 Feb 2023 - The "Make Images" button now shows the correct amount of images it'll create when using operators like `{}` or `|`. For e.g. if the prompt is `Photo of a {woman, man}`, then the button will say `Make 2 Images`. Thanks @JeLuf.
+* 2.5.16 - 10 Feb 2023 - A bunch of UI-related bug fixes. Thanks @patriceac.
+* 2.5.15 - 8 Feb 2023 - Allow using 'balanced' VRAM usage mode on GPUs with 4 GB or less of VRAM. This mode used to be called 'Turbo' in the previous version.
+* 2.5.14 - 8 Feb 2023 - Fix broken auto-save settings. We renamed `sampler` to `sampler_name`, which caused old settings to fail.
+* 2.5.14 - 6 Feb 2023 - Simplify the UI for merging models, and some other minor UI tweaks. Better error reporting if a model failed to load.
+* 2.5.14 - 3 Feb 2023 - Fix the 'Make Similar Images' button, which was producing incorrect images (weren't very similar).
+* 2.5.13 - 1 Feb 2023 - Fix the remaining GPU memory leaks, including a better fix (more comprehensive) for the change in 2.5.12 (27 Jan).
+* 2.5.12 - 27 Jan 2023 - Fix a memory leak, which made the UI unresponsive after an out-of-memory error. The allocated memory is now freed-up after an error.
+* 2.5.11 - 25 Jan 2023 - UI for Merging Models. Thanks @JeLuf. More info: https://github.com/cmdr2/stable-diffusion-ui/wiki/Model-Merging
+* 2.5.10 - 24 Jan 2023 - Reduce the VRAM usage for img2img in 'balanced' mode (without reducing the rendering speed), to make it similar to v2.4 of this UI.
+* 2.5.9 - 23 Jan 2023 - Fix a bug where img2img would produce poorer-quality images for the same settings, as compared to version 2.4 of this UI.
+* 2.5.9 - 23 Jan 2023 - Reduce the VRAM usage for 'balanced' mode (without reducing the rendering speed), to make it similar to v2.4 of the UI.
+* 2.5.8 - 17 Jan 2023 - Fix a bug where 'Low' VRAM usage would consume a LOT of VRAM (on higher-end GPUs). Also fixed a bug that caused out-of-memory errors on SD 2.1-768 models, on 'high' VRAM usage setting.
+* 2.5.7 - 16 Jan 2023 - Fix a bug where VAE files ending with .vae.pt weren't getting displayed. Thanks Madrang, rbertus2000 and JeLuf.
+* 2.5.6 - 10 Jan 2023 - `Fill` tool for the Image Editor, to allow filling areas with color (or the entire image). And some bug fixes to the Image Editor. Thanks @mdiller.
+* 2.5.6 - 10 Jan 2023 - Find Stable Diffusion models in sub-folders inside `models/stable-diffusion`. This allows you to organize your models into sub-folders, instead of keeping them all in a single folder. Thanks @JeLuf.
+* 2.5.5 - 9 Jan 2023 - Lots of bug fixes. Thanks @patriceac and @JeLuf.
+* 2.5.4 - 29 Dec 2022 - Press Esc key on the keyboard to close the Image Editor. Thanks @patriceac.
+* 2.5.4 - 29 Dec 2022 - Lots of bug fixes in the UI. Thanks @patriceac.
+* 2.5.4 - 28 Dec 2022 - Full support for running tasks in parallel on multiple GPUs. Warning: 'Euler Ancestral', 'DPM2 Ancestral' and 'DPM++ 2s Ancestral' may produce slight variations in the image (if run in parallel), so we recommend using the other samplers.
+* 2.5.3 - 27 Dec 2022 - Fix broken drag-and-drop for text metadata files (as well as paste in clipboard).
+* 2.5.3 - 27 Dec 2022 - Allow upscaling by 2x as well as 4x.
+* 2.5.3 - 27 Dec 2022 - Fix broken renders on a second GPU.
+* 2.5.3 - 26 Dec 2022 - Add a `Remove` button on each image. Thanks @JeLuf.
+* 2.5.2 - 26 Dec 2022 - Fix broken inpainting if using non-square target images.
+* 2.5.2 - 26 Dec 2022 - Fix a bug where an incorrect model config would get used for some SD 2.1 models.
+* 2.5.2 - 26 Dec 2022 - Slight performance and memory improvement while rendering using SD 2.1 models.
+* 2.5.1 - 25 Dec 2022 - Allow custom config yaml files for models. You can put a config file (`.yaml`) next to the model file, with the same name as the model. For e.g. if you put `robo-diffusion-v2-base.yaml` next to `robo-diffusion-v2-base.ckpt`, it'll automatically use that config file.
+* 2.5.1 - 25 Dec 2022 - Fix broken rendering for SD 2.1-768 models. Fix broken rendering SD 2.0 safetensor models.
+* 2.5.0 - 25 Dec 2022 - Major new release! Nearly twice as fast, Full support for SD 2.1 (including low GPU RAM optimizations), 6 new samplers, Model Merging, Fast loading/unloading of VAEs, Database of known models, Color correction for img2img, Three GPU Memory Usage Settings, Save metadata as JSON, Major rewrite of the code, Name change.
+
 ## v2.4
 ### Major Changes
 - **Allow reordering the task queue** (by dragging and dropping tasks). Thanks @madrang
@@ -27,6 +99,9 @@
 - Support loading models in the safetensor format, for improved safety
 
 ### Detailed changelog
+* 2.4.24 - 9 Jan 2022 - Urgent fix for failures on old/long-term-support browsers. Thanks @JeLuf.
+* 2.4.23/22 - 29 Dec 2022 - Allow rolling back from the upcoming v2.5 change (in beta).
+* 2.4.21 - 23 Dec 2022 - Speed up image creation, by removing a delay (regression) of 4-5 seconds between clicking the `Make Image` button and calling the server.
 * 2.4.20 - 22 Dec 2022 - `Pause All` button to pause all the pending tasks. Thanks @JeLuf
 * 2.4.20 - 22 Dec 2022 - `Undo`/`Redo` buttons in the image editor. Thanks @JeLuf
 * 2.4.20 - 22 Dec 2022 - Drag handle to reorder the tasks. This fixed a bug where the metadata was no longer selectable (for copying). Thanks @JeLuf

NSIS/.gitignore

@@ -0,0 +1 @@
+*.exe

NSIS/nsisconf.nsh

@@ -0,0 +1 @@
+!define EXISTING_INSTALLATION_DIR "D:\path\to\installed\easy-diffusion"

NSIS/sdui.nsi

@@ -1,20 +1,24 @@
 ; Script generated by the HM NIS Edit Script Wizard.
 
-Target x86-unicode
+Target amd64-unicode
 Unicode True
-!AddPluginDir /x86-unicode "."
+SetCompressor /FINAL lzma
+RequestExecutionLevel user
+!AddPluginDir /amd64-unicode "."
 ; HM NIS Edit Wizard helper defines
-!define PRODUCT_NAME "Stable Diffusion UI"
-!define PRODUCT_VERSION "Installer 2.35"
+!define PRODUCT_NAME "Easy Diffusion"
+!define PRODUCT_VERSION "2.5"
 !define PRODUCT_PUBLISHER "cmdr2 and contributors"
 !define PRODUCT_WEB_SITE "https://stable-diffusion-ui.github.io"
-!define PRODUCT_DIR_REGKEY "Software\Microsoft\Cmdr2\App Paths\installer.exe"
+!define PRODUCT_DIR_REGKEY "Software\Microsoft\Easy Diffusion\App Paths\installer.exe"
 
 ; MUI 1.67 compatible ------
 !include "MUI.nsh"
 !include "LogicLib.nsh"
 !include "nsDialogs.nsh"
 
+!include "nsisconf.nsh"
+
 Var Dialog
 Var Label
 Var Button
@@ -106,7 +110,7 @@ Function DirectoryLeave
    StrCpy $5 $INSTDIR 3
    System::Call 'Kernel32::GetVolumeInformation(t "$5",t,i ${NSIS_MAX_STRLEN},*i,*i,*i,t.r1,i ${NSIS_MAX_STRLEN})i.r0'
    ${If} $0 <> 0
-   ${AndIf} $1 == "NTFS"
+   ${AndIf} $1 != "NTFS"
        MessageBox mb_ok "$5 has filesystem type '$1'.$\nOnly NTFS filesystems are supported.$\nPlease choose a different drive."
        Abort
    ${EndIf}
@@ -140,7 +144,7 @@ Function MediaPackDialog
 	Abort
     ${EndIf}
 
-    ${NSD_CreateLabel} 0 0 100% 48u "The Windows Media Feature Pack is missing on this computer. It is required for the Stable Diffusion UI.$\nYou can continue the installation after installing the Windows Media Feature Pack."
+    ${NSD_CreateLabel} 0 0 100% 48u "The Windows Media Feature Pack is missing on this computer. It is required for Easy Diffusion.$\nYou can continue the installation after installing the Windows Media Feature Pack."
     Pop $Label
  	
     ${NSD_CreateButton} 10% 49u 80% 12u "Download Meda Feature Pack from Microsoft"
@@ -157,12 +161,12 @@ FunctionEnd
 ; MUI Settings
 ;---------------------------------------------------------------------------------------------------------
 !define MUI_ABORTWARNING
-!define MUI_ICON "sd.ico"
+!define MUI_ICON "cyborg_flower_girl.ico"
 
-!define MUI_WELCOMEFINISHPAGE_BITMAP "astro.bmp"
+!define MUI_WELCOMEFINISHPAGE_BITMAP "cyborg_flower_girl.bmp"
 
 ; Welcome page
-!define MUI_WELCOMEPAGE_TEXT "This installer will guide you through the installation of Stable Diffusion UI.$\n$\n\
+!define MUI_WELCOMEPAGE_TEXT "This installer will guide you through the installation of Easy Diffusion.$\n$\n\
 Click Next to continue."
 !insertmacro MUI_PAGE_WELCOME
 Page custom MediaPackDialog
@@ -188,8 +192,8 @@ Page custom MediaPackDialog
 ;---------------------------------------------------------------------------------------------------------
 
 Name "${PRODUCT_NAME} ${PRODUCT_VERSION}"
-OutFile "Install Stable Diffusion UI.exe"
-InstallDir "C:\Stable-Diffusion-UI\"
+OutFile "Install Easy Diffusion.exe"
+InstallDir "C:\EasyDiffusion\"
 InstallDirRegKey HKLM "${PRODUCT_DIR_REGKEY}" ""
 ShowInstDetails show
 
@@ -200,15 +204,40 @@ Section "MainSection" SEC01
   File "..\CreativeML Open RAIL-M License"
   File "..\How to install and run.txt"
   File "..\LICENSE"
-  File "..\Start Stable Diffusion UI.cmd"
+  File "..\scripts\Start Stable Diffusion UI.cmd"
+  File /r "${EXISTING_INSTALLATION_DIR}\installer_files"
+  File /r "${EXISTING_INSTALLATION_DIR}\profile"
+  File /r "${EXISTING_INSTALLATION_DIR}\sd-ui-files"
   SetOutPath "$INSTDIR\scripts"
-  File "..\scripts\bootstrap.bat"
-  File "..\scripts\install_status.txt"
+  File "${EXISTING_INSTALLATION_DIR}\scripts\install_status.txt"
   File "..\scripts\on_env_start.bat"
   File "C:\windows\system32\curl.exe"
-  CreateDirectory "$INSTDIR\profile"
-  CreateDirectory "$SMPROGRAMS\Stable Diffusion UI"
-  CreateShortCut "$SMPROGRAMS\Stable Diffusion UI\Start Stable Diffusion UI.lnk" "$INSTDIR\Start Stable Diffusion UI.cmd"
+  CreateDirectory "$INSTDIR\models"
+  CreateDirectory "$INSTDIR\models\stable-diffusion"
+  CreateDirectory "$INSTDIR\models\gfpgan"
+  CreateDirectory "$INSTDIR\models\realesrgan"
+  CreateDirectory "$INSTDIR\models\vae"
+  CreateDirectory "$SMPROGRAMS\Easy Diffusion"
+  CreateShortCut "$SMPROGRAMS\Easy Diffusion\Easy Diffusion.lnk" "$INSTDIR\Start Stable Diffusion UI.cmd"
+
+  DetailPrint 'Downloading the Stable Diffusion 1.4 model...'
+  NScurl::http get "https://huggingface.co/CompVis/stable-diffusion-v-1-4-original/resolve/main/sd-v1-4.ckpt" "$INSTDIR\models\stable-diffusion\sd-v1-4.ckpt" /CANCEL /INSIST /END
+
+  DetailPrint 'Downloading the GFPGAN model...'
+  NScurl::http get "https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.3.pth" "$INSTDIR\models\gfpgan\GFPGANv1.3.pth" /CANCEL /INSIST /END
+
+  DetailPrint 'Downloading the RealESRGAN_x4plus model...'
+  NScurl::http get "https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth" "$INSTDIR\models\realesrgan\RealESRGAN_x4plus.pth" /CANCEL /INSIST /END
+
+  DetailPrint 'Downloading the RealESRGAN_x4plus_anime model...'
+  NScurl::http get "https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth" "$INSTDIR\models\realesrgan\RealESRGAN_x4plus_anime_6B.pth" /CANCEL /INSIST /END
+
+  DetailPrint 'Downloading the default VAE (sd-vae-ft-mse-original) model...'
+  NScurl::http get "https://huggingface.co/stabilityai/sd-vae-ft-mse-original/resolve/main/vae-ft-mse-840000-ema-pruned.ckpt" "$INSTDIR\models\vae\vae-ft-mse-840000-ema-pruned.ckpt" /CANCEL /INSIST /END
+
+  DetailPrint 'Downloading the CLIP model (clip-vit-large-patch14)...'
+  NScurl::http get "https://huggingface.co/openai/clip-vit-large-patch14/resolve/8d052a0f05efbaefbc9e8786ba291cfdf93e5bff/pytorch_model.bin" "$INSTDIR\profile\.cache\huggingface\hub\models--openai--clip-vit-large-patch14\snapshots\8d052a0f05efbaefbc9e8786ba291cfdf93e5bff\pytorch_model.bin" /CANCEL /INSIST /END
+
 SectionEnd
 
 ;---------------------------------------------------------------------------------------------------------
@@ -254,7 +283,7 @@ Function .onInit
 
    ${If} $4 < "8000"
       MessageBox MB_OK|MB_ICONEXCLAMATION "Warning!$\n$\nYour system has less than 8GB of memory (RAM).$\n$\n\
-You can still try to install Stable Diffusion UI,$\nbut it might have problems to start, or run$\nvery slowly."
+You can still try to install Easy Diffusion,$\nbut it might have problems to start, or run$\nvery slowly."
    ${EndIf}
   
 FunctionEnd

README.md

@@ -1,19 +1,18 @@
-# Stable Diffusion UI
-### The easiest way to install and use [Stable Diffusion](https://github.com/CompVis/stable-diffusion) on your own computer. Does not require technical knowledge, does not require pre-installed software. 1-click install, powerful features, friendly community.
+# Easy Diffusion 2.5
+### The easiest way to install and use [Stable Diffusion](https://github.com/CompVis/stable-diffusion) on your own computer.
 
-[![Discord Server](https://img.shields.io/discord/1014774730907209781?label=Discord)](https://discord.com/invite/u9yhsFmEkB) (for support, and development discussion) | [Troubleshooting guide for common problems](https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting)
+Does not require technical knowledge, does not require pre-installed software. 1-click install, powerful features, friendly community.
 
-### New: 
-Experimental support for Stable Diffusion 2.0 is available in beta!
+[Installation guide](#step-1-download-and-extract-the-installer) | [Troubleshooting guide](https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting) | <sub>[![Discord Server](https://img.shields.io/discord/1014774730907209781?label=Discord)](https://discord.com/invite/u9yhsFmEkB)</sub> <sup>(for support queries, and development discussions)</sup>
 
-----
+![t2i](https://raw.githubusercontent.com/Stability-AI/stablediffusion/main/assets/stable-samples/txt2img/768/merged-0006.png)
 
-# Step 1: Download and prepare the installer
+# Step 1: Download and extract the installer
 Click the download button for your operating system:
 
 <p float="left">
-  <a href="https://github.com/cmdr2/stable-diffusion-ui/releases/download/v2.4.13/stable-diffusion-ui-windows.zip"><img src="https://github.com/cmdr2/stable-diffusion-ui/raw/main/media/download-win.png" width="200" /></a>
-  <a href="https://github.com/cmdr2/stable-diffusion-ui#installation"><img src="https://github.com/cmdr2/stable-diffusion-ui/raw/main/media/download-linux.png" width="200" /></a>
+  <a href="https://github.com/cmdr2/stable-diffusion-ui/releases/download/v2.5.15/stable-diffusion-ui-windows.zip"><img src="https://github.com/cmdr2/stable-diffusion-ui/raw/main/media/download-win.png" width="200" /></a>
+  <a href="https://github.com/cmdr2/stable-diffusion-ui/releases/download/v2.5.15/stable-diffusion-ui-linux.zip"><img src="https://github.com/cmdr2/stable-diffusion-ui/raw/main/media/download-linux.png" width="200" /></a>
 </p>
 
 ## On Windows:
@@ -44,9 +43,18 @@ The installer will take care of whatever is needed. If you face any problems, yo
 ### User experience
 - **Hassle-free installation**: Does not require technical knowledge, does not require pre-installed software. Just download and run!
 - **Clutter-free UI**: A friendly and simple UI, while providing a lot of powerful features.
+- **Task Queue**: Queue up all your ideas, without waiting for the current task to finish.
+- **Intelligent Model Detection**: Automatically figures out the YAML config file to use for the chosen model (via a models database).
+- **Live Preview**: See the image as the AI is drawing it.
+- **Image Modifiers**: A library of *modifier tags* like *"Realistic"*, *"Pencil Sketch"*, *"ArtStation"* etc. Experiment with various styles quickly.
+- **Multiple Prompts File**: Queue multiple prompts by entering one prompt per line, or by running a text file.
+- **Save generated images to disk**: Save your images to your PC!
+- **UI Themes**: Customize the program to your liking.
+- **Searchable models dropdown**: organize your models into sub-folders, and search through them in the UI.
 
 ### Image generation
 - **Supports**: "*Text to Image*" and "*Image to Image*".
+- **19 Samplers**: `ddim`, `plms`, `heun`, `euler`, `euler_a`, `dpm2`, `dpm2_a`, `lms`, `dpm_solver_stability`, `dpmpp_2s_a`, `dpmpp_2m`, `dpmpp_sde`, `dpm_fast`, `dpm_adaptive`, `unipc_snr`, `unipc_tu`, `unipc_tq`, `unipc_snr_2`, `unipc_tu_2`.
 - **In-Painting**: Specify areas of your image to paint into.
 - **Simple Drawing Tool**: Draw basic images to guide the AI, without needing an external drawing program.
 - **Face Correction (GFPGAN)**
@@ -56,21 +64,23 @@ The installer will take care of whatever is needed. If you face any problems, yo
 - **Attention/Emphasis**: () in the prompt increases the model's attention to enclosed words, and [] decreases it.
 - **Weighted Prompts**: Use weights for specific words in your prompt to change their importance, e.g. `red:2.4 dragon:1.2`.
 - **Prompt Matrix**: Quickly create multiple variations of your prompt, e.g. `a photograph of an astronaut riding a horse | illustration | cinematic lighting`.
-- **Lots of Samplers**: ddim, plms, heun, euler, euler_a, dpm2, dpm2_a, lms.
 - **1-click Upscale/Face Correction**: Upscale or correct an image after it has been generated.
 - **Make Similar Images**: Click to generate multiple variations of a generated image.
 - **NSFW Setting**: A setting in the UI to control *NSFW content*.
-- **JPEG/PNG output**: Multiple file formats.
+- **JPEG/PNG/WEBP output**: Multiple file formats.
 
 ### Advanced features
 - **Custom Models**: Use your own `.ckpt` or `.safetensors` file, by placing it inside the `models/stable-diffusion` folder!
-- **Stable Diffusion 2.0 support (experimental)**: available in beta channel.
+- **Stable Diffusion 2.1 support**
+- **Merge Models**
 - **Use custom VAE models**
 - **Use pre-trained Hypernetworks**
+- **Use custom GFPGAN models**
 - **UI Plugins**: Choose from a growing list of [community-generated UI plugins](https://github.com/cmdr2/stable-diffusion-ui/wiki/UI-Plugins), or write your own plugin to add features to the project!
 
 ### Performance and security
-- **Low Memory Usage**: Creates 512x512 images with less than 4GB of GPU RAM!
+- **Fast**: Creates a 512x512 image with euler_a in 5 seconds, on an NVIDIA 3060 12GB.
+- **Low Memory Usage**: Create 512x512 images with less than 3 GB of GPU RAM, and 768x768 images with less than 4 GB of GPU RAM!
 - **Use CPU setting**: If you don't have a compatible graphics card, but still want to run it on your CPU.
 - **Multi-GPU support**: Automatically spreads your tasks across multiple GPUs (if available), for faster performance!
 - **Auto scan for malicious models**: Uses picklescan to prevent malicious models.
@@ -78,23 +88,17 @@ The installer will take care of whatever is needed. If you face any problems, yo
 - **Auto-updater**: Gets you the latest improvements and bug-fixes to a rapidly evolving project.
 - **Developer Console**: A developer-mode for those who want to modify their Stable Diffusion code, and edit the conda environment.
 
-### Usability:
-- **Live Preview**: See the image as the AI is drawing it.
-- **Task Queue**: Queue up all your ideas, without waiting for the current task to finish.
-- **Image Modifiers**: A library of *modifier tags* like *"Realistic"*, *"Pencil Sketch"*, *"ArtStation"* etc. Experiment with various styles quickly.
-- **Multiple Prompts File**: Queue multiple prompts by entering one prompt per line, or by running a text file.
-- **Save generated images to disk**: Save your images to your PC!
-- **UI Themes**: Customize the program to your liking.
-
 **(and a lot more)**
 
 ----
 
 ## Easy for new users:
-![Screenshot of the initial UI](media/shot-v10-simple.jpg?raw=true)
+![Screenshot of the initial UI](https://user-images.githubusercontent.com/844287/217043152-29454d15-0387-4228-b70d-9a4b84aeb8ba.png)
+
 
 ## Powerful features for advanced users:
-![Screenshot of advanced settings](media/shot-v10.jpg?raw=true)
+![Screenshot of advanced settings](https://user-images.githubusercontent.com/844287/217042588-fc53c975-bacd-4a9c-af88-37408734ade3.png)
+
 
 ## Live Preview
 Useful for judging (and stopping) an image quickly, without waiting for it to finish rendering.
@@ -102,7 +106,9 @@ Useful for judging (and stopping) an image quickly, without waiting for it to fi
 ![live-512](https://user-images.githubusercontent.com/844287/192097249-729a0a1e-a677-485e-9ccc-16a9e848fabe.gif)
 
 ## Task Queue
-![Screenshot of task queue](media/task-queue-v1.jpg?raw=true)
+![Screenshot of task queue](https://user-images.githubusercontent.com/844287/217043984-0b35f73b-1318-47cb-9eed-a2a91b430490.png)
+
+
 
 # System Requirements
 1. Windows 10/11, or Linux. Experimental support for Mac is coming soon.

scripts/Developer Console.cmd

@@ -23,23 +23,20 @@ call conda --version
 
 echo.
 
-@rem activate the environment
-call conda activate .\stable-diffusion\env
+@rem activate the legacy environment (if present) and set PYTHONPATH
+if exist "installer_files\env" (
+    set PYTHONPATH=%cd%\installer_files\env\lib\site-packages
+)
+if exist "stable-diffusion\env" (
+    call conda activate .\stable-diffusion\env
+    set PYTHONPATH=%cd%\stable-diffusion\env\lib\site-packages
+)
 
 call where python
 call python --version
 
-@rem set the PYTHONPATH
-cd stable-diffusion
-set SD_DIR=%cd%
-
-cd env\lib\site-packages
-set PYTHONPATH=%SD_DIR%;%cd%
-cd ..\..\..
 echo PYTHONPATH=%PYTHONPATH%
 
-cd ..
-
 @rem done
 echo.
 

scripts/Start Stable Diffusion UI.cmd

@@ -1,20 +1,36 @@
 @echo off
 
 cd /d %~dp0
+echo Install dir: %~dp0
+
 set PATH=C:\Windows\System32;%PATH%
 
+if exist "on_sd_start.bat" (
+    echo ================================================================================
+    echo.
+    echo !!!! WARNING !!!!
+    echo.
+    echo It looks like you're trying to run the installation script from a source code 
+    echo download. This will not work.
+    echo.
+    echo Recommended: Please close this window and download the installer from
+    echo https://stable-diffusion-ui.github.io/docs/installation/
+    echo.
+    echo ================================================================================
+    echo.
+    pause
+    exit /b
+) 
+
 @rem set legacy installer's PATH, if it exists
 if exist "installer" set PATH=%cd%\installer;%cd%\installer\Library\bin;%cd%\installer\Scripts;%cd%\installer\Library\usr\bin;%PATH%
 
-@rem Setup the packages required for the installer
-call scripts\bootstrap.bat
-
 @rem set new installer's PATH, if it downloaded any packages
 if exist "installer_files\env" set PATH=%cd%\installer_files\env;%cd%\installer_files\env\Library\bin;%cd%\installer_files\env\Scripts;%cd%\installer_files\Library\usr\bin;%PATH%
 
 set PYTHONPATH=%cd%\installer;%cd%\installer_files\env
 
-@rem Test the bootstrap
+@rem Test the core requirements
 call where git
 call git --version
 

scripts/bootstrap.bat

@@ -1,4 +1,5 @@
 @echo off
+setlocal enabledelayedexpansion
 
 @rem This script will install git and conda (if not found on the PATH variable)
 @rem  using micromamba (an 8mb static-linked single-file binary, conda replacement).
@@ -24,14 +25,14 @@ if exist "%INSTALL_ENV_DIR%" set PATH=%INSTALL_ENV_DIR%;%INSTALL_ENV_DIR%\Librar
 set PACKAGES_TO_INSTALL=
 
 if not exist "%LEGACY_INSTALL_ENV_DIR%\etc\profile.d\conda.sh" (
-    if not exist "%INSTALL_ENV_DIR%\etc\profile.d\conda.sh" set PACKAGES_TO_INSTALL=%PACKAGES_TO_INSTALL% conda
+    if not exist "%INSTALL_ENV_DIR%\etc\profile.d\conda.sh" set PACKAGES_TO_INSTALL=%PACKAGES_TO_INSTALL% conda python=3.8.5
 )
 
 call git --version >.tmp1 2>.tmp2
-if "%ERRORLEVEL%" NEQ "0" set PACKAGES_TO_INSTALL=%PACKAGES_TO_INSTALL% git
+if "!ERRORLEVEL!" NEQ "0" set PACKAGES_TO_INSTALL=%PACKAGES_TO_INSTALL% git
 
 call "%MAMBA_ROOT_PREFIX%\micromamba.exe" --version >.tmp1 2>.tmp2
-if "%ERRORLEVEL%" EQU "0" set umamba_exists=T
+if "!ERRORLEVEL!" EQU "0" set umamba_exists=T
 
 @rem (if necessary) install git and conda into a contained environment
 if "%PACKAGES_TO_INSTALL%" NEQ "" (
@@ -42,7 +43,7 @@ if "%PACKAGES_TO_INSTALL%" NEQ "" (
         mkdir "%MAMBA_ROOT_PREFIX%"
         call curl -Lk "%MICROMAMBA_DOWNLOAD_URL%" > "%MAMBA_ROOT_PREFIX%\micromamba.exe"
 
-        if "%ERRORLEVEL%" NEQ "0" (
+        if "!ERRORLEVEL!" NEQ "0" (
             echo "There was a problem downloading micromamba. Cannot continue."
             pause
             exit /b

scripts/bootstrap.sh

@@ -21,9 +21,16 @@ OS_ARCH=$(uname -m)
 case "${OS_ARCH}" in
     x86_64*)    OS_ARCH="64";;
     arm64*)     OS_ARCH="arm64";;
+    aarch64*)     OS_ARCH="arm64";;
     *)          echo "Unknown system architecture: $OS_ARCH! This script runs only on x86_64 or arm64" && exit
 esac
 
+if ! which curl; then fail "'curl' not found. Please install curl."; fi
+if ! which tar; then fail "'tar' not found. Please install tar."; fi
+if ! which bzip2; then fail "'bzip2' not found. Please install bzip2."; fi
+
+if pwd | grep ' '; then fail "The installation directory's path contains a space character. Conda will fail to install. Please change the directory."; fi
+
 # https://mamba.readthedocs.io/en/latest/installation.html
 if [ "$OS_NAME" == "linux" ] && [ "$OS_ARCH" == "arm64" ]; then OS_ARCH="aarch64"; fi
 
@@ -39,7 +46,7 @@ if [ -e "$INSTALL_ENV_DIR" ]; then export PATH="$INSTALL_ENV_DIR/bin:$PATH"; fi
 
 PACKAGES_TO_INSTALL=""
 
-if [ ! -e "$LEGACY_INSTALL_ENV_DIR/etc/profile.d/conda.sh" ] && [ ! -e "$INSTALL_ENV_DIR/etc/profile.d/conda.sh" ]; then PACKAGES_TO_INSTALL="$PACKAGES_TO_INSTALL conda"; fi
+if [ ! -e "$LEGACY_INSTALL_ENV_DIR/etc/profile.d/conda.sh" ] && [ ! -e "$INSTALL_ENV_DIR/etc/profile.d/conda.sh" ]; then PACKAGES_TO_INSTALL="$PACKAGES_TO_INSTALL conda python=3.8.5"; fi
 if ! hash "git" &>/dev/null; then PACKAGES_TO_INSTALL="$PACKAGES_TO_INSTALL git"; fi
 
 if "$MAMBA_ROOT_PREFIX/micromamba" --version &>/dev/null; then umamba_exists="T"; fi
@@ -51,7 +58,7 @@ if [ "$PACKAGES_TO_INSTALL" != "" ]; then
         echo "Downloading micromamba from $MICROMAMBA_DOWNLOAD_URL to $MAMBA_ROOT_PREFIX/micromamba"
 
         mkdir -p "$MAMBA_ROOT_PREFIX"
-        curl -L "$MICROMAMBA_DOWNLOAD_URL" | tar -xvj bin/micromamba -O > "$MAMBA_ROOT_PREFIX/micromamba"
+        curl -L "$MICROMAMBA_DOWNLOAD_URL" | tar -xvj -O bin/micromamba > "$MAMBA_ROOT_PREFIX/micromamba"
 
         if [ "$?" != "0" ]; then
             echo

scripts/check_modules.py

@@ -0,0 +1,13 @@
+'''
+This script checks if the given modules exist
+'''
+
+import sys
+import pkgutil
+
+modules = sys.argv[1:]
+missing_modules = []
+for m in modules:
+    if pkgutil.find_loader(m) is None:
+        print('module', m, 'not found')
+        exit(1)

scripts/developer_console.sh

@@ -26,21 +26,23 @@ if [ "$0" == "bash" ]; then
 
   echo ""
 
-  # activate the environment
-  CONDA_BASEPATH=$(conda info --base)
-  source "$CONDA_BASEPATH/etc/profile.d/conda.sh" # otherwise conda complains about 'shell not initialized' (needed when running in a script)
+  # activate the legacy environment (if present) and set PYTHONPATH
+  if [ -e "installer_files/env" ]; then
+    export PYTHONPATH="$(pwd)/installer_files/env/lib/python3.8/site-packages"
+  fi
+  if [ -e "stable-diffusion/env" ]; then
+    CONDA_BASEPATH=$(conda info --base)
+    source "$CONDA_BASEPATH/etc/profile.d/conda.sh" # otherwise conda complains about 'shell not initialized' (needed when running in a script)
 
-  conda activate ./stable-diffusion/env
+    conda activate ./stable-diffusion/env
+
+    export PYTHONPATH="$(pwd)/stable-diffusion/env/lib/python3.8/site-packages"
+  fi
 
   which python
   python --version
 
-  # set the PYTHONPATH
-  cd stable-diffusion
-  SD_PATH=`pwd`
-  export PYTHONPATH="$SD_PATH:$SD_PATH/env/lib/python3.8/site-packages"
   echo "PYTHONPATH=$PYTHONPATH"
-  cd ..
 
   # done
 

scripts/functions.sh

@@ -28,5 +28,12 @@ EOF
 
 }
 
+filesize() {
+    case "$(uname -s)" in
+        Linux*)     stat -c "%s" $1;;
+        Darwin*)    stat -f "%z" $1;;
+        *)          echo "Unknown OS: $OS_NAME! This script runs only on Linux or Mac" && exit
+    esac
+}
 
 

scripts/on_env_start.bat

@@ -1,6 +1,6 @@
 @echo off
 
-@echo. & echo "Stable Diffusion UI - v2" & echo.
+@echo. & echo "Easy Diffusion - v2" & echo.
 
 set PATH=C:\Windows\System32;%PATH%
 
@@ -28,7 +28,7 @@ if "%update_branch%"=="" (
 
 @>nul findstr /m "sd_ui_git_cloned" scripts\install_status.txt
 @if "%ERRORLEVEL%" EQU "0" (
-    @echo "Stable Diffusion UI's git repository was already installed. Updating from %update_branch%.."
+    @echo "Easy Diffusion's git repository was already installed. Updating from %update_branch%.."
 
     @cd sd-ui-files
 
@@ -38,13 +38,13 @@ if "%update_branch%"=="" (
 
     @cd ..
 ) else (
-    @echo. & echo "Downloading Stable Diffusion UI.." & echo.
+    @echo. & echo "Downloading Easy Diffusion..." & echo.
     @echo "Using the %update_branch% channel" & echo.
 
     @call git clone -b "%update_branch%" https://github.com/cmdr2/stable-diffusion-ui.git sd-ui-files && (
         @echo sd_ui_git_cloned >> scripts\install_status.txt
     ) || (
-        @echo "Error downloading Stable Diffusion UI. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
+        @echo "Error downloading Easy Diffusion. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
         pause
         @exit /b
     )
@@ -52,7 +52,7 @@ if "%update_branch%"=="" (
 
 @xcopy sd-ui-files\ui ui /s /i /Y /q
 @copy sd-ui-files\scripts\on_sd_start.bat scripts\ /Y
-@copy sd-ui-files\scripts\bootstrap.bat scripts\ /Y
+@copy sd-ui-files\scripts\check_modules.py scripts\ /Y
 @copy "sd-ui-files\scripts\Start Stable Diffusion UI.cmd" . /Y
 @copy "sd-ui-files\scripts\Developer Console.cmd" . /Y
 

scripts/on_env_start.sh

@@ -2,7 +2,7 @@
 
 source ./scripts/functions.sh
 
-printf "\n\nStable Diffusion UI\n\n"
+printf "\n\nEasy Diffusion\n\n"
 
 if [ -f "scripts/config.sh" ]; then
     source scripts/config.sh
@@ -13,7 +13,7 @@ if [ "$update_branch" == "" ]; then
 fi
 
 if [ -f "scripts/install_status.txt" ] && [ `grep -c sd_ui_git_cloned scripts/install_status.txt` -gt "0" ]; then
-    echo "Stable Diffusion UI's git repository was already installed. Updating from $update_branch.."
+    echo "Easy Diffusion's git repository was already installed. Updating from $update_branch.."
 
     cd sd-ui-files
 
@@ -23,7 +23,7 @@ if [ -f "scripts/install_status.txt" ] && [ `grep -c sd_ui_git_cloned scripts/in
 
     cd ..
 else
-    printf "\n\nDownloading Stable Diffusion UI..\n\n"
+    printf "\n\nDownloading Easy Diffusion..\n\n"
     printf "Using the $update_branch channel\n\n"
 
     if git clone -b "$update_branch" https://github.com/cmdr2/stable-diffusion-ui.git sd-ui-files ; then
@@ -37,9 +37,9 @@ rm -rf ui
 cp -Rf sd-ui-files/ui .
 cp sd-ui-files/scripts/on_sd_start.sh scripts/
 cp sd-ui-files/scripts/bootstrap.sh scripts/
+cp sd-ui-files/scripts/check_modules.py scripts/
 cp sd-ui-files/scripts/start.sh .
 cp sd-ui-files/scripts/developer_console.sh .
+cp sd-ui-files/scripts/functions.sh scripts/
 
-./scripts/on_sd_start.sh
-
-read -p "Press any key to continue"
+exec ./scripts/on_sd_start.sh

scripts/on_sd_start.bat

@@ -5,11 +5,20 @@
 
 @copy sd-ui-files\scripts\on_env_start.bat scripts\ /Y
 @copy sd-ui-files\scripts\bootstrap.bat scripts\ /Y
+@copy sd-ui-files\scripts\check_modules.py scripts\ /Y
 
 if exist "%cd%\profile" (
     set USERPROFILE=%cd%\profile
 )
 
+@rem set the correct installer path (current vs legacy)
+if exist "%cd%\installer_files\env" (
+    set INSTALL_ENV_DIR=%cd%\installer_files\env
+)
+if exist "%cd%\stable-diffusion\env" (
+    set INSTALL_ENV_DIR=%cd%\stable-diffusion\env
+)
+
 @mkdir tmp
 @set TMP=%cd%\tmp
 @set TEMP=%cd%\tmp
@@ -17,7 +26,7 @@ if exist "%cd%\profile" (
 @rem activate the installer env
 call conda activate
 @if "%ERRORLEVEL%" NEQ "0" (
-       @echo. & echo "Error activating conda for Stable Diffusion. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
+       @echo. & echo "Error activating conda for Easy Diffusion. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
        pause
        exit /b
 )
@@ -27,138 +36,121 @@ if exist "Open Developer Console.cmd" del "Open Developer Console.cmd"
 
 @call python -c "import os; import shutil; frm = 'sd-ui-files\\ui\\hotfix\\9c24e6cd9f499d02c4f21a033736dabd365962dc80fe3aeb57a8f85ea45a20a3.26fead7ea4f0f843f6eb4055dfd25693f1a71f3c6871b184042d4b126244e142'; dst = os.path.join(os.path.expanduser('~'), '.cache', 'huggingface', 'transformers', '9c24e6cd9f499d02c4f21a033736dabd365962dc80fe3aeb57a8f85ea45a20a3.26fead7ea4f0f843f6eb4055dfd25693f1a71f3c6871b184042d4b126244e142'); shutil.copyfile(frm, dst) if os.path.exists(dst) else print(''); print('Hotfixed broken JSON file from OpenAI');"
 
-if NOT DEFINED test_sd2 set test_sd2=N
+@rem create the stable-diffusion folder, to work with legacy installations
+if not exist "stable-diffusion" mkdir stable-diffusion
+cd stable-diffusion
 
-@>nul findstr /m "sd_git_cloned" scripts\install_status.txt
-@if "%ERRORLEVEL%" EQU "0" (
-    @echo "Stable Diffusion's git repository was already installed. Updating.."
-
-    @cd stable-diffusion
-
-    @call git remote set-url origin https://github.com/easydiffusion/diffusion-kit.git
-
-    @call git reset --hard
-    @call git pull
-
-    if "%test_sd2%" == "N" (
-        @call git -c advice.detachedHead=false checkout 7f32368ed1030a6e710537047bacd908adea183a
-    )
-    if "%test_sd2%" == "Y" (
-        @call git -c advice.detachedHead=false checkout 733a1f6f9cae9b9a9b83294bf3281b123378cb1f
-    )
-
-    @cd ..
-) else (
-    @echo. & echo "Downloading Stable Diffusion.." & echo.
-
-    @call git clone https://github.com/easydiffusion/diffusion-kit.git stable-diffusion && (
-        @echo sd_git_cloned >> scripts\install_status.txt
-    ) || (
-        @echo "Error downloading Stable Diffusion. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
-        pause
-        @exit /b
-    )
-
-    @cd stable-diffusion
-    @call git -c advice.detachedHead=false checkout 7f32368ed1030a6e710537047bacd908adea183a
-
-    @cd ..
+@rem activate the old stable-diffusion env, if it exists
+if exist "env" (
+    call conda activate .\env
 )
 
-@cd stable-diffusion
+@rem disable the legacy src and ldm folder (otherwise this prevents installing gfpgan and realesrgan)
+if exist src rename src src-old
+if exist ldm rename ldm ldm-old
 
-@>nul findstr /m "conda_sd_env_created" ..\scripts\install_status.txt
-@if "%ERRORLEVEL%" EQU "0" (
-    @echo "Packages necessary for Stable Diffusion were already installed"
+if not exist "..\models\stable-diffusion" mkdir "..\models\stable-diffusion"
+if not exist "..\models\gfpgan" mkdir "..\models\gfpgan"
+if not exist "..\models\realesrgan" mkdir "..\models\realesrgan"
+if not exist "..\models\vae" mkdir "..\models\vae"
 
-    @call conda activate .\env
+@rem migrate the legacy models to the correct path (if already downloaded)
+if exist "sd-v1-4.ckpt" move sd-v1-4.ckpt ..\models\stable-diffusion\
+if exist "custom-model.ckpt" move custom-model.ckpt ..\models\stable-diffusion\
+if exist "GFPGANv1.3.pth" move GFPGANv1.3.pth ..\models\gfpgan\
+if exist "RealESRGAN_x4plus.pth" move RealESRGAN_x4plus.pth ..\models\realesrgan\
+if exist "RealESRGAN_x4plus_anime_6B.pth" move RealESRGAN_x4plus_anime_6B.pth ..\models\realesrgan\
+
+if not exist "%INSTALL_ENV_DIR%\DLLs\libssl-1_1-x64.dll"    copy "%INSTALL_ENV_DIR%\Library\bin\libssl-1_1-x64.dll"    "%INSTALL_ENV_DIR%\DLLs\"
+if not exist "%INSTALL_ENV_DIR%\DLLs\libcrypto-1_1-x64.dll" copy "%INSTALL_ENV_DIR%\Library\bin\libcrypto-1_1-x64.dll" "%INSTALL_ENV_DIR%\DLLs\"
+
+@rem install torch and torchvision
+call python ..\scripts\check_modules.py torch torchvision
+if "%ERRORLEVEL%" EQU "0" (
+    echo "torch and torchvision have already been installed."
 ) else (
-    @echo. & echo "Downloading packages necessary for Stable Diffusion.." & echo. & echo "***** This will take some time (depending on the speed of the Internet connection) and may appear to be stuck, but please be patient ***** .." & echo.
+    echo "Installing torch and torchvision.."
 
-    @rmdir /s /q .\env
+    @REM prevent from using packages from the user's home directory, to avoid conflicts
+    set PYTHONNOUSERSITE=1
+    set PYTHONPATH=%INSTALL_ENV_DIR%\lib\site-packages
 
-    @REM prevent conda from using packages from the user's home directory, to avoid conflicts
-    @set PYTHONNOUSERSITE=1
-
-    set USERPROFILE=%cd%\profile
-
-    set PYTHONPATH=%cd%;%cd%\env\lib\site-packages
-
-    @call conda env create --prefix env -f environment.yaml || (
-        @echo. & echo "Error installing the packages necessary for Stable Diffusion. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
+    call python -m pip install --upgrade torch torchvision --extra-index-url https://download.pytorch.org/whl/cu116 || (
+        echo "Error installing torch. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
         pause
         exit /b
     )
-
-    @call conda activate .\env
-
-    for /f "tokens=*" %%a in ('python -c "import torch; import ldm; import transformers; import numpy; import antlr4; print(42)"') do if "%%a" NEQ "42" (
-        @echo. & echo "Dependency test failed! Error installing the packages necessary for Stable Diffusion. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
-        pause
-        exit /b
-    )
-
-    @echo conda_sd_env_created >> ..\scripts\install_status.txt
 )
 
 set PATH=C:\Windows\System32;%PATH%
 
-@>nul findstr /m "conda_sd_gfpgan_deps_installed" ..\scripts\install_status.txt
-@if "%ERRORLEVEL%" EQU "0" (
-    @echo "Packages necessary for GFPGAN (Face Correction) were already installed"
+@rem install/upgrade sdkit
+call python ..\scripts\check_modules.py sdkit sdkit.models ldm transformers numpy antlr4 gfpgan realesrgan
+if "%ERRORLEVEL%" EQU "0" (
+    echo "sdkit is already installed."
+
+    @rem skip sdkit upgrade if in developer-mode
+    if not exist "..\src\sdkit" (
+        @REM prevent from using packages from the user's home directory, to avoid conflicts
+        set PYTHONNOUSERSITE=1
+        set PYTHONPATH=%INSTALL_ENV_DIR%\lib\site-packages
+
+        call python -m pip install --upgrade sdkit==1.0.47 -q || (
+            echo "Error updating sdkit"
+        )
+    )
 ) else (
-    @echo. & echo "Downloading packages necessary for GFPGAN (Face Correction).." & echo.
+    echo "Installing sdkit: https://pypi.org/project/sdkit/"
 
-    @set PYTHONNOUSERSITE=1
+    @REM prevent from using packages from the user's home directory, to avoid conflicts
+    set PYTHONNOUSERSITE=1
+    set PYTHONPATH=%INSTALL_ENV_DIR%\lib\site-packages
 
-    set USERPROFILE=%cd%\profile
-
-    set PYTHONPATH=%cd%;%cd%\env\lib\site-packages
-
-    for /f "tokens=*" %%a in ('python -c "from gfpgan import GFPGANer; print(42)"') do if "%%a" NEQ "42" (
-        @echo. & echo "Dependency test failed! Error installing the packages necessary for GFPGAN (Face Correction). Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
+    call python -m pip install sdkit==1.0.47 || (
+        echo "Error installing sdkit. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
         pause
         exit /b
     )
-
-    @echo conda_sd_gfpgan_deps_installed >> ..\scripts\install_status.txt
 )
 
-@>nul findstr /m "conda_sd_esrgan_deps_installed" ..\scripts\install_status.txt
-@if "%ERRORLEVEL%" EQU "0" (
-    @echo "Packages necessary for ESRGAN (Resolution Upscaling) were already installed"
+call python -c "from importlib.metadata import version; print('sdkit version:', version('sdkit'))"
+
+@rem upgrade stable-diffusion-sdkit
+call python -m pip install --upgrade stable-diffusion-sdkit==2.1.3 -q || (
+    echo "Error updating stable-diffusion-sdkit"
+)
+call python -c "from importlib.metadata import version; print('stable-diffusion version:', version('stable-diffusion-sdkit'))"
+
+@rem install rich
+call python ..\scripts\check_modules.py rich
+if "%ERRORLEVEL%" EQU "0" (
+    echo "rich has already been installed."
 ) else (
-    @echo. & echo "Downloading packages necessary for ESRGAN (Resolution Upscaling).." & echo.
+    echo "Installing rich.."
 
-    @set PYTHONNOUSERSITE=1
+    set PYTHONNOUSERSITE=1
+    set PYTHONPATH=%INSTALL_ENV_DIR%\lib\site-packages
 
-    set USERPROFILE=%cd%\profile
-
-    set PYTHONPATH=%cd%;%cd%\env\lib\site-packages
-
-    for /f "tokens=*" %%a in ('python -c "from basicsr.archs.rrdbnet_arch import RRDBNet; from realesrgan import RealESRGANer; print(42)"') do if "%%a" NEQ "42" (
-        @echo. & echo "Dependency test failed! Error installing the packages necessary for ESRGAN (Resolution Upscaling). Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
+    call python -m pip install rich || (
+        echo "Error installing rich. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
         pause
         exit /b
     )
-
-    @echo conda_sd_esrgan_deps_installed >> ..\scripts\install_status.txt
 )
 
-@>nul findstr /m "conda_sd_ui_deps_installed" ..\scripts\install_status.txt
+set PATH=C:\Windows\System32;%PATH%
+
+call python ..\scripts\check_modules.py uvicorn fastapi
 @if "%ERRORLEVEL%" EQU "0" (
-    echo "Packages necessary for Stable Diffusion UI were already installed"
+    echo "Packages necessary for Easy Diffusion were already installed"
 ) else (
-    @echo. & echo "Downloading packages necessary for Stable Diffusion UI.." & echo.
+    @echo. & echo "Downloading packages necessary for Easy Diffusion..." & echo.
 
-    @set PYTHONNOUSERSITE=1
+    set PYTHONNOUSERSITE=1
+    set PYTHONPATH=%INSTALL_ENV_DIR%\lib\site-packages
 
-    set USERPROFILE=%cd%\profile
-
-    set PYTHONPATH=%cd%;%cd%\env\lib\site-packages
-
-    @call conda install -c conda-forge -y --prefix env uvicorn fastapi || (
-        echo "Error installing the packages necessary for Stable Diffusion UI. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
+    @call conda install -c conda-forge -y uvicorn fastapi || (
+        echo "Error installing the packages necessary for Easy Diffusion. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
         pause
         exit /b
     )
@@ -172,64 +164,35 @@ call WHERE uvicorn > .tmp
     exit /b
 )
 
-@>nul 2>nul call python -m picklescan --help
-@if "%ERRORLEVEL%" NEQ "0" (
-    @echo. & echo Picklescan not found. Installing
-    @call pip install picklescan || (
-        echo "Error installing the picklescan package necessary for Stable Diffusion UI. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
-        pause
-        exit /b
-    )
-)
-
-@>nul 2>nul call python -c "import safetensors"
-@if "%ERRORLEVEL%" NEQ "0" (
-    @echo. & echo SafeTensors not found. Installing
-    @call pip install safetensors || (
-        echo "Error installing the safetensors package necessary for Stable Diffusion UI. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!"
-        pause
-        exit /b
-    )
-)
-
 @>nul findstr /m "conda_sd_ui_deps_installed" ..\scripts\install_status.txt
 @if "%ERRORLEVEL%" NEQ "0" (
     @echo conda_sd_ui_deps_installed >> ..\scripts\install_status.txt
 )
 
-
-
-if not exist "..\models\stable-diffusion" mkdir "..\models\stable-diffusion"
-if not exist "..\models\vae" mkdir "..\models\vae"
-if not exist "..\models\hypernetwork" mkdir "..\models\hypernetwork"
-echo. > "..\models\stable-diffusion\Put your custom ckpt files here.txt"
-echo. > "..\models\vae\Put your VAE files here.txt"
-echo. > "..\models\hypernetwork\Put your hypernetwork files here.txt"
-
-@if exist "sd-v1-4.ckpt" (
-    for %%I in ("sd-v1-4.ckpt") do if "%%~zI" EQU "4265380512" (
+@if exist "..\models\stable-diffusion\sd-v1-4.ckpt" (
+    for %%I in ("..\models\stable-diffusion\sd-v1-4.ckpt") do if "%%~zI" EQU "4265380512" (
         echo "Data files (weights) necessary for Stable Diffusion were already downloaded. Using the HuggingFace 4 GB Model."
     ) else (
-        for %%J in ("sd-v1-4.ckpt") do if "%%~zJ" EQU "7703807346" (
+        for %%J in ("..\models\stable-diffusion\sd-v1-4.ckpt") do if "%%~zJ" EQU "7703807346" (
             echo "Data files (weights) necessary for Stable Diffusion were already downloaded. Using the HuggingFace 7 GB Model."
         ) else (
-            for %%K in ("sd-v1-4.ckpt") do if "%%~zK" EQU "7703810927" (
+            for %%K in ("..\models\stable-diffusion\sd-v1-4.ckpt") do if "%%~zK" EQU "7703810927" (
                 echo "Data files (weights) necessary for Stable Diffusion were already downloaded. Using the Waifu Model."
             ) else (
-                echo. & echo "The model file present at %cd%\sd-v1-4.ckpt is invalid. It is only %%~zK bytes in size. Re-downloading.." & echo.
-                del "sd-v1-4.ckpt"
+                echo. & echo "The model file present at models\stable-diffusion\sd-v1-4.ckpt is invalid. It is only %%~zK bytes in size. Re-downloading.." & echo.
+                del "..\models\stable-diffusion\sd-v1-4.ckpt"
             )
         )
     )
 )
 
-@if not exist "sd-v1-4.ckpt" (
+@if not exist "..\models\stable-diffusion\sd-v1-4.ckpt" (
     @echo. & echo "Downloading data files (weights) for Stable Diffusion.." & echo.
 
-    @call curl -L -k https://me.cmdr2.org/stable-diffusion-ui/sd-v1-4.ckpt > sd-v1-4.ckpt
+    @call curl -L -k https://huggingface.co/CompVis/stable-diffusion-v-1-4-original/resolve/main/sd-v1-4.ckpt > ..\models\stable-diffusion\sd-v1-4.ckpt
 
-    @if exist "sd-v1-4.ckpt" (
-        for %%I in ("sd-v1-4.ckpt") do if "%%~zI" NEQ "4265380512" (
+    @if exist "..\models\stable-diffusion\sd-v1-4.ckpt" (
+        for %%I in ("..\models\stable-diffusion\sd-v1-4.ckpt") do if "%%~zI" NEQ "4265380512" (
             echo. & echo "Error: The downloaded model file was invalid! Bytes downloaded: %%~zI" & echo.
             echo. & echo "Error downloading the data files (weights) for Stable Diffusion. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
             pause
@@ -244,22 +207,22 @@ echo. > "..\models\hypernetwork\Put your hypernetwork files here.txt"
 
 
 
-@if exist "GFPGANv1.3.pth" (
-    for %%I in ("GFPGANv1.3.pth") do if "%%~zI" EQU "348632874" (
+@if exist "..\models\gfpgan\GFPGANv1.3.pth" (
+    for %%I in ("..\models\gfpgan\GFPGANv1.3.pth") do if "%%~zI" EQU "348632874" (
         echo "Data files (weights) necessary for GFPGAN (Face Correction) were already downloaded"
     ) else (
-        echo. & echo "The GFPGAN model file present at %cd%\GFPGANv1.3.pth is invalid. It is only %%~zI bytes in size. Re-downloading.." & echo.
-        del "GFPGANv1.3.pth"
+        echo. & echo "The GFPGAN model file present at models\gfpgan\GFPGANv1.3.pth is invalid. It is only %%~zI bytes in size. Re-downloading.." & echo.
+        del "..\models\gfpgan\GFPGANv1.3.pth"
     )
 )
 
-@if not exist "GFPGANv1.3.pth" (
+@if not exist "..\models\gfpgan\GFPGANv1.3.pth" (
     @echo. & echo "Downloading data files (weights) for GFPGAN (Face Correction).." & echo.
 
-    @call curl -L -k https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.3.pth > GFPGANv1.3.pth
+    @call curl -L -k https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.3.pth > ..\models\gfpgan\GFPGANv1.3.pth
 
-    @if exist "GFPGANv1.3.pth" (
-        for %%I in ("GFPGANv1.3.pth") do if "%%~zI" NEQ "348632874" (
+    @if exist "..\models\gfpgan\GFPGANv1.3.pth" (
+        for %%I in ("..\models\gfpgan\GFPGANv1.3.pth") do if "%%~zI" NEQ "348632874" (
             echo. & echo "Error: The downloaded GFPGAN model file was invalid! Bytes downloaded: %%~zI" & echo.
             echo. & echo "Error downloading the data files (weights) for GFPGAN (Face Correction). Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
             pause
@@ -274,22 +237,22 @@ echo. > "..\models\hypernetwork\Put your hypernetwork files here.txt"
 
 
 
-@if exist "RealESRGAN_x4plus.pth" (
-    for %%I in ("RealESRGAN_x4plus.pth") do if "%%~zI" EQU "67040989" (
+@if exist "..\models\realesrgan\RealESRGAN_x4plus.pth" (
+    for %%I in ("..\models\realesrgan\RealESRGAN_x4plus.pth") do if "%%~zI" EQU "67040989" (
         echo "Data files (weights) necessary for ESRGAN (Resolution Upscaling) x4plus were already downloaded"
     ) else (
-        echo. & echo "The RealESRGAN model file present at %cd%\RealESRGAN_x4plus.pth is invalid. It is only %%~zI bytes in size. Re-downloading.." & echo.
-        del "RealESRGAN_x4plus.pth"
+        echo. & echo "The RealESRGAN model file present at models\realesrgan\RealESRGAN_x4plus.pth is invalid. It is only %%~zI bytes in size. Re-downloading.." & echo.
+        del "..\models\realesrgan\RealESRGAN_x4plus.pth"
     )
 )
 
-@if not exist "RealESRGAN_x4plus.pth" (
+@if not exist "..\models\realesrgan\RealESRGAN_x4plus.pth" (
     @echo. & echo "Downloading data files (weights) for ESRGAN (Resolution Upscaling) x4plus.." & echo.
 
-    @call curl -L -k https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth > RealESRGAN_x4plus.pth
+    @call curl -L -k https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth > ..\models\realesrgan\RealESRGAN_x4plus.pth
 
-    @if exist "RealESRGAN_x4plus.pth" (
-        for %%I in ("RealESRGAN_x4plus.pth") do if "%%~zI" NEQ "67040989" (
+    @if exist "..\models\realesrgan\RealESRGAN_x4plus.pth" (
+        for %%I in ("..\models\realesrgan\RealESRGAN_x4plus.pth") do if "%%~zI" NEQ "67040989" (
             echo. & echo "Error: The downloaded ESRGAN x4plus model file was invalid! Bytes downloaded: %%~zI" & echo.
             echo. & echo "Error downloading the data files (weights) for ESRGAN (Resolution Upscaling) x4plus. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
             pause
@@ -304,22 +267,22 @@ echo. > "..\models\hypernetwork\Put your hypernetwork files here.txt"
 
 
 
-@if exist "RealESRGAN_x4plus_anime_6B.pth" (
-    for %%I in ("RealESRGAN_x4plus_anime_6B.pth") do if "%%~zI" EQU "17938799" (
+@if exist "..\models\realesrgan\RealESRGAN_x4plus_anime_6B.pth" (
+    for %%I in ("..\models\realesrgan\RealESRGAN_x4plus_anime_6B.pth") do if "%%~zI" EQU "17938799" (
         echo "Data files (weights) necessary for ESRGAN (Resolution Upscaling) x4plus_anime were already downloaded"
     ) else (
-        echo. & echo "The RealESRGAN model file present at %cd%\RealESRGAN_x4plus_anime_6B.pth is invalid. It is only %%~zI bytes in size. Re-downloading.." & echo.
-        del "RealESRGAN_x4plus_anime_6B.pth"
+        echo. & echo "The RealESRGAN model file present at models\realesrgan\RealESRGAN_x4plus_anime_6B.pth is invalid. It is only %%~zI bytes in size. Re-downloading.." & echo.
+        del "..\models\realesrgan\RealESRGAN_x4plus_anime_6B.pth"
     )
 )
 
-@if not exist "RealESRGAN_x4plus_anime_6B.pth" (
+@if not exist "..\models\realesrgan\RealESRGAN_x4plus_anime_6B.pth" (
     @echo. & echo "Downloading data files (weights) for ESRGAN (Resolution Upscaling) x4plus_anime.." & echo.
 
-    @call curl -L -k https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth > RealESRGAN_x4plus_anime_6B.pth
+    @call curl -L -k https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth > ..\models\realesrgan\RealESRGAN_x4plus_anime_6B.pth
 
-    @if exist "RealESRGAN_x4plus_anime_6B.pth" (
-        for %%I in ("RealESRGAN_x4plus_anime_6B.pth") do if "%%~zI" NEQ "17938799" (
+    @if exist "..\models\realesrgan\RealESRGAN_x4plus_anime_6B.pth" (
+        for %%I in ("..\models\realesrgan\RealESRGAN_x4plus_anime_6B.pth") do if "%%~zI" NEQ "17938799" (
             echo. & echo "Error: The downloaded ESRGAN x4plus_anime model file was invalid! Bytes downloaded: %%~zI" & echo.
             echo. & echo "Error downloading the data files (weights) for ESRGAN (Resolution Upscaling) x4plus_anime. Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
             pause
@@ -362,24 +325,18 @@ echo. > "..\models\hypernetwork\Put your hypernetwork files here.txt"
     )
 )
 
-if "%test_sd2%" == "Y" (
-    @call pip install open_clip_torch==2.0.2
-)
-
 @>nul findstr /m "sd_install_complete" ..\scripts\install_status.txt
 @if "%ERRORLEVEL%" NEQ "0" (
     @echo sd_weights_downloaded >> ..\scripts\install_status.txt
     @echo sd_install_complete >> ..\scripts\install_status.txt
 )
 
-@echo. & echo "Stable Diffusion is ready!" & echo.
+@echo. & echo "Easy Diffusion installation complete! Starting the server!" & echo.
 
 @set SD_DIR=%cd%
 
-@cd env\lib\site-packages
-@set PYTHONPATH=%SD_DIR%;%cd%
-@cd ..\..\..
-@echo PYTHONPATH=%PYTHONPATH%
+set PYTHONPATH=%INSTALL_ENV_DIR%\lib\site-packages
+echo PYTHONPATH=%PYTHONPATH%
 
 call where python
 call python --version
@@ -388,17 +345,12 @@ call python --version
 @set SD_UI_PATH=%cd%\ui
 @cd stable-diffusion
 
-@rem
-@rem Rewrite easy-install.pth. This fixes the installation if the user has relocated the SDUI installation
-@rem
->env\Lib\site-packages\easy-install.pth echo %cd%\src\taming-transformers
->>env\Lib\site-packages\easy-install.pth echo %cd%\src\clip
->>env\Lib\site-packages\easy-install.pth echo %cd%\src\gfpgan
->>env\Lib\site-packages\easy-install.pth echo %cd%\src\realesrgan
+@rem set any overrides
+set HF_HUB_DISABLE_SYMLINKS_WARNING=true
 
 @if NOT DEFINED SD_UI_BIND_PORT set SD_UI_BIND_PORT=9000
 @if NOT DEFINED SD_UI_BIND_IP set SD_UI_BIND_IP=0.0.0.0
-@uvicorn server:app --app-dir "%SD_UI_PATH%" --port %SD_UI_BIND_PORT% --host %SD_UI_BIND_IP%
+@uvicorn main:server_api --app-dir "%SD_UI_PATH%" --port %SD_UI_BIND_PORT% --host %SD_UI_BIND_IP% --log-level error
 
 
 @pause

scripts/on_sd_start.sh

@@ -1,9 +1,11 @@
 #!/bin/bash
 
-source ./scripts/functions.sh
-
+cp sd-ui-files/scripts/functions.sh scripts/
 cp sd-ui-files/scripts/on_env_start.sh scripts/
 cp sd-ui-files/scripts/bootstrap.sh scripts/
+cp sd-ui-files/scripts/check_modules.py scripts/
+
+source ./scripts/functions.sh
 
 # activate the installer env
 CONDA_BASEPATH=$(conda info --base)
@@ -21,116 +23,110 @@ python -c "import os; import shutil; frm = 'sd-ui-files/ui/hotfix/9c24e6cd9f499d
 # Caution, this file will make your eyes and brain bleed. It's such an unholy mess.
 # Note to self: Please rewrite this in Python. For the sake of your own sanity.
 
-if [ "$test_sd2" == "" ]; then
-    export test_sd2="N"
-fi
-
-if [ -e "scripts/install_status.txt" ] && [ `grep -c sd_git_cloned scripts/install_status.txt` -gt "0" ]; then
-    echo "Stable Diffusion's git repository was already installed. Updating.."
-
-    cd stable-diffusion
-
-    git remote set-url origin https://github.com/easydiffusion/diffusion-kit.git
-
-    git reset --hard
-    git pull
-
-    if [ "$test_sd2" == "N" ]; then
-        git -c advice.detachedHead=false checkout 7f32368ed1030a6e710537047bacd908adea183a
-    elif [ "$test_sd2" == "Y" ]; then
-        git -c advice.detachedHead=false checkout 733a1f6f9cae9b9a9b83294bf3281b123378cb1f
-    fi
-
-    cd ..
-else
-    printf "\n\nDownloading Stable Diffusion..\n\n"
-
-    if git clone https://github.com/easydiffusion/diffusion-kit.git stable-diffusion ; then
-        echo sd_git_cloned >> scripts/install_status.txt
-    else
-        fail "git clone of basujindal/stable-diffusion.git failed"
-    fi
-
-    cd stable-diffusion
-    git -c advice.detachedHead=false checkout 7f32368ed1030a6e710537047bacd908adea183a
-
-    cd ..
+# set the correct installer path (current vs legacy)
+if [ -e "installer_files/env" ]; then
+    export INSTALL_ENV_DIR="$(pwd)/installer_files/env"
+fi
+if [ -e "stable-diffusion/env" ]; then
+    export INSTALL_ENV_DIR="$(pwd)/stable-diffusion/env"
 fi
 
+# create the stable-diffusion folder, to work with legacy installations
+if [ ! -e "stable-diffusion" ]; then mkdir stable-diffusion; fi
 cd stable-diffusion
 
-if [ `grep -c conda_sd_env_created ../scripts/install_status.txt` -gt "0" ]; then
-    echo "Packages necessary for Stable Diffusion were already installed"
-
+# activate the old stable-diffusion env, if it exists
+if [ -e "env" ]; then
     conda activate ./env || fail "conda activate failed"
+fi
+
+# disable the legacy src and ldm folder (otherwise this prevents installing gfpgan and realesrgan)
+if [ -e "src" ]; then mv src src-old; fi
+if [ -e "ldm" ]; then mv ldm ldm-old; fi
+
+mkdir -p "../models/stable-diffusion"
+mkdir -p "../models/gfpgan"
+mkdir -p "../models/realesrgan"
+mkdir -p "../models/vae"
+
+# migrate the legacy models to the correct path (if already downloaded)
+if [ -e "sd-v1-4.ckpt" ]; then mv sd-v1-4.ckpt ../models/stable-diffusion/; fi
+if [ -e "custom-model.ckpt" ]; then mv custom-model.ckpt ../models/stable-diffusion/; fi
+if [ -e "GFPGANv1.3.pth" ]; then mv GFPGANv1.3.pth ../models/gfpgan/; fi
+if [ -e "RealESRGAN_x4plus.pth" ]; then mv RealESRGAN_x4plus.pth ../models/realesrgan/; fi
+if [ -e "RealESRGAN_x4plus_anime_6B.pth" ]; then mv RealESRGAN_x4plus_anime_6B.pth ../models/realesrgan/; fi
+
+# install torch and torchvision
+if python ../scripts/check_modules.py torch torchvision; then
+    echo "torch and torchvision have already been installed."
 else
-    printf "\n\nDownloading packages necessary for Stable Diffusion..\n"
-    printf "\n\n***** This will take some time (depending on the speed of the Internet connection) and may appear to be stuck, but please be patient ***** ..\n\n"
+    echo "Installing torch and torchvision.."
 
-    # prevent conda from using packages from the user's home directory, to avoid conflicts
     export PYTHONNOUSERSITE=1
-    export PYTHONPATH="$(pwd):$(pwd)/env/lib/site-packages"
+    export PYTHONPATH="$INSTALL_ENV_DIR/lib/python3.8/site-packages"
 
-    if conda env create --prefix env --force -f environment.yaml ; then
-        echo "Installed. Testing.."
+    if python -m pip install --upgrade torch torchvision --extra-index-url https://download.pytorch.org/whl/cu116 ; then
+        echo "Installed."
     else
-        fail "'conda env create' failed"
+        fail "torch install failed" 
     fi
-
-    conda activate ./env || fail "conda activate failed"
-
-    out_test=`python -c "import torch; import ldm; import transformers; import numpy; import antlr4; print(42)"`
-    if [ "$out_test" != "42" ]; then
-        fail "Dependency test failed"
-    fi
-
-    echo conda_sd_env_created >> ../scripts/install_status.txt
 fi
 
-if [ `grep -c conda_sd_gfpgan_deps_installed ../scripts/install_status.txt` -gt "0" ]; then
-    echo "Packages necessary for GFPGAN (Face Correction) were already installed"
+# install/upgrade sdkit
+if python ../scripts/check_modules.py sdkit sdkit.models ldm transformers numpy antlr4 gfpgan realesrgan ; then
+    echo "sdkit is already installed."
+
+    # skip sdkit upgrade if in developer-mode
+    if [ ! -e "../src/sdkit" ]; then
+        export PYTHONNOUSERSITE=1
+        export PYTHONPATH="$INSTALL_ENV_DIR/lib/python3.8/site-packages"
+
+        python -m pip install --upgrade sdkit==1.0.47 -q
+    fi
 else
-    printf "\n\nDownloading packages necessary for GFPGAN (Face Correction)..\n"
+    echo "Installing sdkit: https://pypi.org/project/sdkit/"
 
     export PYTHONNOUSERSITE=1
-    export PYTHONPATH="$(pwd):$(pwd)/env/lib/site-packages"
+    export PYTHONPATH="$INSTALL_ENV_DIR/lib/python3.8/site-packages"
 
-    out_test=`python -c "from gfpgan import GFPGANer; print(42)"`
-    if [ "$out_test" != "42" ]; then
-        echo "EE The dependency check has failed. This usually means that some system libraries are missing."
-	echo "EE On Debian/Ubuntu systems, this are often these packages: libsm6 libxext6 libxrender-dev"
-	echo "EE Other Linux distributions might have different package names for these libraries."
-        fail "GFPGAN dependency test failed"
+    if python -m pip install sdkit==1.0.47 ; then
+        echo "Installed."
+    else
+        fail "sdkit install failed"
     fi
-
-    echo conda_sd_gfpgan_deps_installed >> ../scripts/install_status.txt
 fi
 
-if [ `grep -c conda_sd_esrgan_deps_installed ../scripts/install_status.txt` -gt "0" ]; then
-    echo "Packages necessary for ESRGAN (Resolution Upscaling) were already installed"
+python -c "from importlib.metadata import version; print('sdkit version:', version('sdkit'))"
+
+# upgrade stable-diffusion-sdkit
+python -m pip install --upgrade stable-diffusion-sdkit==2.1.3 -q
+python -c "from importlib.metadata import version; print('stable-diffusion version:', version('stable-diffusion-sdkit'))"
+
+# install rich
+if python ../scripts/check_modules.py rich; then
+    echo "rich has already been installed."
 else
-    printf "\n\nDownloading packages necessary for ESRGAN (Resolution Upscaling)..\n"
+    echo "Installing rich.."
 
     export PYTHONNOUSERSITE=1
-    export PYTHONPATH="$(pwd):$(pwd)/env/lib/site-packages"
+    export PYTHONPATH="$INSTALL_ENV_DIR/lib/python3.8/site-packages"
 
-    out_test=`python -c "from basicsr.archs.rrdbnet_arch import RRDBNet; from realesrgan import RealESRGANer; print(42)"`
-    if [ "$out_test" != "42" ]; then
-        fail "ESRGAN dependency test failed"
+    if python -m pip install rich ; then
+        echo "Installed."
+    else
+        fail "Install failed for rich"
     fi
-
-    echo conda_sd_esrgan_deps_installed >> ../scripts/install_status.txt
 fi
 
-if [ `grep -c conda_sd_ui_deps_installed ../scripts/install_status.txt` -gt "0" ]; then
-    echo "Packages necessary for Stable Diffusion UI were already installed"
+if python ../scripts/check_modules.py uvicorn fastapi ; then
+    echo "Packages necessary for Easy Diffusion were already installed"
 else
-    printf "\n\nDownloading packages necessary for Stable Diffusion UI..\n\n"
+    printf "\n\nDownloading packages necessary for Easy Diffusion..\n\n"
 
     export PYTHONNOUSERSITE=1
-    export PYTHONPATH="$(pwd):$(pwd)/env/lib/site-packages"
+    export PYTHONPATH="$INSTALL_ENV_DIR/lib/python3.8/site-packages"
 
-    if conda install -c conda-forge --prefix ./env -y uvicorn fastapi ; then
+    if conda install -c conda-forge -y uvicorn fastapi ; then
         echo "Installed. Testing.."
     else
         fail "'conda install uvicorn' failed" 
@@ -139,51 +135,26 @@ else
     if ! command -v uvicorn &> /dev/null; then
         fail "UI packages not found!"
     fi
-
-    echo conda_sd_ui_deps_installed >> ../scripts/install_status.txt
 fi
 
-if python -m picklescan --help >/dev/null 2>&1; then
-    echo "Picklescan is already installed."
-else
-    echo "Picklescan not found, installing."
-    pip install picklescan || fail "Picklescan installation failed."
-fi
-
-if python -c "import safetensors" --help >/dev/null 2>&1; then
-    echo "SafeTensors is already installed."
-else
-    echo "SafeTensors not found, installing."
-    pip install safetensors || fail "SafeTensors installation failed."
-fi
-
-
-
-mkdir -p "../models/stable-diffusion"
-mkdir -p "../models/vae"
-mkdir -p "../models/hypernetwork"
-echo "" > "../models/stable-diffusion/Put your custom ckpt files here.txt"
-echo "" > "../models/vae/Put your VAE files here.txt"
-echo "" > "../models/hypernetwork/Put your hypernetwork files here.txt"
-
-if [ -f "sd-v1-4.ckpt" ]; then
-    model_size=`find "sd-v1-4.ckpt" -printf "%s"`
+if [ -f "../models/stable-diffusion/sd-v1-4.ckpt" ]; then
+    model_size=`filesize "../models/stable-diffusion/sd-v1-4.ckpt"`
 
     if [ "$model_size" -eq "4265380512" ] || [ "$model_size" -eq "7703807346" ] || [ "$model_size" -eq "7703810927" ]; then
         echo "Data files (weights) necessary for Stable Diffusion were already downloaded"
     else
-        printf "\n\nThe model file present at $PWD/sd-v1-4.ckpt is invalid. It is only $model_size bytes in size. Re-downloading.."
-        rm sd-v1-4.ckpt
+        printf "\n\nThe model file present at models/stable-diffusion/sd-v1-4.ckpt is invalid. It is only $model_size bytes in size. Re-downloading.."
+        rm ../models/stable-diffusion/sd-v1-4.ckpt
     fi
 fi
 
-if [ ! -f "sd-v1-4.ckpt" ]; then
+if [ ! -f "../models/stable-diffusion/sd-v1-4.ckpt" ]; then
     echo "Downloading data files (weights) for Stable Diffusion.."
 
-    curl -L -k https://me.cmdr2.org/stable-diffusion-ui/sd-v1-4.ckpt > sd-v1-4.ckpt
+    curl -L -k https://huggingface.co/CompVis/stable-diffusion-v-1-4-original/resolve/main/sd-v1-4.ckpt > ../models/stable-diffusion/sd-v1-4.ckpt
 
-    if [ -f "sd-v1-4.ckpt" ]; then
-        model_size=`find "sd-v1-4.ckpt" -printf "%s"`
+    if [ -f "../models/stable-diffusion/sd-v1-4.ckpt" ]; then
+        model_size=`filesize "../models/stable-diffusion/sd-v1-4.ckpt"`
         if [ ! "$model_size" == "4265380512" ]; then
 	    fail "The downloaded model file was invalid! Bytes downloaded: $model_size"
         fi
@@ -193,24 +164,24 @@ if [ ! -f "sd-v1-4.ckpt" ]; then
 fi
 
 
-if [ -f "GFPGANv1.3.pth" ]; then
-    model_size=`find "GFPGANv1.3.pth" -printf "%s"`
+if [ -f "../models/gfpgan/GFPGANv1.3.pth" ]; then
+    model_size=`filesize "../models/gfpgan/GFPGANv1.3.pth"`
 
     if [ "$model_size" -eq "348632874" ]; then
         echo "Data files (weights) necessary for GFPGAN (Face Correction) were already downloaded"
     else
-        printf "\n\nThe model file present at $PWD/GFPGANv1.3.pth is invalid. It is only $model_size bytes in size. Re-downloading.."
-        rm GFPGANv1.3.pth
+        printf "\n\nThe model file present at models/gfpgan/GFPGANv1.3.pth is invalid. It is only $model_size bytes in size. Re-downloading.."
+        rm ../models/gfpgan/GFPGANv1.3.pth
     fi
 fi
 
-if [ ! -f "GFPGANv1.3.pth" ]; then
+if [ ! -f "../models/gfpgan/GFPGANv1.3.pth" ]; then
     echo "Downloading data files (weights) for GFPGAN (Face Correction).."
 
-    curl -L -k https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.3.pth > GFPGANv1.3.pth
+    curl -L -k https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.3.pth > ../models/gfpgan/GFPGANv1.3.pth
 
-    if [ -f "GFPGANv1.3.pth" ]; then
-        model_size=`find "GFPGANv1.3.pth" -printf "%s"`
+    if [ -f "../models/gfpgan/GFPGANv1.3.pth" ]; then
+        model_size=`filesize "../models/gfpgan/GFPGANv1.3.pth"`
         if [ ! "$model_size" -eq "348632874" ]; then
             fail "The downloaded GFPGAN model file was invalid! Bytes downloaded: $model_size"
         fi
@@ -220,24 +191,24 @@ if [ ! -f "GFPGANv1.3.pth" ]; then
 fi
 
 
-if [ -f "RealESRGAN_x4plus.pth" ]; then
-    model_size=`find "RealESRGAN_x4plus.pth" -printf "%s"`
+if [ -f "../models/realesrgan/RealESRGAN_x4plus.pth" ]; then
+    model_size=`filesize "../models/realesrgan/RealESRGAN_x4plus.pth"`
 
     if [ "$model_size" -eq "67040989" ]; then
         echo "Data files (weights) necessary for ESRGAN (Resolution Upscaling) x4plus were already downloaded"
     else
-        printf "\n\nThe model file present at $PWD/RealESRGAN_x4plus.pth is invalid. It is only $model_size bytes in size. Re-downloading.."
-        rm RealESRGAN_x4plus.pth
+        printf "\n\nThe model file present at models/realesrgan/RealESRGAN_x4plus.pth is invalid. It is only $model_size bytes in size. Re-downloading.."
+        rm ../models/realesrgan/RealESRGAN_x4plus.pth
     fi
 fi
 
-if [ ! -f "RealESRGAN_x4plus.pth" ]; then
+if [ ! -f "../models/realesrgan/RealESRGAN_x4plus.pth" ]; then
     echo "Downloading data files (weights) for ESRGAN (Resolution Upscaling) x4plus.."
 
-    curl -L -k https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth > RealESRGAN_x4plus.pth
+    curl -L -k https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth > ../models/realesrgan/RealESRGAN_x4plus.pth
 
-    if [ -f "RealESRGAN_x4plus.pth" ]; then
-        model_size=`find "RealESRGAN_x4plus.pth" -printf "%s"`
+    if [ -f "../models/realesrgan/RealESRGAN_x4plus.pth" ]; then
+        model_size=`filesize "../models/realesrgan/RealESRGAN_x4plus.pth"`
         if [ ! "$model_size" -eq "67040989" ]; then
             fail "The downloaded ESRGAN x4plus model file was invalid! Bytes downloaded: $model_size"
         fi
@@ -247,24 +218,24 @@ if [ ! -f "RealESRGAN_x4plus.pth" ]; then
 fi
 
 
-if [ -f "RealESRGAN_x4plus_anime_6B.pth" ]; then
-    model_size=`find "RealESRGAN_x4plus_anime_6B.pth" -printf "%s"`
+if [ -f "../models/realesrgan/RealESRGAN_x4plus_anime_6B.pth" ]; then
+    model_size=`filesize "../models/realesrgan/RealESRGAN_x4plus_anime_6B.pth"`
 
     if [ "$model_size" -eq "17938799" ]; then
         echo "Data files (weights) necessary for ESRGAN (Resolution Upscaling) x4plus_anime were already downloaded"
     else
-        printf "\n\nThe model file present at $PWD/RealESRGAN_x4plus_anime_6B.pth is invalid. It is only $model_size bytes in size. Re-downloading.."
-        rm RealESRGAN_x4plus_anime_6B.pth
+        printf "\n\nThe model file present at models/realesrgan/RealESRGAN_x4plus_anime_6B.pth is invalid. It is only $model_size bytes in size. Re-downloading.."
+        rm ../models/realesrgan/RealESRGAN_x4plus_anime_6B.pth
     fi
 fi
 
-if [ ! -f "RealESRGAN_x4plus_anime_6B.pth" ]; then
+if [ ! -f "../models/realesrgan/RealESRGAN_x4plus_anime_6B.pth" ]; then
     echo "Downloading data files (weights) for ESRGAN (Resolution Upscaling) x4plus_anime.."
 
-    curl -L -k https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth > RealESRGAN_x4plus_anime_6B.pth
+    curl -L -k https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth > ../models/realesrgan/RealESRGAN_x4plus_anime_6B.pth
 
-    if [ -f "RealESRGAN_x4plus_anime_6B.pth" ]; then
-        model_size=`find "RealESRGAN_x4plus_anime_6B.pth" -printf "%s"`
+    if [ -f "../models/realesrgan/RealESRGAN_x4plus_anime_6B.pth" ]; then
+        model_size=`filesize "../models/realesrgan/RealESRGAN_x4plus_anime_6B.pth"`
         if [ ! "$model_size" -eq "17938799" ]; then
             fail "The downloaded ESRGAN x4plus_anime model file was invalid! Bytes downloaded: $model_size"
         fi
@@ -275,7 +246,7 @@ fi
 
 
 if [ -f "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt" ]; then
-    model_size=`find ../models/vae/vae-ft-mse-840000-ema-pruned.ckpt -printf "%s"`
+    model_size=`filesize "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt"`
 
     if [ "$model_size" -eq "334695179" ]; then
         echo "Data files (weights) necessary for the default VAE (sd-vae-ft-mse-original) were already downloaded"
@@ -291,7 +262,7 @@ if [ ! -f "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt" ]; then
     curl -L -k https://huggingface.co/stabilityai/sd-vae-ft-mse-original/resolve/main/vae-ft-mse-840000-ema-pruned.ckpt > ../models/vae/vae-ft-mse-840000-ema-pruned.ckpt
 
     if [ -f "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt" ]; then
-        model_size=`find ../models/vae/vae-ft-mse-840000-ema-pruned.ckpt -printf "%s"`
+        model_size=`filesize "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt"`
         if [ ! "$model_size" -eq "334695179" ]; then
             printf "\n\nError: The downloaded default VAE (sd-vae-ft-mse-original) file was invalid! Bytes downloaded: $model_size\n\n"
             printf "\n\nError downloading the data files (weights) for the default VAE (sd-vae-ft-mse-original). Sorry about that, please try to:\n  1. Run this installer again.\n  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting\n  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB\n  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues\nThanks!\n\n"
@@ -305,19 +276,17 @@ if [ ! -f "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt" ]; then
     fi
 fi
 
-if [ "$test_sd2" == "Y" ]; then
-    pip install open_clip_torch==2.0.2
-fi
-
 if [ `grep -c sd_install_complete ../scripts/install_status.txt` -gt "0" ]; then
     echo sd_weights_downloaded >> ../scripts/install_status.txt
     echo sd_install_complete >> ../scripts/install_status.txt
 fi
 
-printf "\n\nStable Diffusion is ready!\n\n"
+printf "\n\nEasy Diffusion installation complete, starting the server!\n\n"
 
 SD_PATH=`pwd`
-export PYTHONPATH="$SD_PATH:$SD_PATH/env/lib/python3.8/site-packages"
+
+export PYTORCH_ENABLE_MPS_FALLBACK=1
+export PYTHONPATH="$INSTALL_ENV_DIR/lib/python3.8/site-packages"
 echo "PYTHONPATH=$PYTHONPATH"
 
 which python
@@ -327,6 +296,6 @@ cd ..
 export SD_UI_PATH=`pwd`/ui
 cd stable-diffusion
 
-uvicorn server:app --app-dir "$SD_UI_PATH" --port ${SD_UI_BIND_PORT:-9000} --host ${SD_UI_BIND_IP:-0.0.0.0}
+uvicorn main:server_api --app-dir "$SD_UI_PATH" --port ${SD_UI_BIND_PORT:-9000} --host ${SD_UI_BIND_IP:-0.0.0.0} --log-level error
 
 read -p "Press any key to continue"

scripts/start.sh

@@ -2,6 +2,24 @@
 
 cd "$(dirname "${BASH_SOURCE[0]}")"
 
+if [ -f "on_sd_start.bat" ]; then
+    echo ================================================================================
+    echo
+    echo !!!! WARNING !!!!
+    echo
+    echo It looks like you\'re trying to run the installation script from a source code
+    echo download. This will not work.
+    echo
+    echo Recommended: Please close this window and download the installer from
+    echo https://stable-diffusion-ui.github.io/docs/installation/
+    echo
+    echo ================================================================================
+    echo
+    read
+    exit 1
+fi
+
+
 # set legacy installer's PATH, if it exists
 if [ -e "installer" ]; then export PATH="$(pwd)/installer/bin:$PATH"; fi
 
@@ -19,4 +37,5 @@ which conda
 conda --version || exit 1
 
 # Download the rest of the installer and UI
+chmod +x scripts/*.sh
 scripts/on_env_start.sh

ui/easydiffusion/app.py

@@ -0,0 +1,328 @@
+import os
+import socket
+import sys
+import json
+import traceback
+import logging
+import shlex
+import urllib
+from rich.logging import RichHandler
+
+from sdkit.utils import log as sdkit_log  # hack, so we can overwrite the log config
+
+from easydiffusion import task_manager
+from easydiffusion.utils import log
+
+# Remove all handlers associated with the root logger object.
+for handler in logging.root.handlers[:]:
+    logging.root.removeHandler(handler)
+
+LOG_FORMAT = "%(asctime)s.%(msecs)03d %(levelname)s %(threadName)s %(message)s"
+logging.basicConfig(
+    level=logging.INFO,
+    format=LOG_FORMAT,
+    datefmt="%X",
+    handlers=[RichHandler(markup=True, rich_tracebacks=False, show_time=False, show_level=False)],
+)
+
+SD_DIR = os.getcwd()
+
+SD_UI_DIR = os.getenv("SD_UI_PATH", None)
+
+CONFIG_DIR = os.path.abspath(os.path.join(SD_UI_DIR, "..", "scripts"))
+MODELS_DIR = os.path.abspath(os.path.join(SD_DIR, "..", "models"))
+
+USER_PLUGINS_DIR = os.path.abspath(os.path.join(SD_DIR, "..", "plugins"))
+CORE_PLUGINS_DIR = os.path.abspath(os.path.join(SD_UI_DIR, "plugins"))
+
+USER_UI_PLUGINS_DIR = os.path.join(USER_PLUGINS_DIR, "ui")
+CORE_UI_PLUGINS_DIR = os.path.join(CORE_PLUGINS_DIR, "ui")
+USER_SERVER_PLUGINS_DIR = os.path.join(USER_PLUGINS_DIR, "server")
+UI_PLUGINS_SOURCES = ((CORE_UI_PLUGINS_DIR, "core"), (USER_UI_PLUGINS_DIR, "user"))
+
+sys.path.append(os.path.dirname(SD_UI_DIR))
+sys.path.append(USER_SERVER_PLUGINS_DIR)
+
+OUTPUT_DIRNAME = "Stable Diffusion UI"  # in the user's home folder
+PRESERVE_CONFIG_VARS = ["FORCE_FULL_PRECISION"]
+TASK_TTL = 15 * 60  # Discard last session's task timeout
+APP_CONFIG_DEFAULTS = {
+    # auto: selects the cuda device with the most free memory, cuda: use the currently active cuda device.
+    "render_devices": "auto",  # valid entries: 'auto', 'cpu' or 'cuda:N' (where N is a GPU index)
+    "update_branch": "main",
+    "ui": {
+        "open_browser_on_start": True,
+    },
+}
+
+IMAGE_EXTENSIONS = [".png", ".apng", ".jpg", ".jpeg", ".jfif", ".pjpeg", ".pjp", ".jxl", ".gif", ".webp", ".avif", ".svg"]
+CUSTOM_MODIFIERS_DIR = os.path.abspath(os.path.join(SD_DIR, "..", "modifiers"))
+CUSTOM_MODIFIERS_PORTRAIT_EXTENSIONS=[".portrait", "_portrait", " portrait", "-portrait"]
+CUSTOM_MODIFIERS_LANDSCAPE_EXTENSIONS=[".landscape", "_landscape", " landscape", "-landscape"]
+
+def init():
+    os.makedirs(USER_UI_PLUGINS_DIR, exist_ok=True)
+    os.makedirs(USER_SERVER_PLUGINS_DIR, exist_ok=True)
+
+    load_server_plugins()
+
+    update_render_threads()
+
+
+def getConfig(default_val=APP_CONFIG_DEFAULTS):
+    try:
+        config_json_path = os.path.join(CONFIG_DIR, "config.json")
+        if not os.path.exists(config_json_path):
+            config = default_val
+        else:
+            with open(config_json_path, "r", encoding="utf-8") as f:
+                config = json.load(f)
+        if "net" not in config:
+            config["net"] = {}
+        if os.getenv("SD_UI_BIND_PORT") is not None:
+            config["net"]["listen_port"] = int(os.getenv("SD_UI_BIND_PORT"))
+        else:
+            config["net"]["listen_port"] = 9000
+        if os.getenv("SD_UI_BIND_IP") is not None:
+            config["net"]["listen_to_network"] = os.getenv("SD_UI_BIND_IP") == "0.0.0.0"
+        else:
+            config["net"]["listen_to_network"] = True
+        return config
+    except Exception as e:
+        log.warn(traceback.format_exc())
+        return default_val
+
+
+def setConfig(config):
+    try:  # config.json
+        config_json_path = os.path.join(CONFIG_DIR, "config.json")
+        with open(config_json_path, "w", encoding="utf-8") as f:
+            json.dump(config, f)
+    except:
+        log.error(traceback.format_exc())
+
+    try:  # config.bat
+        config_bat_path = os.path.join(CONFIG_DIR, "config.bat")
+        config_bat = []
+
+        if "update_branch" in config:
+            config_bat.append(f"@set update_branch={config['update_branch']}")
+
+        config_bat.append(f"@set SD_UI_BIND_PORT={config['net']['listen_port']}")
+        bind_ip = "0.0.0.0" if config["net"]["listen_to_network"] else "127.0.0.1"
+        config_bat.append(f"@set SD_UI_BIND_IP={bind_ip}")
+
+        # Preserve these variables if they are set
+        for var in PRESERVE_CONFIG_VARS:
+            if os.getenv(var) is not None:
+                config_bat.append(f"@set {var}={os.getenv(var)}")
+
+        if len(config_bat) > 0:
+            with open(config_bat_path, "w", encoding="utf-8") as f:
+                f.write("\n".join(config_bat))
+    except:
+        log.error(traceback.format_exc())
+
+    try:  # config.sh
+        config_sh_path = os.path.join(CONFIG_DIR, "config.sh")
+        config_sh = ["#!/bin/bash"]
+
+        if "update_branch" in config:
+            config_sh.append(f"export update_branch={config['update_branch']}")
+
+        config_sh.append(f"export SD_UI_BIND_PORT={config['net']['listen_port']}")
+        bind_ip = "0.0.0.0" if config["net"]["listen_to_network"] else "127.0.0.1"
+        config_sh.append(f"export SD_UI_BIND_IP={bind_ip}")
+
+        # Preserve these variables if they are set
+        for var in PRESERVE_CONFIG_VARS:
+            if os.getenv(var) is not None:
+                config_bat.append(f'export {var}="{shlex.quote(os.getenv(var))}"')
+
+        if len(config_sh) > 1:
+            with open(config_sh_path, "w", encoding="utf-8") as f:
+                f.write("\n".join(config_sh))
+    except:
+        log.error(traceback.format_exc())
+
+
+def save_to_config(ckpt_model_name, vae_model_name, hypernetwork_model_name, vram_usage_level):
+    config = getConfig()
+    if "model" not in config:
+        config["model"] = {}
+
+    config["model"]["stable-diffusion"] = ckpt_model_name
+    config["model"]["vae"] = vae_model_name
+    config["model"]["hypernetwork"] = hypernetwork_model_name
+
+    if vae_model_name is None or vae_model_name == "":
+        del config["model"]["vae"]
+    if hypernetwork_model_name is None or hypernetwork_model_name == "":
+        del config["model"]["hypernetwork"]
+
+    config["vram_usage_level"] = vram_usage_level
+
+    setConfig(config)
+
+
+def update_render_threads():
+    config = getConfig()
+    render_devices = config.get("render_devices", "auto")
+    active_devices = task_manager.get_devices()["active"].keys()
+
+    log.debug(f"requesting for render_devices: {render_devices}")
+    task_manager.update_render_threads(render_devices, active_devices)
+
+
+def getUIPlugins():
+    plugins = []
+
+    for plugins_dir, dir_prefix in UI_PLUGINS_SOURCES:
+        for file in os.listdir(plugins_dir):
+            if file.endswith(".plugin.js"):
+                plugins.append(f"/plugins/{dir_prefix}/{file}")
+
+    return plugins
+
+
+def load_server_plugins():
+    if not os.path.exists(USER_SERVER_PLUGINS_DIR):
+        return
+
+    import importlib
+
+    def load_plugin(file):
+        mod_path = file.replace(".py", "")
+        return importlib.import_module(mod_path)
+
+    def apply_plugin(file, plugin):
+        if hasattr(plugin, "get_cond_and_uncond"):
+            import sdkit.generate.image_generator
+
+            sdkit.generate.image_generator.get_cond_and_uncond = plugin.get_cond_and_uncond
+            log.info(f"Overridden get_cond_and_uncond with the one in the server plugin: {file}")
+
+    for file in os.listdir(USER_SERVER_PLUGINS_DIR):
+        file_path = os.path.join(USER_SERVER_PLUGINS_DIR, file)
+        if (not os.path.isdir(file_path) and not file_path.endswith("_plugin.py")) or (
+            os.path.isdir(file_path) and not file_path.endswith("_plugin")
+        ):
+            continue
+
+        try:
+            log.info(f"Loading server plugin: {file}")
+            mod = load_plugin(file)
+
+            log.info(f"Applying server plugin: {file}")
+            apply_plugin(file, mod)
+        except:
+            log.warn(f"Error while loading a server plugin")
+            log.warn(traceback.format_exc())
+
+
+def getIPConfig():
+    try:
+        ips = socket.gethostbyname_ex(socket.gethostname())
+        ips[2].append(ips[0])
+        return ips[2]
+    except Exception as e:
+        log.exception(e)
+        return []
+
+
+def open_browser():
+    config = getConfig()
+    ui = config.get("ui", {})
+    net = config.get("net", {"listen_port": 9000})
+    port = net.get("listen_port", 9000)
+    if ui.get("open_browser_on_start", True):
+        import webbrowser
+
+        webbrowser.open(f"http://localhost:{port}")
+
+def get_image_modifiers():
+    modifiers_json_path = os.path.join(SD_UI_DIR, "modifiers.json")
+
+    modifier_categories = {}
+    original_category_order=[]
+    with open(modifiers_json_path, "r", encoding="utf-8") as f:
+        modifiers_file = json.load(f)
+
+        # The trailing slash is needed to support symlinks
+        if not os.path.isdir(f"{CUSTOM_MODIFIERS_DIR}/"):
+            return modifiers_file
+
+        # convert modifiers from a list of objects to a dict of dicts
+        for category_item in modifiers_file:
+            category_name = category_item['category']
+            original_category_order.append(category_name)
+            category = {}
+            for modifier_item in category_item['modifiers']:
+                modifier = {}
+                for preview_item in modifier_item['previews']:
+                    modifier[preview_item['name']] = preview_item['path']
+                category[modifier_item['modifier']] = modifier
+            modifier_categories[category_name] = category
+
+    def scan_directory(directory_path: str, category_name="Modifiers"):
+        for entry in os.scandir(directory_path):
+            if entry.is_file():
+                file_extension = list(filter(lambda e: entry.name.endswith(e), IMAGE_EXTENSIONS))
+                if len(file_extension) == 0:
+                    continue
+
+                modifier_name = entry.name[: -len(file_extension[0])]
+                modifier_path = f"custom/{entry.path[len(CUSTOM_MODIFIERS_DIR) + 1:]}"
+                # URL encode path segments
+                modifier_path = "/".join(map(lambda segment: urllib.parse.quote(segment), modifier_path.split("/")))
+                is_portrait = True
+                is_landscape = True
+
+                portrait_extension = list(filter(lambda e: modifier_name.lower().endswith(e), CUSTOM_MODIFIERS_PORTRAIT_EXTENSIONS))
+                landscape_extension = list(filter(lambda e: modifier_name.lower().endswith(e), CUSTOM_MODIFIERS_LANDSCAPE_EXTENSIONS))
+
+                if len(portrait_extension) > 0:
+                    is_landscape = False
+                    modifier_name = modifier_name[: -len(portrait_extension[0])]
+                elif len(landscape_extension) > 0:
+                    is_portrait = False
+                    modifier_name = modifier_name[: -len(landscape_extension[0])]
+                
+                if (category_name not in modifier_categories):
+                    modifier_categories[category_name] = {}
+                
+                category = modifier_categories[category_name]
+
+                if (modifier_name not in category):
+                    category[modifier_name] = {}
+
+                if (is_portrait or "portrait" not in category[modifier_name]):
+                    category[modifier_name]["portrait"] = modifier_path
+                
+                if (is_landscape or "landscape" not in category[modifier_name]):
+                    category[modifier_name]["landscape"] = modifier_path
+            elif entry.is_dir():
+                scan_directory(
+                    entry.path,
+                    entry.name if directory_path==CUSTOM_MODIFIERS_DIR else f"{category_name}/{entry.name}",
+                )
+
+    scan_directory(CUSTOM_MODIFIERS_DIR)
+
+    custom_categories = sorted(
+        [cn for cn in modifier_categories.keys() if cn not in original_category_order],
+        key=str.casefold,
+    )
+
+    # convert the modifiers back into a list of objects
+    modifier_categories_list = []
+    for category_name in [*original_category_order, *custom_categories]:
+        category = { 'category': category_name, 'modifiers': [] }
+        for modifier_name in sorted(modifier_categories[category_name].keys(), key=str.casefold):
+            modifier = { 'modifier': modifier_name, 'previews': [] }
+            for preview_name, preview_path in modifier_categories[category_name][modifier_name].items():
+                modifier['previews'].append({ 'name': preview_name, 'path': preview_path })
+            category['modifiers'].append(modifier)
+        modifier_categories_list.append(category)
+
+    return modifier_categories_list

ui/easydiffusion/device_manager.py

@@ -0,0 +1,253 @@
+import os
+import platform
+import torch
+import traceback
+import re
+
+from easydiffusion.utils import log
+
+"""
+Set `FORCE_FULL_PRECISION` in the environment variables, or in `config.bat`/`config.sh` to set full precision (i.e. float32).
+Otherwise the models will load at half-precision (i.e. float16).
+
+Half-precision is fine most of the time. Full precision is only needed for working around GPU bugs (like NVIDIA 16xx GPUs).
+"""
+
+COMPARABLE_GPU_PERCENTILE = (
+    0.65  # if a GPU's free_mem is within this % of the GPU with the most free_mem, it will be picked
+)
+
+mem_free_threshold = 0
+
+
+def get_device_delta(render_devices, active_devices):
+    """
+    render_devices: 'cpu', or 'auto', or 'mps' or ['cuda:N'...]
+    active_devices: ['cpu', 'mps', 'cuda:N'...]
+    """
+
+    if render_devices in ("cpu", "auto", "mps"):
+        render_devices = [render_devices]
+    elif render_devices is not None:
+        if isinstance(render_devices, str):
+            render_devices = [render_devices]
+        if isinstance(render_devices, list) and len(render_devices) > 0:
+            render_devices = list(filter(lambda x: x.startswith("cuda:") or x == "mps", render_devices))
+            if len(render_devices) == 0:
+                raise Exception(
+                    'Invalid render_devices value in config.json. Valid: {"render_devices": ["cuda:0", "cuda:1"...]}, or {"render_devices": "cpu"} or {"render_devices": "mps"} or {"render_devices": "auto"}'
+                )
+
+            render_devices = list(filter(lambda x: is_device_compatible(x), render_devices))
+            if len(render_devices) == 0:
+                raise Exception(
+                    "Sorry, none of the render_devices configured in config.json are compatible with Stable Diffusion"
+                )
+        else:
+            raise Exception(
+                'Invalid render_devices value in config.json. Valid: {"render_devices": ["cuda:0", "cuda:1"...]}, or {"render_devices": "cpu"} or {"render_devices": "auto"}'
+            )
+    else:
+        render_devices = ["auto"]
+
+    if "auto" in render_devices:
+        render_devices = auto_pick_devices(active_devices)
+        if "cpu" in render_devices:
+            log.warn("WARNING: Could not find a compatible GPU. Using the CPU, but this will be very slow!")
+
+    active_devices = set(active_devices)
+    render_devices = set(render_devices)
+
+    devices_to_start = render_devices - active_devices
+    devices_to_stop = active_devices - render_devices
+
+    return devices_to_start, devices_to_stop
+
+
+def is_mps_available():
+    return (
+        platform.system() == "Darwin"
+        and hasattr(torch.backends, "mps")
+        and torch.backends.mps.is_available()
+        and torch.backends.mps.is_built()
+    )
+
+
+def is_cuda_available():
+    return torch.cuda.is_available()
+
+
+def auto_pick_devices(currently_active_devices):
+    global mem_free_threshold
+
+    if is_mps_available():
+        return ["mps"]
+
+    if not is_cuda_available():
+        return ["cpu"]
+
+    device_count = torch.cuda.device_count()
+    if device_count == 1:
+        return ["cuda:0"] if is_device_compatible("cuda:0") else ["cpu"]
+
+    log.debug("Autoselecting GPU. Using most free memory.")
+    devices = []
+    for device in range(device_count):
+        device = f"cuda:{device}"
+        if not is_device_compatible(device):
+            continue
+
+        mem_free, mem_total = torch.cuda.mem_get_info(device)
+        mem_free /= float(10**9)
+        mem_total /= float(10**9)
+        device_name = torch.cuda.get_device_name(device)
+        log.debug(
+            f"{device} detected: {device_name} - Memory (free/total): {round(mem_free, 2)}Gb / {round(mem_total, 2)}Gb"
+        )
+        devices.append({"device": device, "device_name": device_name, "mem_free": mem_free})
+
+    devices.sort(key=lambda x: x["mem_free"], reverse=True)
+    max_mem_free = devices[0]["mem_free"]
+    curr_mem_free_threshold = COMPARABLE_GPU_PERCENTILE * max_mem_free
+    mem_free_threshold = max(curr_mem_free_threshold, mem_free_threshold)
+
+    # Auto-pick algorithm:
+    # 1. Pick the top 75 percentile of the GPUs, sorted by free_mem.
+    # 2. Also include already-running devices (GPU-only), otherwise their free_mem will
+    #    always be very low (since their VRAM contains the model).
+    #    These already-running devices probably aren't terrible, since they were picked in the past.
+    #    Worst case, the user can restart the program and that'll get rid of them.
+    devices = list(
+        filter((lambda x: x["mem_free"] > mem_free_threshold or x["device"] in currently_active_devices), devices)
+    )
+    devices = list(map(lambda x: x["device"], devices))
+    return devices
+
+
+def device_init(context, device):
+    """
+    This function assumes the 'device' has already been verified to be compatible.
+    `get_device_delta()` has already filtered out incompatible devices.
+    """
+
+    validate_device_id(device, log_prefix="device_init")
+
+    if "cuda" not in device:
+        context.device = device
+        context.device_name = get_processor_name()
+        context.half_precision = False
+        log.debug(f"Render device available as {context.device_name}")
+        return
+
+    context.device_name = torch.cuda.get_device_name(device)
+    context.device = device
+
+    # Force full precision on 1660 and 1650 NVIDIA cards to avoid creating green images
+    if needs_to_force_full_precision(context):
+        log.warn(f"forcing full precision on this GPU, to avoid green images. GPU detected: {context.device_name}")
+        # Apply force_full_precision now before models are loaded.
+        context.half_precision = False
+
+    log.info(f'Setting {device} as active, with precision: {"half" if context.half_precision else "full"}')
+    torch.cuda.device(device)
+
+
+def needs_to_force_full_precision(context):
+    if "FORCE_FULL_PRECISION" in os.environ:
+        return True
+
+    device_name = context.device_name.lower()
+    return (
+        ("nvidia" in device_name or "geforce" in device_name or "quadro" in device_name)
+        and (
+            " 1660" in device_name
+            or " 1650" in device_name
+            or " t400" in device_name
+            or " t550" in device_name
+            or " t600" in device_name
+            or " t1000" in device_name
+            or " t1200" in device_name
+            or " t2000" in device_name
+        )
+    ) or ("tesla k40m" in device_name)
+
+
+def get_max_vram_usage_level(device):
+    if "cuda" in device:
+        _, mem_total = torch.cuda.mem_get_info(device)
+    else:
+        return "high"
+
+    mem_total /= float(10**9)
+    if mem_total < 4.5:
+        return "low"
+    elif mem_total < 6.5:
+        return "balanced"
+
+    return "high"
+
+
+def validate_device_id(device, log_prefix=""):
+    def is_valid():
+        if not isinstance(device, str):
+            return False
+        if device == "cpu" or device == "mps":
+            return True
+        if not device.startswith("cuda:") or not device[5:].isnumeric():
+            return False
+        return True
+
+    if not is_valid():
+        raise EnvironmentError(
+            f"{log_prefix}: device id should be 'cpu', 'mps', or 'cuda:N' (where N is an integer index for the GPU). Got: {device}"
+        )
+
+
+def is_device_compatible(device):
+    """
+    Returns True/False, and prints any compatibility errors
+    """
+    # static variable "history".
+    is_device_compatible.history = getattr(is_device_compatible, "history", {})
+    try:
+        validate_device_id(device, log_prefix="is_device_compatible")
+    except:
+        log.error(str(e))
+        return False
+
+    if device in ("cpu", "mps"):
+        return True
+    # Memory check
+    try:
+        _, mem_total = torch.cuda.mem_get_info(device)
+        mem_total /= float(10**9)
+        if mem_total < 3.0:
+            if is_device_compatible.history.get(device) == None:
+                log.warn(f"GPU {device} with less than 3 GB of VRAM is not compatible with Stable Diffusion")
+                is_device_compatible.history[device] = 1
+            return False
+    except RuntimeError as e:
+        log.error(str(e))
+        return False
+    return True
+
+
+def get_processor_name():
+    try:
+        import subprocess
+
+        if platform.system() == "Windows":
+            return platform.processor()
+        elif platform.system() == "Darwin":
+            os.environ["PATH"] = os.environ["PATH"] + os.pathsep + "/usr/sbin"
+            command = "sysctl -n machdep.cpu.brand_string"
+            return subprocess.check_output(command, shell=True).decode().strip()
+        elif platform.system() == "Linux":
+            command = "cat /proc/cpuinfo"
+            all_info = subprocess.check_output(command, shell=True).decode().strip()
+            for line in all_info.split("\n"):
+                if "model name" in line:
+                    return re.sub(".*model name.*:", "", line, 1).strip()
+    except:
+        log.error(traceback.format_exc())
+        return "cpu"

ui/easydiffusion/model_manager.py

@@ -0,0 +1,255 @@
+import os
+
+from easydiffusion import app
+from easydiffusion.types import TaskData
+from easydiffusion.utils import log
+
+from sdkit import Context
+from sdkit.models import load_model, unload_model, scan_model
+
+KNOWN_MODEL_TYPES = ["stable-diffusion", "vae", "hypernetwork", "gfpgan", "realesrgan"]
+MODEL_EXTENSIONS = {
+    "stable-diffusion": [".ckpt", ".safetensors"],
+    "vae": [".vae.pt", ".ckpt", ".safetensors"],
+    "hypernetwork": [".pt", ".safetensors"],
+    "gfpgan": [".pth"],
+    "realesrgan": [".pth"],
+}
+DEFAULT_MODELS = {
+    "stable-diffusion": [  # needed to support the legacy installations
+        "custom-model",  # only one custom model file was supported initially, creatively named 'custom-model'
+        "sd-v1-4",  # Default fallback.
+    ],
+    "gfpgan": ["GFPGANv1.3"],
+    "realesrgan": ["RealESRGAN_x4plus"],
+}
+MODELS_TO_LOAD_ON_START = ["stable-diffusion", "vae", "hypernetwork"]
+
+known_models = {}
+
+
+def init():
+    make_model_folders()
+    getModels()  # run this once, to cache the picklescan results
+
+
+def load_default_models(context: Context):
+    set_vram_optimizations(context)
+
+    # init default model paths
+    for model_type in MODELS_TO_LOAD_ON_START:
+        context.model_paths[model_type] = resolve_model_to_use(model_type=model_type)
+        try:
+            load_model(context, model_type)
+        except Exception as e:
+            log.error(f"[red]Error while loading {model_type} model: {context.model_paths[model_type]}[/red]")
+            log.error(f"[red]Error: {e}[/red]")
+            log.error(f"[red]Consider removing the model from the model folder.[red]")
+
+
+def unload_all(context: Context):
+    for model_type in KNOWN_MODEL_TYPES:
+        unload_model(context, model_type)
+
+
+def resolve_model_to_use(model_name: str = None, model_type: str = None):
+    model_extensions = MODEL_EXTENSIONS.get(model_type, [])
+    default_models = DEFAULT_MODELS.get(model_type, [])
+    config = app.getConfig()
+
+    model_dirs = [os.path.join(app.MODELS_DIR, model_type), app.SD_DIR]
+    if not model_name:  # When None try user configured model.
+        # config = getConfig()
+        if "model" in config and model_type in config["model"]:
+            model_name = config["model"][model_type]
+
+    if model_name:
+        # Check models directory
+        models_dir_path = os.path.join(app.MODELS_DIR, model_type, model_name)
+        for model_extension in model_extensions:
+            if os.path.exists(models_dir_path + model_extension):
+                return models_dir_path + model_extension
+            if os.path.exists(model_name + model_extension):
+                return os.path.abspath(model_name + model_extension)
+
+    # Default locations
+    if model_name in default_models:
+        default_model_path = os.path.join(app.SD_DIR, model_name)
+        for model_extension in model_extensions:
+            if os.path.exists(default_model_path + model_extension):
+                return default_model_path + model_extension
+
+    # Can't find requested model, check the default paths.
+    for default_model in default_models:
+        for model_dir in model_dirs:
+            default_model_path = os.path.join(model_dir, default_model)
+            for model_extension in model_extensions:
+                if os.path.exists(default_model_path + model_extension):
+                    if model_name is not None:
+                        log.warn(
+                            f"Could not find the configured custom model {model_name}{model_extension}. Using the default one: {default_model_path}{model_extension}"
+                        )
+                    return default_model_path + model_extension
+
+    return None
+
+
+def reload_models_if_necessary(context: Context, task_data: TaskData):
+    model_paths_in_req = {
+        "stable-diffusion": task_data.use_stable_diffusion_model,
+        "vae": task_data.use_vae_model,
+        "hypernetwork": task_data.use_hypernetwork_model,
+        "gfpgan": task_data.use_face_correction,
+        "realesrgan": task_data.use_upscale,
+        "nsfw_checker": True if task_data.block_nsfw else None,
+    }
+    models_to_reload = {
+        model_type: path
+        for model_type, path in model_paths_in_req.items()
+        if context.model_paths.get(model_type) != path
+    }
+
+    if set_vram_optimizations(context):  # reload SD
+        models_to_reload["stable-diffusion"] = model_paths_in_req["stable-diffusion"]
+
+    for model_type, model_path_in_req in models_to_reload.items():
+        context.model_paths[model_type] = model_path_in_req
+
+        action_fn = unload_model if context.model_paths[model_type] is None else load_model
+        action_fn(context, model_type, scan_model=False)  # we've scanned them already
+
+
+def resolve_model_paths(task_data: TaskData):
+    task_data.use_stable_diffusion_model = resolve_model_to_use(
+        task_data.use_stable_diffusion_model, model_type="stable-diffusion"
+    )
+    task_data.use_vae_model = resolve_model_to_use(task_data.use_vae_model, model_type="vae")
+    task_data.use_hypernetwork_model = resolve_model_to_use(task_data.use_hypernetwork_model, model_type="hypernetwork")
+
+    if task_data.use_face_correction:
+        task_data.use_face_correction = resolve_model_to_use(task_data.use_face_correction, "gfpgan")
+    if task_data.use_upscale:
+        task_data.use_upscale = resolve_model_to_use(task_data.use_upscale, "realesrgan")
+
+
+def set_vram_optimizations(context: Context):
+    config = app.getConfig()
+    vram_usage_level = config.get("vram_usage_level", "balanced")
+
+    if vram_usage_level != context.vram_usage_level:
+        context.vram_usage_level = vram_usage_level
+        return True
+
+    return False
+
+
+def make_model_folders():
+    for model_type in KNOWN_MODEL_TYPES:
+        model_dir_path = os.path.join(app.MODELS_DIR, model_type)
+
+        os.makedirs(model_dir_path, exist_ok=True)
+
+        help_file_name = f"Place your {model_type} model files here.txt"
+        help_file_contents = f'Supported extensions: {" or ".join(MODEL_EXTENSIONS.get(model_type))}'
+
+        with open(os.path.join(model_dir_path, help_file_name), "w", encoding="utf-8") as f:
+            f.write(help_file_contents)
+
+
+def is_malicious_model(file_path):
+    try:
+        if file_path.endswith(".safetensors"):
+            return False
+        scan_result = scan_model(file_path)
+        if scan_result.issues_count > 0 or scan_result.infected_files > 0:
+            log.warn(
+                ":warning: [bold red]Scan %s: %d scanned, %d issue, %d infected.[/bold red]"
+                % (file_path, scan_result.scanned_files, scan_result.issues_count, scan_result.infected_files)
+            )
+            return True
+        else:
+            log.debug(
+                "Scan %s: [green]%d scanned, %d issue, %d infected.[/green]"
+                % (file_path, scan_result.scanned_files, scan_result.issues_count, scan_result.infected_files)
+            )
+            return False
+    except Exception as e:
+        log.error(f"error while scanning: {file_path}, error: {e}")
+    return False
+
+
+def getModels():
+    models = {
+        "active": {
+            "stable-diffusion": "sd-v1-4",
+            "vae": "",
+            "hypernetwork": "",
+        },
+        "options": {
+            "stable-diffusion": ["sd-v1-4"],
+            "vae": [],
+            "hypernetwork": [],
+        },
+    }
+
+    models_scanned = 0
+
+    class MaliciousModelException(Exception):
+        "Raised when picklescan reports a problem with a model"
+        pass
+
+    def scan_directory(directory, suffixes, directoriesFirst: bool = True):
+        nonlocal models_scanned
+        tree = []
+        for entry in sorted(
+            os.scandir(directory), key=lambda entry: (entry.is_file() == directoriesFirst, entry.name.lower())
+        ):
+            if entry.is_file():
+                matching_suffix = list(filter(lambda s: entry.name.endswith(s), suffixes))
+                if len(matching_suffix) == 0:
+                    continue
+                matching_suffix = matching_suffix[0]
+
+                mtime = entry.stat().st_mtime
+                mod_time = known_models[entry.path] if entry.path in known_models else -1
+                if mod_time != mtime:
+                    models_scanned += 1
+                    if is_malicious_model(entry.path):
+                        raise MaliciousModelException(entry.path)
+                known_models[entry.path] = mtime
+                tree.append(entry.name[: -len(matching_suffix)])
+            elif entry.is_dir():
+                scan = scan_directory(entry.path, suffixes, directoriesFirst=False)
+
+                if len(scan) != 0:
+                    tree.append((entry.name, scan))
+        return tree
+
+    def listModels(model_type):
+        nonlocal models_scanned
+
+        model_extensions = MODEL_EXTENSIONS.get(model_type, [])
+        models_dir = os.path.join(app.MODELS_DIR, model_type)
+        if not os.path.exists(models_dir):
+            os.makedirs(models_dir)
+
+        try:
+            models["options"][model_type] = scan_directory(models_dir, model_extensions)
+        except MaliciousModelException as e:
+            models["scan-error"] = e
+
+    # custom models
+    listModels(model_type="stable-diffusion")
+    listModels(model_type="vae")
+    listModels(model_type="hypernetwork")
+    listModels(model_type="gfpgan")
+
+    if models_scanned > 0:
+        log.info(f"[green]Scanned {models_scanned} models. Nothing infected[/]")
+
+    # legacy
+    custom_weight_path = os.path.join(app.SD_DIR, "custom-model.ckpt")
+    if os.path.exists(custom_weight_path):
+        models["options"]["stable-diffusion"].append("custom-model")
+
+    return models

ui/easydiffusion/renderer.py

@@ -0,0 +1,177 @@
+import queue
+import time
+import json
+import pprint
+
+from easydiffusion import device_manager
+from easydiffusion.types import TaskData, Response, Image as ResponseImage, UserInitiatedStop, GenerateImageRequest
+from easydiffusion.utils import get_printable_request, save_images_to_disk, log
+
+from sdkit import Context
+from sdkit.generate import generate_images
+from sdkit.filter import apply_filters
+from sdkit.utils import img_to_buffer, img_to_base64_str, latent_samples_to_images, gc
+
+context = Context()  # thread-local
+"""
+runtime data (bound locally to this thread), for e.g. device, references to loaded models, optimization flags etc
+"""
+
+
+def init(device):
+    """
+    Initializes the fields that will be bound to this runtime's context, and sets the current torch device
+    """
+    context.stop_processing = False
+    context.temp_images = {}
+    context.partial_x_samples = None
+
+    device_manager.device_init(context, device)
+
+
+def make_images(
+    req: GenerateImageRequest, task_data: TaskData, data_queue: queue.Queue, task_temp_images: list, step_callback
+):
+    context.stop_processing = False
+    print_task_info(req, task_data)
+
+    images, seeds = make_images_internal(req, task_data, data_queue, task_temp_images, step_callback)
+
+    res = Response(req, task_data, images=construct_response(images, seeds, task_data, base_seed=req.seed))
+    res = res.json()
+    data_queue.put(json.dumps(res))
+    log.info("Task completed")
+
+    return res
+
+
+def print_task_info(req: GenerateImageRequest, task_data: TaskData):
+    req_str = pprint.pformat(get_printable_request(req)).replace("[", "\[")
+    task_str = pprint.pformat(task_data.dict()).replace("[", "\[")
+    log.info(f"request: {req_str}")
+    log.info(f"task data: {task_str}")
+
+
+def make_images_internal(
+    req: GenerateImageRequest, task_data: TaskData, data_queue: queue.Queue, task_temp_images: list, step_callback
+):
+
+    images, user_stopped = generate_images_internal(
+        req, task_data, data_queue, task_temp_images, step_callback, task_data.stream_image_progress, task_data.stream_image_progress_interval
+    )
+    filtered_images = filter_images(task_data, images, user_stopped)
+
+    if task_data.save_to_disk_path is not None:
+        save_images_to_disk(images, filtered_images, req, task_data)
+
+    seeds = [*range(req.seed, req.seed + len(images))]
+    if task_data.show_only_filtered_image or filtered_images is images:
+        return filtered_images, seeds
+    else:
+        return images + filtered_images, seeds + seeds
+
+
+def generate_images_internal(
+    req: GenerateImageRequest,
+    task_data: TaskData,
+    data_queue: queue.Queue,
+    task_temp_images: list,
+    step_callback,
+    stream_image_progress: bool,
+    stream_image_progress_interval: int,
+):
+    context.temp_images.clear()
+
+    callback = make_step_callback(req, task_data, data_queue, task_temp_images, step_callback, stream_image_progress, stream_image_progress_interval)
+
+    try:
+        if req.init_image is not None:
+            req.sampler_name = "ddim"
+
+        images = generate_images(context, callback=callback, **req.dict())
+        user_stopped = False
+    except UserInitiatedStop:
+        images = []
+        user_stopped = True
+        if context.partial_x_samples is not None:
+            images = latent_samples_to_images(context, context.partial_x_samples)
+    finally:
+        if hasattr(context, "partial_x_samples") and context.partial_x_samples is not None:
+            del context.partial_x_samples
+            context.partial_x_samples = None
+
+    return images, user_stopped
+
+
+def filter_images(task_data: TaskData, images: list, user_stopped):
+    if user_stopped:
+        return images
+
+    filters_to_apply = []
+    if task_data.block_nsfw:
+        filters_to_apply.append("nsfw_checker")
+    if task_data.use_face_correction and "gfpgan" in task_data.use_face_correction.lower():
+        filters_to_apply.append("gfpgan")
+    if task_data.use_upscale and "realesrgan" in task_data.use_upscale.lower():
+        filters_to_apply.append("realesrgan")
+
+    if len(filters_to_apply) == 0:
+        return images
+
+    return apply_filters(context, filters_to_apply, images, scale=task_data.upscale_amount)
+
+
+def construct_response(images: list, seeds: list, task_data: TaskData, base_seed: int):
+    return [
+        ResponseImage(
+            data=img_to_base64_str(img, task_data.output_format, task_data.output_quality),
+            seed=seed,
+        )
+        for img, seed in zip(images, seeds)
+    ]
+
+
+def make_step_callback(
+    req: GenerateImageRequest,
+    task_data: TaskData,
+    data_queue: queue.Queue,
+    task_temp_images: list,
+    step_callback,
+    stream_image_progress: bool,
+    stream_image_progress_interval: int,
+):
+    n_steps = req.num_inference_steps if req.init_image is None else int(req.num_inference_steps * req.prompt_strength)
+    last_callback_time = -1
+
+    def update_temp_img(x_samples, task_temp_images: list):
+        partial_images = []
+        images = latent_samples_to_images(context, x_samples)
+        for i, img in enumerate(images):
+            buf = img_to_buffer(img, output_format="JPEG")
+
+            context.temp_images[f"{task_data.request_id}/{i}"] = buf
+            task_temp_images[i] = buf
+            partial_images.append({"path": f"/image/tmp/{task_data.request_id}/{i}"})
+        del images
+        return partial_images
+
+    def on_image_step(x_samples, i):
+        nonlocal last_callback_time
+
+        context.partial_x_samples = x_samples
+        step_time = time.time() - last_callback_time if last_callback_time != -1 else -1
+        last_callback_time = time.time()
+
+        progress = {"step": i, "step_time": step_time, "total_steps": n_steps}
+
+        if stream_image_progress and stream_image_progress_interval > 0 and i % stream_image_progress_interval == 0:
+            progress["output"] = update_temp_img(x_samples, task_temp_images)
+
+        data_queue.put(json.dumps(progress))
+
+        step_callback()
+
+        if context.stop_processing:
+            raise UserInitiatedStop("User requested that we stop processing")
+
+    return on_image_step

ui/easydiffusion/server.py

@@ -0,0 +1,299 @@
+"""server.py: FastAPI SD-UI Web Host.
+Notes:
+    async endpoints always run on the main thread. Without they run on the thread pool.
+"""
+import os
+import traceback
+import datetime
+from typing import List, Union
+
+from fastapi import FastAPI, HTTPException
+from fastapi.staticfiles import StaticFiles
+from starlette.responses import FileResponse, JSONResponse, StreamingResponse
+from pydantic import BaseModel
+
+from easydiffusion import app, model_manager, task_manager
+from easydiffusion.types import TaskData, GenerateImageRequest, MergeRequest
+from easydiffusion.utils import log
+
+log.info(f"started in {app.SD_DIR}")
+log.info(f"started at {datetime.datetime.now():%x %X}")
+
+server_api = FastAPI()
+
+NOCACHE_HEADERS = {"Cache-Control": "no-cache, no-store, must-revalidate", "Pragma": "no-cache", "Expires": "0"}
+
+
+class NoCacheStaticFiles(StaticFiles):
+    def __init__(self, directory: str):
+        # follow_symlink is only available on fastapi >= 0.92.0
+        if (os.path.islink(directory)):
+            super().__init__(directory = os.path.realpath(directory))
+        else:
+            super().__init__(directory = directory)
+
+    def is_not_modified(self, response_headers, request_headers) -> bool:
+        if "content-type" in response_headers and (
+            "javascript" in response_headers["content-type"] or "css" in response_headers["content-type"]
+        ):
+            response_headers.update(NOCACHE_HEADERS)
+            return False
+
+        return super().is_not_modified(response_headers, request_headers)
+
+
+class SetAppConfigRequest(BaseModel):
+    update_branch: str = None
+    render_devices: Union[List[str], List[int], str, int] = None
+    model_vae: str = None
+    ui_open_browser_on_start: bool = None
+    listen_to_network: bool = None
+    listen_port: int = None
+
+
+def init():
+    if os.path.isdir(app.CUSTOM_MODIFIERS_DIR):
+        server_api.mount(
+            "/media/modifier-thumbnails/custom",
+            NoCacheStaticFiles(directory=app.CUSTOM_MODIFIERS_DIR),
+            name="custom-thumbnails",
+        )
+
+    server_api.mount("/media", NoCacheStaticFiles(directory=os.path.join(app.SD_UI_DIR, "media")), name="media")
+
+    for plugins_dir, dir_prefix in app.UI_PLUGINS_SOURCES:
+        server_api.mount(
+            f"/plugins/{dir_prefix}", NoCacheStaticFiles(directory=plugins_dir), name=f"plugins-{dir_prefix}"
+        )
+
+    @server_api.post("/app_config")
+    async def set_app_config(req: SetAppConfigRequest):
+        return set_app_config_internal(req)
+
+    @server_api.get("/get/{key:path}")
+    def read_web_data(key: str = None):
+        return read_web_data_internal(key)
+
+    @server_api.get("/ping")  # Get server and optionally session status.
+    def ping(session_id: str = None):
+        return ping_internal(session_id)
+
+    @server_api.post("/render")
+    def render(req: dict):
+        return render_internal(req)
+
+    @server_api.post("/model/merge")
+    def model_merge(req: dict):
+        print(req)
+        return model_merge_internal(req)
+
+    @server_api.get("/image/stream/{task_id:int}")
+    def stream(task_id: int):
+        return stream_internal(task_id)
+
+    @server_api.get("/image/stop")
+    def stop(task: int):
+        return stop_internal(task)
+
+    @server_api.get("/image/tmp/{task_id:int}/{img_id:int}")
+    def get_image(task_id: int, img_id: int):
+        return get_image_internal(task_id, img_id)
+
+    @server_api.get("/")
+    def read_root():
+        return FileResponse(os.path.join(app.SD_UI_DIR, "index.html"), headers=NOCACHE_HEADERS)
+
+    @server_api.on_event("shutdown")
+    def shutdown_event():  # Signal render thread to close on shutdown
+        task_manager.current_state_error = SystemExit("Application shutting down.")
+
+
+# API implementations
+def set_app_config_internal(req: SetAppConfigRequest):
+    config = app.getConfig()
+    if req.update_branch is not None:
+        config["update_branch"] = req.update_branch
+    if req.render_devices is not None:
+        update_render_devices_in_config(config, req.render_devices)
+    if req.ui_open_browser_on_start is not None:
+        if "ui" not in config:
+            config["ui"] = {}
+        config["ui"]["open_browser_on_start"] = req.ui_open_browser_on_start
+    if req.listen_to_network is not None:
+        if "net" not in config:
+            config["net"] = {}
+        config["net"]["listen_to_network"] = bool(req.listen_to_network)
+    if req.listen_port is not None:
+        if "net" not in config:
+            config["net"] = {}
+        config["net"]["listen_port"] = int(req.listen_port)
+    try:
+        app.setConfig(config)
+
+        if req.render_devices:
+            app.update_render_threads()
+
+        return JSONResponse({"status": "OK"}, headers=NOCACHE_HEADERS)
+    except Exception as e:
+        log.error(traceback.format_exc())
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+def update_render_devices_in_config(config, render_devices):
+    if render_devices not in ("cpu", "auto") and not render_devices.startswith("cuda:"):
+        raise HTTPException(status_code=400, detail=f"Invalid render device requested: {render_devices}")
+
+    if render_devices.startswith("cuda:"):
+        render_devices = render_devices.split(",")
+
+    config["render_devices"] = render_devices
+
+
+def read_web_data_internal(key: str = None):
+    if not key:  # /get without parameters, stable-diffusion easter egg.
+        raise HTTPException(status_code=418, detail="StableDiffusion is drawing a teapot!")  # HTTP418 I'm a teapot
+    elif key == "app_config":
+        return JSONResponse(app.getConfig(), headers=NOCACHE_HEADERS)
+    elif key == "system_info":
+        config = app.getConfig()
+
+        output_dir = config.get("force_save_path", os.path.join(os.path.expanduser("~"), app.OUTPUT_DIRNAME))
+
+        system_info = {
+            "devices": task_manager.get_devices(),
+            "hosts": app.getIPConfig(),
+            "default_output_dir": output_dir,
+            "enforce_output_dir": ("force_save_path" in config),
+        }
+        system_info["devices"]["config"] = config.get("render_devices", "auto")
+        return JSONResponse(system_info, headers=NOCACHE_HEADERS)
+    elif key == "models":
+        return JSONResponse(model_manager.getModels(), headers=NOCACHE_HEADERS)
+    elif key == "modifiers":
+        return JSONResponse(app.get_image_modifiers(), headers=NOCACHE_HEADERS)
+    elif key == "ui_plugins":
+        return JSONResponse(app.getUIPlugins(), headers=NOCACHE_HEADERS)
+    else:
+        raise HTTPException(status_code=404, detail=f"Request for unknown {key}")  # HTTP404 Not Found
+
+
+def ping_internal(session_id: str = None):
+    if task_manager.is_alive() <= 0:  # Check that render threads are alive.
+        if task_manager.current_state_error:
+            raise HTTPException(status_code=500, detail=str(task_manager.current_state_error))
+        raise HTTPException(status_code=500, detail="Render thread is dead.")
+    if task_manager.current_state_error and not isinstance(task_manager.current_state_error, StopAsyncIteration):
+        raise HTTPException(status_code=500, detail=str(task_manager.current_state_error))
+    # Alive
+    response = {"status": str(task_manager.current_state)}
+    if session_id:
+        session = task_manager.get_cached_session(session_id, update_ttl=True)
+        response["tasks"] = {id(t): t.status for t in session.tasks}
+    response["devices"] = task_manager.get_devices()
+    return JSONResponse(response, headers=NOCACHE_HEADERS)
+
+
+def render_internal(req: dict):
+    try:
+        # separate out the request data into rendering and task-specific data
+        render_req: GenerateImageRequest = GenerateImageRequest.parse_obj(req)
+        task_data: TaskData = TaskData.parse_obj(req)
+
+        # Overwrite user specified save path
+        config = app.getConfig()
+        if "force_save_path" in config:
+            task_data.save_to_disk_path = config["force_save_path"]
+
+        render_req.init_image_mask = req.get("mask")  # hack: will rename this in the HTTP API in a future revision
+
+        app.save_to_config(
+            task_data.use_stable_diffusion_model,
+            task_data.use_vae_model,
+            task_data.use_hypernetwork_model,
+            task_data.vram_usage_level,
+        )
+
+        # enqueue the task
+        new_task = task_manager.render(render_req, task_data)
+        response = {
+            "status": str(task_manager.current_state),
+            "queue": len(task_manager.tasks_queue),
+            "stream": f"/image/stream/{id(new_task)}",
+            "task": id(new_task),
+        }
+        return JSONResponse(response, headers=NOCACHE_HEADERS)
+    except ChildProcessError as e:  # Render thread is dead
+        raise HTTPException(status_code=500, detail=f"Rendering thread has died.")  # HTTP500 Internal Server Error
+    except ConnectionRefusedError as e:  # Unstarted task pending limit reached, deny queueing too many.
+        raise HTTPException(status_code=503, detail=str(e))  # HTTP503 Service Unavailable
+    except Exception as e:
+        log.error(traceback.format_exc())
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+def model_merge_internal(req: dict):
+    try:
+        from sdkit.train import merge_models
+        from easydiffusion.utils.save_utils import filename_regex
+
+        mergeReq: MergeRequest = MergeRequest.parse_obj(req)
+
+        merge_models(
+            model_manager.resolve_model_to_use(mergeReq.model0, "stable-diffusion"),
+            model_manager.resolve_model_to_use(mergeReq.model1, "stable-diffusion"),
+            mergeReq.ratio,
+            os.path.join(app.MODELS_DIR, "stable-diffusion", filename_regex.sub("_", mergeReq.out_path)),
+            mergeReq.use_fp16,
+        )
+        return JSONResponse({"status": "OK"}, headers=NOCACHE_HEADERS)
+    except Exception as e:
+        log.error(traceback.format_exc())
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+def stream_internal(task_id: int):
+    # TODO Move to WebSockets ??
+    task = task_manager.get_cached_task(task_id, update_ttl=True)
+    if not task:
+        raise HTTPException(status_code=404, detail=f"Request {task_id} not found.")  # HTTP404 NotFound
+    # if (id(task) != task_id): raise HTTPException(status_code=409, detail=f'Wrong task id received. Expected:{id(task)}, Received:{task_id}') # HTTP409 Conflict
+    if task.buffer_queue.empty() and not task.lock.locked():
+        if task.response:
+            # log.info(f'Session {session_id} sending cached response')
+            return JSONResponse(task.response, headers=NOCACHE_HEADERS)
+        raise HTTPException(status_code=425, detail="Too Early, task not started yet.")  # HTTP425 Too Early
+    # log.info(f'Session {session_id} opened live render stream {id(task.buffer_queue)}')
+    return StreamingResponse(task.read_buffer_generator(), media_type="application/json")
+
+
+def stop_internal(task: int):
+    if not task:
+        if (
+            task_manager.current_state == task_manager.ServerStates.Online
+            or task_manager.current_state == task_manager.ServerStates.Unavailable
+        ):
+            raise HTTPException(status_code=409, detail="Not currently running any tasks.")  # HTTP409 Conflict
+        task_manager.current_state_error = StopAsyncIteration("")
+        return {"OK"}
+    task_id = task
+    task = task_manager.get_cached_task(task_id, update_ttl=False)
+    if not task:
+        raise HTTPException(status_code=404, detail=f"Task {task_id} was not found.")  # HTTP404 Not Found
+    if isinstance(task.error, StopAsyncIteration):
+        raise HTTPException(status_code=409, detail=f"Task {task_id} is already stopped.")  # HTTP409 Conflict
+    task.error = StopAsyncIteration(f"Task {task_id} stop requested.")
+    return {"OK"}
+
+
+def get_image_internal(task_id: int, img_id: int):
+    task = task_manager.get_cached_task(task_id, update_ttl=True)
+    if not task:
+        raise HTTPException(status_code=410, detail=f"Task {task_id} could not be found.")  # HTTP404 NotFound
+    if not task.temp_images[img_id]:
+        raise HTTPException(status_code=425, detail="Too Early, task data is not available yet.")  # HTTP425 Too Early
+    try:
+        img_data = task.temp_images[img_id]
+        img_data.seek(0)
+        return StreamingResponse(img_data, media_type="image/jpeg")
+    except KeyError as e:
+        raise HTTPException(status_code=500, detail=str(e))

ui/easydiffusion/task_manager.py

@@ -0,0 +1,565 @@
+"""task_manager.py: manage tasks dispatching and render threads.
+Notes:
+    render_threads should be the only hard reference held by the manager to the threads.
+    Use weak_thread_data to store all other data using weak keys.
+    This will allow for garbage collection after the thread dies.
+"""
+import json
+import traceback
+
+TASK_TTL = 15 * 60  # seconds, Discard last session's task timeout
+
+import torch
+import queue, threading, time, weakref
+from typing import Any, Hashable
+
+from easydiffusion import device_manager
+from easydiffusion.types import TaskData, GenerateImageRequest
+from easydiffusion.utils import log
+
+from sdkit.utils import gc
+
+THREAD_NAME_PREFIX = ""
+ERR_LOCK_FAILED = " failed to acquire lock within timeout."
+LOCK_TIMEOUT = 15  # Maximum locking time in seconds before failing a task.
+# It's better to get an exception than a deadlock... ALWAYS use timeout in critical paths.
+
+DEVICE_START_TIMEOUT = 60  # seconds - Maximum time to wait for a render device to init.
+
+
+class SymbolClass(type):  # Print nicely formatted Symbol names.
+    def __repr__(self):
+        return self.__qualname__
+
+    def __str__(self):
+        return self.__name__
+
+
+class Symbol(metaclass=SymbolClass):
+    pass
+
+
+class ServerStates:
+    class Init(Symbol):
+        pass
+
+    class LoadingModel(Symbol):
+        pass
+
+    class Online(Symbol):
+        pass
+
+    class Rendering(Symbol):
+        pass
+
+    class Unavailable(Symbol):
+        pass
+
+
+class RenderTask:  # Task with output queue and completion lock.
+    def __init__(self, req: GenerateImageRequest, task_data: TaskData):
+        task_data.request_id = id(self)
+        self.render_request: GenerateImageRequest = req  # Initial Request
+        self.task_data: TaskData = task_data
+        self.response: Any = None  # Copy of the last reponse
+        self.render_device = None  # Select the task affinity. (Not used to change active devices).
+        self.temp_images: list = [None] * req.num_outputs * (1 if task_data.show_only_filtered_image else 2)
+        self.error: Exception = None
+        self.lock: threading.Lock = threading.Lock()  # Locks at task start and unlocks when task is completed
+        self.buffer_queue: queue.Queue = queue.Queue()  # Queue of JSON string segments
+
+    async def read_buffer_generator(self):
+        try:
+            while not self.buffer_queue.empty():
+                res = self.buffer_queue.get(block=False)
+                self.buffer_queue.task_done()
+                yield res
+        except queue.Empty as e:
+            yield
+
+    @property
+    def status(self):
+        if self.lock.locked():
+            return "running"
+        if isinstance(self.error, StopAsyncIteration):
+            return "stopped"
+        if self.error:
+            return "error"
+        if not self.buffer_queue.empty():
+            return "buffer"
+        if self.response:
+            return "completed"
+        return "pending"
+
+    @property
+    def is_pending(self):
+        return bool(not self.response and not self.error)
+
+
+# Temporary cache to allow to query tasks results for a short time after they are completed.
+class DataCache:
+    def __init__(self):
+        self._base = dict()
+        self._lock: threading.Lock = threading.Lock()
+
+    def _get_ttl_time(self, ttl: int) -> int:
+        return int(time.time()) + ttl
+
+    def _is_expired(self, timestamp: int) -> bool:
+        return int(time.time()) >= timestamp
+
+    def clean(self) -> None:
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+            raise Exception("DataCache.clean" + ERR_LOCK_FAILED)
+        try:
+            # Create a list of expired keys to delete
+            to_delete = []
+            for key in self._base:
+                ttl, _ = self._base[key]
+                if self._is_expired(ttl):
+                    to_delete.append(key)
+            # Remove Items
+            for key in to_delete:
+                (_, val) = self._base[key]
+                if isinstance(val, RenderTask):
+                    log.debug(f"RenderTask {key} expired. Data removed.")
+                elif isinstance(val, SessionState):
+                    log.debug(f"Session {key} expired. Data removed.")
+                else:
+                    log.debug(f"Key {key} expired. Data removed.")
+                del self._base[key]
+        finally:
+            self._lock.release()
+
+    def clear(self) -> None:
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+            raise Exception("DataCache.clear" + ERR_LOCK_FAILED)
+        try:
+            self._base.clear()
+        finally:
+            self._lock.release()
+
+    def delete(self, key: Hashable) -> bool:
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+            raise Exception("DataCache.delete" + ERR_LOCK_FAILED)
+        try:
+            if key not in self._base:
+                return False
+            del self._base[key]
+            return True
+        finally:
+            self._lock.release()
+
+    def keep(self, key: Hashable, ttl: int) -> bool:
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+            raise Exception("DataCache.keep" + ERR_LOCK_FAILED)
+        try:
+            if key in self._base:
+                _, value = self._base.get(key)
+                self._base[key] = (self._get_ttl_time(ttl), value)
+                return True
+            return False
+        finally:
+            self._lock.release()
+
+    def put(self, key: Hashable, value: Any, ttl: int) -> bool:
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+            raise Exception("DataCache.put" + ERR_LOCK_FAILED)
+        try:
+            self._base[key] = (self._get_ttl_time(ttl), value)
+        except Exception as e:
+            log.error(traceback.format_exc())
+            return False
+        else:
+            return True
+        finally:
+            self._lock.release()
+
+    def tryGet(self, key: Hashable) -> Any:
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+            raise Exception("DataCache.tryGet" + ERR_LOCK_FAILED)
+        try:
+            ttl, value = self._base.get(key, (None, None))
+            if ttl is not None and self._is_expired(ttl):
+                log.debug(f"Session {key} expired. Discarding data.")
+                del self._base[key]
+                return None
+            return value
+        finally:
+            self._lock.release()
+
+
+manager_lock = threading.RLock()
+render_threads = []
+current_state = ServerStates.Init
+current_state_error: Exception = None
+tasks_queue = []
+session_cache = DataCache()
+task_cache = DataCache()
+weak_thread_data = weakref.WeakKeyDictionary()
+idle_event: threading.Event = threading.Event()
+
+
+class SessionState:
+    def __init__(self, id: str):
+        self._id = id
+        self._tasks_ids = []
+
+    @property
+    def id(self):
+        return self._id
+
+    @property
+    def tasks(self):
+        tasks = []
+        for task_id in self._tasks_ids:
+            task = task_cache.tryGet(task_id)
+            if task:
+                tasks.append(task)
+        return tasks
+
+    def put(self, task, ttl=TASK_TTL):
+        task_id = id(task)
+        self._tasks_ids.append(task_id)
+        if not task_cache.put(task_id, task, ttl):
+            return False
+        while len(self._tasks_ids) > len(render_threads) * 2:
+            self._tasks_ids.pop(0)
+        return True
+
+
+def thread_get_next_task():
+    from easydiffusion import renderer
+
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+        log.warn(f"Render thread on device: {renderer.context.device} failed to acquire manager lock.")
+        return None
+    if len(tasks_queue) <= 0:
+        manager_lock.release()
+        return None
+    task = None
+    try:  # Select a render task.
+        for queued_task in tasks_queue:
+            if queued_task.render_device and renderer.context.device != queued_task.render_device:
+                # Is asking for a specific render device.
+                if is_alive(queued_task.render_device) > 0:
+                    continue  # requested device alive, skip current one.
+                else:
+                    # Requested device is not active, return error to UI.
+                    queued_task.error = Exception(queued_task.render_device + " is not currently active.")
+                    task = queued_task
+                    break
+            if not queued_task.render_device and renderer.context.device == "cpu" and is_alive() > 1:
+                # not asking for any specific devices, cpu want to grab task but other render devices are alive.
+                continue  # Skip Tasks, don't run on CPU unless there is nothing else or user asked for it.
+            task = queued_task
+            break
+        if task is not None:
+            del tasks_queue[tasks_queue.index(task)]
+        return task
+    finally:
+        manager_lock.release()
+
+
+def thread_render(device):
+    global current_state, current_state_error
+
+    from easydiffusion import renderer, model_manager
+
+    try:
+        renderer.init(device)
+
+        weak_thread_data[threading.current_thread()] = {
+            "device": renderer.context.device,
+            "device_name": renderer.context.device_name,
+            "alive": True,
+        }
+
+        current_state = ServerStates.LoadingModel
+        model_manager.load_default_models(renderer.context)
+
+        current_state = ServerStates.Online
+    except Exception as e:
+        log.error(traceback.format_exc())
+        weak_thread_data[threading.current_thread()] = {"error": e, "alive": False}
+        return
+
+    while True:
+        session_cache.clean()
+        task_cache.clean()
+        if not weak_thread_data[threading.current_thread()]["alive"]:
+            log.info(f"Shutting down thread for device {renderer.context.device}")
+            model_manager.unload_all(renderer.context)
+            return
+        if isinstance(current_state_error, SystemExit):
+            current_state = ServerStates.Unavailable
+            return
+        task = thread_get_next_task()
+        if task is None:
+            idle_event.clear()
+            idle_event.wait(timeout=1)
+            continue
+        if task.error is not None:
+            log.error(task.error)
+            task.response = {"status": "failed", "detail": str(task.error)}
+            task.buffer_queue.put(json.dumps(task.response))
+            continue
+        if current_state_error:
+            task.error = current_state_error
+            task.response = {"status": "failed", "detail": str(task.error)}
+            task.buffer_queue.put(json.dumps(task.response))
+            continue
+        log.info(f"Session {task.task_data.session_id} starting task {id(task)} on {renderer.context.device_name}")
+        if not task.lock.acquire(blocking=False):
+            raise Exception("Got locked task from queue.")
+        try:
+
+            def step_callback():
+                global current_state_error
+
+                if (
+                    isinstance(current_state_error, SystemExit)
+                    or isinstance(current_state_error, StopAsyncIteration)
+                    or isinstance(task.error, StopAsyncIteration)
+                ):
+                    renderer.context.stop_processing = True
+                    if isinstance(current_state_error, StopAsyncIteration):
+                        task.error = current_state_error
+                        current_state_error = None
+                        log.info(f"Session {task.task_data.session_id} sent cancel signal for task {id(task)}")
+
+            current_state = ServerStates.LoadingModel
+            model_manager.resolve_model_paths(task.task_data)
+            model_manager.reload_models_if_necessary(renderer.context, task.task_data)
+
+            current_state = ServerStates.Rendering
+            task.response = renderer.make_images(
+                task.render_request, task.task_data, task.buffer_queue, task.temp_images, step_callback
+            )
+            # Before looping back to the generator, mark cache as still alive.
+            task_cache.keep(id(task), TASK_TTL)
+            session_cache.keep(task.task_data.session_id, TASK_TTL)
+        except Exception as e:
+            task.error = str(e)
+            task.response = {"status": "failed", "detail": str(task.error)}
+            task.buffer_queue.put(json.dumps(task.response))
+            log.error(traceback.format_exc())
+        finally:
+            gc(renderer.context)
+            task.lock.release()
+        task_cache.keep(id(task), TASK_TTL)
+        session_cache.keep(task.task_data.session_id, TASK_TTL)
+        if isinstance(task.error, StopAsyncIteration):
+            log.info(f"Session {task.task_data.session_id} task {id(task)} cancelled!")
+        elif task.error is not None:
+            log.info(f"Session {task.task_data.session_id} task {id(task)} failed!")
+        else:
+            log.info(
+                f"Session {task.task_data.session_id} task {id(task)} completed by {renderer.context.device_name}."
+            )
+        current_state = ServerStates.Online
+
+
+def get_cached_task(task_id: str, update_ttl: bool = False):
+    # By calling keep before tryGet, wont discard if was expired.
+    if update_ttl and not task_cache.keep(task_id, TASK_TTL):
+        # Failed to keep task, already gone.
+        return None
+    return task_cache.tryGet(task_id)
+
+
+def get_cached_session(session_id: str, update_ttl: bool = False):
+    if update_ttl:
+        session_cache.keep(session_id, TASK_TTL)
+    session = session_cache.tryGet(session_id)
+    if not session:
+        session = SessionState(session_id)
+        session_cache.put(session_id, session, TASK_TTL)
+    return session
+
+
+def get_devices():
+    devices = {
+        "all": {},
+        "active": {},
+    }
+
+    def get_device_info(device):
+        if device in ("cpu", "mps"):
+            return {"name": device_manager.get_processor_name()}
+
+        mem_free, mem_total = torch.cuda.mem_get_info(device)
+        mem_free /= float(10**9)
+        mem_total /= float(10**9)
+
+        return {
+            "name": torch.cuda.get_device_name(device),
+            "mem_free": mem_free,
+            "mem_total": mem_total,
+            "max_vram_usage_level": device_manager.get_max_vram_usage_level(device),
+        }
+
+    # list the compatible devices
+    cuda_count = torch.cuda.device_count()
+    for device in range(cuda_count):
+        device = f"cuda:{device}"
+        if not device_manager.is_device_compatible(device):
+            continue
+
+        devices["all"].update({device: get_device_info(device)})
+
+    if device_manager.is_mps_available():
+        devices["all"].update({"mps": get_device_info("mps")})
+
+    devices["all"].update({"cpu": get_device_info("cpu")})
+
+    # list the activated devices
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+        raise Exception("get_devices" + ERR_LOCK_FAILED)
+    try:
+        for rthread in render_threads:
+            if not rthread.is_alive():
+                continue
+            weak_data = weak_thread_data.get(rthread)
+            if not weak_data or not "device" in weak_data or not "device_name" in weak_data:
+                continue
+            device = weak_data["device"]
+            devices["active"].update({device: get_device_info(device)})
+    finally:
+        manager_lock.release()
+
+    return devices
+
+
+def is_alive(device=None):
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+        raise Exception("is_alive" + ERR_LOCK_FAILED)
+    nbr_alive = 0
+    try:
+        for rthread in render_threads:
+            if device is not None:
+                weak_data = weak_thread_data.get(rthread)
+                if weak_data is None or not "device" in weak_data or weak_data["device"] is None:
+                    continue
+                thread_device = weak_data["device"]
+                if thread_device != device:
+                    continue
+            if rthread.is_alive():
+                nbr_alive += 1
+        return nbr_alive
+    finally:
+        manager_lock.release()
+
+
+def start_render_thread(device):
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+        raise Exception("start_render_thread" + ERR_LOCK_FAILED)
+    log.info(f"Start new Rendering Thread on device: {device}")
+    try:
+        rthread = threading.Thread(target=thread_render, kwargs={"device": device})
+        rthread.daemon = True
+        rthread.name = THREAD_NAME_PREFIX + device
+        rthread.start()
+        render_threads.append(rthread)
+    finally:
+        manager_lock.release()
+    timeout = DEVICE_START_TIMEOUT
+    while not rthread.is_alive() or not rthread in weak_thread_data or not "device" in weak_thread_data[rthread]:
+        if rthread in weak_thread_data and "error" in weak_thread_data[rthread]:
+            log.error(f"{rthread}, {device}, error: {weak_thread_data[rthread]['error']}")
+            return False
+        if timeout <= 0:
+            return False
+        timeout -= 1
+        time.sleep(1)
+    return True
+
+
+def stop_render_thread(device):
+    try:
+        device_manager.validate_device_id(device, log_prefix="stop_render_thread")
+    except:
+        log.error(traceback.format_exc())
+        return False
+
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+        raise Exception("stop_render_thread" + ERR_LOCK_FAILED)
+    log.info(f"Stopping Rendering Thread on device: {device}")
+
+    try:
+        thread_to_remove = None
+        for rthread in render_threads:
+            weak_data = weak_thread_data.get(rthread)
+            if weak_data is None or not "device" in weak_data or weak_data["device"] is None:
+                continue
+            thread_device = weak_data["device"]
+            if thread_device == device:
+                weak_data["alive"] = False
+                thread_to_remove = rthread
+                break
+        if thread_to_remove is not None:
+            render_threads.remove(rthread)
+            return True
+    finally:
+        manager_lock.release()
+
+    return False
+
+
+def update_render_threads(render_devices, active_devices):
+    devices_to_start, devices_to_stop = device_manager.get_device_delta(render_devices, active_devices)
+    log.debug(f"devices_to_start: {devices_to_start}")
+    log.debug(f"devices_to_stop: {devices_to_stop}")
+
+    for device in devices_to_stop:
+        if is_alive(device) <= 0:
+            log.debug(f"{device} is not alive")
+            continue
+        if not stop_render_thread(device):
+            log.warn(f"{device} could not stop render thread")
+
+    for device in devices_to_start:
+        if is_alive(device) >= 1:
+            log.debug(f"{device} already registered.")
+            continue
+        if not start_render_thread(device):
+            log.warn(f"{device} failed to start.")
+
+    if is_alive() <= 0:  # No running devices, probably invalid user config.
+        raise EnvironmentError(
+            'ERROR: No active render devices! Please verify the "render_devices" value in config.json'
+        )
+
+    log.debug(f"active devices: {get_devices()['active']}")
+
+
+def shutdown_event():  # Signal render thread to close on shutdown
+    global current_state_error
+    current_state_error = SystemExit("Application shutting down.")
+
+
+def render(render_req: GenerateImageRequest, task_data: TaskData):
+    current_thread_count = is_alive()
+    if current_thread_count <= 0:  # Render thread is dead
+        raise ChildProcessError("Rendering thread has died.")
+
+    # Alive, check if task in cache
+    session = get_cached_session(task_data.session_id, update_ttl=True)
+    pending_tasks = list(filter(lambda t: t.is_pending, session.tasks))
+    if current_thread_count < len(pending_tasks):
+        raise ConnectionRefusedError(
+            f"Session {task_data.session_id} already has {len(pending_tasks)} pending tasks out of {current_thread_count}."
+        )
+
+    new_task = RenderTask(render_req, task_data)
+    if session.put(new_task, TASK_TTL):
+        # Use twice the normal timeout for adding user requests.
+        # Tries to force session.put to fail before tasks_queue.put would.
+        if manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT * 2):
+            try:
+                tasks_queue.append(new_task)
+                idle_event.set()
+                return new_task
+            finally:
+                manager_lock.release()
+    raise RuntimeError("Failed to add task to cache.")

ui/easydiffusion/types.py

@@ -0,0 +1,103 @@
+from pydantic import BaseModel
+from typing import Any
+
+
+class GenerateImageRequest(BaseModel):
+    prompt: str = ""
+    negative_prompt: str = ""
+
+    seed: int = 42
+    width: int = 512
+    height: int = 512
+
+    num_outputs: int = 1
+    num_inference_steps: int = 50
+    guidance_scale: float = 7.5
+
+    init_image: Any = None
+    init_image_mask: Any = None
+    prompt_strength: float = 0.8
+    preserve_init_image_color_profile = False
+
+    sampler_name: str = None  # "ddim", "plms", "heun", "euler", "euler_a", "dpm2", "dpm2_a", "lms"
+    hypernetwork_strength: float = 0
+
+
+class TaskData(BaseModel):
+    request_id: str = None
+    session_id: str = "session"
+    save_to_disk_path: str = None
+    vram_usage_level: str = "balanced"  # or "low" or "medium"
+
+    use_face_correction: str = None  # or "GFPGANv1.3"
+    use_upscale: str = None  # or "RealESRGAN_x4plus" or "RealESRGAN_x4plus_anime_6B"
+    upscale_amount: int = 4  # or 2
+    use_stable_diffusion_model: str = "sd-v1-4"
+    # use_stable_diffusion_config: str = "v1-inference"
+    use_vae_model: str = None
+    use_hypernetwork_model: str = None
+
+    show_only_filtered_image: bool = False
+    block_nsfw: bool = False
+    output_format: str = "jpeg"  # or "png" or "webp"
+    output_quality: int = 75
+    metadata_output_format: str = "txt"  # or "json"
+    stream_image_progress: bool = False
+    stream_image_progress_interval: int = 5
+
+
+class MergeRequest(BaseModel):
+    model0: str = None
+    model1: str = None
+    ratio: float = None
+    out_path: str = "mix"
+    use_fp16 = True
+
+
+class Image:
+    data: str  # base64
+    seed: int
+    is_nsfw: bool
+    path_abs: str = None
+
+    def __init__(self, data, seed):
+        self.data = data
+        self.seed = seed
+
+    def json(self):
+        return {
+            "data": self.data,
+            "seed": self.seed,
+            "path_abs": self.path_abs,
+        }
+
+
+class Response:
+    render_request: GenerateImageRequest
+    task_data: TaskData
+    images: list
+
+    def __init__(self, render_request: GenerateImageRequest, task_data: TaskData, images: list):
+        self.render_request = render_request
+        self.task_data = task_data
+        self.images = images
+
+    def json(self):
+        del self.render_request.init_image
+        del self.render_request.init_image_mask
+
+        res = {
+            "status": "succeeded",
+            "render_request": self.render_request.dict(),
+            "task_data": self.task_data.dict(),
+            "output": [],
+        }
+
+        for image in self.images:
+            res["output"].append(image.json())
+
+        return res
+
+
+class UserInitiatedStop(Exception):
+    pass

ui/easydiffusion/utils/__init__.py

@@ -0,0 +1,8 @@
+import logging
+
+log = logging.getLogger("easydiffusion")
+
+from .save_utils import (
+    save_images_to_disk,
+    get_printable_request,
+)

ui/easydiffusion/utils/save_utils.py

@@ -0,0 +1,132 @@
+import os
+import time
+import base64
+import re
+
+from easydiffusion.types import TaskData, GenerateImageRequest
+
+from sdkit.utils import save_images, save_dicts
+
+filename_regex = re.compile("[^a-zA-Z0-9._-]")
+
+# keep in sync with `ui/media/js/dnd.js`
+TASK_TEXT_MAPPING = {
+    "prompt": "Prompt",
+    "width": "Width",
+    "height": "Height",
+    "seed": "Seed",
+    "num_inference_steps": "Steps",
+    "guidance_scale": "Guidance Scale",
+    "prompt_strength": "Prompt Strength",
+    "use_face_correction": "Use Face Correction",
+    "use_upscale": "Use Upscaling",
+    "upscale_amount": "Upscale By",
+    "sampler_name": "Sampler",
+    "negative_prompt": "Negative Prompt",
+    "use_stable_diffusion_model": "Stable Diffusion model",
+    "use_vae_model": "VAE model",
+    "use_hypernetwork_model": "Hypernetwork model",
+    "hypernetwork_strength": "Hypernetwork Strength",
+}
+
+
+def save_images_to_disk(images: list, filtered_images: list, req: GenerateImageRequest, task_data: TaskData):
+    now = time.time()
+    save_dir_path = os.path.join(task_data.save_to_disk_path, filename_regex.sub("_", task_data.session_id))
+    metadata_entries = get_metadata_entries_for_request(req, task_data)
+    make_filename = make_filename_callback(req, now=now)
+
+    if task_data.show_only_filtered_image or filtered_images is images:
+        save_images(
+            filtered_images,
+            save_dir_path,
+            file_name=make_filename,
+            output_format=task_data.output_format,
+            output_quality=task_data.output_quality,
+        )
+        if task_data.metadata_output_format.lower() in ["json", "txt", "embed"]:
+            save_dicts(
+                metadata_entries,
+                save_dir_path,
+                file_name=make_filename,
+                output_format=task_data.metadata_output_format,
+                file_format=task_data.output_format,
+            )
+    else:
+        make_filter_filename = make_filename_callback(req, now=now, suffix="filtered")
+
+        save_images(
+            images,
+            save_dir_path,
+            file_name=make_filename,
+            output_format=task_data.output_format,
+            output_quality=task_data.output_quality,
+        )
+        save_images(
+            filtered_images,
+            save_dir_path,
+            file_name=make_filter_filename,
+            output_format=task_data.output_format,
+            output_quality=task_data.output_quality,
+        )
+        if task_data.metadata_output_format.lower() in ["json", "txt", "embed"]:
+            save_dicts(
+                metadata_entries,
+                save_dir_path,
+                file_name=make_filter_filename,
+                output_format=task_data.metadata_output_format,
+                file_format=task_data.output_format,
+            )
+
+
+def get_metadata_entries_for_request(req: GenerateImageRequest, task_data: TaskData):
+    metadata = get_printable_request(req)
+    metadata.update(
+        {
+            "use_stable_diffusion_model": task_data.use_stable_diffusion_model,
+            "use_vae_model": task_data.use_vae_model,
+            "use_hypernetwork_model": task_data.use_hypernetwork_model,
+            "use_face_correction": task_data.use_face_correction,
+            "use_upscale": task_data.use_upscale,
+        }
+    )
+    if metadata["use_upscale"] is not None:
+        metadata["upscale_amount"] = task_data.upscale_amount
+    if task_data.use_hypernetwork_model is None:
+        del metadata["hypernetwork_strength"]
+
+    # if text, format it in the text format expected by the UI
+    is_txt_format = task_data.metadata_output_format.lower() == "txt"
+    if is_txt_format:
+        metadata = {TASK_TEXT_MAPPING[key]: val for key, val in metadata.items() if key in TASK_TEXT_MAPPING}
+
+    entries = [metadata.copy() for _ in range(req.num_outputs)]
+    for i, entry in enumerate(entries):
+        entry["Seed" if is_txt_format else "seed"] = req.seed + i
+
+    return entries
+
+
+def get_printable_request(req: GenerateImageRequest):
+    metadata = req.dict()
+    del metadata["init_image"]
+    del metadata["init_image_mask"]
+    if req.init_image is None:
+        del metadata["prompt_strength"]
+    return metadata
+
+
+def make_filename_callback(req: GenerateImageRequest, suffix=None, now=None):
+    if now is None:
+        now = time.time()
+
+    def make_filename(i):
+        img_id = base64.b64encode(int(now + i).to_bytes(8, "big")).decode()  # Generate unique ID based on time.
+        img_id = img_id.translate({43: None, 47: None, 61: None})[-8:]  # Remove + / = and keep last 8 chars.
+
+        prompt_flattened = filename_regex.sub("_", req.prompt)[:50]
+        name = f"{prompt_flattened}_{img_id}"
+        name = name if suffix is None else f"{name}_{suffix}"
+        return name
+
+    return make_filename

ui/index.html

@@ -1,11 +1,12 @@
 <!DOCTYPE html>
 <html>
 <head>
-    <title>Stable Diffusion UI</title>
+    <title>Easy Diffusion</title>
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
     <meta name="theme-color" content="#673AB6">
     <link rel="icon" type="image/png" href="/media/images/favicon-16x16.png" sizes="16x16">
     <link rel="icon" type="image/png" href="/media/images/favicon-32x32.png" sizes="32x32">
+    <link rel="stylesheet" href="/media/css/jquery-confirm.min.css">
     <link rel="stylesheet" href="/media/css/fonts.css">
     <link rel="stylesheet" href="/media/css/themes.css">
     <link rel="stylesheet" href="/media/css/main.css">
@@ -13,7 +14,7 @@
     <link rel="stylesheet" href="/media/css/modifier-thumbnails.css">
     <link rel="stylesheet" href="/media/css/fontawesome-all.min.css">
     <link rel="stylesheet" href="/media/css/image-editor.css">
-    <link rel="stylesheet" href="/media/css/jquery-confirm.min.css">
+    <link rel="stylesheet" href="/media/css/searchable-models.css">
     <link rel="manifest" href="/media/manifest.webmanifest">
     <script src="/media/js/jquery-3.6.1.min.js"></script>
     <script src="/media/js/jquery-confirm.min.js"></script>
@@ -24,15 +25,16 @@
     <div id="top-nav">
         <div id="logo">
             <h1>
-                Stable Diffusion UI
-                <small>v2.4.20 <span id="updateBranchLabel"></span></small>
+                <img id="logo_img" src="/media/images/icon-512x512.png" >
+                Easy Diffusion
+                <small>v2.5.24 <span id="updateBranchLabel"></span></small>
             </h1>
         </div>
         <div id="server-status">
             <div id="server-status-color">●</div>
             <span id="server-status-msg">Stable Diffusion is starting..</span>
         </div>
-        <div id="tab-container">
+        <div id="tab-container" class="tab-container">
             <span id="tab-main" class="tab active">
                 <span><i class="fa fa-image icon"></i> Generate</span>
             </span>
@@ -50,12 +52,12 @@
         <div id="editor">
             <div id="editor-inputs">
                 <div id="editor-inputs-prompt" class="row">
-                    <label for="prompt"><b>Enter Prompt</b></label> <small>or</small> <button id="promptsFromFileBtn">Load from a file</button>
+                    <label for="prompt"><b>Enter Prompt</b></label> <small>or</small> <button id="promptsFromFileBtn" class="tertiaryButton">Load from a file</button>
                     <textarea id="prompt" class="col-free">a photograph of an astronaut riding a horse</textarea>
                     <input id="prompt_from_file" name="prompt_from_file" type="file" /> <!-- hidden -->
                     <label for="negative_prompt" class="collapsible" id="negative_prompt_handle">
                         Negative Prompt
-                        <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Writing-prompts#negative-prompts" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Click to learn more about Negative Prompts</span></i></a>
+                        <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Writing-prompts#negative-prompts" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top">Click to learn more about Negative Prompts</span></i></a>
                         <small>(optional)</small>
                     </label>
                     <div class="collapsible-content">
@@ -69,7 +71,7 @@
                     <div id="init_image_preview_container" class="image_preview_container">
                         <div id="init_image_wrapper">
                             <img id="init_image_preview" src="" />
-                            <span id="init_image_size_box"></span>
+                            <span id="init_image_size_box" class="img_bottom_label"></span>
                             <button class="init_image_clear image_clear_btn"><i class="fa-solid fa-xmark"></i></button>
                         </div>
                         <div id="init_image_buttons">
@@ -92,10 +94,12 @@
                         </div>
                     </div>
 
+                    <div id="apply_color_correction_setting" class="pl-5"><input id="apply_color_correction" name="apply_color_correction" type="checkbox"> <label for="apply_color_correction">Preserve color profile <small>(helps during inpainting)</small></label></div>
+
                 </div>
 
                 <div id="editor-inputs-tags-container" class="row">
-                    <label>Image Modifiers <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">click an Image Modifier to remove it, use Ctrl+Mouse Wheel to adjust its weight</span></i>:</label>
+                    <label>Image Modifiers <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip right">click an Image Modifier to remove it, right-click to temporarily disable it, use Ctrl+Mouse Wheel to adjust its weight</span></i></label>
                     <div id="editor-inputs-tags-list"></div>
                 </div>
 
@@ -123,28 +127,40 @@
                     <tr><b class="settings-subheader">Image Settings</b></tr>
                     <tr class="pl-5"><td><label for="seed">Seed:</label></td><td><input id="seed" name="seed" size="10" value="0" onkeypress="preventNonNumericalInput(event)"> <input id="random_seed" name="random_seed" type="checkbox" checked><label for="random_seed">Random</label></td></tr>
                     <tr class="pl-5"><td><label for="num_outputs_total">Number of Images:</label></td><td><input id="num_outputs_total" name="num_outputs_total" value="1" size="1" onkeypress="preventNonNumericalInput(event)"> <label><small>(total)</small></label> <input id="num_outputs_parallel" name="num_outputs_parallel" value="1" size="1" onkeypress="preventNonNumericalInput(event)"> <label for="num_outputs_parallel"><small>(in parallel)</small></label></td></tr>
-                    <tr class="pl-5"><td><label for="stable_diffusion_model">Model:</label></td><td>
-                        <select id="stable_diffusion_model" name="stable_diffusion_model">
-                            <!-- <option value="sd-v1-4" selected>sd-v1-4</option> -->
-                        </select>
+                    <tr class="pl-5"><td><label for="stable_diffusion_model">Model:</label></td><td class="model-input">
+                        <input id="stable_diffusion_model" type="text" spellcheck="false" autocomplete="off" class="model-filter" data-path="" />
+                        <button id="reload-models" class="secondaryButton reloadModels"><i class='fa-solid fa-rotate'></i></button>
                         <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Custom-Models" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Click to learn more about custom models</span></i></a>
                     </td></tr>
-                    <tr class="pl-5"><td><label for="vae_model">Custom VAE:</i></label></td><td>
-                        <select id="vae_model" name="vae_model">
-                            <!-- <option value="" selected>None</option> -->
+                    <!-- <tr id="modelConfigSelection" class="pl-5"><td><label for="model_config">Model Config:</i></label></td><td>
+                        <select id="model_config" name="model_config">
                         </select>
+                    </td></tr> -->
+                    <tr class="pl-5"><td><label for="vae_model">Custom VAE:</i></label></td><td>
+                        <input id="vae_model" type="text" spellcheck="false" autocomplete="off" class="model-filter" data-path="" />
                         <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/VAE-Variational-Auto-Encoder" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Click to learn more about VAEs</span></i></a>
                     </td></tr>
-                    <tr id="samplerSelection" class="pl-5"><td><label for="sampler">Sampler:</label></td><td>
-                        <select id="sampler" name="sampler">
-                            <option value="plms">plms</option>
-                            <option value="ddim">ddim</option>
-                            <option value="heun">heun</option>
-                            <option value="euler">euler</option>
-                            <option value="euler_a" selected>euler_a</option>
-                            <option value="dpm2">dpm2</option>
-                            <option value="dpm2_a">dpm2_a</option>
-                            <option value="lms">lms</option>
+                    <tr id="samplerSelection" class="pl-5"><td><label for="sampler_name">Sampler:</label></td><td>
+                        <select id="sampler_name" name="sampler_name">
+                            <option value="plms">PLMS</option>
+                            <option value="ddim">DDIM</option>
+                            <option value="heun">Heun</option>
+                            <option value="euler">Euler</option>
+                            <option value="euler_a" selected>Euler Ancestral</option>
+                            <option value="dpm2">DPM2</option>
+                            <option value="dpm2_a">DPM2 Ancestral</option>
+                            <option value="lms">LMS</option>
+                            <option value="dpm_solver_stability">DPM Solver (Stability AI)</option>
+                            <option value="dpmpp_2s_a">DPM++ 2s Ancestral</option>
+                            <option value="dpmpp_2m">DPM++ 2m</option>
+                            <option value="dpmpp_sde">DPM++ SDE</option>
+                            <option value="dpm_fast">DPM Fast</option>
+                            <option value="dpm_adaptive">DPM Adaptive</option>
+                            <option value="unipc_snr">UniPC SNR</option>
+                            <option value="unipc_tu">UniPC TU</option>
+                            <option value="unipc_snr_2">UniPC SNR 2</option>
+                            <option value="unipc_tu_2">UniPC TC 2</option>
+                            <option value="unipc_tq">UniPC TQ</option>
                         </select>
                         <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/How-to-Use#samplers" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Click to learn more about samplers</span></i></a>
                     </td></tr>
@@ -195,12 +211,10 @@
                         <label for="height"><small>(height)</small></label>
                     </td></tr>
                     <tr class="pl-5"><td><label for="num_inference_steps">Inference Steps:</label></td><td> <input id="num_inference_steps" name="num_inference_steps" size="4" value="25" onkeypress="preventNonNumericalInput(event)"></td></tr>
-                    <tr class="pl-5"><td><label for="guidance_scale_slider">Guidance Scale:</label></td><td> <input id="guidance_scale_slider" name="guidance_scale_slider" class="editor-slider" value="75" type="range" min="10" max="500"> <input id="guidance_scale" name="guidance_scale" size="4" pattern="^[0-9\.]+$" onkeypress="preventNonNumericalInput(event)"></td></tr>
+                    <tr class="pl-5"><td><label for="guidance_scale_slider">Guidance Scale:</label></td><td> <input id="guidance_scale_slider" name="guidance_scale_slider" class="editor-slider" value="75" type="range" min="11" max="500"> <input id="guidance_scale" name="guidance_scale" size="4" pattern="^[0-9\.]+$" onkeypress="preventNonNumericalInput(event)"></td></tr>
                     <tr id="prompt_strength_container" class="pl-5"><td><label for="prompt_strength_slider">Prompt Strength:</label></td><td> <input id="prompt_strength_slider" name="prompt_strength_slider" class="editor-slider" value="80" type="range" min="0" max="99"> <input id="prompt_strength" name="prompt_strength" size="4" pattern="^[0-9\.]+$" onkeypress="preventNonNumericalInput(event)"><br/></td></tr>
                     <tr class="pl-5"><td><label for="hypernetwork_model">Hypernetwork:</i></label></td><td>
-                        <select id="hypernetwork_model" name="hypernetwork_model">
-                            <!-- <option value="" selected>None</option> -->
-                        </select>
+                        <input id="hypernetwork_model" type="text" spellcheck="false" autocomplete="off" class="model-filter" data-path="" />
                     </td></tr>
                     <tr id="hypernetwork_strength_container" class="pl-5">
                         <td><label for="hypernetwork_strength_slider">Hypernetwork Strength:</label></td>
@@ -210,9 +224,10 @@
                         <select id="output_format" name="output_format">
                             <option value="jpeg" selected>jpeg</option>
                             <option value="png">png</option>
+                            <option value="webp">webp</option>
                         </select>
                     </td></tr>
-                    <tr class="pl-5" id="output_quality_row"><td><label for="output_quality">JPEG Quality:</label></td><td>
+                    <tr class="pl-5" id="output_quality_row"><td><label for="output_quality">Image Quality:</label></td><td>
                     <input id="output_quality_slider" name="output_quality" class="editor-slider" value="75" type="range" min="10" max="95"> <input id="output_quality" name="output_quality" size="4" pattern="^[0-9\.]+$" onkeypress="preventNonNumericalInput(event)">
                     </td></tr>
                     </table></div>
@@ -220,9 +235,14 @@
                     <div><ul>
                     <li><b class="settings-subheader">Render Settings</b></li>
                     <li class="pl-5"><input id="stream_image_progress" name="stream_image_progress" type="checkbox"> <label for="stream_image_progress">Show a live preview <small>(uses more VRAM, slower images)</small></label></li>
-                    <li class="pl-5"><input id="use_face_correction" name="use_face_correction" type="checkbox"> <label for="use_face_correction">Fix incorrect faces and eyes <small>(uses GFPGAN)</small></label></li>
+                    <li class="pl-5"><input id="use_face_correction" name="use_face_correction" type="checkbox"> <label for="use_face_correction">Fix incorrect faces and eyes</label> <div style="display:inline-block;"><input id="gfpgan_model" type="text" spellcheck="false" autocomplete="off" class="model-filter" data-path="" /></div></li>
                     <li class="pl-5">
-                        <input id="use_upscale" name="use_upscale" type="checkbox"> <label for="use_upscale">Upscale image by 4x with </label>
+                        <input id="use_upscale" name="use_upscale" type="checkbox"> <label for="use_upscale">Scale up by</label> 
+                        <select id="upscale_amount" name="upscale_amount">
+                            <option value="2">2x</option>
+                            <option value="4" selected>4x</option>
+                        </select>
+                        with 
                         <select id="upscale_model" name="upscale_model">
                             <option value="RealESRGAN_x4plus" selected>RealESRGAN_x4plus</option>
                             <option value="RealESRGAN_x4plus_anime_6B">RealESRGAN_x4plus_anime_6B</option>
@@ -264,8 +284,35 @@
                 and selecting the desired modifiers.<br/><br/>
                 Click "Image Settings" for additional settings like seed, image size, number of images to generate etc.<br/><br/>Enjoy! :)
             </div>
-            <div id="preview-tools">
-                <button id="clear-all-previews" class="secondaryButton"><i class="fa-solid fa-trash-can"></i> Clear All</button>
+
+            <div id="preview-content">
+                <div id="preview-tools">
+                    <button id="clear-all-previews" class="secondaryButton"><i class="fa-solid fa-trash-can icon"></i> Clear All</button>
+                    <button id="save-all-images" class="tertiaryButton"><i class="fa-solid fa-download icon"></i> Download All Images</button>
+                    <div class="display-settings">
+                        <span class="auto-scroll"></span> <!-- hack for Rabbit Hole update -->
+                        <button id="auto_scroll_btn" class="tertiaryButton">
+                            <i class="fa-solid fa-arrows-up-to-line icon"></i>
+                            <input id="auto_scroll" name="auto_scroll" type="checkbox" style="display: none">
+                            <span class="simple-tooltip left">
+                                Scroll to generated image (<span class="state">OFF</span>)
+                            </span>
+                        </button>
+                        <button class="dropdown tertiaryButton">
+                            <i class="fa-solid fa-magnifying-glass-plus icon dropbtn"></i>
+                            <span class="simple-tooltip left">
+                                Image Size
+                            </span>
+                        </button>
+                        <div class="dropdown-content">
+                            <div class="dropdown-item">
+                                <input id="thumbnail_size" name="thumbnail_size" class="editor-slider" type="range" value="70" min="5" max="200" oninput="sliderUpdate(event)"> 
+                                <input id="thumbnail_size-input" name="thumbnail_size-input" size="3" value="70" pattern="^[0-9.]+$" onkeypress="preventNonNumericalInput(event)" oninput="sliderUpdate(event)">&nbsp;%
+                            </div>
+                        </div>
+                    </div>
+                    <div class="clearfix" style="clear: both;"></div>
+                </div>
             </div>
         </div>
         </div>
@@ -408,6 +455,7 @@
 <script src="media/js/image-modifiers.js"></script>
 <script src="media/js/auto-save.js"></script>
 
+<script src="media/js/searchable-models.js"></script>
 <script src="media/js/main.js"></script>
 <script src="media/js/themes.js"></script>
 <script src="media/js/dnd.js"></script>
@@ -416,7 +464,6 @@
 async function init() {
     await initSettings()
     await getModels()
-    await getDiskPath()
     await getAppConfig()
     await loadUIPlugins()
     await loadModifiers()

ui/main.py

@@ -0,0 +1,10 @@
+from easydiffusion import model_manager, app, server
+from easydiffusion.server import server_api # required for uvicorn
+
+# Init the app
+model_manager.init()
+app.init()
+server.init()
+
+# start the browser ui
+app.open_browser()

ui/media/css/image-editor.css

@@ -2,12 +2,12 @@
 	padding-left: 32px;
 	text-align: left;
 	padding-bottom: 20px;
+	max-width: min-content;
 }
 
 .editor-options-container {
 	display: flex;
 	row-gap: 10px;
-	max-width: 210px;
 }
 
 .editor-options-container > * {
@@ -31,7 +31,7 @@
 }
 
 .editor-options-container > * > *.active {
-	border: 2px solid #3584e4;
+	border: 1px solid #3584e4;
 }
 
 .image_editor_opacity .editor-options-container > * > *:not(.active) {
@@ -160,6 +160,7 @@
 	padding: var(--popup-padding);
 	min-height: calc(100vh - (2 * var(--popup-margin)));
 	max-width: none;
+	min-width: fit-content;
 }
 
 .image-editor-popup h1 {
@@ -213,3 +214,10 @@
 .image-editor-popup h4 {
 	text-align: left;
 }
+
+.image-editor-popup .load_mask {
+	display: none;
+}
+.inpainter .load_mask {
+	display: flex;
+}

ui/media/css/main.css

@@ -27,6 +27,11 @@ code {
     padding: 2px 4px;
     border-radius: 4px;
 }
+#logo_img {
+    width: 32px;
+    height: 32px;
+    transform: translateY(4px);
+}
 #prompt {
     width: 100%;
     height: 65pt;
@@ -107,6 +112,7 @@ code {
 .imgContainer {
     display: flex;
     justify-content: flex-end;
+    position: relative;
 }
 .imgItemInfo {
     padding-bottom: 0.5em;
@@ -114,16 +120,35 @@ code {
     align-items: flex-end;
     flex-direction: column;
     position: absolute;
-    padding: 5px;
+    padding-right: 5pt;
+    padding-top: 6pt;
     opacity: 0;
     transition: 0.1s all;
 }
+.imgPreviewItemClearBtn {
+    opacity: 0;
+}
+.imgContainer .img_bottom_label {
+    opacity: 0;
+}
+.imgPreviewItemClearBtn:hover {
+    background: rgb(177, 27, 0);
+}
 .imgContainer:hover > .imgItemInfo {
     opacity: 1;
 }
+.imgContainer:hover > .imgPreviewItemClearBtn {
+    opacity: 1;
+}
+.imgContainer:hover > .img_bottom_label {
+    opacity: 60%;
+}
 .imgItemInfo * {
     margin-bottom: 7px;
 }
+.imgItem .image_clear_btn {
+    transform: translate(40%, -50%);
+}
 #container {
     min-height: 100vh;
     width: 100%;
@@ -179,7 +204,7 @@ code {
     flex: 0 0 70px;
     background: var(--accent-color);
     border: var(--primary-button-border);
-    color: rgb(255, 221, 255);
+    color: var(--accent-text-color);
     width: 100%;
     height: 30pt;
 }
@@ -251,6 +276,11 @@ button#resume {
 img {
     box-shadow: 0 4px 8px 0 rgba(0, 0, 0, 0.15), 0 6px 20px 0 rgba(0, 0, 0, 0.15);
 }
+div.img-preview img {
+    width:100%;
+    height: 100%;
+    max-height: 70vh;
+}
 .line-separator {
     background: var(--background-color3);
     height: 1pt;
@@ -299,6 +329,7 @@ img {
 #logo {
     display: inline;
     padding: 12px;
+    padding-top: 8px;
     white-space: nowrap;
 }
 #logo h1 {
@@ -383,10 +414,8 @@ img {
     display: none;
     position: absolute;
     z-index: 2;
+    width: max-content;
 
-    background: var(--background-color4);
-    border: 2px solid var(--background-color2);
-    border-radius: 7px;
     padding: 5px;
     margin-bottom: 15px;
     box-shadow: 0 20px 28px 0 rgba(0, 0, 0, 0.15), 0 6px 20px 0 rgba(0, 0, 0, 0.15);
@@ -394,6 +423,36 @@ img {
 .dropdown:hover .dropdown-content {
     display: block;
 }
+.dropdown:hover + .dropdown-content {
+    display: block;
+}
+.dropdown-content:hover {
+    display: block;
+}
+
+.display-settings {
+    float: right;
+    position: relative;
+}
+
+.display-settings .dropdown-content {
+    right: 0px;
+    top: 12pt;
+}
+
+.dropdown-item {
+    padding: 4px;
+    background: var(--background-color4);
+    border: 2px solid var(--background-color2);
+}
+
+.dropdown-item:first-child {
+    border-radius: 7px 7px 0px 0px;
+}
+
+.dropdown-item:last-child {
+    border-radius: 0px 0px 7px 7px;
+}
 
 .imageTaskContainer {
     border: 1px solid var(--background-color2);
@@ -449,6 +508,7 @@ img {
     background: var(--accent-color);
     border: var(--primary-button-border);
     color: rgb(255, 221, 255);
+    padding: 3pt 6pt;
 }
 .secondaryButton {
     background: rgb(132, 8, 0);
@@ -460,17 +520,26 @@ img {
 .secondaryButton:hover {
     background: rgb(177, 27, 0);
 }
-.useSettings {
-    background: var(--accent-color);
-    border: 1px solid var(--accent-color);
-    color: rgb(255, 221, 255);
+.tertiaryButton {
+    background: var(--tertiary-background-color);
+    color: var(--tertiary-color);
+    border: 1px solid var(--tertiary-border-color);
     padding: 3pt 6pt;
+    border-radius: 5px;
+}
+.tertiaryButton:hover {
+    background: hsl(var(--accent-hue), 100%, calc(var(--accent-lightness) + 6%));
+    color: var(--accent-text-color);
+}
+.tertiaryButton.pressed {
+    border-style: inset;
+    background: hsl(var(--accent-hue), 100%, calc(var(--accent-lightness) + 6%));
+    color: var(--accent-text-color);
+}
+.useSettings {
     margin-right: 6pt;
     float: right;
 }
-.useSettings:hover {
-    background: hsl(var(--accent-hue), 100%, calc(var(--accent-lightness) + 6%));
-}
 .stopTask {
     float: right;
 }
@@ -558,6 +627,9 @@ img {
 } */
 
 #init_image_size_box {
+    border-radius: 6px 0px;
+}
+.img_bottom_label {
     position: absolute;
     right: 0px;
     bottom: 0px;
@@ -567,7 +639,6 @@ img {
     text-shadow: 0px 0px 4px black;
     opacity: 60%;
     font-size: 12px;
-    border-radius: 6px 0px;
 }
 
 #editor-settings {
@@ -584,7 +655,6 @@ img {
 }
 
 #editor-settings-entries ul {
-    margin: 0px;
     padding: 0px;
 }
 
@@ -731,6 +801,13 @@ input::file-selector-button {
     right: calc(var(--input-border-size) + var(--input-switch-padding));
     opacity: 1;
 }
+.model-filter {
+    width: 90%;
+    padding-right: 20px;
+    white-space: nowrap;
+    overflow: hidden;
+    text-overflow: ellipsis;
+}
 
 /* Small screens */
 @media screen and (max-width: 1265px) {
@@ -768,12 +845,6 @@ input::file-selector-button {
         width: 100%;
         object-fit: contain;
     }
-    .dropdown-content {
-        width: auto !important;
-        transform: none !important;
-        left: 0px;
-        right: 0px;
-    }
     #editor {
         padding: 16px 8px;
     }
@@ -806,6 +877,12 @@ input::file-selector-button {
     .simple-tooltip {
         display: none;
     }
+    #preview-tools button {
+        font-size: 0px;
+    }
+    #preview-tools button .icon {
+        font-size: 12pt;
+    }
 }
 
 @media screen and (max-width: 500px) {
@@ -838,7 +915,7 @@ input::file-selector-button {
 #promptsFromFileBtn {
     font-size: 9pt;
     display: inline;
-    background-color: var(--accent-color);
+    padding: 2pt;
 }
 
 .section-button {
@@ -871,17 +948,19 @@ input::file-selector-button {
 
 /* SIMPLE TOOTIP */
 .simple-tooltip {
-	border-radius: 3px;
-	font-weight: bold;
-	font-size: 12px;
+    border-radius: 3px;
+    font-weight: bold;
+    font-size: 12px;
     background-color: var(--background-color3);
 
-	visibility: hidden;
-	opacity: 0;
-	position: absolute;
-	white-space: nowrap;
-	padding: 8px 12px;
-	transition: 0.3s all;
+    visibility: hidden;
+    opacity: 0;
+    position: absolute;
+    width: max-content;
+    max-width: 300px;
+    padding: 8px 12px;
+    transition: 0.3s all;
+    z-index: 1000;
 
     pointer-events: none;
 }
@@ -895,7 +974,7 @@ input::file-selector-button {
 .simple-tooltip.right {
     right: 0px;
     top: 50%;
-    transform: translate(calc(100% - 15%), -50%);
+    transform: translate(100%, -50%);
 }
 :hover > .simple-tooltip.right {
     transform: translate(100%, -50%);
@@ -1023,7 +1102,7 @@ input::file-selector-button {
 }
 
 /* TABS */
-#tab-container {
+.tab-container {
     display: flex;
     align-items: flex-end;
 }
@@ -1099,11 +1178,11 @@ button:active {
 
 div.task-initimg > img {
     margin-right: 6px;
-    display:      block;
+    display: block;
 }
 div.task-fs-initimage {
-    display:  none;
-#    position: absolute;
+    display: none;
+    position: absolute;
 }
 div.task-initimg:hover div.task-fs-initimage {
     display: block;
@@ -1111,6 +1190,8 @@ div.task-initimg:hover div.task-fs-initimage {
     z-index: 9999;
     box-shadow: 0 0 30px #000;
     margin-top:-64px;
+    max-width: 75vw;
+    max-height: 75vh;
 }
 div.top-right {
     position: absolute;
@@ -1174,3 +1255,14 @@ body.wait-pause {
     50% { border: solid 12px var(--background-color1); }
    100% { border: solid 12px var(--accent-color); }
 }
+
+.jconfirm.jconfirm-modern .jconfirm-box div.jconfirm-title-c {
+    color: var(--button-text-color);
+}
+.jconfirm.jconfirm-modern .jconfirm-box {
+    background-color: var(--background-color1);
+}
+
+.displayNone {
+    display:none !important;
+}

ui/media/css/searchable-models.css

@@ -0,0 +1,99 @@
+.model-list {
+    position: absolute;
+    margin-block-start: 2px;
+    display: none;
+    padding-inline-start: 0;
+    max-height: 200px;
+    overflow: auto;
+    background: var(--input-background-color);
+    border: var(--input-border-size) solid var(--input-border-color);
+    border-radius: var(--input-border-radius);
+    color: var(--input-text-color);
+    z-index: 1;
+    line-height: normal;
+}
+
+.model-list ul {
+    padding-right: 20px;
+    padding-inline-start: 0;
+    margin-top: 3pt;
+}
+
+.model-list li {
+    padding-top: 3px;
+    padding-bottom: 3px;
+}
+
+.model-list .icon {
+    padding-right: 3pt;
+}
+
+.model-result {
+    list-style: none;
+}
+
+.model-no-result {
+    color: var(--text-color);
+    list-style: none;
+    padding: 3px 6px 3px 6px;
+    font-size: 9pt;
+    font-style: italic;
+    display: none;
+}
+
+.model-list li.model-folder {
+    color: var(--text-color);
+    list-style: none;
+    padding: 6px 6px 6px 6px;
+    font-size: 9pt;
+    font-weight: bold;
+    border-top: 1px solid var(--background-color1);
+}
+
+.model-list li.model-file {
+    color: var(--input-text-color);
+    list-style: none;
+    padding-left: 12px;
+    padding-right:20px;
+    font-size: 10pt;
+    font-weight: normal;
+    transition: none;
+    transition:property: none;
+    cursor: default;
+}
+
+.model-list li.model-file.in-root-folder {
+    padding-left: 6px;
+}
+
+.model-list li.model-file.selected {
+    background: grey;
+}
+
+.model-selector {
+    cursor: pointer;
+}
+
+.model-selector-arrow {
+    position: absolute;
+    width: 17px;
+    margin: 5px -17px;
+    padding-top: 3px;
+    cursor: pointer;
+    font-size: 8pt;
+    transition: none;
+}
+
+.model-input {
+    white-space: nowrap;
+}
+
+.reloadModels {
+    background: var(--background-color2);
+    border: none;
+    padding: 0px 0px;
+}
+
+#reload-models.secondaryButton:hover {
+    background: var(--background-color2);
+}

ui/media/css/themes.css

@@ -27,9 +27,13 @@
     --input-border-size: 1px; 
     --accent-color: hsl(var(--accent-hue), 100%, var(--accent-lightness));
     --accent-color-hover: hsl(var(--accent-hue), 100%, var(--accent-lightness-hover));
+    --accent-text-color: rgb(255, 221, 255);
     --primary-button-border: none;
     --input-switch-padding: 1px;
     --input-height: 18px;
+	--tertiary-background-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (2 * var(--value-step))));
+    --tertiary-border-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (3 * var(--value-step))));
+    --tertiary-color: var(--input-text-color)
 
     /* Main theme color, hex color fallback. */
     --theme-color-fallback: #673AB6;
@@ -48,6 +52,11 @@
     --input-border-color: grey; 
 
     --theme-color-fallback: #aaaaaa;
+
+	--tertiary-background-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (16.8 * var(--value-step))));
+    --tertiary-border-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (12 * var(--value-step))));
+
+    --accent-text-color: white;
 }
 
 .theme-discord {
@@ -64,6 +73,10 @@
     --input-border-color: var(--input-background-color); 
 
     --theme-color-fallback: #202225;
+
+    --tertiary-background-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (3.5 * var(--value-step))));
+    --tertiary-border-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (4.5 * var(--value-step))));
+    --accent-text-color: white;
 }
 
 .theme-cool-blue {
@@ -81,6 +94,10 @@
     --accent-hue: 212;
 
     --theme-color-fallback: #0056b8;
+
+    --tertiary-background-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (3.5 * var(--value-step))));
+    --tertiary-border-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (4.5 * var(--value-step))));
+    --accent-text-color: #f7fbff;
 }
 
 
@@ -97,6 +114,9 @@
     --input-background-color: var(--background-color3);
 
     --theme-color-fallback: #5300b8;
+
+    --tertiary-background-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (3.5 * var(--value-step))));
+    --tertiary-border-color: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (4.5 * var(--value-step))));
 }
 
 .theme-super-dark {
@@ -131,6 +151,9 @@
     --input-background-color: hsl(222, var(--main-saturation), calc(var(--value-base) - (2 * var(--value-step))));
     --input-text-color: #FF0000;
     --input-border-color: #005E05;
+
+    --tertiary-color: white;
+    --accent-text-color: #f7fbff;
 }
 
 

ui/media/images/fa-eraser.svg

@@ -0,0 +1,4 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 576 576" width="24" height="24">
+<!--! Font Awesome Pro 6.2.1 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license (Commercial License) Copyright 2022 Fonticons, Inc.-->
+  <path style="filter: drop-shadow(0px 0px 20px white)" d="M290.7 57.4 57.4 290.7c-25 25-25 65.5 0 90.5l80 80c12 12 28.3 18.7 45.3 18.7H512c17.7 0 32-14.3 32-32s-14.3-32-32-32H387.9l130.7-130.6c25-25 25-65.5 0-90.5L381.3 57.4c-25-25-65.5-25-90.5 0zm6.7 358.6H182.6l-80-80 124.7-124.7 137.4 137.4-67.3 67.3z"/>
+</svg>

ui/media/images/fa-eye-dropper.svg

@@ -0,0 +1,4 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" width="24" height="24">
+<!--! Font Awesome Pro 6.2.1 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license (Commercial License) Copyright 2022 Fonticons, Inc.-->
+  <path style="filter: drop-shadow(0px 0px 20px white)" d="M341.6 29.2 240.1 130.8l-9.4-9.4c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3l160 160c12.5 12.5 32.8 12.5 45.3 0s12.5-32.8 0-45.3l-9.4-9.4 101.5-101.6c39-39 39-102.2 0-141.1s-102.2-39-141.1 0zM55.4 323.3c-15 15-23.4 35.4-23.4 56.6v42.4L5.4 462.2c-8.5 12.7-6.8 29.6 4 40.4s27.7 12.5 40.4 4L89.7 480h42.4c21.2 0 41.6-8.4 56.6-23.4l120.7-120.7-45.3-45.3-120.7 120.7c-3 3-7.1 4.7-11.3 4.7H96v-36.1c0-4.2 1.7-8.3 4.7-11.3l120.7-120.7-45.3-45.3L55.4 323.3z"/>
+</svg>

ui/media/images/fa-fill.svg

@@ -0,0 +1,4 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 576 576" width="24" height="24">
+<!--! Font Awesome Pro 6.2.1 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license (Commercial License) Copyright 2022 Fonticons, Inc.-->
+  <path style="filter: drop-shadow(0px 0px 20px white)" d="M118.6 9.4c-12.5-12.5-32.7-12.5-45.2 0s-12.5 32.8 0 45.3l81.3 81.3-92.1 92.1c-37.5 37.5-37.5 98.3 0 135.8l117.5 117.5c37.5 37.5 98.3 37.5 135.8 0l190.4-190.5c28.1-28.1 28.1-73.7 0-101.8L354.9 37.7c-28.1-28.1-73.7-28.1-101.8 0l-53.1 53-81.4-81.3zM200 181.3l49.4 49.4c12.5 12.5 32.8 12.5 45.3 0s12.5-32.8 0-45.3L245.3 136l53.1-53.1c3.1-3.1 8.2-3.1 11.3 0l151.4 151.4c3.1 3.1 3.1 8.2 0 11.3L418.7 288H99.5c1.4-5.4 4.2-10.4 8.4-14.6l92.1-92.1z"/>
+</svg>

ui/media/images/fa-pencil.svg

@@ -0,0 +1,4 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 512 512" width="24" height="24">
+<!--! Font Awesome Pro 6.2.1 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license (Commercial License) Copyright 2022 Fonticons, Inc.-->
+  <path style="filter: drop-shadow(0px 0px 20px white)" d="m410.3 231 11.3-11.3-33.9-33.9-62.1-62.1-33.9-33.9-11.3 11.3-22.6 22.6L58.6 322.9c-10.4 10.4-18 23.3-22.2 37.4L1 480.7c-2.5 8.4-.2 17.5 6.1 23.7s15.3 8.5 23.7 6.1l120.3-35.4c14.1-4.2 27-11.8 37.4-22.2l199.2-199.2 22.6-22.7zM160 399.4l-9.1 22.7c-4 3.1-8.5 5.4-13.3 6.9l-78.2 23 23-78.1c1.4-4.9 3.8-9.4 6.9-13.3l22.7-9.1v32c0 8.8 7.2 16 16 16h32zM362.7 18.7l-14.4 14.5-22.6 22.6-11.4 11.3 33.9 33.9 62.1 62.1 33.9 33.9 11.3-11.3 22.6-22.6 14.5-14.5c25-25 25-65.5 0-90.5l-39.3-39.4c-25-25-65.5-25-90.5 0zm-47.4 168-144 144c-6.2 6.2-16.4 6.2-22.6 0s-6.2-16.4 0-22.6l144-144c6.2-6.2 16.4-6.2 22.6 0s6.2 16.4 0 22.6z"/>
+</svg>

ui/media/js/auto-save.js

@@ -15,7 +15,7 @@ const SETTINGS_IDS_LIST = [
     "stable_diffusion_model",
     "vae_model",
     "hypernetwork_model",
-    "sampler",
+    "sampler_name",
     "width",
     "height",
     "num_inference_steps",
@@ -27,7 +27,10 @@ const SETTINGS_IDS_LIST = [
     "negative_prompt",
     "stream_image_progress",
     "use_face_correction",
+    "gfpgan_model",
     "use_upscale",
+    "upscale_amount",
+    "block_nsfw",
     "show_only_filtered_image",
     "upscale_model",
     "preview-image",
@@ -36,10 +39,14 @@ const SETTINGS_IDS_LIST = [
     "save_to_disk",
     "diskPath",
     "sound_toggle",
-    "turbo",
-    "use_full_precision",
+    "vram_usage_level",
     "confirm_dangerous_actions",
-    "auto_save_settings"
+    "metadata_output_format",
+    "auto_save_settings",
+    "apply_color_correction",
+    "process_order_toggle",
+    "thumbnail_size",
+    "auto_scroll"
 ]
 
 const IGNORE_BY_DEFAULT = [
@@ -89,6 +96,9 @@ async function initSettings() {
 }
 
 function getSetting(element) {
+    if (element.dataset && 'path' in element.dataset) {
+        return element.dataset.path
+    }
     if (typeof element === "string" || element instanceof String) {
         element = SETTINGS[element].element
     }
@@ -98,6 +108,10 @@ function getSetting(element) {
     return element.value
 }
 function setSetting(element, value) {
+    if (element.dataset && 'path' in element.dataset) {
+        element.dataset.path = value
+        return // no need to dispatch any event here because the models are not loaded yet
+    }
     if (typeof element === "string" || element instanceof String) {
         element = SETTINGS[element].element
     }
@@ -259,10 +273,12 @@ function tryLoadOldSettings() {
         var saved_settings = JSON.parse(saved_settings_text)
         Object.keys(saved_settings.should_save).forEach(key => {
             key = key in old_map ? old_map[key] : key
+            if (!(key in SETTINGS)) return
             SETTINGS[key].ignore = !saved_settings.should_save[key]
         });
         Object.keys(saved_settings.values).forEach(key => {
             key = key in old_map ? old_map[key] : key
+            if (!(key in SETTINGS)) return
             var setting = SETTINGS[key]
             if (!setting.ignore) {
                 setting.value = saved_settings.values[key]
@@ -277,8 +293,6 @@ function tryLoadOldSettings() {
         "soundEnabled": "sound_toggle",
         "saveToDisk": "save_to_disk",
         "useCPU": "use_cpu",
-        "useFullPrecision": "use_full_precision",
-        "useTurboMode": "turbo",
         "diskPath": "diskPath",
         "useFaceCorrection": "use_face_correction",
         "useUpscaling": "use_upscale",

ui/media/js/dnd.js

@@ -2,7 +2,7 @@
 
 const EXT_REGEX = /(?:\.([^.]+))?$/
 const TEXT_EXTENSIONS = ['txt', 'json']
-const IMAGE_EXTENSIONS = ['jpg', 'jpeg', 'png', 'bmp', 'tiff', 'tif', 'tga']
+const IMAGE_EXTENSIONS = ['jpg', 'jpeg', 'png', 'bmp', 'tiff', 'tif', 'tga', 'webp']
 
 function parseBoolean(stringValue) {
     if (typeof stringValue === 'boolean') {
@@ -25,6 +25,7 @@ function parseBoolean(stringValue) {
         case "no":
         case "off":
         case "0":
+        case "none":
         case null:
         case undefined:
           return false;
@@ -58,6 +59,13 @@ const TASK_MAPPING = {
         readUI: () => activeTags.map(x => x.name),
         parse: (val) => val
     },
+    inactive_tags: { name: "Inactive Image Modifiers",
+        setUI: (inactive_tags) => {
+            refreshInactiveTags(inactive_tags)
+        },
+        readUI: () => activeTags.filter(tag => tag.inactive === true).map(x => x.name),
+        parse: (val) => val
+    },
     width: { name: 'Width',
         setUI: (width) => {
             const oldVal = widthField.value
@@ -136,23 +144,43 @@ const TASK_MAPPING = {
         readUI: () => (maskSetting.checked ? imageInpainter.getImg() : undefined),
         parse: (val) => val
     },
-
+    preserve_init_image_color_profile: { name: 'Preserve Color Profile',
+        setUI: (preserve_init_image_color_profile) => {
+            applyColorCorrectionField.checked = parseBoolean(preserve_init_image_color_profile)
+        },
+        readUI: () => applyColorCorrectionField.checked,
+        parse: (val) => parseBoolean(val)
+    },
+    
     use_face_correction: { name: 'Use Face Correction',
         setUI: (use_face_correction) => {
-            useFaceCorrectionField.checked = parseBoolean(use_face_correction)
+            const oldVal = gfpganModelField.value
+            gfpganModelField.value = getModelPath(use_face_correction, ['.pth'])
+            if (gfpganModelField.value) { // Is a valid value for the field.
+                useFaceCorrectionField.checked = true
+                gfpganModelField.disabled = false
+            } else { // Not a valid value, restore the old value and disable the filter.
+                gfpganModelField.disabled = true
+                gfpganModelField.value = oldVal
+                useFaceCorrectionField.checked = false
+            }
+
+            //useFaceCorrectionField.checked = parseBoolean(use_face_correction)
         },
-        readUI: () => useFaceCorrectionField.checked,
-        parse: (val) => parseBoolean(val)
+        readUI: () => (useFaceCorrectionField.checked ? gfpganModelField.value : undefined),
+        parse: (val) => val
     },
     use_upscale: { name: 'Use Upscaling',
         setUI: (use_upscale) => {
             const oldVal = upscaleModelField.value
-            upscaleModelField.value = use_upscale
+            upscaleModelField.value = getModelPath(use_upscale, ['.pth'])
             if (upscaleModelField.value) { // Is a valid value for the field.
                 useUpscalingField.checked = true
                 upscaleModelField.disabled = false
+                upscaleAmountField.disabled = false
             } else { // Not a valid value, restore the old value and disable the filter.
                 upscaleModelField.disabled = true
+                upscaleAmountField.disabled = true
                 upscaleModelField.value = oldVal
                 useUpscalingField.checked = false
             }
@@ -160,9 +188,16 @@ const TASK_MAPPING = {
         readUI: () => (useUpscalingField.checked ? upscaleModelField.value : undefined),
         parse: (val) => val
     },
-    sampler: { name: 'Sampler',
-        setUI: (sampler) => {
-            samplerField.value = sampler
+    upscale_amount: { name: 'Upscale By',
+        setUI: (upscale_amount) => {
+            upscaleAmountField.value = upscale_amount
+        },
+        readUI: () => upscaleAmountField.value,
+        parse: (val) => val
+    },
+    sampler_name: { name: 'Sampler',
+        setUI: (sampler_name) => {
+            samplerField.value = sampler_name
         },
         readUI: () => samplerField.value,
         parse: (val) => val
@@ -171,7 +206,7 @@ const TASK_MAPPING = {
         setUI: (use_stable_diffusion_model) => {
             const oldVal = stableDiffusionModelField.value
 
-            use_stable_diffusion_model = getModelPath(use_stable_diffusion_model, ['.ckpt'])
+            use_stable_diffusion_model = getModelPath(use_stable_diffusion_model, ['.ckpt', '.safetensors'])
             stableDiffusionModelField.value = use_stable_diffusion_model
 
             if (!stableDiffusionModelField.value) {
@@ -184,6 +219,7 @@ const TASK_MAPPING = {
     use_vae_model: { name: 'VAE model',
         setUI: (use_vae_model) => {
             const oldVal = vaeModelField.value
+            use_vae_model = (use_vae_model === undefined || use_vae_model === null || use_vae_model === 'None' ? '' : use_vae_model)
 
             if (use_vae_model !== '') {
                 use_vae_model = getModelPath(use_vae_model, ['.vae.pt', '.ckpt'])
@@ -197,6 +233,7 @@ const TASK_MAPPING = {
     use_hypernetwork_model: { name: 'Hypernetwork model',
         setUI: (use_hypernetwork_model) => {
             const oldVal = hypernetworkModelField.value
+            use_hypernetwork_model = (use_hypernetwork_model === undefined || use_hypernetwork_model === null || use_hypernetwork_model === 'None' ? '' : use_hypernetwork_model)
 
             if (use_hypernetwork_model !== '') {
                 use_hypernetwork_model = getModelPath(use_hypernetwork_model, ['.pt'])
@@ -232,20 +269,6 @@ const TASK_MAPPING = {
         readUI: () => useCPUField.checked,
         parse: (val) => val
     },
-    turbo: { name: 'Turbo',
-        setUI: (turbo) => {
-            turboField.checked = turbo
-        },
-        readUI: () => turboField.checked,
-        parse: (val) => Boolean(val)
-    },
-    use_full_precision: { name: 'Use Full Precision',
-        setUI: (use_full_precision) => {
-            useFullPrecisionField.checked = use_full_precision
-        },
-        readUI: () => useFullPrecisionField.checked,
-        parse: (val) => Boolean(val)
-    },
 
     stream_image_progress: { name: 'Stream Image Progress',
         setUI: (stream_image_progress) => {
@@ -277,6 +300,7 @@ const TASK_MAPPING = {
         parse: (val) => val
     }
 }
+
 function restoreTaskToUI(task, fieldsToSkip) {
     fieldsToSkip = fieldsToSkip || []
 
@@ -296,29 +320,47 @@ function restoreTaskToUI(task, fieldsToSkip) {
         }
     }
 
-    // restore the original tag
-    promptField.value = task.reqBody.original_prompt || task.reqBody.prompt
+    // properly reset fields not present in the task
+    if (!('use_hypernetwork_model' in task.reqBody)) {
+        hypernetworkModelField.value = ""
+        hypernetworkModelField.dispatchEvent(new Event("change"))
+    }
 
+    // restore the original prompt if provided (e.g. use settings), fallback to prompt as needed (e.g. copy/paste or d&d)
+    promptField.value = task.reqBody.original_prompt
+    if (!('original_prompt' in task.reqBody)) {
+        promptField.value = task.reqBody.prompt
+    }
+    
     // properly reset checkboxes
     if (!('use_face_correction' in task.reqBody)) {
         useFaceCorrectionField.checked = false
+        gfpganModelField.disabled = true
     }
     if (!('use_upscale' in task.reqBody)) {
         useUpscalingField.checked = false
     }
-    if (!('mask' in task.reqBody)) {
+    if (!('mask' in task.reqBody) && maskSetting.checked) {
         maskSetting.checked = false
+        maskSetting.dispatchEvent(new Event("click"))
     }
     upscaleModelField.disabled = !useUpscalingField.checked
+    upscaleAmountField.disabled = !useUpscalingField.checked
 
-    // Show the source picture if present
-    initImagePreview.src = (task.reqBody.init_image == undefined ? '' : task.reqBody.init_image)
-    if (IMAGE_REGEX.test(initImagePreview.src)) {
-        if (Boolean(task.reqBody.mask)) {
-            setTimeout(() => { // add a delay to insure this happens AFTER the main image loads (which reloads the inpainter)
+    // hide/show source picture as needed
+    if (IMAGE_REGEX.test(initImagePreview.src) && task.reqBody.init_image == undefined) {
+        // hide source image
+        initImageClearBtn.dispatchEvent(new Event("click"))
+    }
+    else if (task.reqBody.init_image !== undefined) {
+        // listen for inpainter loading event, which happens AFTER the main image loads (which reloads the inpainter)
+        initImagePreview.addEventListener('load', function() {
+            if (Boolean(task.reqBody.mask)) {
                 imageInpainter.setImg(task.reqBody.mask)
-            }, 250)
-        }
+                maskSetting.checked = true
+            }
+        }, { once: true })
+        initImagePreview.src = task.reqBody.init_image
     }
 }
 function readUI() {
@@ -334,12 +376,19 @@ function readUI() {
 }
 function getModelPath(filename, extensions)
 {
-    let pathIdx = filename.lastIndexOf('/') // Linux, Mac paths
-    if (pathIdx < 0) {
-        pathIdx = filename.lastIndexOf('\\') // Windows paths.
+    if (typeof filename !== "string") {
+        return
+    }
+    
+    let pathIdx
+    if (filename.includes('/models/stable-diffusion/')) {
+        pathIdx = filename.indexOf('/models/stable-diffusion/') + 25 // Linux, Mac paths
+    }
+    else if (filename.includes('\\models\\stable-diffusion\\')) {
+        pathIdx = filename.indexOf('\\models\\stable-diffusion\\') + 25 // Linux, Mac paths
     }
     if (pathIdx >= 0) {
-        filename = filename.slice(pathIdx + 1)
+        filename = filename.slice(pathIdx)
     }
     extensions.forEach(ext => {
         if (filename.endsWith(ext)) {
@@ -350,6 +399,7 @@ function getModelPath(filename, extensions)
 }
 
 const TASK_TEXT_MAPPING = {
+    prompt: 'Prompt',
     width: 'Width',
     height: 'Height',
     seed: 'Seed',
@@ -358,26 +408,39 @@ const TASK_TEXT_MAPPING = {
     prompt_strength: 'Prompt Strength',
     use_face_correction: 'Use Face Correction',
     use_upscale: 'Use Upscaling',
-    sampler: 'Sampler',
+    upscale_amount: 'Upscale By',
+    sampler_name: 'Sampler',
     negative_prompt: 'Negative Prompt',
     use_stable_diffusion_model: 'Stable Diffusion model',
     use_hypernetwork_model: 'Hypernetwork model',
     hypernetwork_strength: 'Hypernetwork Strength'
 }
-const afterPromptRe = /^\s*Width\s*:\s*\d+\s*(?:\r\n|\r|\n)+\s*Height\s*:\s*\d+\s*(\r\n|\r|\n)+Seed\s*:\s*\d+\s*$/igm
 function parseTaskFromText(str) {
     const taskReqBody = {}
 
+    const lines = str.split('\n')
+    if (lines.length === 0) {
+        return
+    }
+
     // Prompt
-    afterPromptRe.lastIndex = 0
-    const match = afterPromptRe.exec(str)
-    if (match) {
-        let prompt = str.slice(0, match.index)
-        str = str.slice(prompt.length)
-        taskReqBody.prompt = prompt.trim()
+    let knownKeyOnFirstLine = false
+    for (let key in TASK_TEXT_MAPPING) {
+        if (lines[0].startsWith(TASK_TEXT_MAPPING[key] + ':')) {
+            knownKeyOnFirstLine = true
+            break
+        }
+    }
+    if (!knownKeyOnFirstLine) {
+        taskReqBody.prompt = lines[0]
         console.log('Prompt:', taskReqBody.prompt)
     }
+
     for (const key in TASK_TEXT_MAPPING) {
+        if (key in taskReqBody) {
+            continue
+        }
+
         const name = TASK_TEXT_MAPPING[key];
         let val = undefined
 
@@ -410,6 +473,9 @@ async function parseContent(text) {
     if (text.startsWith('{') && text.endsWith('}')) {
         try {
             const task = JSON.parse(text)
+            if (!('reqBody' in task)) { // support the format saved to the disk, by the UI
+                task.reqBody = Object.assign({}, task)
+            }
             restoreTaskToUI(task)
             return true
         } catch (e) {
@@ -419,7 +485,7 @@ async function parseContent(text) {
     }
     // Normal txt file.
     const task = parseTaskFromText(text)
-    if (task) {
+    if (text.toLowerCase().includes('seed:') && task) { // only parse valid task content
         restoreTaskToUI(task)
         return true
     } else {
@@ -467,7 +533,7 @@ function dragOverHandler(ev) {
     ev.dataTransfer.dropEffect = "copy"
 
     let img = new Image()
-    img.src = location.host + '/media/images/favicon-32x32.png'
+    img.src = '//' + location.host + '/media/images/favicon-32x32.png'
     ev.dataTransfer.setDragImage(img, 16, 16)
 }
 
@@ -476,8 +542,6 @@ document.addEventListener("dragover", dragOverHandler)
 
 const TASK_REQ_NO_EXPORT = [
     "use_cpu",
-    "turbo",
-    "use_full_precision",
     "save_to_disk_path"
 ]
 const resetSettings = document.getElementById('reset-image-settings')

ui/media/js/engine.js

@@ -8,7 +8,7 @@
     const SERVER_STATE_VALIDITY_DURATION = 90 * 1000 // ms - 90 seconds to allow ping to timeout more than once before killing tasks.
     const HEALTH_PING_INTERVAL = 5000 // ms
     const IDLE_COOLDOWN = 2500 // ms
-    const CONCURRENT_TASK_INTERVAL = 500 // ms
+    const CONCURRENT_TASK_INTERVAL = 100 // ms
 
     /** Connects to an endpoint and resumes connection after reaching end of stream until all data is received.
      * Allows closing the connection while the server buffers more data.
@@ -718,7 +718,7 @@
         "height": 'number',
         "seed": 'number',
 
-        "sampler": 'string',
+        "sampler_name": 'string',
         "use_stable_diffusion_model": 'string',
         "num_inference_steps": 'number',
         "guidance_scale": 'number',
@@ -727,13 +727,11 @@
         "stream_progress_updates": 'boolean',
         "stream_image_progress": 'boolean',
         "show_only_filtered_image": 'boolean',
-        "turbo": 'boolean',
-        "use_full_precision": 'boolean',
         "output_format": 'string',
         "output_quality": 'number',
     }
     const TASK_DEFAULTS = {
-        "sampler": "plms",
+        "sampler_name": "plms",
         "use_stable_diffusion_model": "sd-v1-4",
         "num_inference_steps": 50,
         "guidance_scale": 7.5,
@@ -743,8 +741,7 @@
         "stream_progress_updates": true,
         "stream_image_progress": true,
         "show_only_filtered_image": true,
-        "turbo": false,
-        "use_full_precision": false,
+        "block_nsfw": false,
         "output_format": "png",
         "output_quality": 75,
     }
@@ -839,6 +836,12 @@
          * @memberof Task
          */
         async post(timeout=-1) {
+            performance.mark('make-render-request')
+            if (performance.getEntriesByName('click-makeImage', 'mark').length > 0) {
+                performance.measure('diff', 'click-makeImage', 'make-render-request')
+                console.log('delay between clicking and making the server request:', performance.getEntriesByName('diff', 'measure')[0].duration + ' ms')
+            }
+
             let jsonResponse = await super.post('/render', timeout)
             if (typeof jsonResponse?.task !== 'number') {
                 console.warn('Endpoint error response: ', jsonResponse)
@@ -1106,9 +1109,9 @@
                 idleEventPromise = makeQuerablePromise(eventSource.fireEvent(EVENT_IDLE, {capacity: serverCapacity, idle: true}))
             }
             // Calling idle could result in task being added to queue.
-            if (task_queue.size <= 0 && concurrent_generators.size <= 0) {
-                return asyncDelay(IDLE_COOLDOWN).then(() => idleEventPromise)
-            }
+            // if (task_queue.size <= 0 && concurrent_generators.size <= 0) {
+            //     return asyncDelay(IDLE_COOLDOWN).then(() => idleEventPromise)
+            // }
         }
         if (task_queue.size < serverCapacity) {
             if (!idleEventPromise?.isPending) {

ui/media/js/image-editor.js

@@ -36,13 +36,14 @@ const defaultToolEnd = (editor, ctx, x, y, is_overlay = false) => {
 		ctx.clearRect(0, 0, editor.width, editor.height)
 	}
 }
+const toolDoNothing = (editor, ctx, x, y, is_overlay = false) => {}
 
 const IMAGE_EDITOR_TOOLS = [
 	{
 		id: "draw",
 		name: "Draw",
 		icon: "fa-solid fa-pencil",
-		cursor: "url(/media/images/fa-pencil.png) 0 24, pointer",
+		cursor: "url(/media/images/fa-pencil.svg) 0 24, pointer",
 		begin: defaultToolBegin,
 		move: defaultToolMove,
 		end: defaultToolEnd
@@ -51,7 +52,7 @@ const IMAGE_EDITOR_TOOLS = [
 		id: "erase",
 		name: "Erase",
 		icon: "fa-solid fa-eraser",
-		cursor: "url(/media/images/fa-eraser.png) 0 18, pointer",
+		cursor: "url(/media/images/fa-eraser.svg) 0 14, pointer",
 		begin: defaultToolBegin,
 		move: (editor, ctx, x, y, is_overlay = false) => {
 			ctx.lineTo(x, y)
@@ -78,27 +79,92 @@ const IMAGE_EDITOR_TOOLS = [
 		}
 	},
 	{
-		id: "colorpicker",
-		name: "Color Picker",
-		icon: "fa-solid fa-eye-dropper",
-		cursor: "url(/media/images/fa-eye-dropper.png) 0 24, pointer",
+		id: "fill",
+		name: "Fill",
+		icon: "fa-solid fa-fill",
+		cursor: "url(/media/images/fa-fill.svg) 20 6, pointer",
 		begin: (editor, ctx, x, y, is_overlay = false) => {
-			var img_rgb = editor.layers.background.ctx.getImageData(x, y, 1, 1).data
-			var drawn_rgb = editor.ctx_current.getImageData(x, y, 1, 1).data
-			var drawn_opacity = drawn_rgb[3] / 255
-			editor.custom_color_input.value = rgbToHex({ 
-				r: (drawn_rgb[0] * drawn_opacity) + (img_rgb[0] * (1 - drawn_opacity)),
-				g: (drawn_rgb[1] * drawn_opacity) + (img_rgb[1] * (1 - drawn_opacity)),
-				b: (drawn_rgb[2] * drawn_opacity) + (img_rgb[2] * (1 - drawn_opacity)),
-			})
-			editor.custom_color_input.dispatchEvent(new Event("change"))
+			if (!is_overlay) {
+				var color = hexToRgb(ctx.fillStyle)
+				color.a = parseInt(ctx.globalAlpha * 255) // layer.ctx.globalAlpha
+				flood_fill(editor, ctx, parseInt(x), parseInt(y), color)
+			}
 		},
-		move: (editor, ctx, x, y, is_overlay = false) => {},
-		end: (editor, ctx, x, y, is_overlay = false) => {}
+		move: toolDoNothing,
+		end: toolDoNothing
+	},
+	{
+		id: "colorpicker",
+		name: "Picker",
+		icon: "fa-solid fa-eye-dropper",
+		cursor: "url(/media/images/fa-eye-dropper.svg) 0 24, pointer",
+		begin: (editor, ctx, x, y, is_overlay = false) => {
+			if (!is_overlay) {
+				var img_rgb = editor.layers.background.ctx.getImageData(x, y, 1, 1).data
+				var drawn_rgb = editor.ctx_current.getImageData(x, y, 1, 1).data
+				var drawn_opacity = drawn_rgb[3] / 255
+				editor.custom_color_input.value = rgbToHex({ 
+					r: (drawn_rgb[0] * drawn_opacity) + (img_rgb[0] * (1 - drawn_opacity)),
+					g: (drawn_rgb[1] * drawn_opacity) + (img_rgb[1] * (1 - drawn_opacity)),
+					b: (drawn_rgb[2] * drawn_opacity) + (img_rgb[2] * (1 - drawn_opacity)),
+				})
+				editor.custom_color_input.dispatchEvent(new Event("change"))
+			}
+		},
+		move: toolDoNothing,
+		end: toolDoNothing
 	}
 ]
 
 const IMAGE_EDITOR_ACTIONS = [
+	{
+		id: "load_mask",
+		name: "Load mask from file",
+		className: "load_mask",
+		icon: "fa-regular fa-folder-open",
+		handler: (editor) => {
+			let el = document.createElement('input')
+			el.setAttribute("type", "file")
+			el.addEventListener("change", function() {
+				if (this.files.length === 0) {
+					return
+				}
+
+				let reader = new FileReader()
+				let file = this.files[0]
+
+				reader.addEventListener('load', function(event) {
+					let maskData = reader.result
+
+					editor.layers.drawing.ctx.clearRect(0, 0, editor.width, editor.height)
+					var image = new Image()
+					image.onload = () => {
+						editor.layers.drawing.ctx.drawImage(image, 0, 0, editor.width, editor.height)
+					}
+					image.src = maskData
+				})
+
+				if (file) {
+					reader.readAsDataURL(file)
+				}
+			})
+
+			el.click()
+		},
+		trackHistory: true
+	},
+	{
+		id: "fill_all",
+		name: "Fill all",
+		icon: "fa-solid fa-paint-roller",
+		handler: (editor) => {
+			editor.ctx_current.globalCompositeOperation = "source-over"
+			editor.ctx_current.rect(0, 0, editor.width, editor.height)
+			editor.ctx_current.fill()
+			editor.setBrush()
+		},
+		trackHistory: true
+	},
 	{
 		id: "clear",
 		name: "Clear",
@@ -214,8 +280,8 @@ var IMAGE_EDITOR_SECTIONS = [
 			var sub_element = document.createElement("div")
 			sub_element.style.background = `var(--background-color3)`
 			sub_element.style.filter = `blur(${blur_amount}px)`
-			sub_element.style.width = `${size - 4}px`
-			sub_element.style.height = `${size - 4}px`
+			sub_element.style.width = `${size - 2}px`
+			sub_element.style.height = `${size - 2}px`
 			sub_element.style['border-radius'] = `${size}px`
 			element.style.background = "none"
 			element.appendChild(sub_element)
@@ -427,6 +493,9 @@ class ImageEditor {
 			var element = document.createElement("div")
 			var icon = document.createElement("i")
 			element.className = "image-editor-button button"
+			if (action.className) {
+				element.className += " " + action.className
+			}
 			icon.className = action.icon
 			element.appendChild(icon)
 			element.append(action.name)
@@ -467,8 +536,8 @@ class ImageEditor {
 			width = (multiplier * width).toFixed()
 			height = (multiplier * height).toFixed()
 		}
-		this.width = width
-		this.height = height
+		this.width = parseInt(width)
+		this.height = parseInt(height)
 	
 		this.container.style.width = width + "px"
 		this.container.style.height = height + "px"
@@ -494,8 +563,10 @@ class ImageEditor {
 	}
 	setImage(url, width, height) {
 		this.setSize(width, height)
-		this.layers.drawing.ctx.clearRect(0, 0, this.width, this.height)
 		this.layers.background.ctx.clearRect(0, 0, this.width, this.height)
+		if (!(url && this.inpainter)) {
+			this.layers.drawing.ctx.clearRect(0, 0, this.width, this.height)
+		}
 		if (url) {
 			var image = new Image()
 			image.onload = () => { 
@@ -604,6 +675,9 @@ class ImageEditor {
 			if (event.key == "y" && event.ctrlKey) {
 				this.history.redo()
 			}
+			if (event.key === "Escape") {
+				this.hide()
+			}
 		}
 		
 		// dropper ctrl holding handler stuff
@@ -682,14 +756,6 @@ class ImageEditor {
 	}
 }
 
-function rgbToHex(rgb) {
-	function componentToHex(c) {
-		var hex = parseInt(c).toString(16)
-		return hex.length == 1 ? "0" + hex : hex
-	}
-	return "#" + componentToHex(rgb.r) + componentToHex(rgb.g) + componentToHex(rgb.b)
-}
-
 const imageEditor = new ImageEditor(document.getElementById("image-editor"))
 const imageInpainter = new ImageEditor(document.getElementById("image-inpainter"), true)
 
@@ -704,3 +770,107 @@ document.getElementById("init_image_button_inpaint").addEventListener("click", (
 })
 
 img2imgUnload() // no init image when the app starts
+
+
+function rgbToHex(rgb) {
+	function componentToHex(c) {
+		var hex = parseInt(c).toString(16)
+		return hex.length == 1 ? "0" + hex : hex
+	}
+	return "#" + componentToHex(rgb.r) + componentToHex(rgb.g) + componentToHex(rgb.b)
+}
+
+function hexToRgb(hex) {
+	var result = /^#?([a-f\d]{2})([a-f\d]{2})([a-f\d]{2})$/i.exec(hex);
+	return result ? {
+		r: parseInt(result[1], 16),
+		g: parseInt(result[2], 16),
+		b: parseInt(result[3], 16)
+	} : null;
+}
+
+function pixelCompare(int1, int2) {
+	return Math.abs(int1 - int2) < 4
+}
+
+// adapted from https://ben.akrin.com/canvas_fill/fill_04.html
+function flood_fill(editor, the_canvas_context, x, y, color) {
+	pixel_stack = [{x:x, y:y}] ;
+	pixels = the_canvas_context.getImageData( 0, 0, editor.width, editor.height ) ;
+	var linear_cords = ( y * editor.width + x ) * 4 ;
+	var original_color = {r:pixels.data[linear_cords],
+						g:pixels.data[linear_cords+1],
+						b:pixels.data[linear_cords+2],
+						a:pixels.data[linear_cords+3]} ;
+	
+	var opacity = color.a / 255;
+	var new_color = {
+		r: parseInt((color.r * opacity) + (original_color.r * (1 - opacity))),
+		g: parseInt((color.g * opacity) + (original_color.g * (1 - opacity))),
+		b: parseInt((color.b * opacity) + (original_color.b * (1 - opacity)))
+	}
+
+	if ((pixelCompare(new_color.r, original_color.r) &&
+		pixelCompare(new_color.g, original_color.g) &&
+		pixelCompare(new_color.b, original_color.b)))
+	{
+		return; // This color is already the color we want, so do nothing
+	}
+	var max_stack_size = editor.width * editor.height;
+	while( pixel_stack.length > 0 && pixel_stack.length < max_stack_size ) {
+		new_pixel = pixel_stack.shift() ;
+		x = new_pixel.x ;
+		y = new_pixel.y ;
+	
+		linear_cords = ( y * editor.width + x ) * 4 ;
+		while( y-->=0 &&
+			   (pixelCompare(pixels.data[linear_cords], original_color.r) &&
+				pixelCompare(pixels.data[linear_cords+1], original_color.g) &&
+				pixelCompare(pixels.data[linear_cords+2], original_color.b))) {
+			linear_cords -= editor.width * 4 ;
+		}
+		linear_cords += editor.width * 4 ;
+		y++ ;
+
+		var reached_left = false ;
+		var reached_right = false ;
+		while( y++<editor.height &&
+			   (pixelCompare(pixels.data[linear_cords], original_color.r) &&
+				pixelCompare(pixels.data[linear_cords+1], original_color.g) &&
+				pixelCompare(pixels.data[linear_cords+2], original_color.b))) {
+			pixels.data[linear_cords]   = new_color.r ;
+			pixels.data[linear_cords+1] = new_color.g ;
+			pixels.data[linear_cords+2] = new_color.b ;
+			pixels.data[linear_cords+3] = 255 ;
+
+			if( x>0 ) {
+				if( pixelCompare(pixels.data[linear_cords-4], original_color.r) &&
+					pixelCompare(pixels.data[linear_cords-4+1], original_color.g) &&
+					pixelCompare(pixels.data[linear_cords-4+2], original_color.b)) {
+					if( !reached_left ) {
+						pixel_stack.push( {x:x-1, y:y} ) ;
+						reached_left = true ;
+					}
+				} else if( reached_left ) {
+					reached_left = false ;
+				}
+			}
+		
+			if( x<editor.width-1 ) {
+				if( pixelCompare(pixels.data[linear_cords+4], original_color.r) &&
+					pixelCompare(pixels.data[linear_cords+4+1], original_color.g) &&
+					pixelCompare(pixels.data[linear_cords+4+2], original_color.b)) {
+					if( !reached_right ) {
+						pixel_stack.push( {x:x+1,y:y} ) ;
+						reached_right = true ;
+					}
+				} else if( reached_right ) {
+					reached_right = false ;
+				}
+			}
+			
+			linear_cords += editor.width * 4 ;
+		}
+	}
+	the_canvas_context.putImageData( pixels, 0, 0 ) ;
+}

ui/media/js/image-modifiers.js

@@ -16,7 +16,7 @@ const modifierThumbnailPath = 'media/modifier-thumbnails'
 const activeCardClass = 'modifier-card-active'
 const CUSTOM_MODIFIERS_KEY = "customModifiers"
 
-function createModifierCard(name, previews) {
+function createModifierCard(name, previews, removeBy) {
     const modifierCard = document.createElement('div')
     modifierCard.className = 'modifier-card'
     modifierCard.innerHTML = `
@@ -44,10 +44,10 @@ function createModifierCard(name, previews) {
     }
 
     const maxLabelLength = 30
-    const nameWithoutBy = name.replace('by ', '')
+    const cardLabel = removeBy ? name.replace('by ', '') : name
 
-    if(nameWithoutBy.length <= maxLabelLength) {
-        label.querySelector('p').innerText = nameWithoutBy
+    if(cardLabel.length <= maxLabelLength) {
+        label.querySelector('p').innerText = cardLabel
     } else {
         const tooltipText = document.createElement('span')
         tooltipText.className = 'tooltip-text'
@@ -56,13 +56,14 @@ function createModifierCard(name, previews) {
         label.classList.add('tooltip')
         label.appendChild(tooltipText)
 
-        label.querySelector('p').innerText = nameWithoutBy.substring(0, maxLabelLength) + '...'
+        label.querySelector('p').innerText = cardLabel.substring(0, maxLabelLength) + '...'
     }
+    label.querySelector('p').dataset.fullName = name // preserve the full name
 
     return modifierCard
 }
 
-function createModifierGroup(modifierGroup, initiallyExpanded) {
+function createModifierGroup(modifierGroup, initiallyExpanded, removeBy) {
     const title = modifierGroup.category
     const modifiers = modifierGroup.modifiers
 
@@ -79,9 +80,9 @@ function createModifierGroup(modifierGroup, initiallyExpanded) {
 
     modifiers.forEach(modObj => {
         const modifierName = modObj.modifier
-        const modifierPreviews = modObj?.previews?.map(preview => `${modifierThumbnailPath}/${preview.path}`)
+        const modifierPreviews = modObj?.previews?.map(preview => `${IMAGE_REGEX.test(preview.image) ? preview.image : modifierThumbnailPath + '/' + preview.path}`)
 
-        const modifierCard = createModifierCard(modifierName, modifierPreviews)
+        const modifierCard = createModifierCard(modifierName, modifierPreviews, removeBy)
 
         if(typeof modifierCard == 'object') {
             modifiersEl.appendChild(modifierCard)
@@ -104,6 +105,7 @@ function createModifierGroup(modifierGroup, initiallyExpanded) {
                 }
 
                 refreshTagsList()
+                document.dispatchEvent(new Event('refreshImageModifiers'))
             })
         }
     })
@@ -113,6 +115,7 @@ function createModifierGroup(modifierGroup, initiallyExpanded) {
     modifiersEl.appendChild(brk)
 
     let e = document.createElement('div')
+    e.className = 'modifier-category'
     e.appendChild(titleEl)
     e.appendChild(modifiersEl)
 
@@ -136,7 +139,7 @@ async function loadModifiers() {
             res.reverse()
 
             res.forEach((modifierGroup, idx) => {
-                createModifierGroup(modifierGroup, idx === res.length - 1)
+                createModifierGroup(modifierGroup, idx === res.length - 1, modifierGroup === 'Artist' ? true : false) // only remove "By " for artists
             })
 
             createCollapsibles(editorModifierEntries)
@@ -146,12 +149,13 @@ async function loadModifiers() {
     }
 
     loadCustomModifiers()
+    document.dispatchEvent(new Event('loadImageModifiers'))
 }
 
 function refreshModifiersState(newTags) {
     // clear existing modifiers
     document.querySelector('#editor-modifiers').querySelectorAll('.modifier-card').forEach(modifierCard => {
-        const modifierName = modifierCard.querySelector('.modifier-card-label').innerText
+        const modifierName = modifierCard.querySelector('.modifier-card-label p').dataset.fullName // pick the full modifier name
         if (activeTags.map(x => x.name).includes(modifierName)) {
             modifierCard.classList.remove(activeCardClass)
             modifierCard.querySelector('.modifier-card-image-overlay').innerText = '+'
@@ -163,13 +167,16 @@ function refreshModifiersState(newTags) {
     newTags.forEach(tag => {
         let found = false
         document.querySelector('#editor-modifiers').querySelectorAll('.modifier-card').forEach(modifierCard => {
-            const modifierName = modifierCard.querySelector('.modifier-card-label').innerText
-            if (tag == modifierName) {
+            const modifierName = modifierCard.querySelector('.modifier-card-label p').dataset.fullName
+            const shortModifierName = modifierCard.querySelector('.modifier-card-label p').innerText
+            if (trimModifiers(tag) == trimModifiers(modifierName)) {
                 // add modifier to active array
                 if (!activeTags.map(x => x.name).includes(tag)) { // only add each tag once even if several custom modifier cards share the same tag
+                    const imageModifierCard = modifierCard.cloneNode(true)
+                    imageModifierCard.querySelector('.modifier-card-label p').innerText = shortModifierName
                     activeTags.push({
                         'name': modifierName,
-                        'element': modifierCard.cloneNode(true),
+                        'element': imageModifierCard,
                         'originElement': modifierCard
                     })
                 }
@@ -179,7 +186,7 @@ function refreshModifiersState(newTags) {
             }
         })
         if (found == false) { // custom tag went missing, create one here
-            let modifierCard = createModifierCard(tag, undefined) // create a modifier card for the missing tag, no image
+            let modifierCard = createModifierCard(tag, undefined, false) // create a modifier card for the missing tag, no image
             
             modifierCard.addEventListener('click', () => {
                 if (activeTags.map(x => x.name).includes(tag)) {
@@ -202,6 +209,26 @@ function refreshModifiersState(newTags) {
     refreshTagsList()
 }
 
+function refreshInactiveTags(inactiveTags) {
+    // update inactive tags
+    if (inactiveTags !== undefined && inactiveTags.length > 0) {
+        activeTags.forEach (tag => {
+            if (inactiveTags.find(element => element === tag.name) !== undefined) {
+                tag.inactive = true
+            }
+        })
+    }
+    
+    // update cards
+    let overlays = document.querySelector('#editor-inputs-tags-list').querySelectorAll('.modifier-card-overlay')
+    overlays.forEach (i => {
+        let modifierName = i.parentElement.getElementsByClassName('modifier-card-label')[0].getElementsByTagName("p")[0].innerText
+        if (inactiveTags.find(element => element === modifierName) !== undefined) {
+            i.parentElement.classList.add('modifier-toggle-inactive')
+        }
+    })
+}
+
 function refreshTagsList() {
     editorModifierTagsList.innerHTML = ''
 
@@ -227,6 +254,7 @@ function refreshTagsList() {
                 activeTags.splice(idx, 1)
                 refreshTagsList()
             }
+            document.dispatchEvent(new Event('refreshImageModifiers'))
         })
     })
 

ui/media/js/main.js

@@ -26,22 +26,30 @@ let initImagePreview = document.querySelector("#init_image_preview")
 let initImageSizeBox = document.querySelector("#init_image_size_box")
 let maskImageSelector = document.querySelector("#mask")
 let maskImagePreview = document.querySelector("#mask_preview")
+let applyColorCorrectionField = document.querySelector('#apply_color_correction')
+let colorCorrectionSetting = document.querySelector('#apply_color_correction_setting')
 let promptStrengthSlider = document.querySelector('#prompt_strength_slider')
 let promptStrengthField = document.querySelector('#prompt_strength')
-let samplerField = document.querySelector('#sampler')
+let samplerField = document.querySelector('#sampler_name')
 let samplerSelectionContainer = document.querySelector("#samplerSelection")
 let useFaceCorrectionField = document.querySelector("#use_face_correction")
+let gfpganModelField = new ModelDropdown(document.querySelector("#gfpgan_model"), 'gfpgan')
 let useUpscalingField = document.querySelector("#use_upscale")
 let upscaleModelField = document.querySelector("#upscale_model")
-let stableDiffusionModelField = document.querySelector('#stable_diffusion_model')
-let vaeModelField = document.querySelector('#vae_model')
-let hypernetworkModelField = document.querySelector('#hypernetwork_model')
+let upscaleAmountField = document.querySelector("#upscale_amount")
+let stableDiffusionModelField = new ModelDropdown(document.querySelector('#stable_diffusion_model'), 'stable-diffusion')
+let vaeModelField = new ModelDropdown(document.querySelector('#vae_model'), 'vae', 'None')
+let hypernetworkModelField = new ModelDropdown(document.querySelector('#hypernetwork_model'), 'hypernetwork', 'None')
 let hypernetworkStrengthSlider = document.querySelector('#hypernetwork_strength_slider')
 let hypernetworkStrengthField = document.querySelector('#hypernetwork_strength')
 let outputFormatField = document.querySelector('#output_format')
+let blockNSFWField = document.querySelector('#block_nsfw')
 let showOnlyFilteredImageField = document.querySelector("#show_only_filtered_image")
 let updateBranchLabel = document.querySelector("#updateBranchLabel")
 let streamImageProgressField = document.querySelector("#stream_image_progress")
+let thumbnailSizeField = document.querySelector("#thumbnail_size-input")
+let autoscrollBtn = document.querySelector("#auto_scroll_btn")
+let autoScroll = document.querySelector("#auto_scroll")
 
 let makeImageBtn = document.querySelector('#makeImage')
 let stopImageBtn = document.querySelector('#stopImage')
@@ -57,12 +65,14 @@ let promptStrengthContainer = document.querySelector('#prompt_strength_container
 let initialText = document.querySelector("#initial-text")
 let previewTools = document.querySelector("#preview-tools")
 let clearAllPreviewsBtn = document.querySelector("#clear-all-previews")
+let saveAllImagesBtn = document.querySelector("#save-all-images")
 
 let maskSetting = document.querySelector('#enable_mask')
 
 const processOrder = document.querySelector('#process_order_toggle')
 
 let imagePreview = document.querySelector("#preview")
+let imagePreviewContent = document.querySelector("#preview-content")
 imagePreview.addEventListener('drop', function(ev) {
     const data = ev.dataTransfer?.getData("text/plain");
     if (!data) {
@@ -74,7 +84,7 @@ imagePreview.addEventListener('drop', function(ev) {
     }
     ev.preventDefault()
     let moveTarget = ev.target
-    while (moveTarget && typeof moveTarget === 'object' && moveTarget.parentNode !== imagePreview) {
+    while (moveTarget && typeof moveTarget === 'object' && moveTarget.parentNode !== imagePreviewContent) {
         moveTarget = moveTarget.parentNode
     }
     if (moveTarget === initialText || moveTarget === previewTools) {
@@ -84,17 +94,17 @@ imagePreview.addEventListener('drop', function(ev) {
         return
     }
     if (moveTarget) {
-        const childs = Array.from(imagePreview.children)
+        const childs = Array.from(imagePreviewContent.children)
         if (moveTarget.nextSibling && childs.indexOf(movedTask) < childs.indexOf(moveTarget)) {
             // Move after the target if lower than current position.
             moveTarget = moveTarget.nextSibling
         }
     }
-    const newNode = imagePreview.insertBefore(movedTask, moveTarget || previewTools.nextSibling)
+    const newNode = imagePreviewContent.insertBefore(movedTask, moveTarget || previewTools.nextSibling)
     if (newNode === movedTask) {
         return
     }
-    imagePreview.removeChild(movedTask)
+    imagePreviewContent.removeChild(movedTask)
     const task = htmlTaskMap.get(movedTask)
     if (task) {
         htmlTaskMap.delete(movedTask)
@@ -260,9 +270,27 @@ function showImages(reqBody, res, outputContainer, livePreview) {
                     <div class="imgItemInfo">
                         <span class="imgSeedLabel"></span>
                     </div>
+                    <button class="imgPreviewItemClearBtn image_clear_btn"><i class="fa-solid fa-xmark"></i></button>
+                    <span class="img_bottom_label"></span>
                 </div>
             `
             outputContainer.appendChild(imageItemElem)
+            const imageRemoveBtn = imageItemElem.querySelector('.imgPreviewItemClearBtn')
+            let parentTaskContainer = imageRemoveBtn.closest('.imageTaskContainer')
+            imageRemoveBtn.addEventListener('click', (e) => {
+                shiftOrConfirm(e, "Remove the image from the results?", () => { 
+                    imageItemElem.style.display = 'none' 
+                    let allHidden = true;
+                    let children = parentTaskContainer.querySelectorAll('.imgItem');
+                    for(let x = 0; x < children.length; x++) {
+                        let child = children[x];
+                        if(child.style.display != "none") {
+                            allHidden = false;
+                        }
+                    }
+                    if(allHidden === true) {parentTaskContainer.classList.add("displayNone")}
+                })
+            })
         }
         const imageElem = imageItemElem.querySelector('img')
         imageElem.src = imageData
@@ -272,6 +300,10 @@ function showImages(reqBody, res, outputContainer, livePreview) {
         imageElem.setAttribute('data-steps', imageInferenceSteps)
         imageElem.setAttribute('data-guidance', imageGuidanceScale)
 
+        imageElem.addEventListener('load', function() {
+            imageItemElem.querySelector('.img_bottom_label').innerText = `${this.naturalWidth} x ${this.naturalHeight}`
+        })
+
 
         const imageInfo = imageItemElem.querySelector('.imgItemInfo')
         imageInfo.style.visibility = (livePreview ? 'hidden' : 'visible')
@@ -302,9 +334,12 @@ function showImages(reqBody, res, outputContainer, livePreview) {
                 const newButton = document.createElement('button')
                 newButton.classList.add('tasksBtns')
                 newButton.innerText = btnInfo.text
-                newButton.addEventListener('click', function() {
-                    btnInfo.on_click(req, img)
+                newButton.addEventListener('click', function(event) {
+                    btnInfo.on_click(req, img, event)
                 })
+                if (btnInfo.class !== undefined) {
+                   newButton.classList.add(btnInfo.class)
+                }
                 imgItemInfo.appendChild(newButton)
             }
             buttons.forEach(btn => {
@@ -401,7 +436,7 @@ function onUpscaleClick(req, img) {
 
 function onFixFacesClick(req, img) {
     enqueueImageVariationTask(req, img, {
-        use_face_correction: 'GFPGANv1.3'
+        use_face_correction: gfpganModelField.value
     })
 }
 
@@ -429,6 +464,10 @@ function getUncompletedTaskEntries() {
 }
 
 function makeImage() {
+    if (typeof performance == "object" && performance.mark) {
+        performance.mark('click-makeImage')
+    }
+
     if (!SD.isServerAvailable()) {
         alert('The server is not available.')
         return
@@ -609,7 +648,7 @@ function onTaskCompleted(task, reqBody, instance, outputContainer, stepUpdate) {
                             <b>Suggestions</b>:
                             <br/>
                             1. If you have set an initial image, please try reducing its dimension to ${MAX_INIT_IMAGE_DIMENSION}x${MAX_INIT_IMAGE_DIMENSION} or smaller.<br/>
-                            2. Try disabling the '<em>Turbo mode</em>' under '<em>Advanced Settings</em>'.<br/>
+                            2. Try picking a lower level in the '<em>GPU Memory Usage</em>' setting (in the '<em>Settings</em>' tab).<br/>
                             3. Try generating a smaller image.<br/>`
                 }
             } else {
@@ -643,7 +682,7 @@ function onTaskCompleted(task, reqBody, instance, outputContainer, stepUpdate) {
         task.progressBar.classList.remove("active")
         setStatus('request', 'done', 'success')
     } else {
-        task.outputMsg.innerText += `Task ended after ${time}`
+        task.outputMsg.innerText += `. Task ended after ${time}`
     }
 
     if (randomSeedField.checked) {
@@ -658,6 +697,9 @@ function onTaskCompleted(task, reqBody, instance, outputContainer, stepUpdate) {
         return
     }
 
+    if (pauseClient) { 
+        resumeBtn.click() 
+    }
     renderButtons.style.display = 'none'
     renameMakeImageButton()
 
@@ -687,12 +729,18 @@ async function onTaskStart(task) {
     if (task.batchCount > 1) {
         // Each output render batch needs it's own task reqBody instance to avoid altering the other runs after they are completed.
         newTaskReqBody = Object.assign({}, task.reqBody)
+        if (task.batchesDone == task.batchCount-1) { 
+            // Last batch of the task
+            // If the number of parallel jobs is no factor of the total number of images, the last batch must create less than "parallel jobs count" images
+            // E.g. with numOutputsTotal = 6 and num_outputs = 5, the last batch shall only generate 1 image.
+            newTaskReqBody.num_outputs = task.numOutputsTotal - task.reqBody.num_outputs * (task.batchCount-1)
+        }
     }
 
     const startSeed = task.seed || newTaskReqBody.seed
     const genSeeds = Boolean(typeof newTaskReqBody.seed !== 'number' || (newTaskReqBody.seed === task.seed && task.numOutputsTotal > 1))
     if (genSeeds) {
-        newTaskReqBody.seed = parseInt(startSeed) + (task.batchesDone * newTaskReqBody.num_outputs)
+        newTaskReqBody.seed = parseInt(startSeed) + (task.batchesDone * task.reqBody.num_outputs)
     }
 
     // Update the seed *before* starting the processing so it's retained if user stops the task
@@ -755,7 +803,10 @@ function createInitImageHover(taskEntry) {
     img.src = taskEntry.querySelector('div.task-initimg > img').src
     $tooltip.append(img)
     $tooltip.append(`<div class="top-right"><button>Use as Input</button></div>`)
-    $tooltip.find('button').on('click', (e) => { onUseAsInputClick(null,img) } )
+    $tooltip.find('button').on('click', (e) => {
+        e.stopPropagation()
+        onUseAsInputClick(null,img) 
+    })
 }
 
 let startX, startY;
@@ -785,10 +836,11 @@ function createTask(task) {
 
     if (task.reqBody.init_image !== undefined) {
         let h = 80
-	let w = task.reqBody.width * h / task.reqBody.height >>0
+        let w = task.reqBody.width * h / task.reqBody.height >>0
         taskConfig += `<div class="task-initimg" style="float:left;"><img style="width:${w}px;height:${h}px;" src="${task.reqBody.init_image}"><div class="task-fs-initimage"></div></div>`
     }
-    taskConfig += `<b>Seed:</b> ${task.seed}, <b>Sampler:</b> ${task.reqBody.sampler}, <b>Inference Steps:</b> ${task.reqBody.num_inference_steps}, <b>Guidance Scale:</b> ${task.reqBody.guidance_scale}, <b>Model:</b> ${task.reqBody.use_stable_diffusion_model}`
+    taskConfig += `<b>Seed:</b> ${task.seed}, <b>Sampler:</b> ${task.reqBody.sampler_name}, <b>Inference Steps:</b> ${task.reqBody.num_inference_steps}, <b>Guidance Scale:</b> ${task.reqBody.guidance_scale}, <b>Model:</b> ${task.reqBody.use_stable_diffusion_model}`
+
     if (task.reqBody.use_vae_model.trim() !== '') {
         taskConfig += `, <b>VAE:</b> ${task.reqBody.use_vae_model}`
     }
@@ -802,12 +854,15 @@ function createTask(task) {
         taskConfig += `, <b>Fix Faces:</b> ${task.reqBody.use_face_correction}`
     }
     if (task.reqBody.use_upscale) {
-        taskConfig += `, <b>Upscale:</b> ${task.reqBody.use_upscale}`
+        taskConfig += `, <b>Upscale:</b> ${task.reqBody.use_upscale} (${task.reqBody.upscale_amount || 4}x)`
     }
     if (task.reqBody.use_hypernetwork_model) {
         taskConfig += `, <b>Hypernetwork:</b> ${task.reqBody.use_hypernetwork_model}`
         taskConfig += `, <b>Hypernetwork Strength:</b> ${task.reqBody.hypernetwork_strength}`
     }
+    if (task.reqBody.preserve_init_image_color_profile) {
+        taskConfig += `, <b>Preserve Color Profile:</b> true`
+    }
 
     let taskEntry = document.createElement('div')
     taskEntry.id = `imageTaskContainer-${Date.now()}`
@@ -816,7 +871,7 @@ function createTask(task) {
                                 <i class="drag-handle fa-solid fa-grip"></i>
                                 <div class="taskStatusLabel">Enqueued</div>
                                 <button class="secondaryButton stopTask"><i class="fa-solid fa-trash-can"></i> Remove</button>
-                                <button class="secondaryButton useSettings"><i class="fa-solid fa-redo"></i> Use these settings</button>
+                                <button class="tertiaryButton useSettings"><i class="fa-solid fa-redo"></i> Use these settings</button>
                                 <div class="preview-prompt"></div>
                                 <div class="taskConfig">${taskConfig}</div>
                                 <div class="outputMsg"></div>
@@ -827,12 +882,20 @@ function createTask(task) {
                             </div>`
 
     createCollapsibles(taskEntry)
-    
+
     let draghandle = taskEntry.querySelector('.drag-handle')
-    draghandle.addEventListener('mousedown', (e) => { taskEntry.setAttribute('draggable',true)})
-    draghandle.addEventListener('mouseup',   (e) => { taskEntry.setAttribute('draggable',false)})
-    taskEntry.addEventListener('dragend',    (e) => { 
-        taskEntry.setAttribute('draggable',false);
+    draghandle.addEventListener('mousedown', (e) => {
+        taskEntry.setAttribute('draggable', true)
+    })
+    // Add a debounce delay to allow mobile to bouble tap.
+    draghandle.addEventListener('mouseup', debounce((e) => {
+        taskEntry.setAttribute('draggable', false)
+    }, 2000))
+    draghandle.addEventListener('click', (e) => {
+        e.preventDefault() // Don't allow the results to be collapsed...
+    })
+    taskEntry.addEventListener('dragend', (e) => {
+        taskEntry.setAttribute('draggable', false);
         imagePreview.querySelectorAll(".imageTaskContainer").forEach(itc => {
             itc.classList.remove('dropTargetBefore','dropTargetAfter');
         });
@@ -845,7 +908,6 @@ function createTask(task) {
         startY = e.target.closest(".imageTaskContainer").offsetTop;
     })
 
-
     if (task.reqBody.init_image !== undefined) {
         createInitImageHover(taskEntry)
     }
@@ -876,14 +938,14 @@ function createTask(task) {
     })
 
     task.isProcessing = true
-    taskEntry = imagePreview.insertBefore(taskEntry, previewTools.nextSibling)
+    taskEntry = imagePreviewContent.insertBefore(taskEntry, previewTools.nextSibling)
     htmlTaskMap.set(taskEntry, task)
 
     task.previewPrompt.innerText = task.reqBody.prompt
     if (task.previewPrompt.innerText.trim() === '') {
         task.previewPrompt.innerHTML = '&nbsp;' // allows the results to be collapsed
     }
-
+    return taskEntry.id
 }
 
 function getCurrentUserRequest() {
@@ -899,6 +961,7 @@ function getCurrentUserRequest() {
 
         reqBody: {
             seed,
+            used_random_seed: randomSeedField.checked,
             negative_prompt: negativePromptField.value.trim(),
             num_outputs: numOutputsParallel,
             num_inference_steps: parseInt(numInferenceStepsField.value),
@@ -906,18 +969,20 @@ function getCurrentUserRequest() {
             width: parseInt(widthField.value),
             height: parseInt(heightField.value),
             // allow_nsfw: allowNSFWField.checked,
-            turbo: turboField.checked,
+            vram_usage_level: vramUsageLevelField.value,
             //render_device: undefined, // Set device affinity. Prefer this device, but wont activate.
-            use_full_precision: useFullPrecisionField.checked,
             use_stable_diffusion_model: stableDiffusionModelField.value,
             use_vae_model: vaeModelField.value,
             stream_progress_updates: true,
             stream_image_progress: (numOutputsTotal > 50 ? false : streamImageProgressField.checked),
             show_only_filtered_image: showOnlyFilteredImageField.checked,
+            block_nsfw: blockNSFWField.checked,
             output_format: outputFormatField.value,
             output_quality: parseInt(outputQualityField.value),
+            metadata_output_format: metadataOutputFormatField.value,
             original_prompt: promptField.value,
-            active_tags: (activeTags.map(x => x.name))
+            active_tags: (activeTags.map(x => x.name)),
+            inactive_tags: (activeTags.filter(tag => tag.inactive === true).map(x => x.name))
         }
     }
     if (IMAGE_REGEX.test(initImagePreview.src)) {
@@ -930,18 +995,20 @@ function getCurrentUserRequest() {
         if (maskSetting.checked) {
             newTask.reqBody.mask = imageInpainter.getImg()
         }
-        newTask.reqBody.sampler = 'ddim'
+        newTask.reqBody.preserve_init_image_color_profile = applyColorCorrectionField.checked
+        newTask.reqBody.sampler_name = 'ddim'
     } else {
-        newTask.reqBody.sampler = samplerField.value
+        newTask.reqBody.sampler_name = samplerField.value
     }
     if (saveToDiskField.checked && diskPathField.value.trim() !== '') {
         newTask.reqBody.save_to_disk_path = diskPathField.value.trim()
     }
     if (useFaceCorrectionField.checked) {
-        newTask.reqBody.use_face_correction = 'GFPGANv1.3'
+        newTask.reqBody.use_face_correction = gfpganModelField.value
     }
     if (useUpscalingField.checked) {
         newTask.reqBody.use_upscale = upscaleModelField.value
+        newTask.reqBody.upscale_amount = upscaleAmountField.value
     }
     if (hypernetworkModelField.value) {
         newTask.reqBody.use_hypernetwork_model = hypernetworkModelField.value
@@ -982,6 +1049,8 @@ function getPrompts(prompts) {
     promptsToMake = applyPermuteOperator(promptsToMake)
     promptsToMake = applySetOperator(promptsToMake)
 
+    PLUGINS['GET_PROMPTS_HOOK'].forEach(fn => { promptsToMake = fn(promptsToMake) })
+
     return promptsToMake
 }
 
@@ -1066,7 +1135,7 @@ function createFileName(prompt, seed, steps, guidance, outputFormat) {
     // fileName += `${tagString}`
 
     // add the file extension
-    fileName += '.' + (outputFormat === 'png' ? 'png' : 'jpeg')
+    fileName += '.' + outputFormat
 
     return fileName
 }
@@ -1101,6 +1170,20 @@ clearAllPreviewsBtn.addEventListener('click', (e) => { shiftOrConfirm(e, "Clear
     taskEntries.forEach(removeTask)
 })})
 
+saveAllImagesBtn.addEventListener('click', (e) => {
+    let i = 0
+    document.querySelectorAll(".imageTaskContainer").forEach(container => {
+        let req = htmlTaskMap.get(container)
+        container.querySelectorAll(".imgContainer img").forEach(img => {
+            if (img.closest('.imgItem').style.display === 'none') {
+                return
+            }
+            setTimeout(() => {onDownloadImageClick(req, img)}, i*200)
+            i = i+1
+        })
+    })
+})
+
 stopImageBtn.addEventListener('click', (e) => { shiftOrConfirm(e, "Stop all the tasks?", async function(e) {
     await stopAllTasks()
 })})
@@ -1109,7 +1192,7 @@ widthField.addEventListener('change', onDimensionChange)
 heightField.addEventListener('change', onDimensionChange)
 
 function renameMakeImageButton() {
-    let totalImages = Math.max(parseInt(numOutputsTotalField.value), parseInt(numOutputsParallelField.value))
+    let totalImages = Math.max(parseInt(numOutputsTotalField.value), parseInt(numOutputsParallelField.value)) * getPrompts().length
     let imageLabel = 'Image'
     if (totalImages > 1) {
         imageLabel = totalImages + ' Images'
@@ -1135,10 +1218,18 @@ function onDimensionChange() {
 }
 
 diskPathField.disabled = !saveToDiskField.checked
+metadataOutputFormatField.disabled = !saveToDiskField.checked
+
+gfpganModelField.disabled = !useFaceCorrectionField.checked
+useFaceCorrectionField.addEventListener('change', function(e) {
+    gfpganModelField.disabled = !this.checked
+})
 
 upscaleModelField.disabled = !useUpscalingField.checked
+upscaleAmountField.disabled = !useUpscalingField.checked
 useUpscalingField.addEventListener('change', function(e) {
     upscaleModelField.disabled = !this.checked
+    upscaleAmountField.disabled = !this.checked
 })
 
 makeImageBtn.addEventListener('click', makeImage)
@@ -1219,7 +1310,7 @@ function updateHypernetworkStrengthContainer() {
 hypernetworkModelField.addEventListener('change', updateHypernetworkStrengthContainer)
 updateHypernetworkStrengthContainer()
 
-/********************* JPEG Quality **********************/
+/********************* JPEG/WEBP Quality **********************/
 function updateOutputQuality() {
     outputQualityField.value =  0 | outputQualitySlider.value
     outputQualityField.dispatchEvent(new Event("change"))
@@ -1241,71 +1332,43 @@ outputQualityField.addEventListener('input', debounce(updateOutputQualitySlider,
 updateOutputQuality()
 
 outputFormatField.addEventListener('change', e => {
-    if (outputFormatField.value == 'jpeg') {
-        outputQualityRow.style.display='table-row'
-    } else {
+    if (outputFormatField.value === 'png') {
         outputQualityRow.style.display='none'
+    } else {
+        outputQualityRow.style.display='table-row'
     }
 })
-
-async function getModels() {
-    try {
-        const sd_model_setting_key = "stable_diffusion_model"
-        const vae_model_setting_key = "vae_model"
-        const hypernetwork_model_key = "hypernetwork_model"
-        const selectedSDModel = SETTINGS[sd_model_setting_key].value
-        const selectedVaeModel = SETTINGS[vae_model_setting_key].value
-        const selectedHypernetworkModel = SETTINGS[hypernetwork_model_key].value
-
-        const models = await SD.getModels()
-        const modelsOptions = models['options']
-        if ("scan-error" in models) {
-            // let previewPane = document.getElementById('tab-content-wrapper')
-            let previewPane = document.getElementById('preview')
-            previewPane.style.background="red"
-            previewPane.style.textAlign="center"
-            previewPane.innerHTML = '<H1>🔥Malware alert!🔥</H1><h2>The file <i>' + models['scan-error'] + '</i> in your <tt>models/stable-diffusion</tt> folder is probably malware infected.</h2><h2>Please delete this file from the folder before proceeding!</h2>After deleting the file, reload this page.<br><br><button onClick="window.location.reload();">Reload Page</button>'
-            makeImageBtn.disabled = true
-        }
-
-        const stableDiffusionOptions = modelsOptions['stable-diffusion']
-        const vaeOptions = modelsOptions['vae']
-        const hypernetworkOptions = modelsOptions['hypernetwork']
-
-        vaeOptions.unshift('') // add a None option
-        hypernetworkOptions.unshift('') // add a None option
-
-        function createModelOptions(modelField, selectedModel) {
-            return function(modelName) {
-                const modelOption = document.createElement('option')
-                modelOption.value = modelName
-                modelOption.innerText = modelName !== '' ? modelName : 'None'
-
-                if (modelName === selectedModel) {
-                    modelOption.selected = true
+/********************* Zoom Slider **********************/
+thumbnailSizeField.addEventListener('change', () => {
+    (function (s) {
+        for (var j =0; j < document.styleSheets.length; j++) {
+            let cssSheet = document.styleSheets[j]
+            for (var i = 0; i < cssSheet.cssRules.length; i++) {
+                var rule = cssSheet.cssRules[i];
+                if (rule.selectorText == "div.img-preview img") {
+                  rule.style['max-height'] = s+'vh';
+                  rule.style['max-width'] = s+'vw';
+                  return;
                 }
-
-                modelField.appendChild(modelOption)
             }
         }
+    })(thumbnailSizeField.value)
+})
 
-        stableDiffusionOptions.forEach(createModelOptions(stableDiffusionModelField, selectedSDModel))
-        vaeOptions.forEach(createModelOptions(vaeModelField, selectedVaeModel))
-        hypernetworkOptions.forEach(createModelOptions(hypernetworkModelField, selectedHypernetworkModel))
-
-        stableDiffusionModelField.dispatchEvent(new Event('change'))
-        vaeModelField.dispatchEvent(new Event('change'))
-        hypernetworkModelField.dispatchEvent(new Event('change'))
-
-        // TODO: set default for model here too
-        SETTINGS[sd_model_setting_key].default = stableDiffusionOptions[0]
-        if (getSetting(sd_model_setting_key) == '' || SETTINGS[sd_model_setting_key].value == '') {
-            setSetting(sd_model_setting_key, stableDiffusionOptions[0])
-        }
-    } catch (e) {
-        console.log('get models error', e)
+function onAutoScrollUpdate() {
+    if (autoScroll.checked) {
+        autoscrollBtn.classList.add('pressed')
+    } else {
+        autoscrollBtn.classList.remove('pressed')
     }
+    autoscrollBtn.querySelector(".state").innerHTML = (autoScroll.checked ? "ON" : "OFF")
 }
+autoscrollBtn.addEventListener('click', function() {
+    autoScroll.checked = !autoScroll.checked
+    autoScroll.dispatchEvent(new Event("change"))
+    onAutoScrollUpdate()
+})
+autoScroll.addEventListener('change', onAutoScrollUpdate)
 
 function checkRandomSeed() {
     if (randomSeedField.checked) {
@@ -1341,6 +1404,7 @@ function img2imgLoad() {
     promptStrengthContainer.style.display = 'table-row'
     samplerSelectionContainer.style.display = "none"
     initImagePreviewContainer.classList.add("has-image")
+    colorCorrectionSetting.style.display = ''
 
     initImageSizeBox.textContent = initImagePreview.naturalWidth + " x " + initImagePreview.naturalHeight
     imageEditor.setImage(this.src, initImagePreview.naturalWidth, initImagePreview.naturalHeight)
@@ -1355,6 +1419,7 @@ function img2imgUnload() {
     promptStrengthContainer.style.display = "none"
     samplerSelectionContainer.style.display = ""
     initImagePreviewContainer.classList.remove("has-image")
+    colorCorrectionSetting.style.display = 'none'
     imageEditor.setImage(null, parseInt(widthField.value), parseInt(heightField.value))
 
 }
@@ -1406,7 +1471,7 @@ function selectTab(tab_id) {
     let tabInfo = tabElements.find(t => t.tab.id == tab_id)
     if (!tabInfo.tab.classList.contains("active")) {
         tabElements.forEach(info => {
-            if (info.tab.classList.contains("active")) {
+            if (info.tab.classList.contains("active") && info.tab.parentNode === tabInfo.tab.parentNode) {
                 info.tab.classList.toggle("active")
                 info.content.classList.toggle("active")
             }
@@ -1427,6 +1492,9 @@ function linkTabContents(tab) {
 
     tab.addEventListener("click", event => selectTab(tab.id))
 }
+function isTabActive(tab) {
+    return tab.classList.contains("active")
+}
 
 let pauseClient = false
 
@@ -1444,6 +1512,9 @@ function resumeClient() {
     })
 }
 
+promptField.addEventListener("input", debounce( renameMakeImageButton, 1000) )
+
+
 pauseBtn.addEventListener("click", function () {
     pauseClient = true
     pauseBtn.style.display="none"
@@ -1476,3 +1547,7 @@ window.addEventListener("beforeunload", function(e) {
 
 createCollapsibles()
 prettifyInputs(document);
+
+// set the textbox as focused on start
+promptField.focus()
+promptField.selectionStart = promptField.value.length

ui/media/js/parameters.js

@@ -7,6 +7,7 @@
     checkbox: "checkbox",
     select: "select",
     select_multiple: "select_multiple",
+    slider: "slider",
     custom: "custom",
 };
 
@@ -53,6 +54,39 @@ var PARAMETERS = [
             return `<input id="${parameter.id}" name="${parameter.id}" size="30" disabled>`
         }
     },
+    {
+        id: "metadata_output_format",
+        type: ParameterType.select,
+        label: "Metadata format",
+        note: "will be saved to disk in this format",
+        default: "txt",
+        options: [
+            {
+                value: "none",
+                label: "none"
+            },
+            {
+                value: "txt",
+                label: "txt"
+            },
+            {
+                value: "json",
+                label: "json"
+            },
+            {
+                value: "embed",
+                label: "embed"
+            }
+        ],
+    },
+    {
+        id: "block_nsfw",
+        type: ParameterType.checkbox,
+        label: "Block NSFW images",
+        note: "blurs out NSFW images",
+        icon: "fa-land-mine-on",
+        default: false,
+    },
     {
         id: "sound_toggle",
         type: ParameterType.checkbox,
@@ -66,6 +100,7 @@ var PARAMETERS = [
         type: ParameterType.checkbox,
         label: "Process newest jobs first",
         note: "reverse the normal processing order",
+        icon: "fa-arrow-down-short-wide",
         default: false,
     },
     {
@@ -77,12 +112,20 @@ var PARAMETERS = [
         default: true,
     },
     {
-        id: "turbo",
-        type: ParameterType.checkbox,
-        label: "Turbo Mode",
-        note: "generates images faster, but uses an additional 1 GB of GPU memory",
+        id: "vram_usage_level",
+        type: ParameterType.select,
+        label: "GPU Memory Usage",
+        note: "Faster performance requires more GPU memory (VRAM)<br/><br/>" +
+              "<b>Balanced:</b> nearly as fast as High, much lower VRAM usage<br/>" +
+              "<b>High:</b> fastest, maximum GPU memory usage</br>" +
+              "<b>Low:</b> slowest, recommended for GPUs with 3 to 4 GB memory",
         icon: "fa-forward",
-        default: true,
+        default: "balanced",
+        options: [
+            {value: "balanced", label: "Balanced"},
+            {value: "high", label: "High"},
+            {value: "low", label: "Low"}
+        ],
     },
     {
         id: "use_cpu",
@@ -105,14 +148,6 @@ var PARAMETERS = [
         note: "to process in parallel",
         default: false,
     },
-    {
-        id: "use_full_precision",
-        type: ParameterType.checkbox,
-        label: "Use Full Precision",
-        note: "for GPU-only. warning: this will consume more VRAM",
-        icon: "fa-crosshairs",
-        default: false,
-    },
     {
         id: "auto_save_settings",
         type: ParameterType.checkbox,
@@ -147,14 +182,6 @@ var PARAMETERS = [
             return `<input id="${parameter.id}" name="${parameter.id}" size="6" value="9000" onkeypress="preventNonNumericalInput(event)">`
         }
     },
-    {
-        id: "test_sd2",
-        type: ParameterType.checkbox,
-        label: "Test SD 2.0",
-        note: "Experimental! High memory usage! GPU-only! Not the final version! Please restart the program after changing this.",
-        icon: "fa-fire",
-        default: false,
-    },
     {
         id: "use_beta_channel",
         type: ParameterType.checkbox,
@@ -173,6 +200,18 @@ function getParameterSettingsEntry(id) {
     return parameter[0].settingsEntry
 }
 
+function sliderUpdate(event) {
+    if (event.srcElement.id.endsWith('-input')) {
+        let slider = document.getElementById(event.srcElement.id.slice(0,-6))
+        slider.value = event.srcElement.value
+        slider.dispatchEvent(new Event("change"))
+    } else {
+        let field = document.getElementById(event.srcElement.id+'-input')
+        field.value = event.srcElement.value
+        field.dispatchEvent(new Event("change"))
+    }
+}
+
 function getParameterElement(parameter) {
     switch (parameter.type) {
         case ParameterType.checkbox:
@@ -183,6 +222,8 @@ function getParameterElement(parameter) {
             var options = (parameter.options || []).map(option => `<option value="${option.value}">${option.label}</option>`).join("")
             var multiple = (parameter.type == ParameterType.select_multiple ? 'multiple' : '')
             return `<select id="${parameter.id}" name="${parameter.id}" ${multiple}>${options}</select>`
+        case ParameterType.slider:
+            return `<input id="${parameter.id}" name="${parameter.id}" class="editor-slider" type="range" value="${parameter.default}" min="${parameter.slider_min}" max="${parameter.slider_max}" oninput="sliderUpdate(event)"> <input id="${parameter.id}-input" name="${parameter.id}-input" size="4" value="${parameter.default}" pattern="^[0-9\.]+$" onkeypress="preventNonNumericalInput(event)" oninput="sliderUpdate(event)">&nbsp;${parameter.slider_unit}`
         case ParameterType.custom:
             return parameter.render(parameter)
         default:
@@ -210,16 +251,15 @@ function initParameters() {
 
 initParameters()
 
-let turboField = document.querySelector('#turbo')
+let vramUsageLevelField = document.querySelector('#vram_usage_level')
 let useCPUField = document.querySelector('#use_cpu')
 let autoPickGPUsField = document.querySelector('#auto_pick_gpus')
 let useGPUsField = document.querySelector('#use_gpus')
-let useFullPrecisionField = document.querySelector('#use_full_precision')
 let saveToDiskField = document.querySelector('#save_to_disk')
 let diskPathField = document.querySelector('#diskPath')
+let metadataOutputFormatField = document.querySelector('#metadata_output_format')
 let listenToNetworkField = document.querySelector("#listen_to_network")
 let listenPortField = document.querySelector("#listen_port")
-let testSD2Field = document.querySelector("#test_sd2")
 let useBetaChannelField = document.querySelector("#use_beta_channel")
 let uiOpenBrowserOnStartField = document.querySelector("#ui_open_browser_on_start")
 let confirmDangerousActionsField = document.querySelector("#confirm_dangerous_actions")
@@ -256,12 +296,6 @@ async function getAppConfig() {
         if (config.ui && config.ui.open_browser_on_start === false) {
             uiOpenBrowserOnStartField.checked = false
         }
-        if ('test_sd2' in config) {
-            testSD2Field.checked = config['test_sd2']
-        }
-
-        let testSD2SettingEntry = getParameterSettingsEntry('test_sd2')
-        testSD2SettingEntry.style.display = (config.update_branch === 'beta' ? '' : 'none')
         if (config.net && config.net.listen_to_network === false) {
             listenToNetworkField.checked = false
         }
@@ -277,6 +311,7 @@ async function getAppConfig() {
 
 saveToDiskField.addEventListener('change', function(e) {
     diskPathField.disabled = !this.checked
+    metadataOutputFormatField.disabled = !this.checked
 })
 
 function getCurrentRenderDeviceSelection() {
@@ -327,20 +362,10 @@ autoPickGPUsField.addEventListener('click', function() {
     gpuSettingEntry.style.display = (this.checked ? 'none' : '')
 })
 
-async function getDiskPath() {
-    try {
-        var diskPath = getSetting("diskPath")
-        if (diskPath == '' || diskPath == undefined || diskPath == "undefined") {
-            let res = await fetch('/get/output_dir')
-            if (res.status === 200) {
-                res = await res.json()
-                res = res.output_dir
-
-                setSetting("diskPath", res)
-            }
-        }
-    } catch (e) {
-        console.log('error fetching output dir path', e)
+async function setDiskPath(defaultDiskPath, force=false) {
+    var diskPath = getSetting("diskPath")
+    if (force || diskPath == '' || diskPath == undefined || diskPath == "undefined") {
+        setSetting("diskPath", defaultDiskPath)
     }
 }
 
@@ -415,6 +440,17 @@ async function getSystemInfo() {
 
         setDeviceInfo(devices)
         setHostInfo(res['hosts'])
+        let force = false
+        if (res['enforce_output_dir'] !== undefined) {
+            force = res['enforce_output_dir']
+            if (force == true) {
+               saveToDiskField.checked = true
+               metadataOutputFormatField.disabled = false
+            }
+            saveToDiskField.disabled = force
+            diskPathField.disabled = force
+        }
+        setDiskPath(res['default_output_dir'], force)
     } catch (e) {
         console.log('error fetching devices', e)
     }
@@ -435,9 +471,9 @@ saveSettingsBtn.addEventListener('click', function() {
         'update_branch': updateBranch,
         'ui_open_browser_on_start': uiOpenBrowserOnStartField.checked,
         'listen_to_network': listenToNetworkField.checked,
-        'listen_port': listenPortField.value,
-        'test_sd2': testSD2Field.checked
+        'listen_port': listenPortField.value
     })
     saveSettingsBtn.classList.add('active')
     asyncDelay(300).then(() => saveSettingsBtn.classList.remove('active'))
 })
+

ui/media/js/plugins.js

@@ -25,11 +25,13 @@ const PLUGINS = {
      * })
      */
     IMAGE_INFO_BUTTONS: [],
+    GET_PROMPTS_HOOK: [],
     MODIFIERS_LOAD: [],
     TASK_CREATE: [],
     OUTPUTS_FORMATS: new ServiceContainer(
         function png() { return (reqBody) => new SD.RenderTask(reqBody) }
         , function jpeg() { return (reqBody) => new SD.RenderTask(reqBody) }
+        , function webp() { return (reqBody) => new SD.RenderTask(reqBody) }
     ),
 }
 PLUGINS.OUTPUTS_FORMATS.register = function(...args) {

ui/media/js/searchable-models.js

@@ -0,0 +1,687 @@
+"use strict"
+
+let modelsCache
+let modelsOptions
+
+/*
+*** SEARCHABLE MODELS ***
+Creates searchable dropdowns for SD, VAE, or HN models.
+Also adds a reload models button (placed next to SD models, reloads everything including VAE and HN models).
+More reload buttons may be added at strategic UI locations as needed.
+Merely calling getModels() makes all the magic happen behind the scene to refresh the dropdowns.
+
+HOW TO CREATE A MODEL DROPDOWN:
+1) Create an input element. Make sure to add a data-path property, as this is how model dropdowns are identified in auto-save.js.
+<input id="stable_diffusion_model" type="text" spellcheck="false" autocomplete="off" class="model-filter" data-path="" />
+
+2) Just declare one of these for your own dropdown (remember to change the element id, e.g. #stable_diffusion_models to your own input's id).
+let stableDiffusionModelField = new ModelDropdown(document.querySelector('#stable_diffusion_model'), 'stable-diffusion')
+let vaeModelField = new ModelDropdown(document.querySelector('#vae_model'), 'vae', 'None')
+let hypernetworkModelField = new ModelDropdown(document.querySelector('#hypernetwork_model'), 'hypernetwork', 'None')
+
+3) Model dropdowns will be refreshed automatically when the reload models button is invoked.
+*/
+class ModelDropdown
+{
+    modelFilter //= document.querySelector("#model-filter")
+    modelFilterArrow //= document.querySelector("#model-filter-arrow")
+    modelList //= document.querySelector("#model-list")
+    modelResult //= document.querySelector("#model-result")
+    modelNoResult //= document.querySelector("#model-no-result")
+    
+    currentSelection //= { elem: undefined, value: '', path: ''}
+    highlightedModelEntry //= undefined
+    activeModel //= undefined
+
+    inputModels //= undefined
+    modelKey //= undefined
+    flatModelList //= []
+    noneEntry //= ''
+    modelFilterInitialized //= undefined
+
+    /* MIMIC A REGULAR INPUT FIELD */
+    get parentElement() {
+        return this.modelFilter.parentElement
+    }
+    get parentNode() {
+        return this.modelFilter.parentNode
+    }
+    get value() {
+        return this.modelFilter.dataset.path
+    }
+    set value(path) {
+        this.modelFilter.dataset.path = path
+        this.selectEntry(path)
+    }
+    get disabled() {
+        return this.modelFilter.disabled
+    }
+    set disabled(state) {
+        this.modelFilter.disabled = state
+        if (this.modelFilterArrow) {
+            this.modelFilterArrow.style.color = state ? 'dimgray' : ''
+        }
+    }
+    get modelElements() {
+        return this.modelList.querySelectorAll('.model-file')
+    }
+    addEventListener(type, listener, options) {
+        return this.modelFilter.addEventListener(type, listener, options)
+    }
+    dispatchEvent(event) {
+        return this.modelFilter.dispatchEvent(event)
+    }
+    appendChild(option) {
+        // do nothing
+    }
+
+    // remember 'this' - http://blog.niftysnippets.org/2008/04/you-must-remember-this.html
+    bind(f, obj) {
+        return function() {
+            return f.apply(obj, arguments)
+        }
+    }
+
+    /* SEARCHABLE INPUT */    
+    constructor (input, modelKey, noneEntry = '') {
+        this.modelFilter = input
+        this.noneEntry = noneEntry
+        this.modelKey = modelKey
+
+        if (modelsOptions !== undefined) { // reuse models from cache (only useful for plugins, which are loaded after models)
+            this.inputModels = modelsOptions[this.modelKey]
+            this.populateModels()
+        }
+        document.addEventListener("refreshModels", this.bind(function(e) {
+            // reload the models
+            this.inputModels = modelsOptions[this.modelKey]
+            this.populateModels()
+        }, this))
+    }
+
+    saveCurrentSelection(elem, value, path) {
+        this.currentSelection.elem = elem
+        this.currentSelection.value = value
+        this.currentSelection.path = path
+        this.modelFilter.dataset.path = path
+        this.modelFilter.value = value
+        this.modelFilter.dispatchEvent(new Event('change'))
+    }
+    
+    processClick(e) {
+        e.preventDefault()
+        if (e.srcElement.classList.contains('model-file') || e.srcElement.classList.contains('fa-file')) {
+            const elem = e.srcElement.classList.contains('model-file') ? e.srcElement : e.srcElement.parentElement
+            this.saveCurrentSelection(elem, elem.innerText, elem.dataset.path)
+            this.hideModelList()
+            this.modelFilter.focus()
+            this.modelFilter.select()
+        }
+    }
+
+    getPreviousVisibleSibling(elem) {
+        const modelElements = Array.from(this.modelElements)
+        const index = modelElements.indexOf(elem)
+        if (index <= 0) {
+            return undefined
+        }
+
+        return modelElements.slice(0, index).reverse().find(e => e.style.display === 'list-item')
+    }
+
+    getLastVisibleChild(elem) {
+        let lastElementChild = elem.lastElementChild
+        if (lastElementChild.style.display == 'list-item') return lastElementChild
+        return this.getPreviousVisibleSibling(lastElementChild)
+    }
+    
+    getNextVisibleSibling(elem) {
+        const modelElements = Array.from(this.modelElements)
+        const index = modelElements.indexOf(elem)
+        return modelElements.slice(index + 1).find(e => e.style.display === 'list-item')
+    }
+    
+    getFirstVisibleChild(elem) {
+        let firstElementChild = elem.firstElementChild
+        if (firstElementChild.style.display == 'list-item') return firstElementChild
+        return this.getNextVisibleSibling(firstElementChild)
+    }
+    
+    selectModelEntry(elem) {
+        if (elem) {
+            if (this.highlightedModelEntry !== undefined) {
+                this.highlightedModelEntry.classList.remove('selected')
+            }
+            this.saveCurrentSelection(elem, elem.innerText, elem.dataset.path)
+            elem.classList.add('selected')
+            elem.scrollIntoView({block: 'nearest'})
+            this.highlightedModelEntry = elem
+        }
+    }
+    
+    selectPreviousFile() {
+        const elem = this.getPreviousVisibleSibling(this.highlightedModelEntry)
+        if (elem) {
+            this.selectModelEntry(elem)
+        }
+        else
+        {
+            //this.highlightedModelEntry.parentElement.parentElement.scrollIntoView({block: 'nearest'})
+            this.highlightedModelEntry.closest('.model-list').scrollTop = 0
+        }
+        this.modelFilter.select()
+    }
+    
+    selectNextFile() {
+        this.selectModelEntry(this.getNextVisibleSibling(this.highlightedModelEntry))
+        this.modelFilter.select()
+    }
+    
+    selectFirstFile() {
+        this.selectModelEntry(this.modelList.querySelector('.model-file'))
+        this.highlightedModelEntry.scrollIntoView({block: 'nearest'})
+        this.modelFilter.select()
+    }
+    
+    selectLastFile() {
+        const elems = this.modelList.querySelectorAll('.model-file:last-child')
+        this.selectModelEntry(elems[elems.length -1])
+        this.modelFilter.select()
+    }
+
+    resetSelection() {
+        this.hideModelList()
+        this.showAllEntries()
+        this.modelFilter.value = this.currentSelection.value
+        this.modelFilter.focus()
+        this.modelFilter.select()
+    }
+
+    validEntrySelected() {
+        return (this.modelNoResult.style.display === 'none')
+    }
+    
+    processKey(e) {
+        switch (e.key) {
+            case 'Escape':
+                e.preventDefault()
+                this.resetSelection()
+                break
+            case 'Enter':
+                e.preventDefault()
+                if (this.validEntrySelected()) {
+                    if (this.modelList.style.display != 'block') {
+                        this.showModelList()
+                    }
+                    else
+                    {
+                        this.saveCurrentSelection(this.highlightedModelEntry, this.highlightedModelEntry.innerText, this.highlightedModelEntry.dataset.path)
+                        this.hideModelList()
+                        this.showAllEntries()
+                    }
+                    this.modelFilter.focus()
+                }
+                else
+                {
+                    this.resetSelection()
+                }
+                break
+            case 'ArrowUp':
+                e.preventDefault()
+                if (this.validEntrySelected()) {
+                    this.selectPreviousFile()
+                }
+                break
+            case 'ArrowDown':
+                e.preventDefault()
+                if (this.validEntrySelected()) {
+                    this.selectNextFile()
+                }
+                break
+            case 'ArrowLeft':
+                if (this.modelList.style.display != 'block') {
+                    e.preventDefault()
+                }
+                break
+            case 'ArrowRight':
+                if (this.modelList.style.display != 'block') {
+                    e.preventDefault()
+                }
+                break
+            case 'PageUp':
+                e.preventDefault()
+                if (this.validEntrySelected()) {
+                    this.selectPreviousFile()
+                    this.selectPreviousFile()
+                    this.selectPreviousFile()
+                    this.selectPreviousFile()
+                    this.selectPreviousFile()
+                    this.selectPreviousFile()
+                    this.selectPreviousFile()
+                    this.selectPreviousFile()
+                }
+                break
+            case 'PageDown':
+                e.preventDefault()
+                if (this.validEntrySelected()) {
+                    this.selectNextFile()
+                    this.selectNextFile()
+                    this.selectNextFile()
+                    this.selectNextFile()
+                    this.selectNextFile()
+                    this.selectNextFile()
+                    this.selectNextFile()
+                    this.selectNextFile()
+                }
+                break
+            case 'Home':
+                //if (this.modelList.style.display != 'block') {
+                    e.preventDefault()
+                    if (this.validEntrySelected()) {
+                        this.selectFirstFile()
+                    }
+                //}
+                break
+            case 'End':
+                //if (this.modelList.style.display != 'block') {
+                    e.preventDefault()
+                    if (this.validEntrySelected()) {
+                        this.selectLastFile()
+                    }
+                //}
+                break
+            default:
+                //console.log(e.key)
+        }
+    }
+    
+    modelListFocus() {
+        this.selectEntry()
+        this.showAllEntries()
+    }
+    
+    showModelList() {
+        this.modelList.style.display = 'block'
+        this.selectEntry()
+        this.showAllEntries()
+        //this.modelFilter.value = ''
+        this.modelFilter.select() // preselect the entire string so user can just start typing.
+        this.modelFilter.focus()
+        this.modelFilter.style.cursor = 'auto'
+    }
+    
+    hideModelList() {
+        this.modelList.style.display = 'none'
+        this.modelFilter.value = this.currentSelection.value
+        this.modelFilter.style.cursor = ''
+    }
+    
+    toggleModelList(e) {
+        e.preventDefault()
+        if (!this.modelFilter.disabled) {
+            if (this.modelList.style.display != 'block') {
+                this.showModelList()
+            }
+            else
+            {
+                this.hideModelList()
+                this.modelFilter.select()
+            }
+        }
+    }
+    
+    selectEntry(path) {
+        if (path !== undefined) {
+            const entries = this.modelElements;
+
+            for (const elem of entries) {
+                if (elem.dataset.path == path) {
+                    this.saveCurrentSelection(elem, elem.innerText, elem.dataset.path)
+                    this.highlightedModelEntry = elem
+                    elem.scrollIntoView({block: 'nearest'})
+                    break
+                }
+            }
+        }
+        
+        if (this.currentSelection.elem !== undefined) {
+            // select the previous element
+            if (this.highlightedModelEntry !== undefined && this.highlightedModelEntry != this.currentSelection.elem) {
+                this.highlightedModelEntry.classList.remove('selected')
+            }
+            this.currentSelection.elem.classList.add('selected')
+            this.highlightedModelEntry = this.currentSelection.elem
+            this.currentSelection.elem.scrollIntoView({block: 'nearest'})
+        }
+        else
+        {
+            this.selectFirstFile()
+        }
+    }
+    
+    highlightModelAtPosition(e) {
+        let elem = document.elementFromPoint(e.clientX, e.clientY)
+        
+        if (elem.classList.contains('model-file')) {
+            this.highlightModel(elem)
+        }
+    }
+    
+    highlightModel(elem) {
+        if (elem.classList.contains('model-file')) {
+            if (this.highlightedModelEntry !== undefined && this.highlightedModelEntry != elem) {
+                this.highlightedModelEntry.classList.remove('selected')
+            }
+            elem.classList.add('selected')
+            this.highlightedModelEntry = elem
+        }
+    }
+    
+    showAllEntries() {
+        this.modelList.querySelectorAll('li').forEach(function(li) {
+            if (li.id !== 'model-no-result') {
+                li.style.display = 'list-item'
+            }
+        })
+        this.modelNoResult.style.display = 'none'
+    }
+    
+    filterList(e) {
+        const filter = this.modelFilter.value.toLowerCase()
+        let found = false
+        let showAllChildren = false
+        
+        this.modelList.querySelectorAll('li').forEach(function(li) {
+            if (li.classList.contains('model-folder')) {
+                showAllChildren = false
+            }
+            if (filter == '') {
+                li.style.display = 'list-item'
+                found = true
+            } else if (showAllChildren || li.textContent.toLowerCase().match(filter)) {
+                li.style.display = 'list-item'
+                if (li.classList.contains('model-folder') && li.firstChild.textContent.toLowerCase().match(filter)) {
+                    showAllChildren = true
+                }
+                found = true
+            } else {
+                li.style.display = 'none'
+            }
+        })
+    
+        if (found) {
+            this.modelResult.style.display = 'list-item'
+            this.modelNoResult.style.display = 'none'
+            const elem = this.getNextVisibleSibling(this.modelList.querySelector('.model-file'))
+            this.highlightModel(elem)
+            elem.scrollIntoView({block: 'nearest'})
+        }
+        else
+        {
+            this.modelResult.style.display = 'none'
+            this.modelNoResult.style.display = 'list-item'
+        }
+        this.modelList.style.display = 'block'
+    }
+
+    /* MODEL LOADER */
+    getElementDimensions(element) {
+        // Clone the element
+        const clone = element.cloneNode(true)
+        
+        // Copy the styles of the original element to the cloned element
+        const originalStyles = window.getComputedStyle(element)
+        for (let i = 0; i < originalStyles.length; i++) {
+            const property = originalStyles[i]
+            clone.style[property] = originalStyles.getPropertyValue(property)
+        }
+        
+        // Set its visibility to hidden and display to inline-block
+        clone.style.visibility = "hidden"
+        clone.style.display = "inline-block"
+        
+        // Put the cloned element next to the original element
+        element.parentNode.insertBefore(clone, element.nextSibling)
+        
+        // Get its width and height
+        const width = clone.offsetWidth
+        const height = clone.offsetHeight
+        
+        // Remove it from the DOM
+        clone.remove()
+        
+        // Return its width and height
+        return { width, height }
+    }
+    
+    /**
+     * @param {Array<string>} models 
+     */
+    sortStringArray(models) {
+        models.sort((a, b) => a.localeCompare(b, undefined, { sensitivity: 'base' }))
+    }
+
+    populateModels() {
+        this.activeModel = this.modelFilter.dataset.path
+        
+        this.currentSelection = { elem: undefined, value: '', path: ''}
+        this.highlightedModelEntry = undefined
+        this.flatModelList = []
+
+        if(this.modelList !== undefined) {
+            this.modelList.remove()
+            this.modelFilterArrow.remove()
+        }
+        this.createDropdown()
+    }
+
+    createDropdown() {
+        // create dropdown entries
+        let rootModelList = this.createRootModelList(this.inputModels)
+        this.modelFilter.insertAdjacentElement('afterend', rootModelList)
+        this.modelFilter.insertAdjacentElement(
+            'afterend',
+            this.createElement(
+                'i',
+                { id: `${this.modelFilter.id}-model-filter-arrow` },
+                ['model-selector-arrow', 'fa-solid', 'fa-angle-down'],
+            ),
+        )
+        this.modelFilter.classList.add('model-selector')
+        this.modelFilterArrow = document.querySelector(`#${this.modelFilter.id}-model-filter-arrow`)
+        if (this.modelFilterArrow) {
+            this.modelFilterArrow.style.color = this.modelFilter.disabled ? 'dimgray' : ''
+        }
+        this.modelList = document.querySelector(`#${this.modelFilter.id}-model-list`)
+        this.modelResult = document.querySelector(`#${this.modelFilter.id}-model-result`)
+        this.modelNoResult = document.querySelector(`#${this.modelFilter.id}-model-no-result`)
+        
+        if (this.modelFilterInitialized !== true) {
+            this.modelFilter.addEventListener('input', this.bind(this.filterList, this))
+            this.modelFilter.addEventListener('focus', this.bind(this.modelListFocus, this))
+            this.modelFilter.addEventListener('blur', this.bind(this.hideModelList, this))
+            this.modelFilter.addEventListener('click', this.bind(this.showModelList, this))
+            this.modelFilter.addEventListener('keydown', this.bind(this.processKey, this))
+
+            this.modelFilterInitialized = true
+        }
+        this.modelFilterArrow.addEventListener('mousedown', this.bind(this.toggleModelList, this))
+        this.modelList.addEventListener('mousemove', this.bind(this.highlightModelAtPosition, this))
+        this.modelList.addEventListener('mousedown', this.bind(this.processClick, this))
+
+        let mf = this.modelFilter
+        this.modelFilter.addEventListener('focus', function() {
+            let modelFilterStyle = window.getComputedStyle(mf)
+            rootModelList.style.minWidth = modelFilterStyle.width
+        })
+
+        this.selectEntry(this.activeModel)
+    }
+
+    /**
+     * 
+     * @param {string} tag 
+     * @param {object} attributes
+     * @param {Array<string>} classes
+     * @returns {HTMLElement}
+     */
+    createElement(tagName, attributes, classes, text, icon) {
+        const element = document.createElement(tagName)
+        if (attributes) {
+            Object.entries(attributes).forEach(([key, value]) => {
+                element.setAttribute(key, value)
+            })
+        }
+        if (classes) {
+            classes.forEach(className => element.classList.add(className))
+        }
+        if (icon) {
+            let iconEl = document.createElement('i')
+            iconEl.className = icon + ' icon'
+            element.appendChild(iconEl)
+        }
+        if (text) {
+            element.appendChild(document.createTextNode(text))
+        }
+        return element
+    }
+
+    /**
+     * @param {Array<string | object} modelTree
+     * @param {string} folderName 
+     * @param {boolean} isRootFolder 
+     * @returns {HTMLElement}
+     */
+    createModelNodeList(folderName, modelTree, isRootFolder) {
+        const listElement = this.createElement('ul')
+
+        const foldersMap = new Map()
+        const modelsMap = new Map()
+
+        modelTree.forEach(model => {
+            if (Array.isArray(model)) {
+                const [childFolderName, childModels] = model
+                foldersMap.set(
+                    childFolderName,
+                    this.createModelNodeList(
+                        `${folderName || ''}/${childFolderName}`,
+                        childModels,
+                        false,
+                    ),
+                )
+            } else {
+                const classes = ['model-file']
+                if (isRootFolder) {
+                    classes.push('in-root-folder')
+                }
+                // Remove the leading slash from the model path
+                const fullPath = folderName ? `${folderName.substring(1)}/${model}` : model
+                modelsMap.set(
+                    model,
+                    this.createElement('li', { 'data-path': fullPath }, classes, model, 'fa-regular fa-file'),
+                )
+            }
+        })
+
+        const childFolderNames = Array.from(foldersMap.keys())
+        this.sortStringArray(childFolderNames)
+        const folderElements = childFolderNames.map(name => foldersMap.get(name))
+
+        const modelNames = Array.from(modelsMap.keys())
+        this.sortStringArray(modelNames)
+        const modelElements = modelNames.map(name => modelsMap.get(name))
+
+        if (modelElements.length && folderName) {
+            listElement.appendChild(this.createElement('li', undefined, ['model-folder'], folderName.substring(1), 'fa-solid fa-folder-open'))
+        }
+
+        // const allModelElements = isRootFolder ? [...folderElements, ...modelElements] : [...modelElements, ...folderElements]
+        const allModelElements = [...modelElements, ...folderElements]
+        allModelElements.forEach(e => listElement.appendChild(e))
+        return listElement
+    }
+
+    /**
+     * @param {object} modelTree
+     * @returns {HTMLElement}
+     */
+    createRootModelList(modelTree) {
+        const rootList = this.createElement(
+            'ul',
+            { id: `${this.modelFilter.id}-model-list` },
+            ['model-list'],
+        )
+        rootList.appendChild(
+            this.createElement(
+                'li',
+                { id: `${this.modelFilter.id}-model-no-result` },
+                ['model-no-result'],
+                'No result'
+            ),
+        )
+
+        if (this.noneEntry) {
+            rootList.appendChild(
+                this.createElement(
+                    'li',
+                    { 'data-path': '' },
+                    ['model-file', 'in-root-folder'],
+                    this.noneEntry,
+                ),
+            )
+        }
+
+        if (modelTree.length > 0) {
+            const containerListItem = this.createElement(
+                'li',
+                { id: `${this.modelFilter.id}-model-result` },
+                ['model-result'],
+            )
+            //console.log(containerListItem)
+            containerListItem.appendChild(this.createModelNodeList(undefined, modelTree, true))
+            rootList.appendChild(containerListItem)
+        }
+
+        return rootList
+    }
+}
+
+/* (RE)LOAD THE MODELS */
+async function getModels() {
+    try {
+        modelsCache = await SD.getModels()
+        modelsOptions = modelsCache['options']
+        if ("scan-error" in modelsCache) {
+            // let previewPane = document.getElementById('tab-content-wrapper')
+            let previewPane = document.getElementById('preview')
+            previewPane.style.background="red"
+            previewPane.style.textAlign="center"
+            previewPane.innerHTML = '<H1>🔥Malware alert!🔥</H1><h2>The file <i>' + modelsCache['scan-error'] + '</i> in your <tt>models/stable-diffusion</tt> folder is probably malware infected.</h2><h2>Please delete this file from the folder before proceeding!</h2>After deleting the file, reload this page.<br><br><button onClick="window.location.reload();">Reload Page</button>'
+            makeImageBtn.disabled = true
+        }
+
+        /* This code should no longer be needed. Commenting out for now, will cleanup later.
+        const sd_model_setting_key = "stable_diffusion_model"
+        const vae_model_setting_key = "vae_model"
+        const hypernetwork_model_key = "hypernetwork_model"
+
+        const stableDiffusionOptions = modelsOptions['stable-diffusion']
+        const vaeOptions = modelsOptions['vae']
+        const hypernetworkOptions = modelsOptions['hypernetwork']
+
+        // TODO: set default for model here too
+        SETTINGS[sd_model_setting_key].default = stableDiffusionOptions[0]
+        if (getSetting(sd_model_setting_key) == '' || SETTINGS[sd_model_setting_key].value == '') {
+            setSetting(sd_model_setting_key, stableDiffusionOptions[0])
+        }
+        */
+
+        // notify ModelDropdown objects to refresh
+        document.dispatchEvent(new Event('refreshModels'))
+    } catch (e) {
+        console.log('get models error', e)
+    }
+}
+
+// reload models button
+document.querySelector('#reload-models').addEventListener('click', getModels)

ui/media/js/themes.js

@@ -13,8 +13,15 @@ function initTheme() {
         .filter(sheet => sheet.href?.startsWith(window.location.origin))
         .flatMap(sheet => Array.from(sheet.cssRules))
         .forEach(rule => {
-            var selector = rule.selectorText; // TODO: also do selector == ":root", re-run un-set props
+            var selector = rule.selectorText;
             if (selector && selector.startsWith(".theme-") && !selector.includes(" ")) {
+                if (DEFAULT_THEME) { // re-add props that dont change (css needs this so they update correctly)
+                    Array.from(DEFAULT_THEME.rule.style)
+                        .filter(cssVariable => !Array.from(rule.style).includes(cssVariable))
+                        .forEach(cssVariable => {
+                        rule.style.setProperty(cssVariable, DEFAULT_THEME.rule.style.getPropertyValue(cssVariable));
+                    });
+                }
                 var theme_key = selector.substring(1);
                 THEMES.push({
                     key: theme_key,
@@ -62,12 +69,6 @@ function themeFieldChanged() {
     var theme = THEMES.find(t => t.key == theme_key);
     let borderColor = undefined
     if (theme) {
-        // refresh variables incase they are back referencing
-        Array.from(DEFAULT_THEME.rule.style)
-            .filter(cssVariable => !Array.from(theme.rule.style).includes(cssVariable))
-            .forEach(cssVariable => {
-            body.style.setProperty(cssVariable, DEFAULT_THEME.rule.style.getPropertyValue(cssVariable));
-        });
         borderColor = theme.rule.style.getPropertyValue('--input-border-color').trim()
         if (!borderColor.startsWith('#')) {
             borderColor = theme.rule.style.getPropertyValue('--theme-color-fallback')

ui/media/js/utils.js

@@ -509,6 +509,9 @@ function makeQuerablePromise(promise) {
 /* inserts custom html to allow prettifying of inputs */
 function prettifyInputs(root_element) {
     root_element.querySelectorAll(`input[type="checkbox"]`).forEach(element => {
+        if (element.style.display === "none") {
+            return
+        }
         var parent = element.parentNode;
         if (!parent.classList.contains("input-toggle")) {
             var wrapper = document.createElement("div");

ui/plugins/ui/Autoscroll.plugin.js

@@ -1,27 +1,7 @@
 (function () {
     "use strict"
 
-    var styleSheet = document.createElement("style");
-    styleSheet.textContent = `
-        .auto-scroll {
-            float: right;
-        }
-    `;
-    document.head.appendChild(styleSheet);
-    
-    const autoScrollControl = document.createElement('div');
-    autoScrollControl.innerHTML = `<input id="auto_scroll" name="auto_scroll" type="checkbox">
-                            <label for="auto_scroll">Scroll to generated image</label>`
-    autoScrollControl.className = "auto-scroll"
-    clearAllPreviewsBtn.parentNode.insertBefore(autoScrollControl, clearAllPreviewsBtn.nextSibling)
-    prettifyInputs(document);
     let autoScroll = document.querySelector("#auto_scroll")
-    
-    // save/restore the toggle state
-    autoScroll.addEventListener('click', (e) => {
-        localStorage.setItem('auto_scroll', autoScroll.checked)
-    })
-    autoScroll.checked = localStorage.getItem('auto_scroll') == "true"
 
     // observe for changes in the preview pane
     var observer = new MutationObserver(function (mutations) {
@@ -39,7 +19,10 @@
 
     function Autoscroll(target) {
         if (autoScroll.checked && target !== null) {
-            target.parentElement.parentElement.parentElement.scrollIntoView();
+            const img = target.querySelector('img')
+            img.addEventListener('load', function() {
+                img.closest('.imageTaskContainer').scrollIntoView()
+            }, { once: true })
         }
     }
 })()

ui/plugins/ui/Modifiers-dnd.plugin.js

@@ -74,6 +74,7 @@
                         // update activeTags
                         const tag = activeTags.splice(currentPos, 1)
                         activeTags.splice(droppedPos, 0, tag[0])
+                        document.dispatchEvent(new Event('refreshImageModifiers'))
                     }
                 }
             };

ui/plugins/ui/Modifiers-wheel.plugin.js

@@ -58,6 +58,7 @@
                             break
                         }
                     }
+                    document.dispatchEvent(new Event('refreshImageModifiers'))
                 }
             }
         })

ui/plugins/ui/merge.plugin.js

@@ -0,0 +1,458 @@
+(function() {
+    "use strict"
+
+    ///////////////////// Function section
+    function smoothstep(x) {
+        return x * x * (3 - 2 * x)
+    }
+
+    function smootherstep(x) {
+        return x * x * x * (x * (x * 6 - 15) + 10)
+    }
+
+    function smootheststep(x) {
+        let y = -20 * Math.pow(x, 7)
+        y += 70 * Math.pow(x, 6)
+        y -= 84 * Math.pow(x, 5)
+        y += 35 * Math.pow(x, 4)
+        return y
+    }
+    function getCurrentTime() {
+        const now = new Date();
+        let hours = now.getHours();
+        let minutes = now.getMinutes();
+        let seconds = now.getSeconds();
+
+        hours = hours < 10 ? `0${hours}` : hours;
+        minutes = minutes < 10 ? `0${minutes}` : minutes;
+        seconds = seconds < 10 ? `0${seconds}` : seconds;
+
+        return `${hours}:${minutes}:${seconds}`;
+    }
+
+    function addLogMessage(message) {
+        const logContainer = document.getElementById('merge-log');
+        logContainer.innerHTML += `<i>${getCurrentTime()}</i> ${message}<br>`;
+
+        // Scroll to the bottom of the log
+        logContainer.scrollTop = logContainer.scrollHeight;
+
+        document.querySelector('#merge-log-container').style.display = 'block'
+    }    
+
+    function addLogSeparator() {
+        const logContainer = document.getElementById('merge-log');
+        logContainer.innerHTML += '<hr>'
+
+        logContainer.scrollTop = logContainer.scrollHeight;
+    }
+
+    function drawDiagram(fn) {
+        const SIZE = 300
+        const canvas = document.getElementById('merge-canvas');
+        canvas.height = canvas.width = SIZE
+        const ctx = canvas.getContext('2d');
+
+        // Draw coordinate system
+        ctx.scale(1, -1);
+        ctx.translate(0, -canvas.height);
+        ctx.lineWidth = 1;
+        ctx.beginPath();
+
+        ctx.strokeStyle = 'white'
+        ctx.moveTo(0,0); ctx.lineTo(0,SIZE); ctx.lineTo(SIZE,SIZE); ctx.lineTo(SIZE,0); ctx.lineTo(0,0); ctx.lineTo(SIZE,SIZE);
+        ctx.stroke()
+        ctx.beginPath()
+        ctx.setLineDash([1,2])
+        const n = SIZE / 10
+        for (let i=n; i<SIZE; i+=n) {
+            ctx.moveTo(0,i)
+            ctx.lineTo(SIZE,i)
+            ctx.moveTo(i,0)
+            ctx.lineTo(i,SIZE)
+        }
+        ctx.stroke()
+        ctx.beginPath()
+        ctx.setLineDash([])
+        ctx.beginPath();
+        ctx.strokeStyle = 'black'
+        ctx.lineWidth = 3;
+        // Plot function
+        const numSamples = 20;
+        for (let i = 0; i <= numSamples; i++) {
+            const x = i / numSamples;
+            const y = fn(x);
+        
+            const canvasX = x * SIZE;
+            const canvasY = y * SIZE;
+
+            if (i === 0) {
+                ctx.moveTo(canvasX, canvasY);
+            } else {
+                ctx.lineTo(canvasX, canvasY);
+            }
+        }
+        ctx.stroke()
+        // Plot alpha values (yellow boxes)
+        let start = parseFloat( document.querySelector('#merge-start').value )
+        let step = parseFloat( document.querySelector('#merge-step').value )
+        let iterations = document.querySelector('#merge-count').value>>0
+        ctx.beginPath()
+        ctx.fillStyle = "yellow"
+        for (let i=0; i< iterations; i++) {
+            const alpha = ( start + i * step ) / 100
+            const x = alpha*SIZE
+            const y = fn(alpha) * SIZE
+            if (x <= SIZE) {
+                ctx.rect(x-3,y-3,6,6)
+                ctx.fill()
+            } else {
+                ctx.strokeStyle = 'red'
+                ctx.moveTo(0,0); ctx.lineTo(0,SIZE); ctx.lineTo(SIZE,SIZE); ctx.lineTo(SIZE,0); ctx.lineTo(0,0); ctx.lineTo(SIZE,SIZE);
+                ctx.stroke()
+                addLogMessage('<i>Warning: maximum ratio is &#8805; 100%</i>')
+            }
+        }
+    }
+
+    function updateChart() {
+        let fn = (x) => x
+        switch (document.querySelector('#merge-interpolation').value) {
+            case 'SmoothStep':
+                fn = smoothstep
+                break
+            case 'SmootherStep':
+                fn = smootherstep
+                break
+            case 'SmoothestStep':
+                fn = smootheststep
+                break
+        }
+        drawDiagram(fn)
+    }
+
+    /////////////////////// Tab implementation
+    document.querySelector('.tab-container')?.insertAdjacentHTML('beforeend', `
+        <span id="tab-merge" class="tab">
+            <span><i class="fa fa-code-merge icon"></i> Merge models</span>
+        </span>
+    `)
+
+    document.querySelector('#tab-content-wrapper')?.insertAdjacentHTML('beforeend', `
+        <div id="tab-content-merge" class="tab-content">
+            <div id="merge" class="tab-content-inner">
+                Loading..
+            </div>
+        </div>
+    `)
+
+    const tabMerge = document.querySelector('#tab-merge')
+    if (tabMerge) {
+        linkTabContents(tabMerge)
+    }
+    const merge = document.querySelector('#merge')
+    if (!merge) {
+        // merge tab not found, dont exec plugin code.
+        return
+    }
+
+    document.querySelector('body').insertAdjacentHTML('beforeend', `
+        <style>
+        #tab-content-merge .tab-content-inner {
+            max-width: 100%;
+            padding: 10pt;
+        }
+        .merge-container {  
+            margin-left: 15%;
+            margin-right: 15%;
+            text-align: left;
+            display: inline-grid;
+            grid-template-columns: 1fr 1fr;
+            grid-template-rows: auto auto auto;
+            gap: 0px 0px;
+            grid-auto-flow: row;
+            grid-template-areas:
+              "merge-input merge-config"
+              "merge-buttons merge-buttons";
+        }
+        .merge-container p {
+            margin-top: 3pt;
+            margin-bottom: 3pt;
+        }
+        .merge-config .tab-content {
+            background: var(--background-color1);
+            border-radius: 3pt;
+        }
+        .merge-config .tab-content-inner {
+            text-align: left;
+        }
+
+        .merge-input { 
+            grid-area: merge-input; 
+            padding-left:1em;
+        }
+        .merge-config { 
+            grid-area: merge-config; 
+            padding:1em;
+        }
+        .merge-config input { 
+            margin-bottom: 3px; 
+        }
+        .merge-config select { 
+            margin-bottom: 3px; 
+        }
+        .merge-buttons { 
+            grid-area: merge-buttons; 
+            padding:1em; 
+            text-align: center;
+        }
+        #merge-button  { 
+            padding: 8px; 
+            width:20em; 
+        }
+        div#merge-log {
+            height:150px; 
+            overflow-x:hidden;
+            overflow-y:scroll;
+            background:var(--background-color1);
+            border-radius: 3pt;
+        }
+        div#merge-log i {
+            color: hsl(var(--accent-hue), 100%, calc(2*var(--accent-lightness))); 
+            font-family: monospace;
+        }
+        .disabled { 
+            background: var(--background-color4); 
+            color: var(--text-color); 
+        }
+        #merge-type-tabs {
+            border-bottom: 1px solid black;
+        }
+        #merge-log-container {
+            display: none;
+        }
+        .merge-container #merge-warning {
+            color: rgb(153, 153, 153);
+        }
+        </style>
+    `)
+
+    merge.innerHTML = `
+    <div class="merge-container panel-box">
+      <div class="merge-input">
+         <p><label for="#mergeModelA">Select Model A:</label></p>
+         <input id="mergeModelA" type="text" spellcheck="false" autocomplete="off" class="model-filter" data-path="" />
+         <p><label for="#mergeModelB">Select Model B:</label></p>
+         <input id="mergeModelB" type="text" spellcheck="false" autocomplete="off" class="model-filter" data-path="" />
+         <br/><br/>
+         <p id="merge-warning"><small><b>Important:</b> Please merge models of similar type.<br/>For e.g. <code>SD 1.4</code> models with only <code>SD 1.4/1.5</code> models,<br/><code>SD 2.0</code> with <code>SD 2.0</code>-type, and <code>SD 2.1</code> with <code>SD 2.1</code>-type models.</small></p>
+         <br/>
+         <table>
+            <tr>
+                <td><label for="#merge-filename">Output file name:</label></td>
+                <td><input id="merge-filename" size=24> <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Base name of the output file.<br>Mix ratio and file suffix will be appended to this.</span></i></td>
+            </tr>
+            <tr>
+                <td><label for="#merge-fp">Output precision:</label></td>
+                <td><select id="merge-fp">
+                    <option value="fp16">fp16 (smaller file size)</option>
+                    <option value="fp32">fp32 (larger file size)</option>
+                </select>
+                <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Image generation uses fp16, so it's a good choice.<br>Use fp32 if you want to use the result models for more mixes</span></i>
+                </td>
+            </tr>
+            <tr>
+                <td><label for="#merge-format">Output file format:</label></td>
+                <td><select id="merge-format">
+                    <option value="safetensors">Safetensors (recommended)</option>
+                    <option value="ckpt">CKPT/Pickle (legacy format)</option>
+                </select>
+                </td>
+            </tr>
+         </table>
+         <br/>
+         <div id="merge-log-container">
+            <p><label for="#merge-log">Log messages:</label></p>
+            <div id="merge-log"></div>
+         </div>
+      </div>
+      <div class="merge-config">
+         <div class="tab-container">
+            <span id="tab-merge-opts-single" class="tab active">
+                <span>Make a single file</small></span>
+            </span>
+            <span id="tab-merge-opts-batch" class="tab">
+                <span>Make multiple variations</small></span>
+            </span>
+         </div>
+         <div>
+            <div id="tab-content-merge-opts-single" class="tab-content active">
+                <div class="tab-content-inner">
+                    <small>Saves a single merged model file, at the specified merge ratio.</small><br/><br/>
+                    <label for="#single-merge-ratio-slider">Merge ratio:</label>
+                    <input id="single-merge-ratio-slider" name="single-merge-ratio-slider" class="editor-slider" value="50" type="range" min="1" max="1000">
+                    <input id="single-merge-ratio" size=2 value="5">%
+                    <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Model A's contribution to the mix. The rest will be from Model B.</span></i>
+                </div>
+            </div>
+            <div id="tab-content-merge-opts-batch" class="tab-content">
+                <div class="tab-content-inner">
+                    <small>Saves multiple variations of the model, at different merge ratios.<br/>Each variation will be saved as a separate file.</small><br/><br/>
+                    <table>
+                        <tr><td><label for="#merge-count">Number of variations:</label></td>
+                            <td> <input id="merge-count" size=2 value="5"></td>
+                            <td> <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Number of models to create</span></i></td></tr>
+                        <tr><td><label for="#merge-start">Starting merge ratio:</label></td>
+                            <td> <input id="merge-start" size=2 value="5">%</td>
+                            <td> <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Smallest share of model A in the mix</span></i></td></tr>
+                        <tr><td><label for="#merge-step">Increment each step:</label></td>
+                            <td> <input id="merge-step" size=2 value="10">%</td>
+                            <td> <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Share of model A added into the mix per step</span></i></td></tr>
+                        <tr><td><label for="#merge-interpolation">Interpolation model:</label></td>
+                            <td> <select id="merge-interpolation">
+                                <option>Exact</option>
+                                <option>SmoothStep</option>
+                                <option>SmootherStep</option>
+                                <option>SmoothestStep</option>
+                            </select></td>
+                            <td> <i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip top-left">Sigmoid function to be applied to the model share before mixing</span></i></td></tr>
+                    </table>
+                    <br/>
+                    <small>Preview of variation ratios:</small><br/>
+                    <canvas id="merge-canvas" width="400" height="400"></canvas>
+                </div>
+            </div>
+         </div>
+      </div>
+      <div class="merge-buttons">
+         <button id="merge-button" class="primaryButton">Merge models</button>
+      </div>
+    </div>`
+
+    const tabSettingsSingle = document.querySelector('#tab-merge-opts-single')
+    const tabSettingsBatch = document.querySelector('#tab-merge-opts-batch')
+    linkTabContents(tabSettingsSingle)
+    linkTabContents(tabSettingsBatch)
+
+    console.log('Activate')
+    let mergeModelAField = new ModelDropdown(document.querySelector('#mergeModelA'), 'stable-diffusion')
+    let mergeModelBField = new ModelDropdown(document.querySelector('#mergeModelB'), 'stable-diffusion')
+    updateChart()
+
+    // slider
+    const singleMergeRatioField = document.querySelector('#single-merge-ratio')
+    const singleMergeRatioSlider = document.querySelector('#single-merge-ratio-slider')
+
+    function updateSingleMergeRatio() {
+        singleMergeRatioField.value = singleMergeRatioSlider.value / 10
+        singleMergeRatioField.dispatchEvent(new Event("change"))
+    }
+
+    function updateSingleMergeRatioSlider() {
+        if (singleMergeRatioField.value < 0) {
+            singleMergeRatioField.value = 0
+        } else if (singleMergeRatioField.value > 100) {
+            singleMergeRatioField.value = 100
+        }
+
+        singleMergeRatioSlider.value = singleMergeRatioField.value * 10
+        singleMergeRatioSlider.dispatchEvent(new Event("change"))
+    }
+
+    singleMergeRatioSlider.addEventListener('input', updateSingleMergeRatio)
+    singleMergeRatioField.addEventListener('input', updateSingleMergeRatioSlider)
+    updateSingleMergeRatio()
+
+    document.querySelector('.merge-config').addEventListener('change', updateChart)
+
+    document.querySelector('#merge-button').addEventListener('click', async function(e) {
+        // Build request template
+        let model0 = document.querySelector('#mergeModelA').value
+        let model1 = document.querySelector('#mergeModelB').value
+        let request = { model0: model0, model1: model1 }
+        request['use_fp16'] = document.querySelector('#merge-fp').value == 'fp16'
+        let iterations = document.querySelector('#merge-count').value>>0
+        let start = parseFloat( document.querySelector('#merge-start').value )
+        let step = parseFloat( document.querySelector('#merge-step').value )
+
+        if (isTabActive(tabSettingsSingle)) {
+            start = parseFloat(singleMergeRatioField.value)
+            step = 0
+            iterations = 1
+            addLogMessage(`merge ratio = ${start}%`)
+        } else {
+            addLogMessage(`start = ${start}%`)
+            addLogMessage(`step  = ${step}%`)
+        }
+
+        if (start + (iterations-1) * step >= 100) {
+            addLogMessage('<i>Aborting: maximum ratio is &#8805; 100%</i>')
+            addLogMessage('Reduce the number of variations or the step size')
+            addLogSeparator()
+            document.querySelector('#merge-count').focus()
+            return
+        }
+
+        if (document.querySelector('#merge-filename').value == "") {
+            addLogMessage('<i>Aborting: No output file name specified</i>')
+            addLogSeparator()
+            document.querySelector('#merge-filename').focus()
+            return
+        }
+        
+        // Disable merge button
+        e.target.disabled=true
+        e.target.classList.add('disabled')
+        let cursor = $("body").css("cursor");
+        let label = document.querySelector('#merge-button').innerHTML
+        $("body").css("cursor", "progress");
+        document.querySelector('#merge-button').innerHTML = 'Merging models ...'
+
+        addLogMessage("Merging models")
+        addLogMessage("Model A: "+model0)
+        addLogMessage("Model B: "+model1)
+
+        // Batch main loop
+        for (let i=0; i<iterations; i++) {
+            let alpha = ( start + i * step ) / 100
+            switch (document.querySelector('#merge-interpolation').value) {
+                case 'SmoothStep':
+                    alpha = smoothstep(alpha)
+                    break
+                case 'SmootherStep':
+                    alpha = smootherstep(alpha)
+                    break
+                case 'SmoothestStep':
+                    alpha = smootheststep(alpha)
+                    break
+            }
+            addLogMessage(`merging batch job ${i+1}/${iterations}, alpha = ${alpha.toFixed(5)}...`)
+
+            request['out_path'] = document.querySelector('#merge-filename').value
+            request['out_path'] += '-' + alpha.toFixed(5) + '.' + document.querySelector('#merge-format').value
+            addLogMessage(`&nbsp;&nbsp;filename: ${request['out_path']}`)
+
+            request['ratio'] = alpha
+            let res = await fetch('/model/merge', {
+                  method: 'POST',
+                  headers: { 'Content-Type': 'application/json' },
+                  body: JSON.stringify(request) })
+            const data = await res.json();
+            addLogMessage(JSON.stringify(data))
+        }
+        addLogMessage("<b>Done.</b> The models have been saved to your <tt>models/stable-diffusion</tt> folder.")
+        addLogSeparator()
+        // Re-enable merge button
+        $("body").css("cursor", cursor);
+        document.querySelector('#merge-button').innerHTML = label
+        e.target.disabled=false
+        e.target.classList.remove('disabled')
+
+        // Update model list
+        stableDiffusionModelField.innerHTML = ''
+        vaeModelField.innerHTML = ''
+        hypernetworkModelField.innerHTML = ''
+        await getModels()
+    })
+
+})()

ui/plugins/ui/modifiers-toggle.plugin.js

@@ -38,15 +38,15 @@
                     i.parentElement.classList.add('modifier-toggle-inactive')
                 }
                 // refresh activeTags
-                let modifierName = i.parentElement.getElementsByClassName('modifier-card-label')[0].getElementsByTagName("p")[0].innerText
+                let modifierName = i.parentElement.getElementsByClassName('modifier-card-label')[0].getElementsByTagName("p")[0].dataset.fullName
                 activeTags = activeTags.map(obj => {
-                    if (obj.name === modifierName) {
+                    if (trimModifiers(obj.name) === trimModifiers(modifierName)) {
                         return {...obj, inactive: (obj.element.classList.contains('modifier-toggle-inactive'))};
                     }
                     
                     return obj;
                 });
-                console.log(activeTags)
+                document.dispatchEvent(new Event('refreshImageModifiers'))
             }
         })
     }

ui/plugins/ui/release-notes.plugin.js

@@ -9,7 +9,7 @@
         }
     }
 
-    document.querySelector('#tab-container')?.insertAdjacentHTML('beforeend', `
+    document.querySelector('.tab-container')?.insertAdjacentHTML('beforeend', `
         <span id="tab-news" class="tab">
             <span><i class="fa fa-bolt icon"></i> What's new?</span>
         </span>

ui/sd_internal/__init__.py

@@ -1,119 +0,0 @@
-import json
-
-class Request:
-    request_id: str = None
-    session_id: str = "session"
-    prompt: str = ""
-    negative_prompt: str = ""
-    init_image: str = None # base64
-    mask: str = None # base64
-    num_outputs: int = 1
-    num_inference_steps: int = 50
-    guidance_scale: float = 7.5
-    width: int = 512
-    height: int = 512
-    seed: int = 42
-    prompt_strength: float = 0.8
-    sampler: str = None # "ddim", "plms", "heun", "euler", "euler_a", "dpm2", "dpm2_a", "lms"
-    # allow_nsfw: bool = False
-    precision: str = "autocast" # or "full"
-    save_to_disk_path: str = None
-    turbo: bool = True
-    use_full_precision: bool = False
-    use_face_correction: str = None # or "GFPGANv1.3"
-    use_upscale: str = None # or "RealESRGAN_x4plus" or "RealESRGAN_x4plus_anime_6B"
-    use_stable_diffusion_model: str = "sd-v1-4"
-    use_vae_model: str = None
-    use_hypernetwork_model: str = None
-    hypernetwork_strength: float = 1
-    show_only_filtered_image: bool = False
-    output_format: str = "jpeg" # or "png"
-    output_quality: int = 75
-
-    stream_progress_updates: bool = False
-    stream_image_progress: bool = False
-
-    def json(self):
-        return {
-            "session_id": self.session_id,
-            "prompt": self.prompt,
-            "negative_prompt": self.negative_prompt,
-            "num_outputs": self.num_outputs,
-            "num_inference_steps": self.num_inference_steps,
-            "guidance_scale": self.guidance_scale,
-            "hypernetwork_strengtgh": self.guidance_scale,
-            "width": self.width,
-            "height": self.height,
-            "seed": self.seed,
-            "prompt_strength": self.prompt_strength,
-            "sampler": self.sampler,
-            "use_face_correction": self.use_face_correction,
-            "use_upscale": self.use_upscale,
-            "use_stable_diffusion_model": self.use_stable_diffusion_model,
-            "use_vae_model": self.use_vae_model,
-            "use_hypernetwork_model": self.use_hypernetwork_model,
-            "hypernetwork_strength": self.hypernetwork_strength,
-            "output_format": self.output_format,
-            "output_quality": self.output_quality,
-        }
-
-    def __str__(self):
-        return f'''
-    session_id: {self.session_id}
-    prompt: {self.prompt}
-    negative_prompt: {self.negative_prompt}
-    seed: {self.seed}
-    num_inference_steps: {self.num_inference_steps}
-    sampler: {self.sampler}
-    guidance_scale: {self.guidance_scale}
-    w: {self.width}
-    h: {self.height}
-    precision: {self.precision}
-    save_to_disk_path: {self.save_to_disk_path}
-    turbo: {self.turbo}
-    use_full_precision: {self.use_full_precision}
-    use_face_correction: {self.use_face_correction}
-    use_upscale: {self.use_upscale}
-    use_stable_diffusion_model: {self.use_stable_diffusion_model}
-    use_vae_model: {self.use_vae_model}
-    use_hypernetwork_model: {self.use_hypernetwork_model}
-    hypernetwork_strength: {self.hypernetwork_strength}
-    show_only_filtered_image: {self.show_only_filtered_image}
-    output_format: {self.output_format}
-    output_quality: {self.output_quality}
-
-    stream_progress_updates: {self.stream_progress_updates}
-    stream_image_progress: {self.stream_image_progress}'''
-
-class Image:
-    data: str # base64
-    seed: int
-    is_nsfw: bool
-    path_abs: str = None
-
-    def __init__(self, data, seed):
-        self.data = data
-        self.seed = seed
-
-    def json(self):
-        return {
-            "data": self.data,
-            "seed": self.seed,
-            "path_abs": self.path_abs,
-        }
-
-class Response:
-    request: Request
-    images: list
-
-    def json(self):
-        res = {
-            "status": 'succeeded',
-            "request": self.request.json(),
-            "output": [],
-        }
-
-        for image in self.images:
-            res["output"].append(image.json())
-
-        return res

ui/sd_internal/ddim_callback.patch

@@ -1,162 +0,0 @@
-diff --git a/optimizedSD/ddpm.py b/optimizedSD/ddpm.py
-index 79058bc..a473411 100644
---- a/optimizedSD/ddpm.py
-+++ b/optimizedSD/ddpm.py
-@@ -564,12 +564,12 @@ class UNet(DDPM):
-                                         unconditional_guidance_scale=unconditional_guidance_scale,
-                                         callback=callback, img_callback=img_callback)
- 
-+        yield from samples
-+
-         if(self.turbo):
-             self.model1.to("cpu")
-             self.model2.to("cpu")
- 
--        return samples
--
-     @torch.no_grad()
-     def plms_sampling(self, cond,b, img,
-                       ddim_use_original_steps=False,
-@@ -608,10 +608,10 @@ class UNet(DDPM):
-             old_eps.append(e_t)
-             if len(old_eps) >= 4:
-                 old_eps.pop(0)
--            if callback: callback(i)
--            if img_callback: img_callback(pred_x0, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(pred_x0, i)
- 
--        return img
-+        yield from img_callback(img, len(iterator)-1)
- 
-     @torch.no_grad()
-     def p_sample_plms(self, x, c, t, index, repeat_noise=False, use_original_steps=False, quantize_denoised=False,
-@@ -740,13 +740,13 @@ class UNet(DDPM):
-                                           unconditional_guidance_scale=unconditional_guidance_scale,
-                                           unconditional_conditioning=unconditional_conditioning)
- 
--            if callback: callback(i)
--            if img_callback: img_callback(x_dec, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(x_dec, i)
-         
-         if mask is not None:
--            return x0 * mask + (1. - mask) * x_dec
-+            x_dec = x0 * mask + (1. - mask) * x_dec
- 
--        return x_dec
-+        yield from img_callback(x_dec, len(iterator)-1)
- 
- 
-     @torch.no_grad()
-@@ -820,12 +820,12 @@ class UNet(DDPM):
- 
- 
-             d = to_d(x, sigma_hat, denoised)
--            if callback: callback(i)
--            if img_callback: img_callback(x, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(x, i)
-             dt = sigmas[i + 1] - sigma_hat
-             # Euler method
-             x = x + d * dt
--        return x
-+        yield from img_callback(x, len(sigmas)-1)
- 
-     @torch.no_grad()
-     def euler_ancestral_sampling(self,ac,x, S, cond, unconditional_conditioning = None, unconditional_guidance_scale = 1,extra_args=None, callback=None, disable=None, img_callback=None):
-@@ -852,14 +852,14 @@ class UNet(DDPM):
-             denoised = e_t_uncond + unconditional_guidance_scale * (e_t - e_t_uncond)
- 
-             sigma_down, sigma_up = get_ancestral_step(sigmas[i], sigmas[i + 1])
--            if callback: callback(i)
--            if img_callback: img_callback(x, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(x, i)
-             d = to_d(x, sigmas[i], denoised)
-             # Euler method
-             dt = sigma_down - sigmas[i]
-             x = x + d * dt
-             x = x + torch.randn_like(x) * sigma_up
--        return x
-+        yield from img_callback(x, len(sigmas)-1)
- 
- 
- 
-@@ -892,8 +892,8 @@ class UNet(DDPM):
-             denoised = e_t_uncond + unconditional_guidance_scale * (e_t - e_t_uncond)
- 
-             d = to_d(x, sigma_hat, denoised)
--            if callback: callback(i)
--            if img_callback: img_callback(x, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(x, i)
-             dt = sigmas[i + 1] - sigma_hat
-             if sigmas[i + 1] == 0:
-                 # Euler method
-@@ -913,7 +913,7 @@ class UNet(DDPM):
-                 d_2 = to_d(x_2, sigmas[i + 1], denoised_2)
-                 d_prime = (d + d_2) / 2
-                 x = x + d_prime * dt
--        return x
-+        yield from img_callback(x, len(sigmas)-1)
- 
- 
-     @torch.no_grad()
-@@ -944,8 +944,8 @@ class UNet(DDPM):
-             e_t_uncond, e_t = (x_in  + eps * c_out).chunk(2)
-             denoised = e_t_uncond + unconditional_guidance_scale * (e_t - e_t_uncond)
- 
--            if callback: callback(i)
--            if img_callback: img_callback(x, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(x, i)
-             
-             d = to_d(x, sigma_hat, denoised)
-             # Midpoint method, where the midpoint is chosen according to a rho=3 Karras schedule
-@@ -966,7 +966,7 @@ class UNet(DDPM):
- 
-             d_2 = to_d(x_2, sigma_mid, denoised_2)
-             x = x + d_2 * dt_2
--        return x
-+        yield from img_callback(x, len(sigmas)-1)
- 
- 
-     @torch.no_grad()
-@@ -994,8 +994,8 @@ class UNet(DDPM):
- 
- 
-             sigma_down, sigma_up = get_ancestral_step(sigmas[i], sigmas[i + 1])
--            if callback: callback(i)
--            if img_callback: img_callback(x, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(x, i)
-             d = to_d(x, sigmas[i], denoised)
-             # Midpoint method, where the midpoint is chosen according to a rho=3 Karras schedule
-             sigma_mid = ((sigmas[i] ** (1 / 3) + sigma_down ** (1 / 3)) / 2) ** 3
-@@ -1016,7 +1016,7 @@ class UNet(DDPM):
-             d_2 = to_d(x_2, sigma_mid, denoised_2)
-             x = x + d_2 * dt_2
-             x = x + torch.randn_like(x) * sigma_up
--        return x
-+        yield from img_callback(x, len(sigmas)-1)
- 
- 
-     @torch.no_grad()
-@@ -1042,8 +1042,8 @@ class UNet(DDPM):
-             e_t_uncond, e_t = (x_in  + eps * c_out).chunk(2)
-             denoised = e_t_uncond + unconditional_guidance_scale * (e_t - e_t_uncond)
- 
--            if callback: callback(i)
--            if img_callback: img_callback(x, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(x, i)
- 
-             d = to_d(x, sigmas[i], denoised)
-             ds.append(d)
-@@ -1054,4 +1054,4 @@ class UNet(DDPM):
-             cur_order = min(i + 1, order)
-             coeffs = [linear_multistep_coeff(cur_order, sigmas.cpu(), i, j) for j in range(cur_order)]
-             x = x + sum(coeff * d for coeff, d in zip(coeffs, reversed(ds)))
--        return x
-+        yield from img_callback(x, len(sigmas)-1)

ui/sd_internal/ddim_callback_sd2.patch

@@ -1,84 +0,0 @@
-diff --git a/ldm/models/diffusion/ddim.py b/ldm/models/diffusion/ddim.py
-index 27ead0e..6215939 100644
---- a/ldm/models/diffusion/ddim.py
-+++ b/ldm/models/diffusion/ddim.py
-@@ -100,7 +100,7 @@ class DDIMSampler(object):
-         size = (batch_size, C, H, W)
-         print(f'Data shape for DDIM sampling is {size}, eta {eta}')
- 
--        samples, intermediates = self.ddim_sampling(conditioning, size,
-+        samples = self.ddim_sampling(conditioning, size,
-                                                     callback=callback,
-                                                     img_callback=img_callback,
-                                                     quantize_denoised=quantize_x0,
-@@ -117,7 +117,8 @@ class DDIMSampler(object):
-                                                     dynamic_threshold=dynamic_threshold,
-                                                     ucg_schedule=ucg_schedule
-                                                     )
--        return samples, intermediates
-+        # return samples, intermediates
-+        yield from samples
- 
-     @torch.no_grad()
-     def ddim_sampling(self, cond, shape,
-@@ -168,14 +169,15 @@ class DDIMSampler(object):
-                                       unconditional_conditioning=unconditional_conditioning,
-                                       dynamic_threshold=dynamic_threshold)
-             img, pred_x0 = outs
--            if callback: callback(i)
--            if img_callback: img_callback(pred_x0, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(pred_x0, i)
- 
-             if index % log_every_t == 0 or index == total_steps - 1:
-                 intermediates['x_inter'].append(img)
-                 intermediates['pred_x0'].append(pred_x0)
- 
--        return img, intermediates
-+        # return img, intermediates
-+        yield from img_callback(pred_x0, len(iterator)-1)
- 
-     @torch.no_grad()
-     def p_sample_ddim(self, x, c, t, index, repeat_noise=False, use_original_steps=False, quantize_denoised=False,
-diff --git a/ldm/models/diffusion/plms.py b/ldm/models/diffusion/plms.py
-index 7002a36..0951f39 100644
---- a/ldm/models/diffusion/plms.py
-+++ b/ldm/models/diffusion/plms.py
-@@ -96,7 +96,7 @@ class PLMSSampler(object):
-         size = (batch_size, C, H, W)
-         print(f'Data shape for PLMS sampling is {size}')
- 
--        samples, intermediates = self.plms_sampling(conditioning, size,
-+        samples = self.plms_sampling(conditioning, size,
-                                                     callback=callback,
-                                                     img_callback=img_callback,
-                                                     quantize_denoised=quantize_x0,
-@@ -112,7 +112,8 @@ class PLMSSampler(object):
-                                                     unconditional_conditioning=unconditional_conditioning,
-                                                     dynamic_threshold=dynamic_threshold,
-                                                     )
--        return samples, intermediates
-+        #return samples, intermediates
-+        yield from samples
- 
-     @torch.no_grad()
-     def plms_sampling(self, cond, shape,
-@@ -165,14 +166,15 @@ class PLMSSampler(object):
-             old_eps.append(e_t)
-             if len(old_eps) >= 4:
-                 old_eps.pop(0)
--            if callback: callback(i)
--            if img_callback: img_callback(pred_x0, i)
-+            if callback: yield from callback(i)
-+            if img_callback: yield from img_callback(pred_x0, i)
- 
-             if index % log_every_t == 0 or index == total_steps - 1:
-                 intermediates['x_inter'].append(img)
-                 intermediates['pred_x0'].append(pred_x0)
- 
--        return img, intermediates
-+        # return img, intermediates
-+        yield from img_callback(pred_x0, len(iterator)-1)
- 
-     @torch.no_grad()
-     def p_sample_plms(self, x, c, t, index, repeat_noise=False, use_original_steps=False, quantize_denoised=False,

ui/sd_internal/device_manager.py

@@ -1,168 +0,0 @@
-import os
-import torch
-import traceback
-import re
-
-COMPARABLE_GPU_PERCENTILE = 0.65 # if a GPU's free_mem is within this % of the GPU with the most free_mem, it will be picked
-
-mem_free_threshold = 0
-
-def get_device_delta(render_devices, active_devices):
-    '''
-    render_devices: 'cpu', or 'auto' or ['cuda:N'...]
-    active_devices: ['cpu', 'cuda:N'...]
-    '''
-
-    if render_devices in ('cpu', 'auto'):
-        render_devices = [render_devices]
-    elif render_devices is not None:
-        if isinstance(render_devices, str):
-            render_devices = [render_devices]
-        if isinstance(render_devices, list) and len(render_devices) > 0:
-            render_devices = list(filter(lambda x: x.startswith('cuda:'), render_devices))
-            if len(render_devices) == 0:
-                raise Exception('Invalid render_devices value in config.json. Valid: {"render_devices": ["cuda:0", "cuda:1"...]}, or {"render_devices": "cpu"} or {"render_devices": "auto"}')
-
-            render_devices = list(filter(lambda x: is_device_compatible(x), render_devices))
-            if len(render_devices) == 0:
-                raise Exception('Sorry, none of the render_devices configured in config.json are compatible with Stable Diffusion')
-        else:
-            raise Exception('Invalid render_devices value in config.json. Valid: {"render_devices": ["cuda:0", "cuda:1"...]}, or {"render_devices": "cpu"} or {"render_devices": "auto"}')
-    else:
-        render_devices = ['auto']
-
-    if 'auto' in render_devices:
-        render_devices = auto_pick_devices(active_devices)
-        if 'cpu' in render_devices:
-            print('WARNING: Could not find a compatible GPU. Using the CPU, but this will be very slow!')
-
-    active_devices = set(active_devices)
-    render_devices = set(render_devices)
-
-    devices_to_start = render_devices - active_devices
-    devices_to_stop = active_devices - render_devices
-
-    return devices_to_start, devices_to_stop
-
-def auto_pick_devices(currently_active_devices):
-    global mem_free_threshold
-
-    if not torch.cuda.is_available(): return ['cpu']
-
-    device_count = torch.cuda.device_count()
-    if device_count == 1:
-        return ['cuda:0'] if is_device_compatible('cuda:0') else ['cpu']
-
-    print('Autoselecting GPU. Using most free memory.')
-    devices = []
-    for device in range(device_count):
-        device = f'cuda:{device}'
-        if not is_device_compatible(device):
-            continue
-
-        mem_free, mem_total = torch.cuda.mem_get_info(device)
-        mem_free /= float(10**9)
-        mem_total /= float(10**9)
-        device_name = torch.cuda.get_device_name(device)
-        print(f'{device} detected: {device_name} - Memory (free/total): {round(mem_free, 2)}Gb / {round(mem_total, 2)}Gb')
-        devices.append({'device': device, 'device_name': device_name, 'mem_free': mem_free})
-
-    devices.sort(key=lambda x:x['mem_free'], reverse=True)
-    max_mem_free = devices[0]['mem_free']
-    curr_mem_free_threshold = COMPARABLE_GPU_PERCENTILE * max_mem_free
-    mem_free_threshold = max(curr_mem_free_threshold, mem_free_threshold)
-
-    # Auto-pick algorithm:
-    # 1. Pick the top 75 percentile of the GPUs, sorted by free_mem.
-    # 2. Also include already-running devices (GPU-only), otherwise their free_mem will
-    #    always be very low (since their VRAM contains the model).
-    #    These already-running devices probably aren't terrible, since they were picked in the past.
-    #    Worst case, the user can restart the program and that'll get rid of them.
-    devices = list(filter((lambda x: x['mem_free'] > mem_free_threshold or x['device'] in currently_active_devices), devices))
-    devices = list(map(lambda x: x['device'], devices))
-    return devices
-
-def device_init(thread_data, device):
-    '''
-    This function assumes the 'device' has already been verified to be compatible.
-    `get_device_delta()` has already filtered out incompatible devices.
-    '''
-
-    validate_device_id(device, log_prefix='device_init')
-
-    if device == 'cpu':
-        thread_data.device = 'cpu'
-        thread_data.device_name = get_processor_name()
-        print('Render device CPU available as', thread_data.device_name)
-        return
-
-    thread_data.device_name = torch.cuda.get_device_name(device)
-    thread_data.device = device
-
-    # Force full precision on 1660 and 1650 NVIDIA cards to avoid creating green images
-    device_name = thread_data.device_name.lower()
-    thread_data.force_full_precision = (('nvidia' in device_name or 'geforce' in device_name) and (' 1660' in device_name or ' 1650' in device_name)) or ('Quadro T2000' in device_name)
-    if thread_data.force_full_precision:
-        print('forcing full precision on NVIDIA 16xx cards, to avoid green images. GPU detected: ', thread_data.device_name)
-        # Apply force_full_precision now before models are loaded.
-        thread_data.precision = 'full'
-
-    print(f'Setting {device} as active')
-    torch.cuda.device(device)
-
-    return
-
-def validate_device_id(device, log_prefix=''):
-    def is_valid():
-        if not isinstance(device, str):
-            return False
-        if device == 'cpu':
-            return True
-        if not device.startswith('cuda:') or not device[5:].isnumeric():
-            return False
-        return True
-
-    if not is_valid():
-        raise EnvironmentError(f"{log_prefix}: device id should be 'cpu', or 'cuda:N' (where N is an integer index for the GPU). Got: {device}")
-
-def is_device_compatible(device):
-    '''
-    Returns True/False, and prints any compatibility errors
-    '''
-    try:
-        validate_device_id(device, log_prefix='is_device_compatible')
-    except:
-        print(str(e))
-        return False
-
-    if device == 'cpu': return True
-    # Memory check
-    try:
-        _, mem_total = torch.cuda.mem_get_info(device)
-        mem_total /= float(10**9)
-        if mem_total < 3.0:
-            print(f'GPU {device} with less than 3 GB of VRAM is not compatible with Stable Diffusion')
-            return False
-    except RuntimeError as e:
-        print(str(e))
-        return False
-    return True
-
-def get_processor_name():
-    try:
-        import platform, subprocess
-        if platform.system() == "Windows":
-            return platform.processor()
-        elif platform.system() == "Darwin":
-            os.environ['PATH'] = os.environ['PATH'] + os.pathsep + '/usr/sbin'
-            command = "sysctl -n machdep.cpu.brand_string"
-            return subprocess.check_output(command).strip()
-        elif platform.system() == "Linux":
-            command = "cat /proc/cpuinfo"
-            all_info = subprocess.check_output(command, shell=True).decode().strip()
-            for line in all_info.split("\n"):
-                if "model name" in line:
-                    return re.sub(".*model name.*:", "", line, 1).strip()
-    except:
-        print(traceback.format_exc())
-        return "cpu"

ui/sd_internal/hypernetwork.py

@@ -1,198 +0,0 @@
-# this is basically a cut down version of https://github.com/AUTOMATIC1111/stable-diffusion-webui/blob/c9a2cfdf2a53d37c2de1908423e4f548088667ef/modules/hypernetworks/hypernetwork.py, mostly for feature parity
-# I, c0bra5, don't really understand how deep learning works. I just know how to port stuff.
-
-import inspect
-import torch
-import optimizedSD.splitAttention
-from . import runtime
-from einops import rearrange
-
-optimizer_dict = {optim_name : cls_obj for optim_name, cls_obj in inspect.getmembers(torch.optim, inspect.isclass) if optim_name != "Optimizer"}
-
-loaded_hypernetwork = None
-
-class HypernetworkModule(torch.nn.Module):
-    multiplier = 0.5
-    activation_dict = {
-        "linear": torch.nn.Identity,
-        "relu": torch.nn.ReLU,
-        "leakyrelu": torch.nn.LeakyReLU,
-        "elu": torch.nn.ELU,
-        "swish": torch.nn.Hardswish,
-        "tanh": torch.nn.Tanh,
-        "sigmoid": torch.nn.Sigmoid,
-    }
-    activation_dict.update({cls_name.lower(): cls_obj for cls_name, cls_obj in inspect.getmembers(torch.nn.modules.activation) if inspect.isclass(cls_obj) and cls_obj.__module__ == 'torch.nn.modules.activation'})
-
-    def __init__(self, dim, state_dict=None, layer_structure=None, activation_func=None, weight_init='Normal',
-                 add_layer_norm=False, use_dropout=False, activate_output=False, last_layer_dropout=False):
-        super().__init__()
-
-        assert layer_structure is not None, "layer_structure must not be None"
-        assert layer_structure[0] == 1, "Multiplier Sequence should start with size 1!"
-        assert layer_structure[-1] == 1, "Multiplier Sequence should end with size 1!"
-
-        linears = []
-        for i in range(len(layer_structure) - 1):
-
-            # Add a fully-connected layer
-            linears.append(torch.nn.Linear(int(dim * layer_structure[i]), int(dim * layer_structure[i+1])))
-
-            # Add an activation func except last layer
-            if activation_func == "linear" or activation_func is None or (i >= len(layer_structure) - 2 and not activate_output):
-                pass
-            elif activation_func in self.activation_dict:
-                linears.append(self.activation_dict[activation_func]())
-            else:
-                raise RuntimeError(f'hypernetwork uses an unsupported activation function: {activation_func}')
-
-            # Add layer normalization
-            if add_layer_norm:
-                linears.append(torch.nn.LayerNorm(int(dim * layer_structure[i+1])))
-
-            # Add dropout except last layer
-            if use_dropout and (i < len(layer_structure) - 3 or last_layer_dropout and i < len(layer_structure) - 2):
-                linears.append(torch.nn.Dropout(p=0.3))
-
-        self.linear = torch.nn.Sequential(*linears)
-
-        self.fix_old_state_dict(state_dict)
-        self.load_state_dict(state_dict)
-
-        self.to(runtime.thread_data.device)
-
-    def fix_old_state_dict(self, state_dict):
-        changes = {
-            'linear1.bias': 'linear.0.bias',
-            'linear1.weight': 'linear.0.weight',
-            'linear2.bias': 'linear.1.bias',
-            'linear2.weight': 'linear.1.weight',
-        }
-
-        for fr, to in changes.items():
-            x = state_dict.get(fr, None)
-            if x is None:
-                continue
-
-            del state_dict[fr]
-            state_dict[to] = x
-
-    def forward(self, x: torch.Tensor):
-        return x + self.linear(x) * runtime.thread_data.hypernetwork_strength
-
-def apply_hypernetwork(hypernetwork, context, layer=None):
-    hypernetwork_layers = hypernetwork.get(context.shape[2], None)
-
-    if hypernetwork_layers is None:
-        return context, context
-
-    if layer is not None:
-        layer.hyper_k = hypernetwork_layers[0]
-        layer.hyper_v = hypernetwork_layers[1]
-
-    context_k = hypernetwork_layers[0](context)
-    context_v = hypernetwork_layers[1](context)
-    return context_k, context_v
-
-def get_kv(context, hypernetwork):
-    if hypernetwork is None:
-        return context, context
-    else:
-        return apply_hypernetwork(runtime.thread_data.hypernetwork, context)
-
-# This might need updating as the optimisedSD code changes
-# I think yall have a system for this (patch files in sd_internal) but idk how it works and no amount of searching gave me any clue
-# just in case for attribution https://github.com/easydiffusion/diffusion-kit/blob/e8ea0cadd543056059cd951e76d4744de76327d2/optimizedSD/splitAttention.py#L171
-def new_cross_attention_forward(self, x, context=None, mask=None):
-    h = self.heads
-
-    q = self.to_q(x)
-    # default context
-    context = context if context is not None else x() if inspect.isfunction(x) else x
-    # hypernetwork!
-    context_k, context_v =  get_kv(context, runtime.thread_data.hypernetwork)
-    k = self.to_k(context_k)
-    v = self.to_v(context_v)
-    del context, x
-
-    q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v))
-
-
-    limit = k.shape[0]
-    att_step = self.att_step
-    q_chunks = list(torch.tensor_split(q, limit//att_step, dim=0))
-    k_chunks = list(torch.tensor_split(k, limit//att_step, dim=0))
-    v_chunks = list(torch.tensor_split(v, limit//att_step, dim=0))
-
-    q_chunks.reverse()
-    k_chunks.reverse()
-    v_chunks.reverse()
-    sim = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device)
-    del k, q, v
-    for i in range (0, limit, att_step):
-
-        q_buffer = q_chunks.pop()
-        k_buffer = k_chunks.pop()
-        v_buffer = v_chunks.pop()
-        sim_buffer = torch.einsum('b i d, b j d -> b i j', q_buffer, k_buffer) * self.scale
-
-        del k_buffer, q_buffer
-        # attention, what we cannot get enough of, by chunks
-
-        sim_buffer = sim_buffer.softmax(dim=-1)
-
-        sim_buffer = torch.einsum('b i j, b j d -> b i d', sim_buffer, v_buffer)
-        del v_buffer
-        sim[i:i+att_step,:,:] = sim_buffer
-
-        del sim_buffer
-    sim = rearrange(sim, '(b h) n d -> b n (h d)', h=h)
-    return self.to_out(sim)
-
-
-def load_hypernetwork(path: str):
-
-    state_dict = torch.load(path, map_location='cpu')
-
-    layer_structure = state_dict.get('layer_structure', [1, 2, 1])
-    activation_func = state_dict.get('activation_func', None)
-    weight_init = state_dict.get('weight_initialization', 'Normal')
-    add_layer_norm = state_dict.get('is_layer_norm', False)
-    use_dropout = state_dict.get('use_dropout', False)
-    activate_output = state_dict.get('activate_output', True)
-    last_layer_dropout = state_dict.get('last_layer_dropout', False)
-    # this is a bit verbose so leaving it commented out for the poor soul who ever has to debug this
-    # print(f"layer_structure: {layer_structure}")
-    # print(f"activation_func: {activation_func}")
-    # print(f"weight_init: {weight_init}")
-    # print(f"add_layer_norm: {add_layer_norm}")
-    # print(f"use_dropout: {use_dropout}")
-    # print(f"activate_output: {activate_output}")
-    # print(f"last_layer_dropout: {last_layer_dropout}")
-
-    layers = {}
-    for size, sd in state_dict.items():
-        if type(size) == int:
-            layers[size] = (
-                HypernetworkModule(size, sd[0], layer_structure, activation_func, weight_init, add_layer_norm,
-                                   use_dropout, activate_output, last_layer_dropout=last_layer_dropout),
-                HypernetworkModule(size, sd[1], layer_structure, activation_func, weight_init, add_layer_norm,
-                                   use_dropout, activate_output, last_layer_dropout=last_layer_dropout),
-            )
-    print(f"hypernetwork loaded")
-    return layers
-
-
-
-# overriding of original function
-old_cross_attention_forward = optimizedSD.splitAttention.CrossAttention.forward
-# hijacks the cross attention forward function to add hyper network support
-def hijack_cross_attention():
-    print("hypernetwork functionality added to cross attention")
-    optimizedSD.splitAttention.CrossAttention.forward = new_cross_attention_forward
-# there was a cop on board
-def unhijack_cross_attention_forward():
-    print("hypernetwork functionality removed from cross attention")
-    optimizedSD.splitAttention.CrossAttention.forward = old_cross_attention_forward
-
-hijack_cross_attention()

ui/sd_internal/task_manager.py

@@ -1,609 +0,0 @@
-"""task_manager.py: manage tasks dispatching and render threads.
-Notes:
-    render_threads should be the only hard reference held by the manager to the threads.
-    Use weak_thread_data to store all other data using weak keys.
-    This will allow for garbage collection after the thread dies.
-"""
-import json
-import traceback
-
-TASK_TTL = 15 * 60 # seconds, Discard last session's task timeout
-
-import torch
-import queue, threading, time, weakref
-from typing import Any, Generator, Hashable, Optional, Union
-
-from pydantic import BaseModel
-from sd_internal import Request, Response, runtime, device_manager
-
-THREAD_NAME_PREFIX = 'Runtime-Render/'
-ERR_LOCK_FAILED = ' failed to acquire lock within timeout.'
-LOCK_TIMEOUT = 15 # Maximum locking time in seconds before failing a task.
-# It's better to get an exception than a deadlock... ALWAYS use timeout in critical paths.
-
-DEVICE_START_TIMEOUT = 60 # seconds - Maximum time to wait for a render device to init.
-
-class SymbolClass(type): # Print nicely formatted Symbol names.
-    def __repr__(self): return self.__qualname__
-    def __str__(self): return self.__name__
-class Symbol(metaclass=SymbolClass): pass
-
-class ServerStates:
-    class Init(Symbol): pass
-    class LoadingModel(Symbol): pass
-    class Online(Symbol): pass
-    class Rendering(Symbol): pass
-    class Unavailable(Symbol): pass
-
-class RenderTask(): # Task with output queue and completion lock.
-    def __init__(self, req: Request):
-        req.request_id = id(self)
-        self.request: Request = req  # Initial Request
-        self.response: Any = None # Copy of the last reponse
-        self.render_device = None # Select the task affinity. (Not used to change active devices).
-        self.temp_images:list = [None] * req.num_outputs * (1 if req.show_only_filtered_image else 2)
-        self.error: Exception = None
-        self.lock: threading.Lock = threading.Lock() # Locks at task start and unlocks when task is completed
-        self.buffer_queue: queue.Queue = queue.Queue() # Queue of JSON string segments
-    async def read_buffer_generator(self):
-        try:
-            while not self.buffer_queue.empty():
-                res = self.buffer_queue.get(block=False)
-                self.buffer_queue.task_done()
-                yield res
-        except queue.Empty as e: yield
-    @property
-    def status(self):
-        if self.lock.locked():
-            return 'running'
-        if isinstance(self.error, StopAsyncIteration):
-            return 'stopped'
-        if self.error:
-            return 'error'
-        if not self.buffer_queue.empty():
-            return 'buffer'
-        if self.response:
-            return 'completed'
-        return 'pending'
-    @property
-    def is_pending(self):
-        return bool(not self.response and not self.error)
-
-# defaults from https://huggingface.co/blog/stable_diffusion
-class ImageRequest(BaseModel):
-    session_id: str = "session"
-    prompt: str = ""
-    negative_prompt: str = ""
-    init_image: str = None # base64
-    mask: str = None # base64
-    num_outputs: int = 1
-    num_inference_steps: int = 50
-    guidance_scale: float = 7.5
-    width: int = 512
-    height: int = 512
-    seed: int = 42
-    prompt_strength: float = 0.8
-    sampler: str = None # "ddim", "plms", "heun", "euler", "euler_a", "dpm2", "dpm2_a", "lms"
-    # allow_nsfw: bool = False
-    save_to_disk_path: str = None
-    turbo: bool = True
-    use_cpu: bool = False ##TODO Remove after UI and plugins transition.
-    render_device: str = None # Select the task affinity. (Not used to change active devices).
-    use_full_precision: bool = False
-    use_face_correction: str = None # or "GFPGANv1.3"
-    use_upscale: str = None # or "RealESRGAN_x4plus" or "RealESRGAN_x4plus_anime_6B"
-    use_stable_diffusion_model: str = "sd-v1-4"
-    use_vae_model: str = None
-    use_hypernetwork_model: str = None
-    hypernetwork_strength: float = None
-    show_only_filtered_image: bool = False
-    output_format: str = "jpeg" # or "png"
-    output_quality: int = 75
-
-    stream_progress_updates: bool = False
-    stream_image_progress: bool = False
-
-class FilterRequest(BaseModel):
-    session_id: str = "session"
-    model: str = None
-    name: str = ""
-    init_image: str = None # base64
-    width: int = 512
-    height: int = 512
-    save_to_disk_path: str = None
-    turbo: bool = True
-    render_device: str = None
-    use_full_precision: bool = False
-    output_format: str = "jpeg" # or "png"
-    output_quality: int = 75
-
-# Temporary cache to allow to query tasks results for a short time after they are completed.
-class DataCache():
-    def __init__(self):
-        self._base = dict()
-        self._lock: threading.Lock = threading.Lock()
-    def _get_ttl_time(self, ttl: int) -> int:
-        return int(time.time()) + ttl
-    def _is_expired(self, timestamp: int) -> bool:
-        return int(time.time()) >= timestamp
-    def clean(self) -> None:
-        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('DataCache.clean' + ERR_LOCK_FAILED)
-        try:
-            # Create a list of expired keys to delete
-            to_delete = []
-            for key in self._base:
-                ttl, _ = self._base[key]
-                if self._is_expired(ttl):
-                    to_delete.append(key)
-            # Remove Items
-            for key in to_delete:
-                (_, val) = self._base[key]
-                if isinstance(val, RenderTask):
-                    print(f'RenderTask {key} expired. Data removed.')
-                elif isinstance(val, SessionState):
-                    print(f'Session {key} expired. Data removed.')
-                else:
-                    print(f'Key {key} expired. Data removed.')
-                del self._base[key]
-        finally:
-            self._lock.release()
-    def clear(self) -> None:
-        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('DataCache.clear' + ERR_LOCK_FAILED)
-        try: self._base.clear()
-        finally: self._lock.release()
-    def delete(self, key: Hashable) -> bool:
-        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('DataCache.delete' + ERR_LOCK_FAILED)
-        try:
-            if key not in self._base:
-                return False
-            del self._base[key]
-            return True
-        finally:
-            self._lock.release()
-    def keep(self, key: Hashable, ttl: int) -> bool:
-        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('DataCache.keep' + ERR_LOCK_FAILED)
-        try:
-            if key in self._base:
-                _, value = self._base.get(key)
-                self._base[key] = (self._get_ttl_time(ttl), value)
-                return True
-            return False
-        finally:
-            self._lock.release()
-    def put(self, key: Hashable, value: Any, ttl: int) -> bool:
-        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('DataCache.put' + ERR_LOCK_FAILED)
-        try:
-            self._base[key] = (
-                self._get_ttl_time(ttl), value
-            )
-        except Exception as e:
-            print(str(e))
-            print(traceback.format_exc())
-            return False
-        else:
-            return True
-        finally:
-            self._lock.release()
-    def tryGet(self, key: Hashable) -> Any:
-        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('DataCache.tryGet' + ERR_LOCK_FAILED)
-        try:
-            ttl, value = self._base.get(key, (None, None))
-            if ttl is not None and self._is_expired(ttl):
-                print(f'Session {key} expired. Discarding data.')
-                del self._base[key]
-                return None
-            return value
-        finally:
-            self._lock.release()
-
-manager_lock = threading.RLock()
-render_threads = []
-current_state = ServerStates.Init
-current_state_error:Exception = None
-current_model_path = None
-current_vae_path = None
-current_hypernetwork_path = None
-tasks_queue = []
-session_cache = DataCache()
-task_cache = DataCache()
-default_model_to_load = None
-default_vae_to_load = None
-default_hypernetwork_to_load = None
-weak_thread_data = weakref.WeakKeyDictionary()
-idle_event: threading.Event = threading.Event()
-
-class SessionState():
-    def __init__(self, id: str):
-        self._id = id
-        self._tasks_ids = []
-    @property
-    def id(self):
-        return self._id
-    @property
-    def tasks(self):
-        tasks = []
-        for task_id in self._tasks_ids:
-            task = task_cache.tryGet(task_id)
-            if task:
-                tasks.append(task)
-        return tasks
-    def put(self, task, ttl=TASK_TTL):
-        task_id = id(task)
-        self._tasks_ids.append(task_id)
-        if not task_cache.put(task_id, task, ttl):
-            return False
-        while len(self._tasks_ids) > len(render_threads) * 2:
-            self._tasks_ids.pop(0)
-        return True
-
-def preload_model(ckpt_file_path=None, vae_file_path=None, hypernetwork_file_path=None):
-    global current_state, current_state_error, current_model_path, current_vae_path, current_hypernetwork_path
-    if ckpt_file_path == None:
-        ckpt_file_path = default_model_to_load
-    if vae_file_path == None:
-        vae_file_path = default_vae_to_load
-    if hypernetwork_file_path == None:
-        hypernetwork_file_path = default_hypernetwork_to_load
-    if ckpt_file_path == current_model_path and vae_file_path == current_vae_path:
-        return
-    current_state = ServerStates.LoadingModel
-    try:
-        from . import runtime
-        runtime.thread_data.hypernetwork_file = hypernetwork_file_path
-        runtime.thread_data.ckpt_file = ckpt_file_path
-        runtime.thread_data.vae_file = vae_file_path
-        runtime.load_model_ckpt()
-        runtime.load_hypernetwork()
-        current_model_path = ckpt_file_path
-        current_vae_path = vae_file_path
-        current_hypernetwork_path = hypernetwork_file_path
-        current_state_error = None
-        current_state = ServerStates.Online
-    except Exception as e:
-        current_model_path = None
-        current_vae_path = None
-        current_state_error = e
-        current_state = ServerStates.Unavailable
-        print(traceback.format_exc())
-
-def thread_get_next_task():
-    from . import runtime
-    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
-        print('Render thread on device', runtime.thread_data.device, 'failed to acquire manager lock.')
-        return None
-    if len(tasks_queue) <= 0:
-        manager_lock.release()
-        return None
-    task = None
-    try:  # Select a render task.
-        for queued_task in tasks_queue:
-            if queued_task.render_device and runtime.thread_data.device != queued_task.render_device:
-                # Is asking for a specific render device.
-                if is_alive(queued_task.render_device) > 0:
-                    continue  # requested device alive, skip current one.
-                else:
-                    # Requested device is not active, return error to UI.
-                    queued_task.error = Exception(queued_task.render_device + ' is not currently active.')
-                    task = queued_task
-                    break
-            if not queued_task.render_device and runtime.thread_data.device == 'cpu' and is_alive() > 1:
-                # not asking for any specific devices, cpu want to grab task but other render devices are alive.
-                continue  # Skip Tasks, don't run on CPU unless there is nothing else or user asked for it.
-            task = queued_task
-            break
-        if task is not None:
-            del tasks_queue[tasks_queue.index(task)]
-        return task
-    finally:
-        manager_lock.release()
-
-def thread_render(device):
-    global current_state, current_state_error, current_model_path, current_vae_path, current_hypernetwork_path
-    from . import runtime
-    try:
-        runtime.thread_init(device)
-    except Exception as e:
-        print(traceback.format_exc())
-        weak_thread_data[threading.current_thread()] = {
-            'error': e
-        }
-        return
-    weak_thread_data[threading.current_thread()] = {
-        'device': runtime.thread_data.device,
-        'device_name': runtime.thread_data.device_name,
-        'alive': True
-    }
-    if runtime.thread_data.device != 'cpu' or is_alive() == 1:
-        preload_model()
-        current_state = ServerStates.Online
-    while True:
-        session_cache.clean()
-        task_cache.clean()
-        if not weak_thread_data[threading.current_thread()]['alive']:
-            print(f'Shutting down thread for device {runtime.thread_data.device}')
-            runtime.unload_models()
-            runtime.unload_filters()
-            return
-        if isinstance(current_state_error, SystemExit):
-            current_state = ServerStates.Unavailable
-            return
-        task = thread_get_next_task()
-        if task is None:
-            idle_event.clear()
-            idle_event.wait(timeout=1)
-            continue
-        if task.error is not None:
-            print(task.error)
-            task.response = {"status": 'failed', "detail": str(task.error)}
-            task.buffer_queue.put(json.dumps(task.response))
-            continue
-        if current_state_error:
-            task.error = current_state_error
-            task.response = {"status": 'failed', "detail": str(task.error)}
-            task.buffer_queue.put(json.dumps(task.response))
-            continue
-        print(f'Session {task.request.session_id} starting task {id(task)} on {runtime.thread_data.device_name}')
-        if not task.lock.acquire(blocking=False): raise Exception('Got locked task from queue.')
-        try:
-            if runtime.is_hypernetwork_reload_necessary(task.request):
-                runtime.reload_hypernetwork()
-                current_hypernetwork_path = task.request.use_hypernetwork_model
-                
-            if runtime.is_model_reload_necessary(task.request):
-                current_state = ServerStates.LoadingModel
-                runtime.reload_model()
-                current_model_path = task.request.use_stable_diffusion_model
-                current_vae_path = task.request.use_vae_model
-
-            def step_callback():
-                global current_state_error
-
-                if isinstance(current_state_error, SystemExit) or isinstance(current_state_error, StopAsyncIteration) or isinstance(task.error, StopAsyncIteration):
-                    runtime.thread_data.stop_processing = True
-                    if isinstance(current_state_error, StopAsyncIteration):
-                        task.error = current_state_error
-                        current_state_error = None
-                        print(f'Session {task.request.session_id} sent cancel signal for task {id(task)}')
-
-            current_state = ServerStates.Rendering
-            task.response = runtime.mk_img(task.request, task.buffer_queue, task.temp_images, step_callback)
-            # Before looping back to the generator, mark cache as still alive.
-            task_cache.keep(id(task), TASK_TTL)
-            session_cache.keep(task.request.session_id, TASK_TTL)
-        except Exception as e:
-            task.error = e
-            task.response = {"status": 'failed', "detail": str(task.error)}
-            task.buffer_queue.put(json.dumps(task.response))
-            print(traceback.format_exc())
-            continue
-        finally:
-            # Task completed
-            task.lock.release()
-        task_cache.keep(id(task), TASK_TTL)
-        session_cache.keep(task.request.session_id, TASK_TTL)
-        if isinstance(task.error, StopAsyncIteration):
-            print(f'Session {task.request.session_id} task {id(task)} cancelled!')
-        elif task.error is not None:
-            print(f'Session {task.request.session_id} task {id(task)} failed!')
-        else:
-            print(f'Session {task.request.session_id} task {id(task)} completed by {runtime.thread_data.device_name}.')
-        current_state = ServerStates.Online
-
-def get_cached_task(task_id:str, update_ttl:bool=False):
-    # By calling keep before tryGet, wont discard if was expired.
-    if update_ttl and not task_cache.keep(task_id, TASK_TTL):
-        # Failed to keep task, already gone.
-        return None
-    return task_cache.tryGet(task_id)
-
-def get_cached_session(session_id:str, update_ttl:bool=False):
-    if update_ttl:
-        session_cache.keep(session_id, TASK_TTL)
-    session = session_cache.tryGet(session_id)
-    if not session:
-        session = SessionState(session_id)
-        session_cache.put(session_id, session, TASK_TTL)
-    return session
-
-def get_devices():
-    devices = {
-        'all': {},
-        'active': {},
-    }
-
-    def get_device_info(device):
-        if device == 'cpu':
-            return {'name': device_manager.get_processor_name()}
-            
-        mem_free, mem_total = torch.cuda.mem_get_info(device)
-        mem_free /= float(10**9)
-        mem_total /= float(10**9)
-
-        return {
-            'name': torch.cuda.get_device_name(device),
-            'mem_free': mem_free,
-            'mem_total': mem_total,
-        }
-
-    # list the compatible devices
-    gpu_count = torch.cuda.device_count()
-    for device in range(gpu_count):
-        device = f'cuda:{device}'
-        if not device_manager.is_device_compatible(device):
-            continue
-
-        devices['all'].update({device: get_device_info(device)})
-
-    devices['all'].update({'cpu': get_device_info('cpu')})
-
-    # list the activated devices
-    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('get_devices' + ERR_LOCK_FAILED)
-    try:
-        for rthread in render_threads:
-            if not rthread.is_alive():
-                continue
-            weak_data = weak_thread_data.get(rthread)
-            if not weak_data or not 'device' in weak_data or not 'device_name' in weak_data:
-                continue
-            device = weak_data['device']
-            devices['active'].update({device: get_device_info(device)})
-    finally:
-        manager_lock.release()
-
-    return devices
-
-def is_alive(device=None):
-    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('is_alive' + ERR_LOCK_FAILED)
-    nbr_alive = 0
-    try:
-        for rthread in render_threads:
-            if device is not None:
-                weak_data = weak_thread_data.get(rthread)
-                if weak_data is None or not 'device' in weak_data or weak_data['device'] is None:
-                    continue
-                thread_device = weak_data['device']
-                if thread_device != device:
-                    continue
-            if rthread.is_alive():
-                nbr_alive += 1
-        return nbr_alive
-    finally:
-        manager_lock.release()
-
-def start_render_thread(device):
-    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('start_render_thread' + ERR_LOCK_FAILED)
-    print('Start new Rendering Thread on device', device)
-    try:
-        rthread = threading.Thread(target=thread_render, kwargs={'device': device})
-        rthread.daemon = True
-        rthread.name = THREAD_NAME_PREFIX + device
-        rthread.start()
-        render_threads.append(rthread)
-    finally:
-        manager_lock.release()
-    timeout = DEVICE_START_TIMEOUT
-    while not rthread.is_alive() or not rthread in weak_thread_data or not 'device' in weak_thread_data[rthread]:
-        if rthread in weak_thread_data and 'error' in weak_thread_data[rthread]:
-            print(rthread, device, 'error:', weak_thread_data[rthread]['error'])
-            return False
-        if timeout <= 0:
-            return False
-        timeout -= 1
-        time.sleep(1)
-    return True
-
-def stop_render_thread(device):
-    try:
-        device_manager.validate_device_id(device, log_prefix='stop_render_thread')
-    except:
-        print(traceback.format_exc())
-        return False
-
-    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('stop_render_thread' + ERR_LOCK_FAILED)
-    print('Stopping Rendering Thread on device', device)
-
-    try:
-        thread_to_remove = None
-        for rthread in render_threads:
-            weak_data = weak_thread_data.get(rthread)
-            if weak_data is None or not 'device' in weak_data or weak_data['device'] is None:
-                continue
-            thread_device = weak_data['device']
-            if thread_device == device:
-                weak_data['alive'] = False
-                thread_to_remove = rthread
-                break
-        if thread_to_remove is not None:
-            render_threads.remove(rthread)
-            return True
-    finally:
-        manager_lock.release()
-
-    return False
-
-def update_render_threads(render_devices, active_devices):
-    devices_to_start, devices_to_stop = device_manager.get_device_delta(render_devices, active_devices)
-    print('devices_to_start', devices_to_start)
-    print('devices_to_stop', devices_to_stop)
-
-    for device in devices_to_stop:
-        if is_alive(device) <= 0:
-            print(device, 'is not alive')
-            continue
-        if not stop_render_thread(device):
-            print(device, 'could not stop render thread')
-
-    for device in devices_to_start:
-        if is_alive(device) >= 1:
-            print(device, 'already registered.')
-            continue
-        if not start_render_thread(device):
-            print(device, 'failed to start.')
-
-    if is_alive() <= 0: # No running devices, probably invalid user config.
-        raise EnvironmentError('ERROR: No active render devices! Please verify the "render_devices" value in config.json')
-
-    print('active devices', get_devices()['active'])
-
-def shutdown_event(): # Signal render thread to close on shutdown
-    global current_state_error
-    current_state_error = SystemExit('Application shutting down.')
-
-def render(req : ImageRequest):
-    current_thread_count = is_alive()
-    if current_thread_count <= 0:  # Render thread is dead
-        raise ChildProcessError('Rendering thread has died.')
-
-    # Alive, check if task in cache
-    session = get_cached_session(req.session_id, update_ttl=True)
-    pending_tasks = list(filter(lambda t: t.is_pending, session.tasks))
-    if current_thread_count < len(pending_tasks):
-        raise ConnectionRefusedError(f'Session {req.session_id} already has {len(pending_tasks)} pending tasks out of {current_thread_count}.')
-
-    from . import runtime
-    r = Request()
-    r.session_id = req.session_id
-    r.prompt = req.prompt
-    r.negative_prompt = req.negative_prompt
-    r.init_image = req.init_image
-    r.mask = req.mask
-    r.num_outputs = req.num_outputs
-    r.num_inference_steps = req.num_inference_steps
-    r.guidance_scale = req.guidance_scale
-    r.width = req.width
-    r.height = req.height
-    r.seed = req.seed
-    r.prompt_strength = req.prompt_strength
-    r.sampler = req.sampler
-    # r.allow_nsfw = req.allow_nsfw
-    r.turbo = req.turbo
-    r.use_full_precision = req.use_full_precision
-    r.save_to_disk_path = req.save_to_disk_path
-    r.use_upscale: str = req.use_upscale
-    r.use_face_correction = req.use_face_correction
-    r.use_stable_diffusion_model = req.use_stable_diffusion_model
-    r.use_vae_model = req.use_vae_model
-    r.use_hypernetwork_model = req.use_hypernetwork_model
-    r.hypernetwork_strength = req.hypernetwork_strength
-    r.show_only_filtered_image = req.show_only_filtered_image
-    r.output_format = req.output_format
-    r.output_quality = req.output_quality
-
-    r.stream_progress_updates = True # the underlying implementation only supports streaming
-    r.stream_image_progress = req.stream_image_progress
-
-    if not req.stream_progress_updates:
-        r.stream_image_progress = False
-
-    new_task = RenderTask(r)
-    if session.put(new_task, TASK_TTL):
-        # Use twice the normal timeout for adding user requests.
-        # Tries to force session.put to fail before tasks_queue.put would.
-        if manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT * 2):
-            try:
-                tasks_queue.append(new_task)
-                idle_event.set()
-                return new_task
-            finally:
-                manager_lock.release()
-    raise RuntimeError('Failed to add task to cache.')

ui/server.py

@@ -1,500 +0,0 @@
-"""server.py: FastAPI SD-UI Web Host.
-Notes:
-    async endpoints always run on the main thread. Without they run on the thread pool.
-"""
-import json
-import traceback
-
-import sys
-import os
-import socket
-import picklescan.scanner
-import rich
-
-SD_DIR = os.getcwd()
-print('started in ', SD_DIR)
-
-SD_UI_DIR = os.getenv('SD_UI_PATH', None)
-sys.path.append(os.path.dirname(SD_UI_DIR))
-
-CONFIG_DIR = os.path.abspath(os.path.join(SD_UI_DIR, '..', 'scripts'))
-MODELS_DIR = os.path.abspath(os.path.join(SD_DIR, '..', 'models'))
-
-USER_UI_PLUGINS_DIR = os.path.abspath(os.path.join(SD_DIR, '..', 'plugins', 'ui'))
-CORE_UI_PLUGINS_DIR = os.path.abspath(os.path.join(SD_UI_DIR, 'plugins', 'ui'))
-UI_PLUGINS_SOURCES = ((CORE_UI_PLUGINS_DIR, 'core'), (USER_UI_PLUGINS_DIR, 'user'))
-
-STABLE_DIFFUSION_MODEL_EXTENSIONS = ['.ckpt', '.safetensors']
-VAE_MODEL_EXTENSIONS = ['.vae.pt', '.ckpt']
-HYPERNETWORK_MODEL_EXTENSIONS = ['.pt']
-
-OUTPUT_DIRNAME = "Stable Diffusion UI" # in the user's home folder
-TASK_TTL = 15 * 60 # Discard last session's task timeout
-APP_CONFIG_DEFAULTS = {
-    # auto: selects the cuda device with the most free memory, cuda: use the currently active cuda device.
-    'render_devices': 'auto', # valid entries: 'auto', 'cpu' or 'cuda:N' (where N is a GPU index)
-    'update_branch': 'main',
-    'ui': {
-        'open_browser_on_start': True,
-    },
-}
-APP_CONFIG_DEFAULT_MODELS = [
-    # needed to support the legacy installations
-    'custom-model', # Check if user has a custom model, use it first.
-    'sd-v1-4', # Default fallback.
-]
-
-from fastapi import FastAPI, HTTPException
-from fastapi.staticfiles import StaticFiles
-from starlette.responses import FileResponse, JSONResponse, StreamingResponse
-from pydantic import BaseModel
-import logging
-from typing import Any, Generator, Hashable, List, Optional, Union
-
-from sd_internal import Request, Response, task_manager
-
-app = FastAPI()
-
-outpath = os.path.join(os.path.expanduser("~"), OUTPUT_DIRNAME)
-
-os.makedirs(USER_UI_PLUGINS_DIR, exist_ok=True)
-
-# don't show access log entries for URLs that start with the given prefix
-ACCESS_LOG_SUPPRESS_PATH_PREFIXES = ['/ping', '/image', '/modifier-thumbnails']
-
-NOCACHE_HEADERS={"Cache-Control": "no-cache, no-store, must-revalidate", "Pragma": "no-cache", "Expires": "0"}
-
-class NoCacheStaticFiles(StaticFiles):
-    def is_not_modified(self, response_headers, request_headers) -> bool:
-        if 'content-type' in response_headers and ('javascript' in response_headers['content-type'] or 'css' in response_headers['content-type']):
-            response_headers.update(NOCACHE_HEADERS)
-            return False
-
-        return super().is_not_modified(response_headers, request_headers)
-
-app.mount('/media', NoCacheStaticFiles(directory=os.path.join(SD_UI_DIR, 'media')), name="media")
-
-for plugins_dir, dir_prefix in UI_PLUGINS_SOURCES:
-    app.mount(f'/plugins/{dir_prefix}', NoCacheStaticFiles(directory=plugins_dir), name=f"plugins-{dir_prefix}")
-
-def getConfig(default_val=APP_CONFIG_DEFAULTS):
-    try:
-        config_json_path = os.path.join(CONFIG_DIR, 'config.json')
-        if not os.path.exists(config_json_path):
-            return default_val
-        with open(config_json_path, 'r', encoding='utf-8') as f:
-            config = json.load(f)
-            if 'net' not in config:
-                config['net'] = {}
-            if os.getenv('SD_UI_BIND_PORT') is not None:
-                config['net']['listen_port'] = int(os.getenv('SD_UI_BIND_PORT'))
-            if os.getenv('SD_UI_BIND_IP') is not None:
-                config['net']['listen_to_network'] = ( os.getenv('SD_UI_BIND_IP') == '0.0.0.0' )
-            return config
-    except Exception as e:
-        print(str(e))
-        print(traceback.format_exc())
-        return default_val
-
-def setConfig(config):
-    print( json.dumps(config) )
-    try: # config.json
-        config_json_path = os.path.join(CONFIG_DIR, 'config.json')
-        with open(config_json_path, 'w', encoding='utf-8') as f:
-            json.dump(config, f)
-    except:
-        print(traceback.format_exc())
-
-    try: # config.bat
-        config_bat_path = os.path.join(CONFIG_DIR, 'config.bat')
-        config_bat = []
-
-        if 'update_branch' in config:
-            config_bat.append(f"@set update_branch={config['update_branch']}")
-
-        config_bat.append(f"@set SD_UI_BIND_PORT={config['net']['listen_port']}")
-        bind_ip = '0.0.0.0' if config['net']['listen_to_network'] else '127.0.0.1'
-        config_bat.append(f"@set SD_UI_BIND_IP={bind_ip}")
-
-        config_bat.append(f"@set test_sd2={'Y' if config.get('test_sd2', False) else 'N'}")
-
-        if len(config_bat) > 0:
-            with open(config_bat_path, 'w', encoding='utf-8') as f:
-                f.write('\r\n'.join(config_bat))
-    except:
-        print(traceback.format_exc())
-
-    try: # config.sh
-        config_sh_path = os.path.join(CONFIG_DIR, 'config.sh')
-        config_sh = ['#!/bin/bash']
-
-        if 'update_branch' in config:
-            config_sh.append(f"export update_branch={config['update_branch']}")
-
-        config_sh.append(f"export SD_UI_BIND_PORT={config['net']['listen_port']}")
-        bind_ip = '0.0.0.0' if config['net']['listen_to_network'] else '127.0.0.1'
-        config_sh.append(f"export SD_UI_BIND_IP={bind_ip}")
-
-        config_sh.append(f"export test_sd2=\"{'Y' if config.get('test_sd2', False) else 'N'}\"")
-
-        if len(config_sh) > 1:
-            with open(config_sh_path, 'w', encoding='utf-8') as f:
-                f.write('\n'.join(config_sh))
-    except:
-        print(traceback.format_exc())
-
-def resolve_model_to_use(model_name:str, model_type:str, model_dir:str, model_extensions:list, default_models=[]):
-    config = getConfig()
-
-    model_dirs = [os.path.join(MODELS_DIR, model_dir), SD_DIR]
-    if not model_name: # When None try user configured model.
-        # config = getConfig()
-        if 'model' in config and model_type in config['model']:
-            model_name = config['model'][model_type]
-    if model_name:
-        is_sd2 = config.get('test_sd2', False)
-        if model_name.startswith('sd2_') and not is_sd2: # temp hack, until SD2 is unified with 1.4
-            print('ERROR: Cannot use SD 2.0 models with SD 1.0 code. Using the sd-v1-4 model instead!')
-            model_name = 'sd-v1-4'
-
-        # Check models directory
-        models_dir_path = os.path.join(MODELS_DIR, model_dir, model_name)
-        for model_extension in model_extensions:
-            if os.path.exists(models_dir_path + model_extension):
-                return models_dir_path
-            if os.path.exists(model_name + model_extension):
-                # Direct Path to file
-                model_name = os.path.abspath(model_name)
-                return model_name
-    # Default locations
-    if model_name in default_models:
-        default_model_path = os.path.join(SD_DIR, model_name)
-        for model_extension in model_extensions:
-            if os.path.exists(default_model_path + model_extension):
-                return default_model_path
-    # Can't find requested model, check the default paths.
-    for default_model in default_models:
-        for model_dir in model_dirs:
-            default_model_path = os.path.join(model_dir, default_model)
-            for model_extension in model_extensions:
-                if os.path.exists(default_model_path + model_extension):
-                    if model_name is not None:
-                        print(f'Could not find the configured custom model {model_name}{model_extension}. Using the default one: {default_model_path}{model_extension}')
-                    return default_model_path
-    raise Exception('No valid models found.')
-
-def resolve_ckpt_to_use(model_name:str=None):
-    return resolve_model_to_use(model_name, model_type='stable-diffusion', model_dir='stable-diffusion', model_extensions=STABLE_DIFFUSION_MODEL_EXTENSIONS, default_models=APP_CONFIG_DEFAULT_MODELS)
-
-def resolve_vae_to_use(model_name:str=None):
-    try:
-        return resolve_model_to_use(model_name, model_type='vae', model_dir='vae', model_extensions=VAE_MODEL_EXTENSIONS, default_models=[])
-    except:
-        return None
-
-def resolve_hypernetwork_to_use(model_name:str=None):
-    try:
-        return resolve_model_to_use(model_name, model_type='hypernetwork', model_dir='hypernetwork', model_extensions=HYPERNETWORK_MODEL_EXTENSIONS, default_models=[])
-    except:
-        return None
-
-class SetAppConfigRequest(BaseModel):
-    update_branch: str = None
-    render_devices: Union[List[str], List[int], str, int] = None
-    model_vae: str = None
-    ui_open_browser_on_start: bool = None
-    listen_to_network: bool = None
-    listen_port: int = None
-    test_sd2: bool = None
-
-@app.post('/app_config')
-async def setAppConfig(req : SetAppConfigRequest):
-    config = getConfig()
-    if req.update_branch is not None:
-        config['update_branch'] = req.update_branch
-    if req.render_devices is not None:
-        update_render_devices_in_config(config, req.render_devices)
-    if req.ui_open_browser_on_start is not None:
-        if 'ui' not in config:
-            config['ui'] = {}
-        config['ui']['open_browser_on_start'] = req.ui_open_browser_on_start
-    if req.listen_to_network is not None:
-       if 'net' not in config:
-           config['net'] = {}
-       config['net']['listen_to_network'] = bool(req.listen_to_network)
-    if req.listen_port is not None:
-       if 'net' not in config:
-           config['net'] = {}
-       config['net']['listen_port'] = int(req.listen_port)
-    if req.test_sd2 is not None:
-        config['test_sd2'] = req.test_sd2
-    try:
-        setConfig(config)
-
-        if req.render_devices:
-            update_render_threads()
-
-        return JSONResponse({'status': 'OK'}, headers=NOCACHE_HEADERS)
-    except Exception as e:
-        print(traceback.format_exc())
-        raise HTTPException(status_code=500, detail=str(e))
-
-def is_malicious_model(file_path):
-    try:
-        scan_result = picklescan.scanner.scan_file_path(file_path)
-        if scan_result.issues_count > 0 or scan_result.infected_files > 0:
-            rich.print(":warning: [bold red]Scan %s: %d scanned, %d issue, %d infected.[/bold red]" % (file_path, scan_result.scanned_files, scan_result.issues_count, scan_result.infected_files))
-            return True
-        else:
-            rich.print("Scan %s: [green]%d scanned, %d issue, %d infected.[/green]" % (file_path, scan_result.scanned_files, scan_result.issues_count, scan_result.infected_files))
-            return False
-    except Exception as e:
-        print('error while scanning', file_path, 'error:', e)
-    return False
-
-known_models = {}
-def getModels():
-    models = {
-        'active': {
-            'stable-diffusion': 'sd-v1-4',
-            'vae': '',
-            'hypernetwork': '',
-        },
-        'options': {
-            'stable-diffusion': ['sd-v1-4'],
-            'vae': [],
-            'hypernetwork': [],
-        },
-    }
-
-    def listModels(models_dirname, model_type, model_extensions):
-        models_dir = os.path.join(MODELS_DIR, models_dirname)
-        if not os.path.exists(models_dir):
-            os.makedirs(models_dir)
-
-        for file in os.listdir(models_dir):
-            for model_extension in model_extensions:
-                if not file.endswith(model_extension):
-                    continue
-
-                model_path = os.path.join(models_dir, file)
-                mtime = os.path.getmtime(model_path)
-                mod_time = known_models[model_path] if model_path in known_models else -1
-                if mod_time != mtime:
-                    if is_malicious_model(model_path):
-                        models['scan-error'] = file
-                        return
-                known_models[model_path] = mtime
-
-                model_name = file[:-len(model_extension)]
-                models['options'][model_type].append(model_name)
-
-        models['options'][model_type] = [*set(models['options'][model_type])] # remove duplicates
-        models['options'][model_type].sort()
-
-    # custom models
-    listModels(models_dirname='stable-diffusion', model_type='stable-diffusion', model_extensions=STABLE_DIFFUSION_MODEL_EXTENSIONS)
-    listModels(models_dirname='vae', model_type='vae', model_extensions=VAE_MODEL_EXTENSIONS)
-    listModels(models_dirname='hypernetwork', model_type='hypernetwork', model_extensions=HYPERNETWORK_MODEL_EXTENSIONS)
-    # legacy
-    custom_weight_path = os.path.join(SD_DIR, 'custom-model.ckpt')
-    if os.path.exists(custom_weight_path):
-        models['options']['stable-diffusion'].append('custom-model')
-
-    return models
-
-def getUIPlugins():
-    plugins = []
-
-    for plugins_dir, dir_prefix in UI_PLUGINS_SOURCES:
-        for file in os.listdir(plugins_dir):
-            if file.endswith('.plugin.js'):
-                plugins.append(f'/plugins/{dir_prefix}/{file}')
-
-    return plugins
-
-def getIPConfig():
-    try: 
-        ips = socket.gethostbyname_ex(socket.gethostname())
-        ips[2].append(ips[0])
-        return ips[2]
-    except Exception as e:
-        print(e)
-        print(traceback.format_exc())
-        return []
-        
-
-@app.get('/get/{key:path}')
-def read_web_data(key:str=None):
-    if not key: # /get without parameters, stable-diffusion easter egg.
-        raise HTTPException(status_code=418, detail="StableDiffusion is drawing a teapot!") # HTTP418 I'm a teapot
-    elif key == 'app_config':
-        config = getConfig(default_val=None)
-        if config is None:
-            config = APP_CONFIG_DEFAULTS
-        return JSONResponse(config, headers=NOCACHE_HEADERS)
-    elif key == 'system_info':
-        config = getConfig()
-        system_info = {
-            'devices': task_manager.get_devices(),
-            'hosts': getIPConfig(),
-        }
-        system_info['devices']['config'] = config.get('render_devices', "auto")
-        return JSONResponse(system_info, headers=NOCACHE_HEADERS)
-    elif key == 'models':
-        return JSONResponse(getModels(), headers=NOCACHE_HEADERS)
-    elif key == 'modifiers': return FileResponse(os.path.join(SD_UI_DIR, 'modifiers.json'), headers=NOCACHE_HEADERS)
-    elif key == 'output_dir': return JSONResponse({ 'output_dir': outpath }, headers=NOCACHE_HEADERS)
-    elif key == 'ui_plugins': return JSONResponse(getUIPlugins(), headers=NOCACHE_HEADERS)
-    else:
-        raise HTTPException(status_code=404, detail=f'Request for unknown {key}') # HTTP404 Not Found
-
-@app.get('/ping') # Get server and optionally session status.
-def ping(session_id:str=None):
-    if task_manager.is_alive() <= 0: # Check that render threads are alive.
-        if task_manager.current_state_error: raise HTTPException(status_code=500, detail=str(task_manager.current_state_error))
-        raise HTTPException(status_code=500, detail='Render thread is dead.')
-    if task_manager.current_state_error and not isinstance(task_manager.current_state_error, StopAsyncIteration): raise HTTPException(status_code=500, detail=str(task_manager.current_state_error))
-    # Alive
-    response = {'status': str(task_manager.current_state)}
-    if session_id:
-        session = task_manager.get_cached_session(session_id, update_ttl=True)
-        response['tasks'] = {id(t): t.status for t in session.tasks}
-    response['devices'] = task_manager.get_devices()
-    return JSONResponse(response, headers=NOCACHE_HEADERS)
-
-def save_model_to_config(ckpt_model_name, vae_model_name, hypernetwork_model_name):
-    config = getConfig()
-    if 'model' not in config:
-        config['model'] = {}
-
-    config['model']['stable-diffusion'] = ckpt_model_name
-    config['model']['vae'] = vae_model_name
-    config['model']['hypernetwork'] = hypernetwork_model_name
-
-    if vae_model_name is None or vae_model_name == "":
-        del config['model']['vae']
-    if hypernetwork_model_name is None or hypernetwork_model_name == "":
-        del config['model']['hypernetwork']
-
-    setConfig(config)
-
-def update_render_devices_in_config(config, render_devices):
-    if render_devices not in ('cpu', 'auto') and not render_devices.startswith('cuda:'):
-        raise HTTPException(status_code=400, detail=f'Invalid render device requested: {render_devices}')
-
-    if render_devices.startswith('cuda:'):
-        render_devices = render_devices.split(',')
-
-    config['render_devices'] = render_devices
-
-@app.post('/render')
-def render(req : task_manager.ImageRequest):
-    try:
-        save_model_to_config(req.use_stable_diffusion_model, req.use_vae_model, req.use_hypernetwork_model)
-        req.use_stable_diffusion_model = resolve_ckpt_to_use(req.use_stable_diffusion_model)
-        req.use_vae_model = resolve_vae_to_use(req.use_vae_model)
-        req.use_hypernetwork_model = resolve_hypernetwork_to_use(req.use_hypernetwork_model)
-        new_task = task_manager.render(req)
-        response = {
-            'status': str(task_manager.current_state), 
-            'queue': len(task_manager.tasks_queue),
-            'stream': f'/image/stream/{id(new_task)}',
-            'task': id(new_task)
-        }
-        return JSONResponse(response, headers=NOCACHE_HEADERS)
-    except ChildProcessError as e: # Render thread is dead
-        raise HTTPException(status_code=500, detail=f'Rendering thread has died.') # HTTP500 Internal Server Error
-    except ConnectionRefusedError as e: # Unstarted task pending limit reached, deny queueing too many.
-        raise HTTPException(status_code=503, detail=str(e)) # HTTP503 Service Unavailable
-    except Exception as e:
-        print(e)
-        print(traceback.format_exc())
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.get('/image/stream/{task_id:int}')
-def stream(task_id:int):
-    #TODO Move to WebSockets ??
-    task = task_manager.get_cached_task(task_id, update_ttl=True)
-    if not task: raise HTTPException(status_code=404, detail=f'Request {task_id} not found.') # HTTP404 NotFound
-    #if (id(task) != task_id): raise HTTPException(status_code=409, detail=f'Wrong task id received. Expected:{id(task)}, Received:{task_id}') # HTTP409 Conflict
-    if task.buffer_queue.empty() and not task.lock.locked():
-        if task.response:
-            #print(f'Session {session_id} sending cached response')
-            return JSONResponse(task.response, headers=NOCACHE_HEADERS)
-        raise HTTPException(status_code=425, detail='Too Early, task not started yet.') # HTTP425 Too Early
-    #print(f'Session {session_id} opened live render stream {id(task.buffer_queue)}')
-    return StreamingResponse(task.read_buffer_generator(), media_type='application/json')
-
-@app.get('/image/stop')
-def stop(task: int):
-    if not task:
-        if task_manager.current_state == task_manager.ServerStates.Online or task_manager.current_state == task_manager.ServerStates.Unavailable:
-            raise HTTPException(status_code=409, detail='Not currently running any tasks.') # HTTP409 Conflict
-        task_manager.current_state_error = StopAsyncIteration('')
-        return {'OK'}
-    task_id = task
-    task = task_manager.get_cached_task(task_id, update_ttl=False)
-    if not task: raise HTTPException(status_code=404, detail=f'Task {task_id} was not found.') # HTTP404 Not Found
-    if isinstance(task.error, StopAsyncIteration): raise HTTPException(status_code=409, detail=f'Task {task_id} is already stopped.') # HTTP409 Conflict
-    task.error = StopAsyncIteration(f'Task {task_id} stop requested.')
-    return {'OK'}
-
-@app.get('/image/tmp/{task_id:int}/{img_id:int}')
-def get_image(task_id: int, img_id: int):
-    task = task_manager.get_cached_task(task_id, update_ttl=True)
-    if not task: raise HTTPException(status_code=410, detail=f'Task {task_id} could not be found.') # HTTP404 NotFound
-    if not task.temp_images[img_id]: raise HTTPException(status_code=425, detail='Too Early, task data is not available yet.') # HTTP425 Too Early
-    try:
-        img_data = task.temp_images[img_id]
-        img_data.seek(0)
-        return StreamingResponse(img_data, media_type='image/jpeg')
-    except KeyError as e:
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.get('/')
-def read_root():
-    return FileResponse(os.path.join(SD_UI_DIR, 'index.html'), headers=NOCACHE_HEADERS)
-
-@app.on_event("shutdown")
-def shutdown_event(): # Signal render thread to close on shutdown
-    task_manager.current_state_error = SystemExit('Application shutting down.')
-
-# don't log certain requests
-class LogSuppressFilter(logging.Filter):
-    def filter(self, record: logging.LogRecord) -> bool:
-        path = record.getMessage()
-        for prefix in ACCESS_LOG_SUPPRESS_PATH_PREFIXES:
-            if path.find(prefix) != -1:
-                return False
-        return True
-logging.getLogger('uvicorn.access').addFilter(LogSuppressFilter())
-
-# Check models and prepare cache for UI open
-getModels()
-
-# Start the task_manager
-task_manager.default_model_to_load = resolve_ckpt_to_use()
-task_manager.default_vae_to_load = resolve_vae_to_use()
-task_manager.default_hypernetwork_to_load = resolve_hypernetwork_to_use()
-
-def update_render_threads():
-    config = getConfig()
-    render_devices = config.get('render_devices', 'auto')
-    active_devices = task_manager.get_devices()['active'].keys()
-
-    print('requesting for render_devices', render_devices)
-    task_manager.update_render_threads(render_devices, active_devices)
-
-update_render_threads()
-
-# start the browser ui
-def open_browser():
-    config = getConfig()
-    ui = config.get('ui', {})
-    net = config.get('net', {'listen_port':9000})
-    port = net.get('listen_port', 9000)
-    if ui.get('open_browser_on_start', True):
-        import webbrowser; webbrowser.open(f"http://localhost:{port}")
-
-open_browser()