diff --git a/.eslintrc.js b/.eslintrc.js
index 4777c276e9b..2e7258f6b13 100644
--- a/.eslintrc.js
+++ b/.eslintrc.js
@@ -74,9 +74,12 @@ module.exports = {
         create_submit_args: "readonly",
         restart_reload: "readonly",
         updateInput: "readonly",
+        onEdit: "readonly",
         //extraNetworks.js
         requestGet: "readonly",
         popup: "readonly",
+        // profilerVisualization.js
+        createVisualizationTable: "readonly",
         // from python
         localization: "readonly",
         // progrssbar.js
@@ -85,8 +88,6 @@ module.exports = {
         // imageviewer.js
         modalPrevImage: "readonly",
         modalNextImage: "readonly",
-        // token-counters.js
-        setupTokenCounters: "readonly",
         // localStorage.js
         localSet: "readonly",
         localGet: "readonly",
diff --git a/.github/ISSUE_TEMPLATE/bug_report.yml b/.github/ISSUE_TEMPLATE/bug_report.yml
index cf6a2be86fa..c86bd8a680b 100644
--- a/.github/ISSUE_TEMPLATE/bug_report.yml
+++ b/.github/ISSUE_TEMPLATE/bug_report.yml
@@ -1,25 +1,45 @@
 name: Bug Report
-description: You think somethings is broken in the UI
+description: You think something is broken in the UI
 title: "[Bug]: "
 labels: ["bug-report"]
 
 body:
+  - type: markdown
+    attributes:
+      value: |
+        > The title of the bug report should be short and descriptive.
+        > Use relevant keywords for searchability.
+        > Do not leave it blank, but also do not put an entire error log in it.
   - type: checkboxes
     attributes:
-      label: Is there an existing issue for this?
-      description: Please search to see if an issue already exists for the bug you encountered, and that it hasn't been fixed in a recent build/commit.
+      label: Checklist
+      description: |
+        Please perform basic debugging to see if extensions or configuration is the cause of the issue.
+        Basic debug procedure
+        　1. Disable all third-party extensions - check if extension is the cause
+        　2. Update extensions and webui - sometimes things just need to be updated
+        　3. Backup and remove your config.json and ui-config.json - check if the issue is caused by bad configuration
+        　4. Delete venv with third-party extensions disabled - sometimes extensions might cause wrong libraries to be installed
+        　5. Try a fresh installation webui in a different directory - see if a clean installation solves the issue
+        Before making a issue report please, check that the issue hasn't been reported recently.
       options:
-        - label: I have searched the existing issues and checked the recent builds/commits
-          required: true
+        - label: The issue exists after disabling all extensions
+        - label: The issue exists on a clean installation of webui
+        - label: The issue is caused by an extension, but I believe it is caused by a bug in the webui
+        - label: The issue exists in the current version of the webui
+        - label: The issue has not been reported before recently
+        - label: The issue has been reported before but has not been fixed yet
   - type: markdown
     attributes:
       value: |
-        *Please fill this form with as much information as possible, don't forget to fill "What OS..." and "What browsers" and *provide screenshots if possible**
+        > Please fill this form with as much information as possible. Don't forget to "Upload Sysinfo" and "What browsers" and provide screenshots if possible
   - type: textarea
     id: what-did
     attributes:
       label: What happened?
       description: Tell us what happened in a very clear and simple way
+      placeholder: |
+        txt2img is not working as intended.
     validations:
       required: true
   - type: textarea
@@ -27,9 +47,9 @@ body:
     attributes:
       label: Steps to reproduce the problem
       description: Please provide us with precise step by step instructions on how to reproduce the bug
-      value: |
-        1. Go to .... 
-        2. Press ....
+      placeholder: |
+        1. Go to ...
+        2. Press ...
         3. ...
     validations:
       required: true
@@ -38,13 +58,8 @@ body:
     attributes:
       label: What should have happened?
       description: Tell us what you think the normal behavior should be
-    validations:
-      required: true
-  - type: textarea
-    id: sysinfo
-    attributes:
-      label: Sysinfo
-      description: System info file, generated by WebUI. You can generate it in settings, on the Sysinfo page. Drag the file into the field to upload it. If you submit your report without including the sysinfo file, the report will be closed. If needed, review the report to make sure it includes no personal information you don't want to share. If you can't start WebUI, you can use --dump-sysinfo commandline argument to generate the file.
+      placeholder: |
+        WebUI should ...
     validations:
       required: true
   - type: dropdown
@@ -58,12 +73,25 @@ body:
         - Brave
         - Apple Safari
         - Microsoft Edge
+        - Android
+        - iOS
         - Other
+  - type: textarea
+    id: sysinfo
+    attributes:
+      label: Sysinfo
+      description: System info file, generated by WebUI. You can generate it in settings, on the Sysinfo page. Drag the file into the field to upload it. If you submit your report without including the sysinfo file, the report will be closed. If needed, review the report to make sure it includes no personal information you don't want to share. If you can't start WebUI, you can use --dump-sysinfo commandline argument to generate the file.
+      placeholder: |
+        1. Go to WebUI Settings -> Sysinfo -> Download system info.
+            If WebUI fails to launch, use --dump-sysinfo commandline argument to generate the file
+        2. Upload the Sysinfo as a attached file, Do NOT paste it in as plain text.
+    validations:
+      required: true
   - type: textarea
     id: logs
     attributes:
       label: Console logs
-      description: Please provide **full** cmd/terminal logs from the moment you started UI to the end of it, after your bug happened. If it's very long, provide a link to pastebin or similar service.
+      description: Please provide **full** cmd/terminal logs from the moment you started UI to the end of it, after the bug occurred. If it's very long, provide a link to pastebin or similar service.
       render: Shell
     validations:
       required: true
@@ -71,4 +99,7 @@ body:
     id: misc
     attributes:
       label: Additional information
-      description: Please provide us with any relevant additional info or context.
+      description: | 
+        Please provide us with any relevant additional info or context.
+        Examples:
+        　I have updated my GPU driver recently.
diff --git a/.github/workflows/on_pull_request.yaml b/.github/workflows/on_pull_request.yaml
index 78e608ee945..9326c6a45e9 100644
--- a/.github/workflows/on_pull_request.yaml
+++ b/.github/workflows/on_pull_request.yaml
@@ -11,8 +11,8 @@ jobs:
     if: github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name != github.event.pull_request.base.repo.full_name
     steps:
       - name: Checkout Code
-        uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+        uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
         with:
           python-version: 3.11
           # NB: there's no cache: pip here since we're not installing anything
@@ -20,7 +20,7 @@ jobs:
           #     not to have GHA download an (at the time of writing) 4 GB cache
           #     of PyTorch and other dependencies.
       - name: Install Ruff
-        run: pip install ruff==0.0.272
+        run: pip install ruff==0.3.3
       - name: Run Ruff
         run: ruff .
   lint-js:
@@ -29,9 +29,9 @@ jobs:
     if: github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name != github.event.pull_request.base.repo.full_name
     steps:
       - name: Checkout Code
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
       - name: Install Node.js
-        uses: actions/setup-node@v3
+        uses: actions/setup-node@v4
         with:
           node-version: 18
       - run: npm i --ci
diff --git a/.github/workflows/run_tests.yaml b/.github/workflows/run_tests.yaml
index 3dafaf8dcfc..0610f4f5436 100644
--- a/.github/workflows/run_tests.yaml
+++ b/.github/workflows/run_tests.yaml
@@ -11,15 +11,21 @@ jobs:
     if: github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name != github.event.pull_request.base.repo.full_name
     steps:
       - name: Checkout Code
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
       - name: Set up Python 3.10
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         with:
           python-version: 3.10.6
           cache: pip
           cache-dependency-path: |
             **/requirements*txt
             launch.py
+      - name: Cache models
+        id: cache-models
+        uses: actions/cache@v4
+        with:
+          path: models
+          key: "2023-12-30"
       - name: Install test dependencies
         run: pip install wait-for-it -r requirements-test.txt
         env:
@@ -33,6 +39,8 @@ jobs:
           TORCH_INDEX_URL: https://download.pytorch.org/whl/cpu
           WEBUI_LAUNCH_LIVE_OUTPUT: "1"
           PYTHONUNBUFFERED: "1"
+      - name: Print installed packages
+        run: pip freeze
       - name: Start test server
         run: >
           python -m coverage run
@@ -49,7 +57,7 @@ jobs:
           2>&1 | tee output.txt &
       - name: Run tests
         run: |
-          wait-for-it --service 127.0.0.1:7860 -t 600
+          wait-for-it --service 127.0.0.1:7860 -t 20
           python -m pytest -vv --junitxml=test/results.xml --cov . --cov-report=xml --verify-base-url test
       - name: Kill test server
         if: always()
@@ -60,13 +68,13 @@ jobs:
           python -m coverage report -i
           python -m coverage html -i
       - name: Upload main app output
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         if: always()
         with:
           name: output
           path: output.txt
       - name: Upload coverage HTML
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         if: always()
         with:
           name: htmlcov
diff --git a/.gitignore b/.gitignore
index 09734267ff5..e81ad31f53a 100644
--- a/.gitignore
+++ b/.gitignore
@@ -2,6 +2,7 @@ __pycache__
 *.ckpt
 *.safetensors
 *.pth
+.DS_Store
 /ESRGAN/*
 /SwinIR/*
 /repositories
@@ -37,3 +38,7 @@ notification.mp3
 /node_modules
 /package-lock.json
 /.coverage*
+/test/test_outputs
+/cache
+trace.json
+/sysinfo-????-??-??-??-??.json
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 1cd3572c8e0..5f7e59351cb 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,581 @@
+## 1.10.1
+
+### Bug Fixes:
+* fix image upscale on cpu ([#16275](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16275))
+
+
+## 1.10.0
+
+### Features:
+* A lot of performance improvements (see below in Performance section)
+* Stable Diffusion 3 support ([#16030](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16030), [#16164](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16164), [#16212](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16212))
+  * Recommended Euler sampler; DDIM and other timestamp samplers currently not supported
+  * T5 text model is disabled by default, enable it in settings
+* New schedulers:
+  * Align Your Steps ([#15751](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15751))
+  * KL Optimal ([#15608](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15608))
+  * Normal ([#16149](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16149))
+  * DDIM ([#16149](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16149))
+  * Simple ([#16142](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16142))
+  * Beta ([#16235](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16235))
+* New sampler: DDIM CFG++ ([#16035](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16035))
+
+### Minor:
+* Option to skip CFG on early steps ([#15607](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15607))
+* Add --models-dir option ([#15742](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15742))
+* Allow mobile users to open context menu by using two fingers press ([#15682](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15682))
+* Infotext: add Lora name as TI hashes for bundled Textual Inversion ([#15679](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15679))
+* Check model's hash after downloading it to prevent corruped downloads ([#15602](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15602))
+* More extension tag filtering options ([#15627](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15627))
+* When saving AVIF, use JPEG's quality setting ([#15610](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15610))
+* Add filename pattern: `[basename]` ([#15978](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15978))
+* Add option to enable clip skip for clip L on SDXL ([#15992](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15992))
+* Option to prevent screen sleep during generation ([#16001](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16001))
+* ToggleLivePriview button in image viewer ([#16065](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16065))
+* Remove ui flashing on reloading and fast scrollong ([#16153](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16153))
+* option to disable save button log.csv ([#16242](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16242))
+
+### Extensions and API:
+* Add process_before_every_sampling hook ([#15984](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15984))
+* Return HTTP 400 instead of 404 on invalid sampler error ([#16140](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16140))
+
+### Performance:
+* [Performance 1/6] use_checkpoint = False ([#15803](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15803))
+* [Performance 2/6] Replace einops.rearrange with torch native ops ([#15804](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15804))
+* [Performance 4/6] Precompute is_sdxl_inpaint flag ([#15806](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15806))
+* [Performance 5/6] Prevent unnecessary extra networks bias backup ([#15816](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15816))
+* [Performance 6/6] Add --precision half option to avoid casting during inference ([#15820](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15820))
+* [Performance] LDM optimization patches ([#15824](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15824))
+* [Performance] Keep sigmas on CPU ([#15823](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15823))
+* Check for nans in unet only once, after all steps have been completed
+* Added pption to run torch profiler for image generation
+
+### Bug Fixes:
+* Fix for grids without comprehensive infotexts ([#15958](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15958))
+* feat: lora partial update precede full update ([#15943](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15943))
+* Fix bug where file extension had an extra '.' under some circumstances ([#15893](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15893))
+* Fix corrupt model initial load loop ([#15600](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15600))
+* Allow old sampler names in API ([#15656](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15656))
+* more old sampler scheduler compatibility ([#15681](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15681))
+* Fix Hypertile xyz ([#15831](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15831))
+* XYZ CSV skipinitialspace ([#15832](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15832))
+* fix soft inpainting on mps and xpu, torch_utils.float64 ([#15815](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15815))
+* fix extention update when not on main branch ([#15797](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15797))
+* update pickle safe filenames
+* use relative path for webui-assets css ([#15757](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15757))
+* When creating a virtual environment, upgrade pip in webui.bat/webui.sh ([#15750](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15750))
+* Fix AttributeError ([#15738](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15738))
+* use script_path for webui root in launch_utils ([#15705](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15705))
+* fix extra batch mode P Transparency ([#15664](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15664))
+* use gradio theme colors in css ([#15680](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15680))
+* Fix dragging text within prompt input ([#15657](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15657))
+* Add correct mimetype for .mjs files ([#15654](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15654))
+* QOL Items - handle metadata issues more cleanly for SD models, Loras and embeddings ([#15632](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15632))
+* replace wsl-open with wslpath and explorer.exe ([#15968](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15968))
+* Fix SDXL Inpaint ([#15976](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15976))
+* multi size grid ([#15988](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15988))
+* fix Replace preview ([#16118](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16118))
+* Possible fix of wrong scale in weight decomposition ([#16151](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16151))
+* Ensure use of python from venv on Mac and Linux ([#16116](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16116))
+* Prioritize python3.10 over python3 if both are available on Linux and Mac (with fallback) ([#16092](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16092))
+* stoping generation extras ([#16085](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16085))
+* Fix SD2 loading ([#16078](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16078), [#16079](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16079))
+* fix infotext Lora hashes for hires fix different lora ([#16062](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16062))
+* Fix sampler scheduler autocorrection warning ([#16054](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16054))
+* fix ui flashing on reloading and fast scrollong ([#16153](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16153))
+* fix upscale logic ([#16239](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16239))
+* [bug] do not break progressbar on non-job actions (add wrap_gradio_call_no_job) ([#16202](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16202))
+* fix OSError: cannot write mode P as JPEG ([#16194](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16194))
+
+### Other:
+* fix changelog #15883 -> #15882 ([#15907](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15907))
+* ReloadUI backgroundColor --background-fill-primary ([#15864](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15864))
+* Use different torch versions for Intel and ARM Macs ([#15851](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15851))
+* XYZ override rework ([#15836](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15836))
+* scroll extensions table on overflow ([#15830](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15830))
+* img2img batch upload method ([#15817](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15817))
+* chore: sync v1.8.0 packages according to changelog ([#15783](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15783))
+* Add AVIF MIME type support to mimetype definitions ([#15739](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15739))
+* Update imageviewer.js ([#15730](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15730))
+* no-referrer ([#15641](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15641))
+* .gitignore trace.json ([#15980](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15980))
+* Bump spandrel to 0.3.4 ([#16144](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16144))
+* Defunct --max-batch-count ([#16119](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16119))
+* docs: update bug_report.yml ([#16102](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16102))
+* Maintaining Project Compatibility for Python 3.9 Users Without Upgrade Requirements. ([#16088](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16088), [#16169](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16169), [#16192](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16192))
+* Update torch for ARM Macs to 2.3.1 ([#16059](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16059))
+* remove deprecated setting dont_fix_second_order_samplers_schedule ([#16061](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16061))
+* chore: fix typos ([#16060](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16060))
+* shlex.join launch args in console log ([#16170](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16170))
+* activate venv .bat ([#16231](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16231))
+* add ids to the resize tabs in img2img ([#16218](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16218))
+* update installation guide linux ([#16178](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16178))
+* Robust sysinfo ([#16173](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16173))
+* do not send image size on paste inpaint ([#16180](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16180))
+* Fix noisy DS_Store files for MacOS ([#16166](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16166))
+
+
+## 1.9.4
+
+### Bug Fixes:
+*  pin setuptools version to fix the startup error ([#15882](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15882)) 
+
+## 1.9.3
+
+### Bug Fixes:
+*  fix get_crop_region_v2 ([#15594](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15594)) 
+
+## 1.9.2
+
+### Extensions and API:
+* restore 1.8.0-style naming of scripts
+
+## 1.9.1
+
+### Minor:
+* Add avif support ([#15582](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15582))
+* Add filename patterns: `[sampler_scheduler]` and `[scheduler]` ([#15581](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15581))
+
+### Extensions and API:
+* undo adding scripts to sys.modules
+* Add schedulers API endpoint ([#15577](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15577))
+* Remove API upscaling factor limits ([#15560](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15560))
+
+### Bug Fixes:
+* Fix images do not match / Coordinate 'right' is less than 'left' ([#15534](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15534))
+* fix: remove_callbacks_for_function should also remove from the ordered map ([#15533](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15533))
+* fix x1 upscalers ([#15555](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15555))
+* Fix cls.__module__ value in extension script ([#15532](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15532))
+* fix typo in function call (eror -> error) ([#15531](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15531))
+
+### Other:
+* Hide 'No Image data blocks found.' message ([#15567](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15567))
+* Allow webui.sh to be runnable from arbitrary directories containing a .git file ([#15561](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15561))
+* Compatibility with Debian 11, Fedora 34+ and openSUSE 15.4+ ([#15544](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15544))
+* numpy DeprecationWarning product -> prod ([#15547](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15547))
+* get_crop_region_v2 ([#15583](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15583), [#15587](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15587))
+
+
+## 1.9.0
+
+### Features:
+* Make refiner switchover based on model timesteps instead of sampling steps ([#14978](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14978))
+* add an option to have old-style directory view instead of tree view; stylistic changes for extra network sorting/search controls
+* add UI for reordering callbacks, support for specifying callback order in extension metadata ([#15205](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15205))
+* Sgm uniform scheduler for SDXL-Lightning models ([#15325](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15325))
+* Scheduler selection in main UI ([#15333](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15333), [#15361](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15361), [#15394](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15394))
+
+### Minor:
+* "open images directory" button now opens the actual dir ([#14947](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14947))
+* Support inference with LyCORIS BOFT networks ([#14871](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14871), [#14973](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14973))
+* make extra network card description plaintext by default, with an option to re-enable HTML as it was
+* resize handle for extra networks ([#15041](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15041))
+* cmd args: `--unix-filenames-sanitization` and `--filenames-max-length` ([#15031](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15031))
+* show extra networks parameters in HTML table rather than raw JSON ([#15131](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15131))
+* Add DoRA (weight-decompose) support for LoRA/LoHa/LoKr ([#15160](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15160), [#15283](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15283))
+* Add '--no-prompt-history' cmd args for disable last generation prompt history ([#15189](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15189))
+* update preview on Replace Preview ([#15201](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15201))
+* only fetch updates for extensions' active git branches ([#15233](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15233))
+* put upscale postprocessing UI into an accordion ([#15223](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15223))
+* Support dragdrop for URLs to read infotext ([#15262](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15262))
+* use diskcache library for caching ([#15287](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15287), [#15299](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15299))
+* Allow PNG-RGBA for Extras Tab ([#15334](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15334))
+* Support cover images embedded in safetensors metadata ([#15319](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15319))
+* faster interrupt when using NN upscale ([#15380](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15380))
+* Extras upscaler: an input field to limit maximul side length for the output image ([#15293](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15293), [#15415](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15415), [#15417](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15417), [#15425](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15425))
+* add an option to hide postprocessing options in Extras tab
+
+### Extensions and API:
+* ResizeHandleRow - allow overriden column scale parametr ([#15004](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15004))
+* call script_callbacks.ui_settings_callback earlier; fix extra-options-section built-in extension killing the ui if using a setting that doesn't exist
+* make it possible to use zoom.js outside webui context ([#15286](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15286), [#15288](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15288))
+* allow variants for extension name in metadata.ini ([#15290](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15290))
+* make reloading UI scripts optional when doing Reload UI, and off by default
+* put request: gr.Request at start of img2img function similar to txt2img
+* open_folder as util ([#15442](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15442))
+* make it possible to import extensions' script files as `import scripts.<filename>` ([#15423](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15423))
+
+### Performance:
+* performance optimization for extra networks HTML pages
+* optimization for extra networks filtering
+* optimization for extra networks sorting
+
+### Bug Fixes:
+* prevent escape button causing an interrupt when no generation has been made yet
+* [bug] avoid doble upscaling in inpaint ([#14966](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14966))
+* possible fix for reload button not appearing in some cases for extra networks.
+* fix: the `split_threshold` parameter does not work when running Split oversized images ([#15006](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15006))
+* Fix resize-handle visability for vertical layout (mobile) ([#15010](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15010))
+* register_tmp_file also for mtime ([#15012](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15012))
+* Protect alphas_cumprod during refiner switchover ([#14979](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14979))
+* Fix EXIF orientation in API image loading ([#15062](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15062))
+* Only override emphasis if actually used in prompt ([#15141](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15141))
+* Fix emphasis infotext missing from `params.txt` ([#15142](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15142))
+* fix extract_style_text_from_prompt #15132 ([#15135](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15135))
+* Fix Soft Inpaint for AnimateDiff ([#15148](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15148))
+* edit-attention: deselect surrounding whitespace ([#15178](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15178))
+* chore: fix font not loaded ([#15183](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15183))
+* use natural sort in extra networks when ordering by path
+* Fix built-in lora system bugs caused by torch.nn.MultiheadAttention ([#15190](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15190))
+* Avoid error from None in get_learned_conditioning ([#15191](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15191))
+* Add entry to MassFileLister after writing metadata ([#15199](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15199))
+* fix issue with Styles when Hires prompt is used ([#15269](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15269), [#15276](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15276))
+* Strip comments from hires fix prompt ([#15263](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15263))
+* Make imageviewer event listeners browser consistent ([#15261](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15261))
+* Fix AttributeError in OFT when trying to get MultiheadAttention weight ([#15260](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15260))
+* Add missing .mean() back ([#15239](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15239))
+* fix "Restore progress" button ([#15221](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15221))
+* fix ui-config for InputAccordion [custom_script_source] ([#15231](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15231))
+* handle 0 wheel deltaY ([#15268](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15268))
+* prevent alt menu for firefox ([#15267](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15267))
+* fix: fix syntax errors ([#15179](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15179))
+* restore outputs path ([#15307](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15307))
+* Escape btn_copy_path filename ([#15316](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15316))
+* Fix extra networks buttons when filename contains an apostrophe ([#15331](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15331))
+* escape brackets in lora random prompt generator ([#15343](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15343))
+* fix: Python version check for PyTorch installation compatibility ([#15390](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15390))
+* fix typo in call_queue.py ([#15386](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15386))
+* fix: when find already_loaded model, remove loaded by array index ([#15382](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15382))
+* minor bug fix of sd model memory management ([#15350](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15350))
+* Fix CodeFormer weight ([#15414](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15414))
+* Fix: Remove script callbacks in ordered_callbacks_map ([#15428](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15428))
+* fix limited file write (thanks, Sylwia)
+* Fix extra-single-image API not doing upscale failed ([#15465](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15465))
+* error handling paste_field callables ([#15470](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15470))
+
+### Hardware:
+* Add training support and change lspci for Ascend NPU ([#14981](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14981))
+* Update to ROCm5.7 and PyTorch ([#14820](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14820))
+* Better workaround for Navi1, removing --pre for Navi3 ([#15224](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15224))
+* Ascend NPU wiki page ([#15228](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15228))
+
+### Other:
+* Update comment for Pad prompt/negative prompt v0 to add a warning about truncation, make it override the v1 implementation
+* support resizable columns for touch (tablets) ([#15002](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15002))
+* Fix #14591 using translated content to do categories mapping ([#14995](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14995))
+* Use `absolute` path for normalized filepath ([#15035](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15035))
+* resizeHandle handle double tap ([#15065](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15065))
+* --dat-models-path cmd flag ([#15039](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15039))
+* Add a direct link to the binary release ([#15059](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15059))
+* upscaler_utils: Reduce logging ([#15084](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15084))
+* Fix various typos with crate-ci/typos ([#15116](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15116))
+* fix_jpeg_live_preview ([#15102](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15102))
+* [alternative fix] can't load webui if selected wrong extra option in ui ([#15121](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15121))
+* Error handling for unsupported transparency ([#14958](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14958))
+* Add model description to searched terms ([#15198](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15198))
+* bump action version ([#15272](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15272))
+* PEP 604 annotations ([#15259](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15259))
+* Automatically Set the Scale by value when user selects an Upscale Model ([#15244](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15244))
+* move postprocessing-for-training into builtin extensions ([#15222](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15222))
+* type hinting in shared.py ([#15211](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15211))
+* update ruff to 0.3.3
+* Update pytorch lightning utilities ([#15310](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15310))
+* Add Size as an XYZ Grid option ([#15354](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15354))
+* Use HF_ENDPOINT variable for HuggingFace domain with default ([#15443](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15443))
+* re-add update_file_entry ([#15446](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15446))
+* create_infotext allow index and callable, re-work Hires prompt infotext ([#15460](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15460))
+* update restricted_opts to include more options for --hide-ui-dir-config ([#15492](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15492))
+
+
+## 1.8.0
+
+### Features:
+* Update torch to version 2.1.2
+* Soft Inpainting ([#14208](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14208))
+* FP8 support ([#14031](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14031), [#14327](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14327))
+* Support for SDXL-Inpaint Model ([#14390](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14390))
+* Use Spandrel for upscaling and face restoration architectures ([#14425](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14425), [#14467](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14467), [#14473](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14473), [#14474](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14474), [#14477](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14477), [#14476](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14476), [#14484](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14484), [#14500](https://github.com/AUTOMATIC1111/stable-difusion-webui/pull/14500), [#14501](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14501), [#14504](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14504), [#14524](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14524), [#14809](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14809))
+* Automatic backwards version compatibility (when loading infotexts from old images with program version specified, will add compatibility settings)
+* Implement zero terminal SNR noise schedule option (**[SEED BREAKING CHANGE](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Seed-breaking-changes#180-dev-170-225-2024-01-01---zero-terminal-snr-noise-schedule-option)**, [#14145](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14145), [#14979](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14979))
+* Add a [✨] button to run hires fix on selected image in the gallery (with help from [#14598](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14598), [#14626](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14626), [#14728](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14728))
+* [Separate assets repository](https://github.com/AUTOMATIC1111/stable-diffusion-webui-assets); serve fonts locally rather than from google's servers
+* Official LCM Sampler Support ([#14583](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14583))
+* Add support for DAT upscaler models ([#14690](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14690), [#15039](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15039))
+* Extra Networks Tree View ([#14588](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14588), [#14900](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14900))
+* NPU Support ([#14801](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14801))
+* Prompt comments support
+
+### Minor:
+* Allow pasting in WIDTHxHEIGHT strings into the width/height fields ([#14296](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14296))
+* add option: Live preview in full page image viewer ([#14230](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14230), [#14307](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14307))
+* Add keyboard shortcuts for generate/skip/interrupt ([#14269](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14269))
+* Better TCMALLOC support on different platforms ([#14227](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14227), [#14883](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14883), [#14910](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14910))
+* Lora not found warning ([#14464](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14464))
+* Adding negative prompts to Loras in extra networks ([#14475](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14475))
+* xyz_grid: allow varying the seed along an axis separate from axis options ([#12180](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12180))
+* option to convert VAE to bfloat16 (implementation of [#9295](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/9295))
+* Better IPEX support ([#14229](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14229), [#14353](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14353), [#14559](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14559), [#14562](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14562), [#14597](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14597))
+* Option to interrupt after current generation rather than immediately ([#13653](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13653), [#14659](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14659))
+* Fullscreen Preview control fading/disable ([#14291](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14291))
+* Finer settings freezing control ([#13789](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13789))
+* Increase Upscaler Limits ([#14589](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14589))
+* Adjust brush size with hotkeys ([#14638](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14638))
+* Add checkpoint info to csv log file when saving images ([#14663](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14663))
+* Make more columns resizable ([#14740](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14740), [#14884](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14884))
+* Add an option to not overlay original image for inpainting for #14727
+* Add Pad conds v0 option to support same generation with DDIM as before 1.6.0
+* Add "Interrupting..." placeholder.
+* Button for refresh extensions list ([#14857](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14857))
+* Add an option to disable normalization after calculating emphasis. ([#14874](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14874))
+* When counting tokens, also include enabled styles (can be disabled in settings to revert to previous behavior)
+* Configuration for the [📂] button for image gallery ([#14947](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14947))
+* Support inference with LyCORIS BOFT networks ([#14871](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14871), [#14973](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14973))
+* support resizable columns for touch (tablets) ([#15002](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15002))
+
+### Extensions and API:
+* Removed packages from requirements: basicsr, gfpgan, realesrgan; as well as their dependencies: absl-py, addict, beautifulsoup4, future, gdown, grpcio, importlib-metadata, lmdb, lpips, Markdown, platformdirs, PySocks, soupsieve, tb-nightly, tensorboard-data-server, tomli, Werkzeug, yapf, zipp, soupsieve
+* Enable task ids for API ([#14314](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14314))
+* add override_settings support for infotext API
+* rename generation_parameters_copypaste module to infotext_utils
+* prevent crash due to Script __init__ exception ([#14407](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14407))
+* Bump numpy to 1.26.2 ([#14471](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14471))
+* Add utility to inspect a model's dtype/device ([#14478](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14478))
+* Implement general forward method for all method in built-in lora ext ([#14547](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14547))
+* Execute model_loaded_callback after moving to target device ([#14563](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14563))
+* Add self to CFGDenoiserParams ([#14573](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14573))
+* Allow TLS with API only mode (--nowebui) ([#14593](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14593))
+* New callback: postprocess_image_after_composite ([#14657](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14657))
+* modules/api/api.py: add api endpoint to refresh embeddings list ([#14715](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14715))
+* set_named_arg ([#14773](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14773))
+* add before_token_counter callback and use it for prompt comments
+* ResizeHandleRow - allow overridden column scale parameter ([#15004](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15004))
+
+### Performance:
+* Massive performance improvement for extra networks directories with a huge number of files in them in an attempt to tackle #14507 ([#14528](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14528))
+* Reduce unnecessary re-indexing extra networks directory ([#14512](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14512))
+* Avoid unnecessary `isfile`/`exists` calls ([#14527](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14527))
+
+### Bug Fixes:
+* fix multiple bugs related to styles multi-file support ([#14203](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14203), [#14276](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14276), [#14707](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14707))
+* Lora fixes ([#14300](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14300), [#14237](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14237), [#14546](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14546), [#14726](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14726))
+* Re-add setting lost as part of e294e46 ([#14266](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14266))
+* fix extras caption BLIP ([#14330](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14330))
+* include infotext into saved init image for img2img ([#14452](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14452))
+* xyz grid handle axis_type is None ([#14394](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14394))
+* Update Added (Fixed) IPV6 Functionality When there is No Webui Argument Passed webui.py ([#14354](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14354))
+* fix API thread safe issues of txt2img and img2img ([#14421](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14421))
+* handle selectable script_index is None ([#14487](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14487))
+* handle config.json failed to load ([#14525](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14525), [#14767](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14767))
+* paste infotext cast int as float ([#14523](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14523))
+* Ensure GRADIO_ANALYTICS_ENABLED is set early enough ([#14537](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14537))
+* Fix logging configuration again ([#14538](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14538))
+* Handle CondFunc exception when resolving attributes ([#14560](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14560))
+* Fix extras big batch crashes ([#14699](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14699))
+* Fix using wrong model caused by alias ([#14655](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14655))
+* Add # to the invalid_filename_chars list ([#14640](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14640))
+* Fix extension check for requirements ([#14639](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14639))
+* Fix tab indexes are reset after restart UI ([#14637](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14637))
+* Fix nested manual cast ([#14689](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14689))
+* Keep postprocessing upscale selected tab after restart ([#14702](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14702))
+* XYZ grid: filter out blank vals when axis is int or float type (like int axis seed) ([#14754](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14754))
+* fix CLIP Interrogator topN regex ([#14775](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14775))
+* Fix dtype error in MHA layer/change dtype checking mechanism for manual cast ([#14791](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14791))
+* catch load style.csv error ([#14814](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14814))
+* fix error when editing extra networks card
+* fix extra networks metadata failing to work properly when you create the .json file with metadata for the first time.
+* util.walk_files extensions case insensitive ([#14879](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14879))
+* if extensions page not loaded, prevent apply ([#14873](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14873))
+* call the right function for token counter in img2img
+* Fix the bugs that search/reload will disappear when using other ExtraNetworks extensions ([#14939](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14939))
+* Gracefully handle mtime read exception from cache ([#14933](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14933))
+* Only trigger interrupt on `Esc` when interrupt button visible ([#14932](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14932))
+* Disable prompt token counters option actually disables token counting rather than just hiding results.
+* avoid double upscaling in inpaint ([#14966](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14966))
+* Fix #14591 using translated content to do categories mapping ([#14995](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14995))
+* fix: the `split_threshold` parameter does not work when running Split oversized images ([#15006](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15006))
+* Fix resize-handle for mobile ([#15010](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15010), [#15065](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15065))
+
+### Other:
+* Assign id for "extra_options". Replace numeric field with slider. ([#14270](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14270))
+* change state dict comparison to ref compare ([#14216](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14216))
+* Bump torch-rocm to 5.6/5.7 ([#14293](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14293))
+* Base output path off data path ([#14446](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14446))
+* reorder training preprocessing modules in extras tab ([#14367](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14367))
+* Remove `cleanup_models` code ([#14472](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14472))
+* only rewrite ui-config when there is change ([#14352](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14352))
+* Fix lint issue from 501993eb ([#14495](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14495))
+* Update README.md ([#14548](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14548))
+* hires button, fix seeds ()
+* Logging: set formatter correctly for fallback logger too ([#14618](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14618))
+* Read generation info from infotexts rather than json for internal needs (save, extract seed from generated pic) ([#14645](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14645))
+* improve get_crop_region ([#14709](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14709))
+* Bump safetensors' version to 0.4.2 ([#14782](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14782))
+* add tooltip create_submit_box ([#14803](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14803))
+* extensions tab table row hover highlight ([#14885](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14885))
+* Always add timestamp to displayed image ([#14890](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14890))
+* Added core.filemode=false so doesn't track changes in file permission… ([#14930](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14930))
+* Normalize command-line argument paths ([#14934](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14934), [#15035](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15035))
+* Use original App Title in progress bar ([#14916](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14916))
+* register_tmp_file also for mtime ([#15012](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15012))
+
+## 1.7.0
+
+### Features:
+* settings tab rework: add search field, add categories, split UI settings page into many
+* add altdiffusion-m18 support ([#13364](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13364))
+* support inference with LyCORIS GLora networks ([#13610](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13610))
+* add lora-embedding bundle system ([#13568](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13568))
+* option to move prompt from top row into generation parameters
+* add support for SSD-1B ([#13865](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13865))
+* support inference with OFT networks ([#13692](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13692))
+* script metadata and DAG sorting mechanism ([#13944](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13944))
+* support HyperTile optimization ([#13948](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13948))
+* add support for SD 2.1 Turbo ([#14170](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14170))
+* remove Train->Preprocessing tab and put all its functionality into Extras tab
+* initial IPEX support for Intel Arc GPU ([#14171](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14171))
+
+### Minor:
+* allow reading model hash from images in img2img batch mode ([#12767](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12767))
+* add option to align with sgm repo's sampling implementation ([#12818](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12818))
+* extra field for lora metadata viewer: `ss_output_name` ([#12838](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12838))
+* add action in settings page to calculate all SD checkpoint hashes ([#12909](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12909))
+* add button to copy prompt to style editor ([#12975](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12975))
+* add --skip-load-model-at-start option ([#13253](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13253))
+* write infotext to gif images
+* read infotext from gif images ([#13068](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13068))
+* allow configuring the initial state of InputAccordion in ui-config.json ([#13189](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13189))
+* allow editing whitespace delimiters for ctrl+up/ctrl+down prompt editing ([#13444](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13444))
+* prevent accidentally closing popup dialogs ([#13480](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13480))
+* added option to play notification sound or not ([#13631](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13631))
+* show the preview image in the full screen image viewer if available ([#13459](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13459))
+* support for webui.settings.bat ([#13638](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13638))
+* add an option to not print stack traces on ctrl+c
+* start/restart generation by Ctrl (Alt) + Enter ([#13644](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13644))
+* update prompts_from_file script to allow concatenating entries with the general prompt ([#13733](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13733))
+* added a visible checkbox to input accordion
+* added an option to hide all txt2img/img2img parameters in an accordion ([#13826](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13826))
+* added 'Path' sorting option for Extra network cards ([#13968](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13968))
+* enable prompt hotkeys in style editor ([#13931](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13931))
+* option to show batch img2img results in UI ([#14009](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14009))
+* infotext updates: add option to disregard certain infotext fields, add option to not include VAE in infotext, add explanation to infotext settings page, move some options to infotext settings page
+* add FP32 fallback support on sd_vae_approx ([#14046](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14046))
+* support XYZ scripts / split hires path from unet ([#14126](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14126))
+* allow use of multiple styles csv files ([#14125](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14125))
+* make extra network card description plaintext by default, with an option (Treat card description as HTML) to re-enable HTML as it was (originally by [#13241](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13241))
+
+### Extensions and API:
+* update gradio to 3.41.2
+* support installed extensions list api ([#12774](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12774))
+* update pnginfo API to return dict with parsed values
+* add noisy latent to `ExtraNoiseParams` for callback ([#12856](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12856))
+* show extension datetime in UTC ([#12864](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12864), [#12865](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12865), [#13281](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13281))
+* add an option to choose how to combine hires fix and refiner
+* include program version in info response. ([#13135](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13135))
+* sd_unet support for SDXL
+* patch DDPM.register_betas so that users can put given_betas in model yaml ([#13276](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13276))
+* xyz_grid: add prepare ([#13266](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13266))
+* allow multiple localization files with same language in extensions ([#13077](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13077))
+* add onEdit function for js and rework token-counter.js to use it
+* fix the key error exception when processing override_settings keys ([#13567](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13567))
+* ability for extensions to return custom data via api in response.images ([#13463](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13463))
+* call state.jobnext() before postproces*() ([#13762](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13762))
+* add option to set notification sound volume ([#13884](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13884))
+* update Ruff to 0.1.6 ([#14059](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14059))
+* add Block component creation callback ([#14119](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14119))
+* catch uncaught exception with ui creation scripts ([#14120](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14120))
+* use extension name for determining an extension is installed in the index ([#14063](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14063))
+* update is_installed() from launch_utils.py to fix reinstalling already installed packages ([#14192](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14192))
+
+### Bug Fixes:
+* fix pix2pix producing bad results
+* fix defaults settings page breaking when any of main UI tabs are hidden
+* fix error that causes some extra networks to be disabled if both <lora:> and <lyco:> are present in the prompt
+* fix for Reload UI function: if you reload UI on one tab, other opened tabs will no longer stop working
+* prevent duplicate resize handler ([#12795](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12795))
+* small typo: vae resolve bug ([#12797](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12797))
+* hide broken image crop tool ([#12792](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12792))
+* don't show hidden samplers in dropdown for XYZ script ([#12780](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12780))
+* fix style editing dialog breaking if it's opened in both img2img and txt2img tabs
+* hide --gradio-auth and --api-auth values from /internal/sysinfo report
+* add missing infotext for RNG in options ([#12819](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12819))
+* fix notification not playing when built-in webui tab is inactive ([#12834](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12834))
+* honor `--skip-install` for extension installers ([#12832](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12832))
+* don't print blank stdout in extension installers ([#12833](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12833), [#12855](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12855))
+* get progressbar to display correctly in extensions tab
+* keep order in list of checkpoints when loading model that doesn't have a checksum
+* fix inpainting models in txt2img creating black pictures
+* fix generation params regex ([#12876](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12876))
+* fix batch img2img output dir with script ([#12926](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12926))
+* fix #13080 - Hypernetwork/TI preview generation ([#13084](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13084))
+* fix bug with sigma min/max overrides. ([#12995](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12995))
+* more accurate check for enabling cuDNN benchmark on 16XX cards ([#12924](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12924))
+* don't use multicond parser for negative prompt counter ([#13118](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13118))
+* fix data-sort-name containing spaces ([#13412](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13412))
+* update card on correct tab when editing metadata ([#13411](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13411))
+* fix viewing/editing metadata when filename contains an apostrophe ([#13395](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13395))
+* fix: --sd_model in "Prompts from file or textbox" script is not working ([#13302](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13302))
+* better Support for Portable Git ([#13231](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13231))
+* fix issues when webui_dir is not work_dir ([#13210](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13210))
+* fix: lora-bias-backup don't reset cache ([#13178](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13178))
+* account for customizable extra network separators whyen removing extra network text from the prompt ([#12877](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12877))
+* re fix batch img2img output dir with script ([#13170](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13170))
+* fix `--ckpt-dir` path separator and option use `short name` for checkpoint dropdown ([#13139](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13139))
+* consolidated allowed preview formats, Fix extra network `.gif` not woking as preview ([#13121](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13121))
+* fix venv_dir=- environment variable not working as expected on linux ([#13469](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13469))
+* repair unload sd checkpoint button
+* edit-attention fixes ([#13533](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13533))
+* fix bug when using --gfpgan-models-path ([#13718](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13718))
+* properly apply sort order for extra network cards when selected from dropdown
+* fixes generation restart not working for some users when 'Ctrl+Enter' is pressed ([#13962](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13962))
+* thread safe extra network list_items ([#13014](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13014))
+* fix not able to exit metadata popup when pop up is too big ([#14156](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14156))
+* fix auto focal point crop for opencv >= 4.8 ([#14121](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14121))
+* make 'use-cpu all' actually apply to 'all' ([#14131](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14131))
+* extras tab batch: actually use original filename
+* make webui not crash when running with --disable-all-extensions option
+
+### Other:
+* non-local condition ([#12814](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12814))
+* fix minor typos ([#12827](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12827))
+* remove xformers Python version check ([#12842](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12842))
+* style: file-metadata word-break ([#12837](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12837))
+* revert SGM noise multiplier change for img2img because it breaks hires fix
+* do not change quicksettings dropdown option when value returned is `None` ([#12854](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12854))
+* [RC 1.6.0 - zoom is partly hidden] Update style.css ([#12839](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12839))
+* chore: change extension time format ([#12851](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12851))
+* WEBUI.SH - Use torch 2.1.0 release candidate for Navi 3 ([#12929](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12929))
+* add Fallback at images.read_info_from_image if exif data was invalid ([#13028](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13028))
+* update cmd arg description ([#12986](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12986))
+* fix: update shared.opts.data when add_option ([#12957](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12957), [#13213](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13213))
+* restore missing tooltips ([#12976](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12976))
+* use default dropdown padding on mobile ([#12880](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12880))
+* put enable console prompts option into settings from commandline args ([#13119](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13119))
+* fix some deprecated types ([#12846](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12846))
+* bump to torchsde==0.2.6 ([#13418](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13418))
+* update dragdrop.js ([#13372](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13372))
+* use orderdict as lru cache:opt/bug ([#13313](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13313))
+* XYZ if not include sub grids do not save sub grid ([#13282](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13282))
+* initialize state.time_start befroe state.job_count ([#13229](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13229))
+* fix fieldname regex ([#13458](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13458))
+* change denoising_strength default to None. ([#13466](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13466))
+* fix regression ([#13475](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13475))
+* fix IndexError ([#13630](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13630))
+* fix: checkpoints_loaded:{checkpoint:state_dict}, model.load_state_dict issue in dict value empty ([#13535](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13535))
+* update bug_report.yml ([#12991](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12991))
+* requirements_versions httpx==0.24.1 ([#13839](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13839))
+* fix parenthesis auto selection ([#13829](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13829))
+* fix #13796 ([#13797](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13797))
+* corrected a typo in `modules/cmd_args.py` ([#13855](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13855))
+* feat: fix randn found element of type float at pos 2 ([#14004](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14004))
+* adds tqdm handler to logging_config.py for progress bar integration ([#13996](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13996))
+* hotfix: call shared.state.end() after postprocessing done ([#13977](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13977))
+* fix dependency address patch 1 ([#13929](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13929))
+* save sysinfo as .json ([#14035](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14035))
+* move exception_records related methods to errors.py ([#14084](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14084))
+* compatibility ([#13936](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13936))
+* json.dump(ensure_ascii=False) ([#14108](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14108))
+* dir buttons start with / so only the correct dir will be shown and no… ([#13957](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13957))
+* alternate implementation for unet forward replacement that does not depend on hijack being applied
+* re-add `keyedit_delimiters_whitespace` setting lost as part of commit e294e46 ([#14178](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14178))
+* fix `save_samples` being checked early when saving masked composite ([#14177](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14177))
+* slight optimization for mask and mask_composite ([#14181](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14181))
+* add import_hook hack to work around basicsr/torchvision incompatibility ([#14186](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14186))
+
+## 1.6.1
+
+### Bug Fixes:
+ * fix an error causing the webui to fail to start ([#13839](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13839))
+
 ## 1.6.0
 
 ### Features:
@@ -9,7 +587,7 @@
  * new samplers: Restart, DPM++ 2M SDE Exponential, DPM++ 2M SDE Heun, DPM++ 2M SDE Heun Karras, DPM++ 2M SDE Heun Exponential, DPM++ 3M SDE, DPM++ 3M SDE Karras, DPM++ 3M SDE Exponential ([#12300](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12300), [#12519](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12519), [#12542](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12542))
  * rework DDIM, PLMS, UniPC to use CFG denoiser same as in k-diffusion samplers:
    * makes all of them work with img2img
-   * makes prompt composition posssible (AND)
+   * makes prompt composition possible (AND)
    * makes them available for SDXL
  * always show extra networks tabs in the UI ([#11808](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/11808))
  * use less RAM when creating models ([#11958](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/11958), [#12599](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12599))
@@ -185,7 +763,7 @@
  * user metadata system for custom networks
  * extended Lora metadata editor: set activation text, default weight, view tags, training info
  * Lora extension rework to include other types of networks (all that were previously handled by LyCORIS extension)
- * show github stars for extenstions
+ * show github stars for extensions
  * img2img batch mode can read extra stuff from png info
  * img2img batch works with subdirectories
  * hotkeys to move prompt elements: alt+left/right
@@ -404,7 +982,7 @@
  * do not wait for Stable Diffusion model to load at startup
  * add filename patterns: `[denoising]`
  * directory hiding for extra networks: dirs starting with `.` will hide their cards on extra network tabs unless specifically searched for
- * LoRA: for the `<...>` text in prompt, use name of LoRA that is in the metdata of the file, if present, instead of filename (both can be used to activate LoRA)
+ * LoRA: for the `<...>` text in prompt, use name of LoRA that is in the metadata of the file, if present, instead of filename (both can be used to activate LoRA)
  * LoRA: read infotext params from kohya-ss's extension parameters if they are present and if his extension is not active
  * LoRA: fix some LoRAs not working (ones that have 3x3 convolution layer)
  * LoRA: add an option to use old method of applying LoRAs (producing same results as with kohya-ss)
@@ -434,7 +1012,7 @@
  * fix gamepad navigation
  * make the lightbox fullscreen image function properly
  * fix squished thumbnails in extras tab
- * keep "search" filter for extra networks when user refreshes the tab (previously it showed everthing after you refreshed)
+ * keep "search" filter for extra networks when user refreshes the tab (previously it showed everything after you refreshed)
  * fix webui showing the same image if you configure the generation to always save results into same file
  * fix bug with upscalers not working properly
  * fix MPS on PyTorch 2.0.1, Intel Macs
@@ -452,7 +1030,7 @@
  * switch to PyTorch 2.0.0 (except for AMD GPUs)
  * visual improvements to custom code scripts
  * add filename patterns: `[clip_skip]`, `[hasprompt<>]`, `[batch_number]`, `[generation_number]`
- * add support for saving init images in img2img, and record their hashes in infotext for reproducability
+ * add support for saving init images in img2img, and record their hashes in infotext for reproducibility
  * automatically select current word when adjusting weight with ctrl+up/down
  * add dropdowns for X/Y/Z plot
  * add setting: Stable Diffusion/Random number generator source: makes it possible to make images generated from a given manual seed consistent across different GPUs
diff --git a/README.md b/README.md
index 4e08344008c..bc62945c0c5 100644
--- a/README.md
+++ b/README.md
@@ -1,5 +1,5 @@
 # Stable Diffusion web UI
-A browser interface based on Gradio library for Stable Diffusion.
+A web interface for Stable Diffusion, implemented using Gradio library.
 
 ![](screenshot.png)
 
@@ -78,7 +78,7 @@ A browser interface based on Gradio library for Stable Diffusion.
 - Clip skip
 - Hypernetworks
 - Loras (same as Hypernetworks but more pretty)
-- A separate UI where you can choose, with preview, which embeddings, hypernetworks or Loras to add to your prompt 
+- A separate UI where you can choose, with preview, which embeddings, hypernetworks or Loras to add to your prompt
 - Can select to load a different VAE from settings screen
 - Estimated completion time in progress bar
 - API
@@ -88,22 +88,24 @@ A browser interface based on Gradio library for Stable Diffusion.
 - [Alt-Diffusion](https://arxiv.org/abs/2211.06679) support - see [wiki](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Features#alt-diffusion) for instructions
 - Now without any bad letters!
 - Load checkpoints in safetensors format
-- Eased resolution restriction: generated image's dimension must be a multiple of 8 rather than 64
+- Eased resolution restriction: generated image's dimensions must be a multiple of 8 rather than 64
 - Now with a license!
 - Reorder elements in the UI from settings screen
+- [Segmind Stable Diffusion](https://huggingface.co/segmind/SSD-1B) support
 
 ## Installation and Running
 Make sure the required [dependencies](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Dependencies) are met and follow the instructions available for:
 - [NVidia](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-NVidia-GPUs) (recommended)
 - [AMD](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-AMD-GPUs) GPUs.
 - [Intel CPUs, Intel GPUs (both integrated and discrete)](https://github.com/openvinotoolkit/stable-diffusion-webui/wiki/Installation-on-Intel-Silicon) (external wiki page)
+- [Ascend NPUs](https://github.com/wangshuai09/stable-diffusion-webui/wiki/Install-and-run-on-Ascend-NPUs) (external wiki page)
 
 Alternatively, use online services (like Google Colab):
 
 - [List of Online Services](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Online-Services)
 
 ### Installation on Windows 10/11 with NVidia-GPUs using release package
-1. Download `sd.webui.zip` from [v1.0.0-pre](https://github.com/AUTOMATIC1111/stable-diffusion-webui/releases/tag/v1.0.0-pre) and extract it's contents.
+1. Download `sd.webui.zip` from [v1.0.0-pre](https://github.com/AUTOMATIC1111/stable-diffusion-webui/releases/tag/v1.0.0-pre) and extract its contents.
 2. Run `update.bat`.
 3. Run `run.bat`.
 > For more details see [Install-and-Run-on-NVidia-GPUs](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-NVidia-GPUs)
@@ -120,14 +122,38 @@ Alternatively, use online services (like Google Colab):
 # Debian-based:
 sudo apt install wget git python3 python3-venv libgl1 libglib2.0-0
 # Red Hat-based:
-sudo dnf install wget git python3
+sudo dnf install wget git python3 gperftools-libs libglvnd-glx
+# openSUSE-based:
+sudo zypper install wget git python3 libtcmalloc4 libglvnd
 # Arch-based:
 sudo pacman -S wget git python3
 ```
+If your system is very new, you need to install python3.11 or python3.10:
+```bash
+# Ubuntu 24.04
+sudo add-apt-repository ppa:deadsnakes/ppa
+sudo apt update
+sudo apt install python3.11
+
+# Manjaro/Arch
+sudo pacman -S yay
+yay -S python311 # do not confuse with python3.11 package
+
+# Only for 3.11
+# Then set up env variable in launch script
+export python_cmd="python3.11"
+# or in webui-user.sh
+python_cmd="python3.11"
+```
 2. Navigate to the directory you would like the webui to be installed and execute the following command:
 ```bash
 wget -q https://raw.githubusercontent.com/AUTOMATIC1111/stable-diffusion-webui/master/webui.sh
 ```
+Or just clone the repo wherever you want:
+```bash
+git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui
+```
+
 3. Run `webui.sh`.
 4. Check `webui-user.sh` for options.
 ### Installation on Apple Silicon
@@ -146,13 +172,14 @@ For the purposes of getting Google and other search engines to crawl the wiki, h
 ## Credits
 Licenses for borrowed code can be found in `Settings -> Licenses` screen, and also in `html/licenses.html` file.
 
-- Stable Diffusion - https://github.com/CompVis/stable-diffusion, https://github.com/CompVis/taming-transformers
+- Stable Diffusion - https://github.com/Stability-AI/stablediffusion, https://github.com/CompVis/taming-transformers, https://github.com/mcmonkey4eva/sd3-ref
 - k-diffusion - https://github.com/crowsonkb/k-diffusion.git
-- GFPGAN - https://github.com/TencentARC/GFPGAN.git
-- CodeFormer - https://github.com/sczhou/CodeFormer
-- ESRGAN - https://github.com/xinntao/ESRGAN
-- SwinIR - https://github.com/JingyunLiang/SwinIR
-- Swin2SR - https://github.com/mv-lab/swin2sr
+- Spandrel - https://github.com/chaiNNer-org/spandrel implementing
+  - GFPGAN - https://github.com/TencentARC/GFPGAN.git
+  - CodeFormer - https://github.com/sczhou/CodeFormer
+  - ESRGAN - https://github.com/xinntao/ESRGAN
+  - SwinIR - https://github.com/JingyunLiang/SwinIR
+  - Swin2SR - https://github.com/mv-lab/swin2sr
 - LDSR - https://github.com/Hafiidz/latent-diffusion
 - MiDaS - https://github.com/isl-org/MiDaS
 - Ideas for optimizations - https://github.com/basujindal/stable-diffusion
@@ -173,5 +200,6 @@ Licenses for borrowed code can be found in `Settings -> Licenses` screen, and al
 - TAESD - Ollin Boer Bohan - https://github.com/madebyollin/taesd
 - LyCORIS - KohakuBlueleaf
 - Restart sampling - lambertae - https://github.com/Newbeeer/diffusion_restart_sampling
+- Hypertile - tfernd - https://github.com/tfernd/HyperTile
 - Initial Gradio script - posted on 4chan by an Anonymous user. Thank you Anonymous user.
 - (You)
diff --git a/_typos.toml b/_typos.toml
new file mode 100644
index 00000000000..1c63fe70331
--- /dev/null
+++ b/_typos.toml
@@ -0,0 +1,5 @@
+[default.extend-words]
+# Part of "RGBa" (Pillow's pre-multiplied alpha RGB mode)
+Ba = "Ba"
+# HSA is something AMD uses for their GPUs
+HSA = "HSA"
diff --git a/configs/alt-diffusion-inference.yaml b/configs/alt-diffusion-inference.yaml
index cfbee72d71b..4944ab5c8dc 100644
--- a/configs/alt-diffusion-inference.yaml
+++ b/configs/alt-diffusion-inference.yaml
@@ -40,7 +40,7 @@ model:
         use_spatial_transformer: True
         transformer_depth: 1
         context_dim: 768
-        use_checkpoint: True
+        use_checkpoint: False
         legacy: False
 
     first_stage_config:
diff --git a/configs/alt-diffusion-m18-inference.yaml b/configs/alt-diffusion-m18-inference.yaml
new file mode 100644
index 00000000000..c60dca8c7b3
--- /dev/null
+++ b/configs/alt-diffusion-m18-inference.yaml
@@ -0,0 +1,73 @@
+model:
+  base_learning_rate: 1.0e-04
+  target: ldm.models.diffusion.ddpm.LatentDiffusion
+  params:
+    linear_start: 0.00085
+    linear_end: 0.0120
+    num_timesteps_cond: 1
+    log_every_t: 200
+    timesteps: 1000
+    first_stage_key: "jpg"
+    cond_stage_key: "txt"
+    image_size: 64
+    channels: 4
+    cond_stage_trainable: false   # Note: different from the one we trained before
+    conditioning_key: crossattn
+    monitor: val/loss_simple_ema
+    scale_factor: 0.18215
+    use_ema: False
+
+    scheduler_config: # 10000 warmup steps
+      target: ldm.lr_scheduler.LambdaLinearScheduler
+      params:
+        warm_up_steps: [ 10000 ]
+        cycle_lengths: [ 10000000000000 ] # incredibly large number to prevent corner cases
+        f_start: [ 1.e-6 ]
+        f_max: [ 1. ]
+        f_min: [ 1. ]
+
+    unet_config:
+      target: ldm.modules.diffusionmodules.openaimodel.UNetModel
+      params:
+        image_size: 32 # unused
+        in_channels: 4
+        out_channels: 4
+        model_channels: 320
+        attention_resolutions: [ 4, 2, 1 ]
+        num_res_blocks: 2
+        channel_mult: [ 1, 2, 4, 4 ]
+        num_head_channels: 64
+        use_spatial_transformer: True
+        use_linear_in_transformer: True
+        transformer_depth: 1
+        context_dim: 1024
+        use_checkpoint: False
+        legacy: False
+
+    first_stage_config:
+      target: ldm.models.autoencoder.AutoencoderKL
+      params:
+        embed_dim: 4
+        monitor: val/rec_loss
+        ddconfig:
+          double_z: true
+          z_channels: 4
+          resolution: 256
+          in_channels: 3
+          out_ch: 3
+          ch: 128
+          ch_mult:
+          - 1
+          - 2
+          - 4
+          - 4
+          num_res_blocks: 2
+          attn_resolutions: []
+          dropout: 0.0
+        lossconfig:
+          target: torch.nn.Identity
+
+    cond_stage_config:
+      target: modules.xlmr_m18.BertSeriesModelWithTransformation
+      params:
+        name: "XLMR-Large"
diff --git a/configs/instruct-pix2pix.yaml b/configs/instruct-pix2pix.yaml
index 4e896879dd7..564e50ae246 100644
--- a/configs/instruct-pix2pix.yaml
+++ b/configs/instruct-pix2pix.yaml
@@ -45,7 +45,7 @@ model:
         use_spatial_transformer: True
         transformer_depth: 1
         context_dim: 768
-        use_checkpoint: True
+        use_checkpoint: False
         legacy: False
 
     first_stage_config:
diff --git a/configs/sd3-inference.yaml b/configs/sd3-inference.yaml
new file mode 100644
index 00000000000..bccb69d2ea3
--- /dev/null
+++ b/configs/sd3-inference.yaml
@@ -0,0 +1,5 @@
+model:
+  target: modules.models.sd3.sd3_model.SD3Inferencer
+  params:
+    shift: 3
+    state_dict: null
diff --git a/configs/sd_xl_inpaint.yaml b/configs/sd_xl_inpaint.yaml
new file mode 100644
index 00000000000..f40f45e3316
--- /dev/null
+++ b/configs/sd_xl_inpaint.yaml
@@ -0,0 +1,98 @@
+model:
+  target: sgm.models.diffusion.DiffusionEngine
+  params:
+    scale_factor: 0.13025
+    disable_first_stage_autocast: True
+
+    denoiser_config:
+      target: sgm.modules.diffusionmodules.denoiser.DiscreteDenoiser
+      params:
+        num_idx: 1000
+
+        weighting_config:
+          target: sgm.modules.diffusionmodules.denoiser_weighting.EpsWeighting
+        scaling_config:
+          target: sgm.modules.diffusionmodules.denoiser_scaling.EpsScaling
+        discretization_config:
+          target: sgm.modules.diffusionmodules.discretizer.LegacyDDPMDiscretization
+
+    network_config:
+      target: sgm.modules.diffusionmodules.openaimodel.UNetModel
+      params:
+        adm_in_channels: 2816
+        num_classes: sequential
+        use_checkpoint: False
+        in_channels: 9
+        out_channels: 4
+        model_channels: 320
+        attention_resolutions: [4, 2]
+        num_res_blocks: 2
+        channel_mult: [1, 2, 4]
+        num_head_channels: 64
+        use_spatial_transformer: True
+        use_linear_in_transformer: True
+        transformer_depth: [1, 2, 10]  # note: the first is unused (due to attn_res starting at 2) 32, 16, 8 --> 64, 32, 16
+        context_dim: 2048
+        spatial_transformer_attn_type: softmax-xformers
+        legacy: False
+
+    conditioner_config:
+      target: sgm.modules.GeneralConditioner
+      params:
+        emb_models:
+          # crossattn cond
+          - is_trainable: False
+            input_key: txt
+            target: sgm.modules.encoders.modules.FrozenCLIPEmbedder
+            params:
+              layer: hidden
+              layer_idx: 11
+          # crossattn and vector cond
+          - is_trainable: False
+            input_key: txt
+            target: sgm.modules.encoders.modules.FrozenOpenCLIPEmbedder2
+            params:
+              arch: ViT-bigG-14
+              version: laion2b_s39b_b160k
+              freeze: True
+              layer: penultimate
+              always_return_pooled: True
+              legacy: False
+          # vector cond
+          - is_trainable: False
+            input_key: original_size_as_tuple
+            target: sgm.modules.encoders.modules.ConcatTimestepEmbedderND
+            params:
+              outdim: 256  # multiplied by two
+          # vector cond
+          - is_trainable: False
+            input_key: crop_coords_top_left
+            target: sgm.modules.encoders.modules.ConcatTimestepEmbedderND
+            params:
+              outdim: 256  # multiplied by two
+          # vector cond
+          - is_trainable: False
+            input_key: target_size_as_tuple
+            target: sgm.modules.encoders.modules.ConcatTimestepEmbedderND
+            params:
+              outdim: 256  # multiplied by two
+
+    first_stage_config:
+      target: sgm.models.autoencoder.AutoencoderKLInferenceWrapper
+      params:
+        embed_dim: 4
+        monitor: val/rec_loss
+        ddconfig:
+          attn_type: vanilla-xformers
+          double_z: true
+          z_channels: 4
+          resolution: 256
+          in_channels: 3
+          out_ch: 3
+          ch: 128
+          ch_mult: [1, 2, 4, 4]
+          num_res_blocks: 2
+          attn_resolutions: []
+          dropout: 0.0
+        lossconfig:
+          target: torch.nn.Identity
diff --git a/configs/v1-inference.yaml b/configs/v1-inference.yaml
index d4effe569e8..25c4d9ed066 100644
--- a/configs/v1-inference.yaml
+++ b/configs/v1-inference.yaml
@@ -40,7 +40,7 @@ model:
         use_spatial_transformer: True
         transformer_depth: 1
         context_dim: 768
-        use_checkpoint: True
+        use_checkpoint: False
         legacy: False
 
     first_stage_config:
diff --git a/configs/v1-inpainting-inference.yaml b/configs/v1-inpainting-inference.yaml
index f9eec37d24b..68c199f99c3 100644
--- a/configs/v1-inpainting-inference.yaml
+++ b/configs/v1-inpainting-inference.yaml
@@ -40,7 +40,7 @@ model:
         use_spatial_transformer: True
         transformer_depth: 1
         context_dim: 768
-        use_checkpoint: True
+        use_checkpoint: False
         legacy: False
 
     first_stage_config:
diff --git a/extensions-builtin/LDSR/sd_hijack_ddpm_v1.py b/extensions-builtin/LDSR/sd_hijack_ddpm_v1.py
index 04adc5eb2cf..51ab1821282 100644
--- a/extensions-builtin/LDSR/sd_hijack_ddpm_v1.py
+++ b/extensions-builtin/LDSR/sd_hijack_ddpm_v1.py
@@ -301,7 +301,7 @@ def p_losses(self, x_start, t, noise=None):
         elif self.parameterization == "x0":
             target = x_start
         else:
-            raise NotImplementedError(f"Paramterization {self.parameterization} not yet supported")
+            raise NotImplementedError(f"Parameterization {self.parameterization} not yet supported")
 
         loss = self.get_loss(model_out, target, mean=False).mean(dim=[1, 2, 3])
 
@@ -572,7 +572,7 @@ def delta_border(self, h, w):
         :param h: height
         :param w: width
         :return: normalized distance to image border,
-         wtith min distance = 0 at border and max dist = 0.5 at image center
+         with min distance = 0 at border and max dist = 0.5 at image center
         """
         lower_right_corner = torch.tensor([h - 1, w - 1]).view(1, 1, 2)
         arr = self.meshgrid(h, w) / lower_right_corner
@@ -880,7 +880,7 @@ def forward(self, x, c, *args, **kwargs):
     def apply_model(self, x_noisy, t, cond, return_ids=False):
 
         if isinstance(cond, dict):
-            # hybrid case, cond is exptected to be a dict
+            # hybrid case, cond is expected to be a dict
             pass
         else:
             if not isinstance(cond, list):
@@ -916,7 +916,7 @@ def apply_model(self, x_noisy, t, cond, return_ids=False):
                 cond_list = [{c_key: [c[:, :, :, :, i]]} for i in range(c.shape[-1])]
 
             elif self.cond_stage_key == 'coordinates_bbox':
-                assert 'original_image_size' in self.split_input_params, 'BoudingBoxRescaling is missing original_image_size'
+                assert 'original_image_size' in self.split_input_params, 'BoundingBoxRescaling is missing original_image_size'
 
                 # assuming padding of unfold is always 0 and its dilation is always 1
                 n_patches_per_row = int((w - ks[0]) / stride[0] + 1)
@@ -926,7 +926,7 @@ def apply_model(self, x_noisy, t, cond, return_ids=False):
                 num_downs = self.first_stage_model.encoder.num_resolutions - 1
                 rescale_latent = 2 ** (num_downs)
 
-                # get top left postions of patches as conforming for the bbbox tokenizer, therefore we
+                # get top left positions of patches as conforming for the bbbox tokenizer, therefore we
                 # need to rescale the tl patch coordinates to be in between (0,1)
                 tl_patch_coordinates = [(rescale_latent * stride[0] * (patch_nr % n_patches_per_row) / full_img_w,
                                          rescale_latent * stride[1] * (patch_nr // n_patches_per_row) / full_img_h)
diff --git a/extensions-builtin/Lora/extra_networks_lora.py b/extensions-builtin/Lora/extra_networks_lora.py
index 005ff32cbe3..17a620f77e3 100644
--- a/extensions-builtin/Lora/extra_networks_lora.py
+++ b/extensions-builtin/Lora/extra_networks_lora.py
@@ -9,6 +9,8 @@ def __init__(self):
         self.errors = {}
         """mapping of network names to the number of errors the network had during operation"""
 
+    remove_symbols = str.maketrans('', '', ":,")
+
     def activate(self, p, params_list):
         additional = shared.opts.sd_lora
 
@@ -43,22 +45,15 @@ def activate(self, p, params_list):
         networks.load_networks(names, te_multipliers, unet_multipliers, dyn_dims)
 
         if shared.opts.lora_add_hashes_to_infotext:
-            network_hashes = []
-            for item in networks.loaded_networks:
-                shorthash = item.network_on_disk.shorthash
-                if not shorthash:
-                    continue
-
-                alias = item.mentioned_name
-                if not alias:
-                    continue
+            if not getattr(p, "is_hr_pass", False) or not hasattr(p, "lora_hashes"):
+                p.lora_hashes = {}
 
-                alias = alias.replace(":", "").replace(",", "")
-
-                network_hashes.append(f"{alias}: {shorthash}")
+            for item in networks.loaded_networks:
+                if item.network_on_disk.shorthash and item.mentioned_name:
+                    p.lora_hashes[item.mentioned_name.translate(self.remove_symbols)] = item.network_on_disk.shorthash
 
-            if network_hashes:
-                p.extra_generation_params["Lora hashes"] = ", ".join(network_hashes)
+            if p.lora_hashes:
+                p.extra_generation_params["Lora hashes"] = ', '.join(f'{k}: {v}' for k, v in p.lora_hashes.items())
 
     def deactivate(self, p):
         if self.errors:
diff --git a/extensions-builtin/Lora/lora_logger.py b/extensions-builtin/Lora/lora_logger.py
new file mode 100644
index 00000000000..d51de29704f
--- /dev/null
+++ b/extensions-builtin/Lora/lora_logger.py
@@ -0,0 +1,33 @@
+import sys
+import copy
+import logging
+
+
+class ColoredFormatter(logging.Formatter):
+    COLORS = {
+        "DEBUG": "\033[0;36m",  # CYAN
+        "INFO": "\033[0;32m",  # GREEN
+        "WARNING": "\033[0;33m",  # YELLOW
+        "ERROR": "\033[0;31m",  # RED
+        "CRITICAL": "\033[0;37;41m",  # WHITE ON RED
+        "RESET": "\033[0m",  # RESET COLOR
+    }
+
+    def format(self, record):
+        colored_record = copy.copy(record)
+        levelname = colored_record.levelname
+        seq = self.COLORS.get(levelname, self.COLORS["RESET"])
+        colored_record.levelname = f"{seq}{levelname}{self.COLORS['RESET']}"
+        return super().format(colored_record)
+
+
+logger = logging.getLogger("lora")
+logger.propagate = False
+
+
+if not logger.handlers:
+    handler = logging.StreamHandler(sys.stdout)
+    handler.setFormatter(
+        ColoredFormatter("[%(name)s]-%(levelname)s: %(message)s")
+    )
+    logger.addHandler(handler)
diff --git a/extensions-builtin/Lora/lyco_helpers.py b/extensions-builtin/Lora/lyco_helpers.py
index 279b34bc928..6f134d54eb1 100644
--- a/extensions-builtin/Lora/lyco_helpers.py
+++ b/extensions-builtin/Lora/lyco_helpers.py
@@ -19,3 +19,50 @@ def rebuild_cp_decomposition(up, down, mid):
     up = up.reshape(up.size(0), -1)
     down = down.reshape(down.size(0), -1)
     return torch.einsum('n m k l, i n, m j -> i j k l', mid, up, down)
+
+
+# copied from https://github.com/KohakuBlueleaf/LyCORIS/blob/dev/lycoris/modules/lokr.py
+def factorization(dimension: int, factor:int=-1) -> tuple[int, int]:
+    '''
+    return a tuple of two value of input dimension decomposed by the number closest to factor
+    second value is higher or equal than first value.
+
+    In LoRA with Kroneckor Product, first value is a value for weight scale.
+    secon value is a value for weight.
+
+    Because of non-commutative property, A⊗B ≠ B⊗A. Meaning of two matrices is slightly different.
+
+    examples)
+    factor
+        -1               2                4               8               16               ...
+    127 -> 1, 127   127 -> 1, 127    127 -> 1, 127   127 -> 1, 127   127 -> 1, 127
+    128 -> 8, 16    128 -> 2, 64     128 -> 4, 32    128 -> 8, 16    128 -> 8, 16
+    250 -> 10, 25   250 -> 2, 125    250 -> 2, 125   250 -> 5, 50    250 -> 10, 25
+    360 -> 8, 45    360 -> 2, 180    360 -> 4, 90    360 -> 8, 45    360 -> 12, 30
+    512 -> 16, 32   512 -> 2, 256    512 -> 4, 128   512 -> 8, 64    512 -> 16, 32
+    1024 -> 32, 32  1024 -> 2, 512   1024 -> 4, 256  1024 -> 8, 128  1024 -> 16, 64
+    '''
+
+    if factor > 0 and (dimension % factor) == 0:
+        m = factor
+        n = dimension // factor
+        if m > n:
+            n, m = m, n
+        return m, n
+    if factor < 0:
+        factor = dimension
+    m, n = 1, dimension
+    length = m + n
+    while m<n:
+        new_m = m + 1
+        while dimension%new_m != 0:
+            new_m += 1
+        new_n = dimension // new_m
+        if new_m + new_n > length or new_m>factor:
+            break
+        else:
+            m, n = new_m, new_n
+    if m > n:
+        n, m = m, n
+    return m, n
+
diff --git a/extensions-builtin/Lora/network.py b/extensions-builtin/Lora/network.py
index d8e8dfb7ff0..98ff367fd8a 100644
--- a/extensions-builtin/Lora/network.py
+++ b/extensions-builtin/Lora/network.py
@@ -3,7 +3,11 @@
 from collections import namedtuple
 import enum
 
+import torch.nn as nn
+import torch.nn.functional as F
+
 from modules import sd_models, cache, errors, hashes, shared
+import modules.models.sd3.mmdit
 
 NetworkWeights = namedtuple('NetworkWeights', ['network_key', 'sd_key', 'w', 'sd_module'])
 
@@ -26,7 +30,6 @@ def __init__(self, name, filename):
 
         def read_metadata():
             metadata = sd_models.read_metadata_from_safetensors(filename)
-            metadata.pop('ssmd_cover_images', None)  # those are cover images, and they are too big to display in UI as text
 
             return metadata
 
@@ -93,6 +96,7 @@ def __init__(self, name, network_on_disk: NetworkOnDisk):
         self.unet_multiplier = 1.0
         self.dyn_dim = None
         self.modules = {}
+        self.bundle_embeddings = {}
         self.mtime = None
 
         self.mentioned_name = None
@@ -111,14 +115,49 @@ def __init__(self, net: Network, weights: NetworkWeights):
         self.sd_key = weights.sd_key
         self.sd_module = weights.sd_module
 
-        if hasattr(self.sd_module, 'weight'):
+        if isinstance(self.sd_module, modules.models.sd3.mmdit.QkvLinear):
+            s = self.sd_module.weight.shape
+            self.shape = (s[0] // 3, s[1])
+        elif hasattr(self.sd_module, 'weight'):
             self.shape = self.sd_module.weight.shape
+        elif isinstance(self.sd_module, nn.MultiheadAttention):
+            # For now, only self-attn use Pytorch's MHA
+            # So assume all qkvo proj have same shape
+            self.shape = self.sd_module.out_proj.weight.shape
+        else:
+            self.shape = None
+
+        self.ops = None
+        self.extra_kwargs = {}
+        if isinstance(self.sd_module, nn.Conv2d):
+            self.ops = F.conv2d
+            self.extra_kwargs = {
+                'stride': self.sd_module.stride,
+                'padding': self.sd_module.padding
+            }
+        elif isinstance(self.sd_module, nn.Linear):
+            self.ops = F.linear
+        elif isinstance(self.sd_module, nn.LayerNorm):
+            self.ops = F.layer_norm
+            self.extra_kwargs = {
+                'normalized_shape': self.sd_module.normalized_shape,
+                'eps': self.sd_module.eps
+            }
+        elif isinstance(self.sd_module, nn.GroupNorm):
+            self.ops = F.group_norm
+            self.extra_kwargs = {
+                'num_groups': self.sd_module.num_groups,
+                'eps': self.sd_module.eps
+            }
 
         self.dim = None
         self.bias = weights.w.get("bias")
         self.alpha = weights.w["alpha"].item() if "alpha" in weights.w else None
         self.scale = weights.w["scale"].item() if "scale" in weights.w else None
 
+        self.dora_scale = weights.w.get("dora_scale", None)
+        self.dora_norm_dims = len(self.shape) - 1
+
     def multiplier(self):
         if 'transformer' in self.sd_key[:20]:
             return self.network.te_multiplier
@@ -133,10 +172,31 @@ def calc_scale(self):
 
         return 1.0
 
+    def apply_weight_decompose(self, updown, orig_weight):
+        # Match the device/dtype
+        orig_weight = orig_weight.to(updown.dtype)
+        dora_scale = self.dora_scale.to(device=orig_weight.device, dtype=updown.dtype)
+        updown = updown.to(orig_weight.device)
+
+        merged_scale1 = updown + orig_weight
+        merged_scale1_norm = (
+            merged_scale1.transpose(0, 1)
+            .reshape(merged_scale1.shape[1], -1)
+            .norm(dim=1, keepdim=True)
+            .reshape(merged_scale1.shape[1], *[1] * self.dora_norm_dims)
+            .transpose(0, 1)
+        )
+
+        dora_merged = (
+            merged_scale1 * (dora_scale / merged_scale1_norm)
+        )
+        final_updown = dora_merged - orig_weight
+        return final_updown
+
     def finalize_updown(self, updown, orig_weight, output_shape, ex_bias=None):
         if self.bias is not None:
             updown = updown.reshape(self.bias.shape)
-            updown += self.bias.to(orig_weight.device, dtype=orig_weight.dtype)
+            updown += self.bias.to(orig_weight.device, dtype=updown.dtype)
             updown = updown.reshape(output_shape)
 
         if len(output_shape) == 4:
@@ -148,11 +208,21 @@ def finalize_updown(self, updown, orig_weight, output_shape, ex_bias=None):
         if ex_bias is not None:
             ex_bias = ex_bias * self.multiplier()
 
-        return updown * self.calc_scale() * self.multiplier(), ex_bias
+        updown = updown * self.calc_scale()
+
+        if self.dora_scale is not None:
+            updown = self.apply_weight_decompose(updown, orig_weight)
+
+        return updown * self.multiplier(), ex_bias
 
     def calc_updown(self, target):
         raise NotImplementedError()
 
     def forward(self, x, y):
-        raise NotImplementedError()
+        """A general forward implementation for all modules"""
+        if self.ops is None:
+            raise NotImplementedError()
+        else:
+            updown, ex_bias = self.calc_updown(self.sd_module.weight)
+            return y + self.ops(x, weight=updown, bias=ex_bias, **self.extra_kwargs)
 
diff --git a/extensions-builtin/Lora/network_full.py b/extensions-builtin/Lora/network_full.py
index bf6930e96c0..f221c95f3b5 100644
--- a/extensions-builtin/Lora/network_full.py
+++ b/extensions-builtin/Lora/network_full.py
@@ -18,9 +18,9 @@ def __init__(self,  net: network.Network, weights: network.NetworkWeights):
 
     def calc_updown(self, orig_weight):
         output_shape = self.weight.shape
-        updown = self.weight.to(orig_weight.device, dtype=orig_weight.dtype)
+        updown = self.weight.to(orig_weight.device)
         if self.ex_bias is not None:
-            ex_bias = self.ex_bias.to(orig_weight.device, dtype=orig_weight.dtype)
+            ex_bias = self.ex_bias.to(orig_weight.device)
         else:
             ex_bias = None
 
diff --git a/extensions-builtin/Lora/network_glora.py b/extensions-builtin/Lora/network_glora.py
new file mode 100644
index 00000000000..efe5c6814fa
--- /dev/null
+++ b/extensions-builtin/Lora/network_glora.py
@@ -0,0 +1,33 @@
+
+import network
+
+class ModuleTypeGLora(network.ModuleType):
+    def create_module(self, net: network.Network, weights: network.NetworkWeights):
+        if all(x in weights.w for x in ["a1.weight", "a2.weight", "alpha", "b1.weight", "b2.weight"]):
+            return NetworkModuleGLora(net, weights)
+
+        return None
+
+# adapted from https://github.com/KohakuBlueleaf/LyCORIS
+class NetworkModuleGLora(network.NetworkModule):
+    def __init__(self,  net: network.Network, weights: network.NetworkWeights):
+        super().__init__(net, weights)
+
+        if hasattr(self.sd_module, 'weight'):
+            self.shape = self.sd_module.weight.shape
+
+        self.w1a = weights.w["a1.weight"]
+        self.w1b = weights.w["b1.weight"]
+        self.w2a = weights.w["a2.weight"]
+        self.w2b = weights.w["b2.weight"]
+
+    def calc_updown(self, orig_weight):
+        w1a = self.w1a.to(orig_weight.device)
+        w1b = self.w1b.to(orig_weight.device)
+        w2a = self.w2a.to(orig_weight.device)
+        w2b = self.w2b.to(orig_weight.device)
+
+        output_shape = [w1a.size(0), w1b.size(1)]
+        updown = ((w2b @ w1b) + ((orig_weight.to(dtype = w1a.dtype) @ w2a) @ w1a))
+
+        return self.finalize_updown(updown, orig_weight, output_shape)
diff --git a/extensions-builtin/Lora/network_hada.py b/extensions-builtin/Lora/network_hada.py
index 5fcb0695fbb..d95a0fd18e3 100644
--- a/extensions-builtin/Lora/network_hada.py
+++ b/extensions-builtin/Lora/network_hada.py
@@ -27,16 +27,16 @@ def __init__(self,  net: network.Network, weights: network.NetworkWeights):
         self.t2 = weights.w.get("hada_t2")
 
     def calc_updown(self, orig_weight):
-        w1a = self.w1a.to(orig_weight.device, dtype=orig_weight.dtype)
-        w1b = self.w1b.to(orig_weight.device, dtype=orig_weight.dtype)
-        w2a = self.w2a.to(orig_weight.device, dtype=orig_weight.dtype)
-        w2b = self.w2b.to(orig_weight.device, dtype=orig_weight.dtype)
+        w1a = self.w1a.to(orig_weight.device)
+        w1b = self.w1b.to(orig_weight.device)
+        w2a = self.w2a.to(orig_weight.device)
+        w2b = self.w2b.to(orig_weight.device)
 
         output_shape = [w1a.size(0), w1b.size(1)]
 
         if self.t1 is not None:
             output_shape = [w1a.size(1), w1b.size(1)]
-            t1 = self.t1.to(orig_weight.device, dtype=orig_weight.dtype)
+            t1 = self.t1.to(orig_weight.device)
             updown1 = lyco_helpers.make_weight_cp(t1, w1a, w1b)
             output_shape += t1.shape[2:]
         else:
@@ -45,7 +45,7 @@ def calc_updown(self, orig_weight):
             updown1 = lyco_helpers.rebuild_conventional(w1a, w1b, output_shape)
 
         if self.t2 is not None:
-            t2 = self.t2.to(orig_weight.device, dtype=orig_weight.dtype)
+            t2 = self.t2.to(orig_weight.device)
             updown2 = lyco_helpers.make_weight_cp(t2, w2a, w2b)
         else:
             updown2 = lyco_helpers.rebuild_conventional(w2a, w2b, output_shape)
diff --git a/extensions-builtin/Lora/network_ia3.py b/extensions-builtin/Lora/network_ia3.py
index 7edc4249791..96faeaf3ede 100644
--- a/extensions-builtin/Lora/network_ia3.py
+++ b/extensions-builtin/Lora/network_ia3.py
@@ -17,7 +17,7 @@ def __init__(self,  net: network.Network, weights: network.NetworkWeights):
         self.on_input = weights.w["on_input"].item()
 
     def calc_updown(self, orig_weight):
-        w = self.w.to(orig_weight.device, dtype=orig_weight.dtype)
+        w = self.w.to(orig_weight.device)
 
         output_shape = [w.size(0), orig_weight.size(1)]
         if self.on_input:
diff --git a/extensions-builtin/Lora/network_lokr.py b/extensions-builtin/Lora/network_lokr.py
index 340acdab3d0..fcdaeafd896 100644
--- a/extensions-builtin/Lora/network_lokr.py
+++ b/extensions-builtin/Lora/network_lokr.py
@@ -37,22 +37,22 @@ def __init__(self,  net: network.Network, weights: network.NetworkWeights):
 
     def calc_updown(self, orig_weight):
         if self.w1 is not None:
-            w1 = self.w1.to(orig_weight.device, dtype=orig_weight.dtype)
+            w1 = self.w1.to(orig_weight.device)
         else:
-            w1a = self.w1a.to(orig_weight.device, dtype=orig_weight.dtype)
-            w1b = self.w1b.to(orig_weight.device, dtype=orig_weight.dtype)
+            w1a = self.w1a.to(orig_weight.device)
+            w1b = self.w1b.to(orig_weight.device)
             w1 = w1a @ w1b
 
         if self.w2 is not None:
-            w2 = self.w2.to(orig_weight.device, dtype=orig_weight.dtype)
+            w2 = self.w2.to(orig_weight.device)
         elif self.t2 is None:
-            w2a = self.w2a.to(orig_weight.device, dtype=orig_weight.dtype)
-            w2b = self.w2b.to(orig_weight.device, dtype=orig_weight.dtype)
+            w2a = self.w2a.to(orig_weight.device)
+            w2b = self.w2b.to(orig_weight.device)
             w2 = w2a @ w2b
         else:
-            t2 = self.t2.to(orig_weight.device, dtype=orig_weight.dtype)
-            w2a = self.w2a.to(orig_weight.device, dtype=orig_weight.dtype)
-            w2b = self.w2b.to(orig_weight.device, dtype=orig_weight.dtype)
+            t2 = self.t2.to(orig_weight.device)
+            w2a = self.w2a.to(orig_weight.device)
+            w2b = self.w2b.to(orig_weight.device)
             w2 = lyco_helpers.make_weight_cp(t2, w2a, w2b)
 
         output_shape = [w1.size(0) * w2.size(0), w1.size(1) * w2.size(1)]
diff --git a/extensions-builtin/Lora/network_lora.py b/extensions-builtin/Lora/network_lora.py
index 26c0a72c237..a7a088949ea 100644
--- a/extensions-builtin/Lora/network_lora.py
+++ b/extensions-builtin/Lora/network_lora.py
@@ -1,6 +1,7 @@
 import torch
 
 import lyco_helpers
+import modules.models.sd3.mmdit
 import network
 from modules import devices
 
@@ -10,6 +11,13 @@ def create_module(self, net: network.Network, weights: network.NetworkWeights):
         if all(x in weights.w for x in ["lora_up.weight", "lora_down.weight"]):
             return NetworkModuleLora(net, weights)
 
+        if all(x in weights.w for x in ["lora_A.weight", "lora_B.weight"]):
+            w = weights.w.copy()
+            weights.w.clear()
+            weights.w.update({"lora_up.weight": w["lora_B.weight"], "lora_down.weight": w["lora_A.weight"]})
+
+            return NetworkModuleLora(net, weights)
+
         return None
 
 
@@ -29,7 +37,7 @@ def create_module(self, weights, key, none_ok=False):
         if weight is None and none_ok:
             return None
 
-        is_linear = type(self.sd_module) in [torch.nn.Linear, torch.nn.modules.linear.NonDynamicallyQuantizableLinear, torch.nn.MultiheadAttention]
+        is_linear = type(self.sd_module) in [torch.nn.Linear, torch.nn.modules.linear.NonDynamicallyQuantizableLinear, torch.nn.MultiheadAttention, modules.models.sd3.mmdit.QkvLinear]
         is_conv = type(self.sd_module) in [torch.nn.Conv2d]
 
         if is_linear:
@@ -61,13 +69,13 @@ def create_module(self, weights, key, none_ok=False):
         return module
 
     def calc_updown(self, orig_weight):
-        up = self.up_model.weight.to(orig_weight.device, dtype=orig_weight.dtype)
-        down = self.down_model.weight.to(orig_weight.device, dtype=orig_weight.dtype)
+        up = self.up_model.weight.to(orig_weight.device)
+        down = self.down_model.weight.to(orig_weight.device)
 
         output_shape = [up.size(0), down.size(1)]
         if self.mid_model is not None:
             # cp-decomposition
-            mid = self.mid_model.weight.to(orig_weight.device, dtype=orig_weight.dtype)
+            mid = self.mid_model.weight.to(orig_weight.device)
             updown = lyco_helpers.rebuild_cp_decomposition(up, down, mid)
             output_shape += mid.shape[2:]
         else:
diff --git a/extensions-builtin/Lora/network_norm.py b/extensions-builtin/Lora/network_norm.py
index ce450158068..d25afcbb928 100644
--- a/extensions-builtin/Lora/network_norm.py
+++ b/extensions-builtin/Lora/network_norm.py
@@ -18,10 +18,10 @@ def __init__(self,  net: network.Network, weights: network.NetworkWeights):
 
     def calc_updown(self, orig_weight):
         output_shape = self.w_norm.shape
-        updown = self.w_norm.to(orig_weight.device, dtype=orig_weight.dtype)
+        updown = self.w_norm.to(orig_weight.device)
 
         if self.b_norm is not None:
-            ex_bias = self.b_norm.to(orig_weight.device, dtype=orig_weight.dtype)
+            ex_bias = self.b_norm.to(orig_weight.device)
         else:
             ex_bias = None
 
diff --git a/extensions-builtin/Lora/network_oft.py b/extensions-builtin/Lora/network_oft.py
new file mode 100644
index 00000000000..1c515ebb739
--- /dev/null
+++ b/extensions-builtin/Lora/network_oft.py
@@ -0,0 +1,118 @@
+import torch
+import network
+from einops import rearrange
+
+
+class ModuleTypeOFT(network.ModuleType):
+    def create_module(self, net: network.Network, weights: network.NetworkWeights):
+        if all(x in weights.w for x in ["oft_blocks"]) or all(x in weights.w for x in ["oft_diag"]):
+            return NetworkModuleOFT(net, weights)
+
+        return None
+
+# Supports both kohya-ss' implementation of COFT  https://github.com/kohya-ss/sd-scripts/blob/main/networks/oft.py
+# and KohakuBlueleaf's implementation of OFT/COFT https://github.com/KohakuBlueleaf/LyCORIS/blob/dev/lycoris/modules/diag_oft.py
+class NetworkModuleOFT(network.NetworkModule):
+    def __init__(self,  net: network.Network, weights: network.NetworkWeights):
+
+        super().__init__(net, weights)
+
+        self.lin_module = None
+        self.org_module: list[torch.Module] = [self.sd_module]
+
+        self.scale = 1.0
+        self.is_R = False
+        self.is_boft = False
+
+        # kohya-ss/New LyCORIS OFT/BOFT
+        if "oft_blocks" in weights.w.keys():
+            self.oft_blocks = weights.w["oft_blocks"] # (num_blocks, block_size, block_size)
+            self.alpha = weights.w.get("alpha", None) # alpha is constraint
+            self.dim = self.oft_blocks.shape[0] # lora dim
+        # Old LyCORIS OFT
+        elif "oft_diag" in weights.w.keys():
+            self.is_R = True
+            self.oft_blocks = weights.w["oft_diag"]
+            # self.alpha is unused
+            self.dim = self.oft_blocks.shape[1] # (num_blocks, block_size, block_size)
+
+        is_linear = type(self.sd_module) in [torch.nn.Linear, torch.nn.modules.linear.NonDynamicallyQuantizableLinear]
+        is_conv = type(self.sd_module) in [torch.nn.Conv2d]
+        is_other_linear = type(self.sd_module) in [torch.nn.MultiheadAttention] # unsupported
+
+        if is_linear:
+            self.out_dim = self.sd_module.out_features
+        elif is_conv:
+            self.out_dim = self.sd_module.out_channels
+        elif is_other_linear:
+            self.out_dim = self.sd_module.embed_dim
+
+        # LyCORIS BOFT
+        if self.oft_blocks.dim() == 4:
+            self.is_boft = True
+        self.rescale = weights.w.get('rescale', None)
+        if self.rescale is not None and not is_other_linear:
+            self.rescale = self.rescale.reshape(-1, *[1]*(self.org_module[0].weight.dim() - 1))
+
+        self.num_blocks = self.dim
+        self.block_size = self.out_dim // self.dim
+        self.constraint = (0 if self.alpha is None else self.alpha) * self.out_dim
+        if self.is_R:
+            self.constraint = None
+            self.block_size = self.dim
+            self.num_blocks = self.out_dim // self.dim
+        elif self.is_boft:
+            self.boft_m = self.oft_blocks.shape[0]
+            self.num_blocks = self.oft_blocks.shape[1]
+            self.block_size = self.oft_blocks.shape[2]
+            self.boft_b = self.block_size
+
+    def calc_updown(self, orig_weight):
+        oft_blocks = self.oft_blocks.to(orig_weight.device)
+        eye = torch.eye(self.block_size, device=oft_blocks.device)
+
+        if not self.is_R:
+            block_Q = oft_blocks - oft_blocks.transpose(-1, -2) # ensure skew-symmetric orthogonal matrix
+            if self.constraint != 0:
+                norm_Q = torch.norm(block_Q.flatten())
+                new_norm_Q = torch.clamp(norm_Q, max=self.constraint.to(oft_blocks.device))
+                block_Q = block_Q * ((new_norm_Q + 1e-8) / (norm_Q + 1e-8))
+            oft_blocks = torch.matmul(eye + block_Q, (eye - block_Q).float().inverse())
+
+        R = oft_blocks.to(orig_weight.device)
+
+        if not self.is_boft:
+            # This errors out for MultiheadAttention, might need to be handled up-stream
+            merged_weight = rearrange(orig_weight, '(k n) ... -> k n ...', k=self.num_blocks, n=self.block_size)
+            merged_weight = torch.einsum(
+                'k n m, k n ... -> k m ...',
+                R,
+                merged_weight
+            )
+            merged_weight = rearrange(merged_weight, 'k m ... -> (k m) ...')
+        else:
+            # TODO: determine correct value for scale
+            scale = 1.0
+            m = self.boft_m
+            b = self.boft_b
+            r_b = b // 2
+            inp = orig_weight
+            for i in range(m):
+                bi = R[i] # b_num, b_size, b_size
+                if i == 0:
+                    # Apply multiplier/scale and rescale into first weight
+                    bi = bi * scale + (1 - scale) * eye
+                inp = rearrange(inp, "(c g k) ... -> (c k g) ...", g=2, k=2**i * r_b)
+                inp = rearrange(inp, "(d b) ... -> d b ...", b=b)
+                inp = torch.einsum("b i j, b j ... -> b i ...", bi, inp)
+                inp = rearrange(inp, "d b ... -> (d b) ...")
+                inp = rearrange(inp, "(c k g) ... -> (c g k) ...", g=2, k=2**i * r_b)
+            merged_weight = inp
+
+        # Rescale mechanism
+        if self.rescale is not None:
+            merged_weight = self.rescale.to(merged_weight) * merged_weight
+
+        updown = merged_weight.to(orig_weight.device) - orig_weight.to(merged_weight.dtype)
+        output_shape = orig_weight.shape
+        return self.finalize_updown(updown, orig_weight, output_shape)
diff --git a/extensions-builtin/Lora/networks.py b/extensions-builtin/Lora/networks.py
index 96f935b236f..67f9abe2a37 100644
--- a/extensions-builtin/Lora/networks.py
+++ b/extensions-builtin/Lora/networks.py
@@ -1,3 +1,5 @@
+from __future__ import annotations
+import gradio as gr
 import logging
 import os
 import re
@@ -5,16 +7,22 @@
 import lora_patches
 import network
 import network_lora
+import network_glora
 import network_hada
 import network_ia3
 import network_lokr
 import network_full
 import network_norm
+import network_oft
 
 import torch
 from typing import Union
 
 from modules import shared, devices, sd_models, errors, scripts, sd_hijack
+import modules.textual_inversion.textual_inversion as textual_inversion
+import modules.models.sd3.mmdit
+
+from lora_logger import logger
 
 module_types = [
     network_lora.ModuleTypeLora(),
@@ -23,6 +31,8 @@
     network_lokr.ModuleTypeLokr(),
     network_full.ModuleTypeFull(),
     network_norm.ModuleTypeNorm(),
+    network_glora.ModuleTypeGLora(),
+    network_oft.ModuleTypeOFT(),
 ]
 
 
@@ -122,7 +132,9 @@ def assign_network_names_to_compvis_modules(sd_model):
                 network_layer_mapping[network_name] = module
                 module.network_layer_name = network_name
     else:
-        for name, module in shared.sd_model.cond_stage_model.wrapped.named_modules():
+        cond_stage_model = getattr(shared.sd_model.cond_stage_model, 'wrapped', shared.sd_model.cond_stage_model)
+
+        for name, module in cond_stage_model.named_modules():
             network_name = name.replace(".", "_")
             network_layer_mapping[network_name] = module
             module.network_layer_name = network_name
@@ -135,6 +147,14 @@ def assign_network_names_to_compvis_modules(sd_model):
     sd_model.network_layer_mapping = network_layer_mapping
 
 
+class BundledTIHash(str):
+    def __init__(self, hash_str):
+        self.hash = hash_str
+
+    def __str__(self):
+        return self.hash if shared.opts.lora_bundled_ti_to_infotext else ''
+
+
 def load_network(name, network_on_disk):
     net = network.Network(name, network_on_disk)
     net.mtime = os.path.getmtime(network_on_disk.filename)
@@ -147,13 +167,42 @@ def load_network(name, network_on_disk):
 
     keys_failed_to_match = {}
     is_sd2 = 'model_transformer_resblocks' in shared.sd_model.network_layer_mapping
+    if hasattr(shared.sd_model, 'diffusers_weight_map'):
+        diffusers_weight_map = shared.sd_model.diffusers_weight_map
+    elif hasattr(shared.sd_model, 'diffusers_weight_mapping'):
+        diffusers_weight_map = {}
+        for k, v in shared.sd_model.diffusers_weight_mapping():
+            diffusers_weight_map[k] = v
+        shared.sd_model.diffusers_weight_map = diffusers_weight_map
+    else:
+        diffusers_weight_map = None
 
     matched_networks = {}
+    bundle_embeddings = {}
 
     for key_network, weight in sd.items():
-        key_network_without_network_parts, network_part = key_network.split(".", 1)
 
-        key = convert_diffusers_name_to_compvis(key_network_without_network_parts, is_sd2)
+        if diffusers_weight_map:
+            key_network_without_network_parts, network_name, network_weight = key_network.rsplit(".", 2)
+            network_part = network_name + '.' + network_weight
+        else:
+            key_network_without_network_parts, _, network_part = key_network.partition(".")
+
+        if key_network_without_network_parts == "bundle_emb":
+            emb_name, vec_name = network_part.split(".", 1)
+            emb_dict = bundle_embeddings.get(emb_name, {})
+            if vec_name.split('.')[0] == 'string_to_param':
+                _, k2 = vec_name.split('.', 1)
+                emb_dict['string_to_param'] = {k2: weight}
+            else:
+                emb_dict[vec_name] = weight
+            bundle_embeddings[emb_name] = emb_dict
+
+        if diffusers_weight_map:
+            key = diffusers_weight_map.get(key_network_without_network_parts, key_network_without_network_parts)
+        else:
+            key = convert_diffusers_name_to_compvis(key_network_without_network_parts, is_sd2)
+
         sd_module = shared.sd_model.network_layer_mapping.get(key, None)
 
         if sd_module is None:
@@ -174,6 +223,17 @@ def load_network(name, network_on_disk):
                 key = key_network_without_network_parts.replace("lora_te1_text_model", "transformer_text_model")
                 sd_module = shared.sd_model.network_layer_mapping.get(key, None)
 
+        # kohya_ss OFT module
+        elif sd_module is None and "oft_unet" in key_network_without_network_parts:
+            key = key_network_without_network_parts.replace("oft_unet", "diffusion_model")
+            sd_module = shared.sd_model.network_layer_mapping.get(key, None)
+
+        # KohakuBlueLeaf OFT module
+        if sd_module is None and "oft_diag" in key:
+            key = key_network_without_network_parts.replace("lora_unet", "diffusion_model")
+            key = key_network_without_network_parts.replace("lora_te1_text_model", "0_transformer_text_model")
+            sd_module = shared.sd_model.network_layer_mapping.get(key, None)
+
         if sd_module is None:
             keys_failed_to_match[key_network] = key
             continue
@@ -195,6 +255,15 @@ def load_network(name, network_on_disk):
 
         net.modules[key] = net_module
 
+    embeddings = {}
+    for emb_name, data in bundle_embeddings.items():
+        embedding = textual_inversion.create_embedding_from_data(data, emb_name, filename=network_on_disk.filename + "/" + emb_name)
+        embedding.loaded = None
+        embedding.shorthash = BundledTIHash(name)
+        embeddings[emb_name] = embedding
+
+    net.bundle_embeddings = embeddings
+
     if keys_failed_to_match:
         logging.debug(f"Network {network_on_disk.filename} didn't match keys: {keys_failed_to_match}")
 
@@ -210,19 +279,33 @@ def purge_networks_from_memory():
 
 
 def load_networks(names, te_multipliers=None, unet_multipliers=None, dyn_dims=None):
+    emb_db = sd_hijack.model_hijack.embedding_db
     already_loaded = {}
 
     for net in loaded_networks:
         if net.name in names:
             already_loaded[net.name] = net
+        for emb_name, embedding in net.bundle_embeddings.items():
+            if embedding.loaded:
+                emb_db.register_embedding_by_name(None, shared.sd_model, emb_name)
 
     loaded_networks.clear()
 
-    networks_on_disk = [available_network_aliases.get(name, None) for name in names]
+    unavailable_networks = []
+    for name in names:
+        if name.lower() in forbidden_network_aliases and available_networks.get(name) is None:
+            unavailable_networks.append(name)
+        elif available_network_aliases.get(name) is None:
+            unavailable_networks.append(name)
+
+    if unavailable_networks:
+        update_available_networks_by_names(unavailable_networks)
+
+    networks_on_disk = [available_networks.get(name, None) if name.lower() in forbidden_network_aliases else available_network_aliases.get(name, None) for name in names]
     if any(x is None for x in networks_on_disk):
         list_available_networks()
 
-        networks_on_disk = [available_network_aliases.get(name, None) for name in names]
+        networks_on_disk = [available_networks.get(name, None) if name.lower() in forbidden_network_aliases else available_network_aliases.get(name, None) for name in names]
 
     failed_to_load_networks = []
 
@@ -257,12 +340,54 @@ def load_networks(names, te_multipliers=None, unet_multipliers=None, dyn_dims=No
         net.dyn_dim = dyn_dims[i] if dyn_dims else 1.0
         loaded_networks.append(net)
 
+        for emb_name, embedding in net.bundle_embeddings.items():
+            if embedding.loaded is None and emb_name in emb_db.word_embeddings:
+                logger.warning(
+                    f'Skip bundle embedding: "{emb_name}"'
+                    ' as it was already loaded from embeddings folder'
+                )
+                continue
+
+            embedding.loaded = False
+            if emb_db.expected_shape == -1 or emb_db.expected_shape == embedding.shape:
+                embedding.loaded = True
+                emb_db.register_embedding(embedding, shared.sd_model)
+            else:
+                emb_db.skipped_embeddings[name] = embedding
+
     if failed_to_load_networks:
-        sd_hijack.model_hijack.comments.append("Networks not found: " + ", ".join(failed_to_load_networks))
+        lora_not_found_message = f'Lora not found: {", ".join(failed_to_load_networks)}'
+        sd_hijack.model_hijack.comments.append(lora_not_found_message)
+        if shared.opts.lora_not_found_warning_console:
+            print(f'\n{lora_not_found_message}\n')
+        if shared.opts.lora_not_found_gradio_warning:
+            gr.Warning(lora_not_found_message)
 
     purge_networks_from_memory()
 
 
+def allowed_layer_without_weight(layer):
+    if isinstance(layer, torch.nn.LayerNorm) and not layer.elementwise_affine:
+        return True
+
+    return False
+
+
+def store_weights_backup(weight):
+    if weight is None:
+        return None
+
+    return weight.to(devices.cpu, copy=True)
+
+
+def restore_weights_backup(obj, field, weight):
+    if weight is None:
+        setattr(obj, field, None)
+        return
+
+    getattr(obj, field).copy_(weight)
+
+
 def network_restore_weights_from_backup(self: Union[torch.nn.Conv2d, torch.nn.Linear, torch.nn.GroupNorm, torch.nn.LayerNorm, torch.nn.MultiheadAttention]):
     weights_backup = getattr(self, "network_weights_backup", None)
     bias_backup = getattr(self, "network_bias_backup", None)
@@ -272,28 +397,22 @@ def network_restore_weights_from_backup(self: Union[torch.nn.Conv2d, torch.nn.Li
 
     if weights_backup is not None:
         if isinstance(self, torch.nn.MultiheadAttention):
-            self.in_proj_weight.copy_(weights_backup[0])
-            self.out_proj.weight.copy_(weights_backup[1])
+            restore_weights_backup(self, 'in_proj_weight', weights_backup[0])
+            restore_weights_backup(self.out_proj, 'weight', weights_backup[1])
         else:
-            self.weight.copy_(weights_backup)
+            restore_weights_backup(self, 'weight', weights_backup)
 
-    if bias_backup is not None:
-        if isinstance(self, torch.nn.MultiheadAttention):
-            self.out_proj.bias.copy_(bias_backup)
-        else:
-            self.bias.copy_(bias_backup)
+    if isinstance(self, torch.nn.MultiheadAttention):
+        restore_weights_backup(self.out_proj, 'bias', bias_backup)
     else:
-        if isinstance(self, torch.nn.MultiheadAttention):
-            self.out_proj.bias = None
-        else:
-            self.bias = None
+        restore_weights_backup(self, 'bias', bias_backup)
 
 
 def network_apply_weights(self: Union[torch.nn.Conv2d, torch.nn.Linear, torch.nn.GroupNorm, torch.nn.LayerNorm, torch.nn.MultiheadAttention]):
     """
     Applies the currently selected set of networks to the weights of torch layer self.
     If weights already have this particular set of networks applied, does nothing.
-    If not, restores orginal weights from backup and alters weights according to networks.
+    If not, restores original weights from backup and alters weights according to networks.
     """
 
     network_layer_name = getattr(self, 'network_layer_name', None)
@@ -305,24 +424,30 @@ def network_apply_weights(self: Union[torch.nn.Conv2d, torch.nn.Linear, torch.nn
 
     weights_backup = getattr(self, "network_weights_backup", None)
     if weights_backup is None and wanted_names != ():
-        if current_names != ():
-            raise RuntimeError("no backup weights found and current weights are not unchanged")
+        if current_names != () and not allowed_layer_without_weight(self):
+            raise RuntimeError(f"{network_layer_name} - no backup weights found and current weights are not unchanged")
 
         if isinstance(self, torch.nn.MultiheadAttention):
-            weights_backup = (self.in_proj_weight.to(devices.cpu, copy=True), self.out_proj.weight.to(devices.cpu, copy=True))
+            weights_backup = (store_weights_backup(self.in_proj_weight), store_weights_backup(self.out_proj.weight))
         else:
-            weights_backup = self.weight.to(devices.cpu, copy=True)
+            weights_backup = store_weights_backup(self.weight)
 
         self.network_weights_backup = weights_backup
 
     bias_backup = getattr(self, "network_bias_backup", None)
-    if bias_backup is None:
+    if bias_backup is None and wanted_names != ():
         if isinstance(self, torch.nn.MultiheadAttention) and self.out_proj.bias is not None:
-            bias_backup = self.out_proj.bias.to(devices.cpu, copy=True)
+            bias_backup = store_weights_backup(self.out_proj.bias)
         elif getattr(self, 'bias', None) is not None:
-            bias_backup = self.bias.to(devices.cpu, copy=True)
+            bias_backup = store_weights_backup(self.bias)
         else:
             bias_backup = None
+
+        # Unlike weight which always has value, some modules don't have bias.
+        # Only report if bias is not None and current bias are not unchanged.
+        if bias_backup is not None and current_names != ():
+            raise RuntimeError("no backup bias found and current bias are not unchanged")
+
         self.network_bias_backup = bias_backup
 
     if current_names != wanted_names:
@@ -330,21 +455,29 @@ def network_apply_weights(self: Union[torch.nn.Conv2d, torch.nn.Linear, torch.nn
 
         for net in loaded_networks:
             module = net.modules.get(network_layer_name, None)
-            if module is not None and hasattr(self, 'weight'):
+            if module is not None and hasattr(self, 'weight') and not isinstance(module, modules.models.sd3.mmdit.QkvLinear):
                 try:
                     with torch.no_grad():
-                        updown, ex_bias = module.calc_updown(self.weight)
+                        if getattr(self, 'fp16_weight', None) is None:
+                            weight = self.weight
+                            bias = self.bias
+                        else:
+                            weight = self.fp16_weight.clone().to(self.weight.device)
+                            bias = getattr(self, 'fp16_bias', None)
+                            if bias is not None:
+                                bias = bias.clone().to(self.bias.device)
+                        updown, ex_bias = module.calc_updown(weight)
 
-                        if len(self.weight.shape) == 4 and self.weight.shape[1] == 9:
+                        if len(weight.shape) == 4 and weight.shape[1] == 9:
                             # inpainting model. zero pad updown to make channel[1]  4 to 9
                             updown = torch.nn.functional.pad(updown, (0, 0, 0, 0, 0, 5))
 
-                        self.weight += updown
+                        self.weight.copy_((weight.to(dtype=updown.dtype) + updown).to(dtype=self.weight.dtype))
                         if ex_bias is not None and hasattr(self, 'bias'):
                             if self.bias is None:
-                                self.bias = torch.nn.Parameter(ex_bias)
+                                self.bias = torch.nn.Parameter(ex_bias).to(self.weight.dtype)
                             else:
-                                self.bias += ex_bias
+                                self.bias.copy_((bias + ex_bias).to(dtype=self.bias.dtype))
                 except RuntimeError as e:
                     logging.debug(f"Network {net.name} layer {network_layer_name}: {e}")
                     extra_network_lora.errors[net.name] = extra_network_lora.errors.get(net.name, 0) + 1
@@ -359,9 +492,12 @@ def network_apply_weights(self: Union[torch.nn.Conv2d, torch.nn.Linear, torch.nn
             if isinstance(self, torch.nn.MultiheadAttention) and module_q and module_k and module_v and module_out:
                 try:
                     with torch.no_grad():
-                        updown_q, _ = module_q.calc_updown(self.in_proj_weight)
-                        updown_k, _ = module_k.calc_updown(self.in_proj_weight)
-                        updown_v, _ = module_v.calc_updown(self.in_proj_weight)
+                        # Send "real" orig_weight into MHA's lora module
+                        qw, kw, vw = self.in_proj_weight.chunk(3, 0)
+                        updown_q, _ = module_q.calc_updown(qw)
+                        updown_k, _ = module_k.calc_updown(kw)
+                        updown_v, _ = module_v.calc_updown(vw)
+                        del qw, kw, vw
                         updown_qkv = torch.vstack([updown_q, updown_k, updown_v])
                         updown_out, ex_bias = module_out.calc_updown(self.out_proj.weight)
 
@@ -379,6 +515,24 @@ def network_apply_weights(self: Union[torch.nn.Conv2d, torch.nn.Linear, torch.nn
 
                 continue
 
+            if isinstance(self, modules.models.sd3.mmdit.QkvLinear) and module_q and module_k and module_v:
+                try:
+                    with torch.no_grad():
+                        # Send "real" orig_weight into MHA's lora module
+                        qw, kw, vw = self.weight.chunk(3, 0)
+                        updown_q, _ = module_q.calc_updown(qw)
+                        updown_k, _ = module_k.calc_updown(kw)
+                        updown_v, _ = module_v.calc_updown(vw)
+                        del qw, kw, vw
+                        updown_qkv = torch.vstack([updown_q, updown_k, updown_v])
+                        self.weight += updown_qkv
+
+                except RuntimeError as e:
+                    logging.debug(f"Network {net.name} layer {network_layer_name}: {e}")
+                    extra_network_lora.errors[net.name] = extra_network_lora.errors.get(net.name, 0) + 1
+
+                continue
+
             if module is None:
                 continue
 
@@ -388,23 +542,23 @@ def network_apply_weights(self: Union[torch.nn.Conv2d, torch.nn.Linear, torch.nn
         self.network_current_names = wanted_names
 
 
-def network_forward(module, input, original_forward):
+def network_forward(org_module, input, original_forward):
     """
     Old way of applying Lora by executing operations during layer's forward.
     Stacking many loras this way results in big performance degradation.
     """
 
     if len(loaded_networks) == 0:
-        return original_forward(module, input)
+        return original_forward(org_module, input)
 
     input = devices.cond_cast_unet(input)
 
-    network_restore_weights_from_backup(module)
-    network_reset_cached_weight(module)
+    network_restore_weights_from_backup(org_module)
+    network_reset_cached_weight(org_module)
 
-    y = original_forward(module, input)
+    y = original_forward(org_module, input)
 
-    network_layer_name = getattr(module, 'network_layer_name', None)
+    network_layer_name = getattr(org_module, 'network_layer_name', None)
     for lora in loaded_networks:
         module = lora.modules.get(network_layer_name, None)
         if module is None:
@@ -418,6 +572,7 @@ def network_forward(module, input, original_forward):
 def network_reset_cached_weight(self: Union[torch.nn.Conv2d, torch.nn.Linear]):
     self.network_current_names = ()
     self.network_weights_backup = None
+    self.network_bias_backup = None
 
 
 def network_Linear_forward(self, input):
@@ -492,22 +647,16 @@ def network_MultiheadAttention_load_state_dict(self, *args, **kwargs):
     return originals.MultiheadAttention_load_state_dict(self, *args, **kwargs)
 
 
-def list_available_networks():
-    available_networks.clear()
-    available_network_aliases.clear()
-    forbidden_network_aliases.clear()
-    available_network_hash_lookup.clear()
-    forbidden_network_aliases.update({"none": 1, "Addams": 1})
-
-    os.makedirs(shared.cmd_opts.lora_dir, exist_ok=True)
-
+def process_network_files(names: list[str] | None = None):
     candidates = list(shared.walk_files(shared.cmd_opts.lora_dir, allowed_extensions=[".pt", ".ckpt", ".safetensors"]))
     candidates += list(shared.walk_files(shared.cmd_opts.lyco_dir_backcompat, allowed_extensions=[".pt", ".ckpt", ".safetensors"]))
     for filename in candidates:
         if os.path.isdir(filename):
             continue
-
         name = os.path.splitext(os.path.basename(filename))[0]
+        # if names is provided, only load networks with names in the list
+        if names and name not in names:
+            continue
         try:
             entry = network.NetworkOnDisk(name, filename)
         except OSError:  # should catch FileNotFoundError and PermissionError etc.
@@ -523,6 +672,22 @@ def list_available_networks():
         available_network_aliases[entry.alias] = entry
 
 
+def update_available_networks_by_names(names: list[str]):
+    process_network_files(names)
+
+
+def list_available_networks():
+    available_networks.clear()
+    available_network_aliases.clear()
+    forbidden_network_aliases.clear()
+    available_network_hash_lookup.clear()
+    forbidden_network_aliases.update({"none": 1, "Addams": 1})
+
+    os.makedirs(shared.cmd_opts.lora_dir, exist_ok=True)
+
+    process_network_files()
+
+
 re_network_name = re.compile(r"(.*)\s*\([0-9a-fA-F]+\)")
 
 
@@ -564,6 +729,7 @@ def infotext_pasted(infotext, params):
 available_networks = {}
 available_network_aliases = {}
 loaded_networks = []
+loaded_bundle_embeddings = {}
 networks_in_memory = {}
 available_network_hash_lookup = {}
 forbidden_network_aliases = {}
diff --git a/extensions-builtin/Lora/preload.py b/extensions-builtin/Lora/preload.py
index 50961be33d7..52fab29b08b 100644
--- a/extensions-builtin/Lora/preload.py
+++ b/extensions-builtin/Lora/preload.py
@@ -1,7 +1,8 @@
 import os
 from modules import paths
+from modules.paths_internal import normalized_filepath
 
 
 def preload(parser):
-    parser.add_argument("--lora-dir", type=str, help="Path to directory with Lora networks.", default=os.path.join(paths.models_path, 'Lora'))
-    parser.add_argument("--lyco-dir-backcompat", type=str, help="Path to directory with LyCORIS networks (for backawards compatibility; can also use --lyco-dir).", default=os.path.join(paths.models_path, 'LyCORIS'))
+    parser.add_argument("--lora-dir", type=normalized_filepath, help="Path to directory with Lora networks.", default=os.path.join(paths.models_path, 'Lora'))
+    parser.add_argument("--lyco-dir-backcompat", type=normalized_filepath, help="Path to directory with LyCORIS networks (for backawards compatibility; can also use --lyco-dir).", default=os.path.join(paths.models_path, 'LyCORIS'))
diff --git a/extensions-builtin/Lora/scripts/lora_script.py b/extensions-builtin/Lora/scripts/lora_script.py
index ef23968c563..d3ea369ae26 100644
--- a/extensions-builtin/Lora/scripts/lora_script.py
+++ b/extensions-builtin/Lora/scripts/lora_script.py
@@ -36,9 +36,12 @@ def before_ui():
     "sd_lora": shared.OptionInfo("None", "Add network to prompt", gr.Dropdown, lambda: {"choices": ["None", *networks.available_networks]}, refresh=networks.list_available_networks),
     "lora_preferred_name": shared.OptionInfo("Alias from file", "When adding to prompt, refer to Lora by", gr.Radio, {"choices": ["Alias from file", "Filename"]}),
     "lora_add_hashes_to_infotext": shared.OptionInfo(True, "Add Lora hashes to infotext"),
+    "lora_bundled_ti_to_infotext": shared.OptionInfo(True, "Add Lora name as TI hashes for bundled Textual Inversion").info('"Add Textual Inversion hashes to infotext" needs to be enabled'),
     "lora_show_all": shared.OptionInfo(False, "Always show all networks on the Lora page").info("otherwise, those detected as for incompatible version of Stable Diffusion will be hidden"),
     "lora_hide_unknown_for_versions": shared.OptionInfo([], "Hide networks of unknown versions for model versions", gr.CheckboxGroup, {"choices": ["SD1", "SD2", "SDXL"]}),
     "lora_in_memory_limit": shared.OptionInfo(0, "Number of Lora networks to keep cached in memory", gr.Number, {"precision": 0}),
+    "lora_not_found_warning_console": shared.OptionInfo(False, "Lora not found warning in console"),
+    "lora_not_found_gradio_warning": shared.OptionInfo(False, "Lora not found warning popup in webui"),
 }))
 
 
diff --git a/extensions-builtin/Lora/ui_edit_user_metadata.py b/extensions-builtin/Lora/ui_edit_user_metadata.py
index c7011909055..b6c4d1c6acb 100644
--- a/extensions-builtin/Lora/ui_edit_user_metadata.py
+++ b/extensions-builtin/Lora/ui_edit_user_metadata.py
@@ -21,10 +21,12 @@ def is_non_comma_tagset(tags):
 def build_tags(metadata):
     tags = {}
 
-    for _, tags_dict in metadata.get("ss_tag_frequency", {}).items():
-        for tag, tag_count in tags_dict.items():
-            tag = tag.strip()
-            tags[tag] = tags.get(tag, 0) + int(tag_count)
+    ss_tag_frequency = metadata.get("ss_tag_frequency", {})
+    if ss_tag_frequency is not None and hasattr(ss_tag_frequency, 'items'):
+        for _, tags_dict in ss_tag_frequency.items():
+            for tag, tag_count in tags_dict.items():
+                tag = tag.strip()
+                tags[tag] = tags.get(tag, 0) + int(tag_count)
 
     if tags and is_non_comma_tagset(tags):
         new_tags = {}
@@ -54,12 +56,13 @@ def __init__(self, ui, tabname, page):
         self.slider_preferred_weight = None
         self.edit_notes = None
 
-    def save_lora_user_metadata(self, name, desc, sd_version, activation_text, preferred_weight, notes):
+    def save_lora_user_metadata(self, name, desc, sd_version, activation_text, preferred_weight, negative_text, notes):
         user_metadata = self.get_user_metadata(name)
         user_metadata["description"] = desc
         user_metadata["sd version"] = sd_version
         user_metadata["activation text"] = activation_text
         user_metadata["preferred weight"] = preferred_weight
+        user_metadata["negative text"] = negative_text
         user_metadata["notes"] = notes
 
         self.write_user_metadata(name, user_metadata)
@@ -127,6 +130,7 @@ def put_values_into_components(self, name):
             gr.HighlightedText.update(value=gradio_tags, visible=True if tags else False),
             user_metadata.get('activation text', ''),
             float(user_metadata.get('preferred weight', 0.0)),
+            user_metadata.get('negative text', ''),
             gr.update(visible=True if tags else False),
             gr.update(value=self.generate_random_prompt_from_tags(tags), visible=True if tags else False),
         ]
@@ -147,6 +151,8 @@ def generate_random_prompt_from_tags(self, tags):
 
             v = random.random() * max_count
             if count > v:
+                for x in "({[]})":
+                    tag = tag.replace(x, '\\' + x)
                 res.append(tag)
 
         return ", ".join(sorted(res))
@@ -162,7 +168,7 @@ def create_editor(self):
         self.taginfo = gr.HighlightedText(label="Training dataset tags")
         self.edit_activation_text = gr.Text(label='Activation text', info="Will be added to prompt along with Lora")
         self.slider_preferred_weight = gr.Slider(label='Preferred weight', info="Set to 0 to disable", minimum=0.0, maximum=2.0, step=0.01)
-
+        self.edit_negative_text = gr.Text(label='Negative prompt', info="Will be added to negative prompts")
         with gr.Row() as row_random_prompt:
             with gr.Column(scale=8):
                 random_prompt = gr.Textbox(label='Random prompt', lines=4, max_lines=4, interactive=False)
@@ -198,6 +204,7 @@ def select_tag(activation_text, evt: gr.SelectData):
             self.taginfo,
             self.edit_activation_text,
             self.slider_preferred_weight,
+            self.edit_negative_text,
             row_random_prompt,
             random_prompt,
         ]
@@ -211,7 +218,9 @@ def select_tag(activation_text, evt: gr.SelectData):
             self.select_sd_version,
             self.edit_activation_text,
             self.slider_preferred_weight,
+            self.edit_negative_text,
             self.edit_notes,
         ]
 
+
         self.setup_save_handler(self.button_save, self.save_lora_user_metadata, edited_components)
diff --git a/extensions-builtin/Lora/ui_extra_networks_lora.py b/extensions-builtin/Lora/ui_extra_networks_lora.py
index 55409a7829d..3e34d69dca4 100644
--- a/extensions-builtin/Lora/ui_extra_networks_lora.py
+++ b/extensions-builtin/Lora/ui_extra_networks_lora.py
@@ -17,18 +17,23 @@ def refresh(self):
 
     def create_item(self, name, index=None, enable_filter=True):
         lora_on_disk = networks.available_networks.get(name)
+        if lora_on_disk is None:
+            return
 
         path, ext = os.path.splitext(lora_on_disk.filename)
 
         alias = lora_on_disk.get_alias()
 
+        search_terms = [self.search_terms_from_path(lora_on_disk.filename)]
+        if lora_on_disk.hash:
+            search_terms.append(lora_on_disk.hash)
         item = {
             "name": name,
             "filename": lora_on_disk.filename,
             "shorthash": lora_on_disk.shorthash,
-            "preview": self.find_preview(path),
+            "preview": self.find_preview(path) or self.find_embedded_preview(path, name, lora_on_disk.metadata),
             "description": self.find_description(path),
-            "search_term": self.search_terms_from_path(lora_on_disk.filename) + " " + (lora_on_disk.hash or ""),
+            "search_terms": search_terms,
             "local_preview": f"{path}.{shared.opts.samples_format}",
             "metadata": lora_on_disk.metadata,
             "sort_keys": {'default': index, **self.get_sort_keys(lora_on_disk.filename)},
@@ -43,6 +48,11 @@ def create_item(self, name, index=None, enable_filter=True):
         if activation_text:
             item["prompt"] += " + " + quote_js(" " + activation_text)
 
+        negative_prompt = item["user_metadata"].get("negative text")
+        item["negative_prompt"] = quote_js("")
+        if negative_prompt:
+            item["negative_prompt"] = quote_js('(' + negative_prompt + ':1)')
+
         sd_version = item["user_metadata"].get("sd version")
         if sd_version in network.SdVersion.__members__:
             item["sd_version"] = sd_version
@@ -50,7 +60,7 @@ def create_item(self, name, index=None, enable_filter=True):
         else:
             sd_version = lora_on_disk.sd_version
 
-        if shared.opts.lora_show_all or not enable_filter:
+        if shared.opts.lora_show_all or not enable_filter or not shared.sd_model:
             pass
         elif sd_version == network.SdVersion.Unknown:
             model_version = network.SdVersion.SDXL if shared.sd_model.is_sdxl else network.SdVersion.SD2 if shared.sd_model.is_sd2 else network.SdVersion.SD1
@@ -66,9 +76,10 @@ def create_item(self, name, index=None, enable_filter=True):
         return item
 
     def list_items(self):
-        for index, name in enumerate(networks.available_networks):
+        # instantiate a list to protect against concurrent modification
+        names = list(networks.available_networks)
+        for index, name in enumerate(names):
             item = self.create_item(name, index)
-
             if item is not None:
                 yield item
 
diff --git a/extensions-builtin/ScuNET/scripts/scunet_model.py b/extensions-builtin/ScuNET/scripts/scunet_model.py
index 167d2f64b8e..fe5e5a19265 100644
--- a/extensions-builtin/ScuNET/scripts/scunet_model.py
+++ b/extensions-builtin/ScuNET/scripts/scunet_model.py
@@ -1,16 +1,9 @@
 import sys
 
 import PIL.Image
-import numpy as np
-import torch
-from tqdm import tqdm
 
 import modules.upscaler
-from modules import devices, modelloader, script_callbacks, errors
-from scunet_model_arch import SCUNet
-
-from modules.modelloader import load_file_from_url
-from modules.shared import opts
+from modules import devices, errors, modelloader, script_callbacks, shared, upscaler_utils
 
 
 class UpscalerScuNET(modules.upscaler.Upscaler):
@@ -42,100 +35,37 @@ def __init__(self, dirname):
             scalers.append(scaler_data2)
         self.scalers = scalers
 
-    @staticmethod
-    @torch.no_grad()
-    def tiled_inference(img, model):
-        # test the image tile by tile
-        h, w = img.shape[2:]
-        tile = opts.SCUNET_tile
-        tile_overlap = opts.SCUNET_tile_overlap
-        if tile == 0:
-            return model(img)
-
-        device = devices.get_device_for('scunet')
-        assert tile % 8 == 0, "tile size should be a multiple of window_size"
-        sf = 1
-
-        stride = tile - tile_overlap
-        h_idx_list = list(range(0, h - tile, stride)) + [h - tile]
-        w_idx_list = list(range(0, w - tile, stride)) + [w - tile]
-        E = torch.zeros(1, 3, h * sf, w * sf, dtype=img.dtype, device=device)
-        W = torch.zeros_like(E, dtype=devices.dtype, device=device)
-
-        with tqdm(total=len(h_idx_list) * len(w_idx_list), desc="ScuNET tiles") as pbar:
-            for h_idx in h_idx_list:
-
-                for w_idx in w_idx_list:
-
-                    in_patch = img[..., h_idx: h_idx + tile, w_idx: w_idx + tile]
-
-                    out_patch = model(in_patch)
-                    out_patch_mask = torch.ones_like(out_patch)
-
-                    E[
-                        ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf
-                    ].add_(out_patch)
-                    W[
-                        ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf
-                    ].add_(out_patch_mask)
-                    pbar.update(1)
-        output = E.div_(W)
-
-        return output
-
     def do_upscale(self, img: PIL.Image.Image, selected_file):
-
         devices.torch_gc()
-
         try:
             model = self.load_model(selected_file)
         except Exception as e:
             print(f"ScuNET: Unable to load model from {selected_file}: {e}", file=sys.stderr)
             return img
 
-        device = devices.get_device_for('scunet')
-        tile = opts.SCUNET_tile
-        h, w = img.height, img.width
-        np_img = np.array(img)
-        np_img = np_img[:, :, ::-1]  # RGB to BGR
-        np_img = np_img.transpose((2, 0, 1)) / 255  # HWC to CHW
-        torch_img = torch.from_numpy(np_img).float().unsqueeze(0).to(device)  # type: ignore
-
-        if tile > h or tile > w:
-            _img = torch.zeros(1, 3, max(h, tile), max(w, tile), dtype=torch_img.dtype, device=torch_img.device)
-            _img[:, :, :h, :w] = torch_img # pad image
-            torch_img = _img
-
-        torch_output = self.tiled_inference(torch_img, model).squeeze(0)
-        torch_output = torch_output[:, :h * 1, :w * 1] # remove padding, if any
-        np_output: np.ndarray = torch_output.float().cpu().clamp_(0, 1).numpy()
-        del torch_img, torch_output
+        img = upscaler_utils.upscale_2(
+            img,
+            model,
+            tile_size=shared.opts.SCUNET_tile,
+            tile_overlap=shared.opts.SCUNET_tile_overlap,
+            scale=1,  # ScuNET is a denoising model, not an upscaler
+            desc='ScuNET',
+        )
         devices.torch_gc()
-
-        output = np_output.transpose((1, 2, 0))  # CHW to HWC
-        output = output[:, :, ::-1]  # BGR to RGB
-        return PIL.Image.fromarray((output * 255).astype(np.uint8))
+        return img
 
     def load_model(self, path: str):
         device = devices.get_device_for('scunet')
         if path.startswith("http"):
             # TODO: this doesn't use `path` at all?
-            filename = load_file_from_url(self.model_url, model_dir=self.model_download_path, file_name=f"{self.name}.pth")
+            filename = modelloader.load_file_from_url(self.model_url, model_dir=self.model_download_path, file_name=f"{self.name}.pth")
         else:
             filename = path
-        model = SCUNet(in_nc=3, config=[4, 4, 4, 4, 4, 4, 4], dim=64)
-        model.load_state_dict(torch.load(filename), strict=True)
-        model.eval()
-        for _, v in model.named_parameters():
-            v.requires_grad = False
-        model = model.to(device)
-
-        return model
+        return modelloader.load_spandrel_model(filename, device=device, expected_architecture='SCUNet')
 
 
 def on_ui_settings():
     import gradio as gr
-    from modules import shared
 
     shared.opts.add_option("SCUNET_tile", shared.OptionInfo(256, "Tile size for SCUNET upscalers.", gr.Slider, {"minimum": 0, "maximum": 512, "step": 16}, section=('upscaling', "Upscaling")).info("0 = no tiling"))
     shared.opts.add_option("SCUNET_tile_overlap", shared.OptionInfo(8, "Tile overlap for SCUNET upscalers.", gr.Slider, {"minimum": 0, "maximum": 64, "step": 1}, section=('upscaling', "Upscaling")).info("Low values = visible seam"))
diff --git a/extensions-builtin/ScuNET/scunet_model_arch.py b/extensions-builtin/ScuNET/scunet_model_arch.py
deleted file mode 100644
index b51a880629b..00000000000
--- a/extensions-builtin/ScuNET/scunet_model_arch.py
+++ /dev/null
@@ -1,268 +0,0 @@
-# -*- coding: utf-8 -*-
-import numpy as np
-import torch
-import torch.nn as nn
-from einops import rearrange
-from einops.layers.torch import Rearrange
-from timm.models.layers import trunc_normal_, DropPath
-
-
-class WMSA(nn.Module):
-    """ Self-attention module in Swin Transformer
-    """
-
-    def __init__(self, input_dim, output_dim, head_dim, window_size, type):
-        super(WMSA, self).__init__()
-        self.input_dim = input_dim
-        self.output_dim = output_dim
-        self.head_dim = head_dim
-        self.scale = self.head_dim ** -0.5
-        self.n_heads = input_dim // head_dim
-        self.window_size = window_size
-        self.type = type
-        self.embedding_layer = nn.Linear(self.input_dim, 3 * self.input_dim, bias=True)
-
-        self.relative_position_params = nn.Parameter(
-            torch.zeros((2 * window_size - 1) * (2 * window_size - 1), self.n_heads))
-
-        self.linear = nn.Linear(self.input_dim, self.output_dim)
-
-        trunc_normal_(self.relative_position_params, std=.02)
-        self.relative_position_params = torch.nn.Parameter(
-            self.relative_position_params.view(2 * window_size - 1, 2 * window_size - 1, self.n_heads).transpose(1,
-                                                                                                                 2).transpose(
-                0, 1))
-
-    def generate_mask(self, h, w, p, shift):
-        """ generating the mask of SW-MSA
-        Args:
-            shift: shift parameters in CyclicShift.
-        Returns:
-            attn_mask: should be (1 1 w p p),
-        """
-        # supporting square.
-        attn_mask = torch.zeros(h, w, p, p, p, p, dtype=torch.bool, device=self.relative_position_params.device)
-        if self.type == 'W':
-            return attn_mask
-
-        s = p - shift
-        attn_mask[-1, :, :s, :, s:, :] = True
-        attn_mask[-1, :, s:, :, :s, :] = True
-        attn_mask[:, -1, :, :s, :, s:] = True
-        attn_mask[:, -1, :, s:, :, :s] = True
-        attn_mask = rearrange(attn_mask, 'w1 w2 p1 p2 p3 p4 -> 1 1 (w1 w2) (p1 p2) (p3 p4)')
-        return attn_mask
-
-    def forward(self, x):
-        """ Forward pass of Window Multi-head Self-attention module.
-        Args:
-            x: input tensor with shape of [b h w c];
-            attn_mask: attention mask, fill -inf where the value is True;
-        Returns:
-            output: tensor shape [b h w c]
-        """
-        if self.type != 'W':
-            x = torch.roll(x, shifts=(-(self.window_size // 2), -(self.window_size // 2)), dims=(1, 2))
-
-        x = rearrange(x, 'b (w1 p1) (w2 p2) c -> b w1 w2 p1 p2 c', p1=self.window_size, p2=self.window_size)
-        h_windows = x.size(1)
-        w_windows = x.size(2)
-        # square validation
-        # assert h_windows == w_windows
-
-        x = rearrange(x, 'b w1 w2 p1 p2 c -> b (w1 w2) (p1 p2) c', p1=self.window_size, p2=self.window_size)
-        qkv = self.embedding_layer(x)
-        q, k, v = rearrange(qkv, 'b nw np (threeh c) -> threeh b nw np c', c=self.head_dim).chunk(3, dim=0)
-        sim = torch.einsum('hbwpc,hbwqc->hbwpq', q, k) * self.scale
-        # Adding learnable relative embedding
-        sim = sim + rearrange(self.relative_embedding(), 'h p q -> h 1 1 p q')
-        # Using Attn Mask to distinguish different subwindows.
-        if self.type != 'W':
-            attn_mask = self.generate_mask(h_windows, w_windows, self.window_size, shift=self.window_size // 2)
-            sim = sim.masked_fill_(attn_mask, float("-inf"))
-
-        probs = nn.functional.softmax(sim, dim=-1)
-        output = torch.einsum('hbwij,hbwjc->hbwic', probs, v)
-        output = rearrange(output, 'h b w p c -> b w p (h c)')
-        output = self.linear(output)
-        output = rearrange(output, 'b (w1 w2) (p1 p2) c -> b (w1 p1) (w2 p2) c', w1=h_windows, p1=self.window_size)
-
-        if self.type != 'W':
-            output = torch.roll(output, shifts=(self.window_size // 2, self.window_size // 2), dims=(1, 2))
-
-        return output
-
-    def relative_embedding(self):
-        cord = torch.tensor(np.array([[i, j] for i in range(self.window_size) for j in range(self.window_size)]))
-        relation = cord[:, None, :] - cord[None, :, :] + self.window_size - 1
-        # negative is allowed
-        return self.relative_position_params[:, relation[:, :, 0].long(), relation[:, :, 1].long()]
-
-
-class Block(nn.Module):
-    def __init__(self, input_dim, output_dim, head_dim, window_size, drop_path, type='W', input_resolution=None):
-        """ SwinTransformer Block
-        """
-        super(Block, self).__init__()
-        self.input_dim = input_dim
-        self.output_dim = output_dim
-        assert type in ['W', 'SW']
-        self.type = type
-        if input_resolution <= window_size:
-            self.type = 'W'
-
-        self.ln1 = nn.LayerNorm(input_dim)
-        self.msa = WMSA(input_dim, input_dim, head_dim, window_size, self.type)
-        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
-        self.ln2 = nn.LayerNorm(input_dim)
-        self.mlp = nn.Sequential(
-            nn.Linear(input_dim, 4 * input_dim),
-            nn.GELU(),
-            nn.Linear(4 * input_dim, output_dim),
-        )
-
-    def forward(self, x):
-        x = x + self.drop_path(self.msa(self.ln1(x)))
-        x = x + self.drop_path(self.mlp(self.ln2(x)))
-        return x
-
-
-class ConvTransBlock(nn.Module):
-    def __init__(self, conv_dim, trans_dim, head_dim, window_size, drop_path, type='W', input_resolution=None):
-        """ SwinTransformer and Conv Block
-        """
-        super(ConvTransBlock, self).__init__()
-        self.conv_dim = conv_dim
-        self.trans_dim = trans_dim
-        self.head_dim = head_dim
-        self.window_size = window_size
-        self.drop_path = drop_path
-        self.type = type
-        self.input_resolution = input_resolution
-
-        assert self.type in ['W', 'SW']
-        if self.input_resolution <= self.window_size:
-            self.type = 'W'
-
-        self.trans_block = Block(self.trans_dim, self.trans_dim, self.head_dim, self.window_size, self.drop_path,
-                                 self.type, self.input_resolution)
-        self.conv1_1 = nn.Conv2d(self.conv_dim + self.trans_dim, self.conv_dim + self.trans_dim, 1, 1, 0, bias=True)
-        self.conv1_2 = nn.Conv2d(self.conv_dim + self.trans_dim, self.conv_dim + self.trans_dim, 1, 1, 0, bias=True)
-
-        self.conv_block = nn.Sequential(
-            nn.Conv2d(self.conv_dim, self.conv_dim, 3, 1, 1, bias=False),
-            nn.ReLU(True),
-            nn.Conv2d(self.conv_dim, self.conv_dim, 3, 1, 1, bias=False)
-        )
-
-    def forward(self, x):
-        conv_x, trans_x = torch.split(self.conv1_1(x), (self.conv_dim, self.trans_dim), dim=1)
-        conv_x = self.conv_block(conv_x) + conv_x
-        trans_x = Rearrange('b c h w -> b h w c')(trans_x)
-        trans_x = self.trans_block(trans_x)
-        trans_x = Rearrange('b h w c -> b c h w')(trans_x)
-        res = self.conv1_2(torch.cat((conv_x, trans_x), dim=1))
-        x = x + res
-
-        return x
-
-
-class SCUNet(nn.Module):
-    # def __init__(self, in_nc=3, config=[2, 2, 2, 2, 2, 2, 2], dim=64, drop_path_rate=0.0, input_resolution=256):
-    def __init__(self, in_nc=3, config=None, dim=64, drop_path_rate=0.0, input_resolution=256):
-        super(SCUNet, self).__init__()
-        if config is None:
-            config = [2, 2, 2, 2, 2, 2, 2]
-        self.config = config
-        self.dim = dim
-        self.head_dim = 32
-        self.window_size = 8
-
-        # drop path rate for each layer
-        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(config))]
-
-        self.m_head = [nn.Conv2d(in_nc, dim, 3, 1, 1, bias=False)]
-
-        begin = 0
-        self.m_down1 = [ConvTransBlock(dim // 2, dim // 2, self.head_dim, self.window_size, dpr[i + begin],
-                                       'W' if not i % 2 else 'SW', input_resolution)
-                        for i in range(config[0])] + \
-                       [nn.Conv2d(dim, 2 * dim, 2, 2, 0, bias=False)]
-
-        begin += config[0]
-        self.m_down2 = [ConvTransBlock(dim, dim, self.head_dim, self.window_size, dpr[i + begin],
-                                       'W' if not i % 2 else 'SW', input_resolution // 2)
-                        for i in range(config[1])] + \
-                       [nn.Conv2d(2 * dim, 4 * dim, 2, 2, 0, bias=False)]
-
-        begin += config[1]
-        self.m_down3 = [ConvTransBlock(2 * dim, 2 * dim, self.head_dim, self.window_size, dpr[i + begin],
-                                       'W' if not i % 2 else 'SW', input_resolution // 4)
-                        for i in range(config[2])] + \
-                       [nn.Conv2d(4 * dim, 8 * dim, 2, 2, 0, bias=False)]
-
-        begin += config[2]
-        self.m_body = [ConvTransBlock(4 * dim, 4 * dim, self.head_dim, self.window_size, dpr[i + begin],
-                                      'W' if not i % 2 else 'SW', input_resolution // 8)
-                       for i in range(config[3])]
-
-        begin += config[3]
-        self.m_up3 = [nn.ConvTranspose2d(8 * dim, 4 * dim, 2, 2, 0, bias=False), ] + \
-                     [ConvTransBlock(2 * dim, 2 * dim, self.head_dim, self.window_size, dpr[i + begin],
-                                     'W' if not i % 2 else 'SW', input_resolution // 4)
-                      for i in range(config[4])]
-
-        begin += config[4]
-        self.m_up2 = [nn.ConvTranspose2d(4 * dim, 2 * dim, 2, 2, 0, bias=False), ] + \
-                     [ConvTransBlock(dim, dim, self.head_dim, self.window_size, dpr[i + begin],
-                                     'W' if not i % 2 else 'SW', input_resolution // 2)
-                      for i in range(config[5])]
-
-        begin += config[5]
-        self.m_up1 = [nn.ConvTranspose2d(2 * dim, dim, 2, 2, 0, bias=False), ] + \
-                     [ConvTransBlock(dim // 2, dim // 2, self.head_dim, self.window_size, dpr[i + begin],
-                                     'W' if not i % 2 else 'SW', input_resolution)
-                      for i in range(config[6])]
-
-        self.m_tail = [nn.Conv2d(dim, in_nc, 3, 1, 1, bias=False)]
-
-        self.m_head = nn.Sequential(*self.m_head)
-        self.m_down1 = nn.Sequential(*self.m_down1)
-        self.m_down2 = nn.Sequential(*self.m_down2)
-        self.m_down3 = nn.Sequential(*self.m_down3)
-        self.m_body = nn.Sequential(*self.m_body)
-        self.m_up3 = nn.Sequential(*self.m_up3)
-        self.m_up2 = nn.Sequential(*self.m_up2)
-        self.m_up1 = nn.Sequential(*self.m_up1)
-        self.m_tail = nn.Sequential(*self.m_tail)
-        # self.apply(self._init_weights)
-
-    def forward(self, x0):
-
-        h, w = x0.size()[-2:]
-        paddingBottom = int(np.ceil(h / 64) * 64 - h)
-        paddingRight = int(np.ceil(w / 64) * 64 - w)
-        x0 = nn.ReplicationPad2d((0, paddingRight, 0, paddingBottom))(x0)
-
-        x1 = self.m_head(x0)
-        x2 = self.m_down1(x1)
-        x3 = self.m_down2(x2)
-        x4 = self.m_down3(x3)
-        x = self.m_body(x4)
-        x = self.m_up3(x + x4)
-        x = self.m_up2(x + x3)
-        x = self.m_up1(x + x2)
-        x = self.m_tail(x + x1)
-
-        x = x[..., :h, :w]
-
-        return x
-
-    def _init_weights(self, m):
-        if isinstance(m, nn.Linear):
-            trunc_normal_(m.weight, std=.02)
-            if m.bias is not None:
-                nn.init.constant_(m.bias, 0)
-        elif isinstance(m, nn.LayerNorm):
-            nn.init.constant_(m.bias, 0)
-            nn.init.constant_(m.weight, 1.0)
diff --git a/extensions-builtin/SwinIR/scripts/swinir_model.py b/extensions-builtin/SwinIR/scripts/swinir_model.py
index ae0d0e6a8ea..16bf9b792fc 100644
--- a/extensions-builtin/SwinIR/scripts/swinir_model.py
+++ b/extensions-builtin/SwinIR/scripts/swinir_model.py
@@ -1,20 +1,15 @@
+import logging
 import sys
-import platform
 
-import numpy as np
 import torch
 from PIL import Image
-from tqdm import tqdm
 
-from modules import modelloader, devices, script_callbacks, shared
-from modules.shared import opts, state
-from swinir_model_arch import SwinIR
-from swinir_model_arch_v2 import Swin2SR
+from modules import devices, modelloader, script_callbacks, shared, upscaler_utils
 from modules.upscaler import Upscaler, UpscalerData
 
 SWINIR_MODEL_URL = "https://github.com/JingyunLiang/SwinIR/releases/download/v0.0/003_realSR_BSRGAN_DFOWMFC_s64w8_SwinIR-L_x4_GAN.pth"
 
-device_swinir = devices.get_device_for('swinir')
+logger = logging.getLogger(__name__)
 
 
 class UpscalerSwinIR(Upscaler):
@@ -37,26 +32,28 @@ def __init__(self, dirname):
             scalers.append(model_data)
         self.scalers = scalers
 
-    def do_upscale(self, img, model_file):
-        use_compile = hasattr(opts, 'SWIN_torch_compile') and opts.SWIN_torch_compile \
-            and int(torch.__version__.split('.')[0]) >= 2 and platform.system() != "Windows"
-        current_config = (model_file, opts.SWIN_tile)
+    def do_upscale(self, img: Image.Image, model_file: str) -> Image.Image:
+        current_config = (model_file, shared.opts.SWIN_tile)
 
-        if use_compile and self._cached_model_config == current_config:
+        if self._cached_model_config == current_config:
             model = self._cached_model
         else:
-            self._cached_model = None
             try:
                 model = self.load_model(model_file)
             except Exception as e:
                 print(f"Failed loading SwinIR model {model_file}: {e}", file=sys.stderr)
                 return img
-            model = model.to(device_swinir, dtype=devices.dtype)
-            if use_compile:
-                model = torch.compile(model)
-                self._cached_model = model
-                self._cached_model_config = current_config
-        img = upscale(img, model)
+            self._cached_model = model
+            self._cached_model_config = current_config
+
+        img = upscaler_utils.upscale_2(
+            img,
+            model,
+            tile_size=shared.opts.SWIN_tile,
+            tile_overlap=shared.opts.SWIN_tile_overlap,
+            scale=model.scale,
+            desc="SwinIR",
+        )
         devices.torch_gc()
         return img
 
@@ -69,115 +66,22 @@ def load_model(self, path, scale=4):
             )
         else:
             filename = path
-        if filename.endswith(".v2.pth"):
-            model = Swin2SR(
-                upscale=scale,
-                in_chans=3,
-                img_size=64,
-                window_size=8,
-                img_range=1.0,
-                depths=[6, 6, 6, 6, 6, 6],
-                embed_dim=180,
-                num_heads=[6, 6, 6, 6, 6, 6],
-                mlp_ratio=2,
-                upsampler="nearest+conv",
-                resi_connection="1conv",
-            )
-            params = None
-        else:
-            model = SwinIR(
-                upscale=scale,
-                in_chans=3,
-                img_size=64,
-                window_size=8,
-                img_range=1.0,
-                depths=[6, 6, 6, 6, 6, 6, 6, 6, 6],
-                embed_dim=240,
-                num_heads=[8, 8, 8, 8, 8, 8, 8, 8, 8],
-                mlp_ratio=2,
-                upsampler="nearest+conv",
-                resi_connection="3conv",
-            )
-            params = "params_ema"
 
-        pretrained_model = torch.load(filename)
-        if params is not None:
-            model.load_state_dict(pretrained_model[params], strict=True)
-        else:
-            model.load_state_dict(pretrained_model, strict=True)
-        return model
-
-
-def upscale(
-        img,
-        model,
-        tile=None,
-        tile_overlap=None,
-        window_size=8,
-        scale=4,
-):
-    tile = tile or opts.SWIN_tile
-    tile_overlap = tile_overlap or opts.SWIN_tile_overlap
-
-
-    img = np.array(img)
-    img = img[:, :, ::-1]
-    img = np.moveaxis(img, 2, 0) / 255
-    img = torch.from_numpy(img).float()
-    img = img.unsqueeze(0).to(device_swinir, dtype=devices.dtype)
-    with torch.no_grad(), devices.autocast():
-        _, _, h_old, w_old = img.size()
-        h_pad = (h_old // window_size + 1) * window_size - h_old
-        w_pad = (w_old // window_size + 1) * window_size - w_old
-        img = torch.cat([img, torch.flip(img, [2])], 2)[:, :, : h_old + h_pad, :]
-        img = torch.cat([img, torch.flip(img, [3])], 3)[:, :, :, : w_old + w_pad]
-        output = inference(img, model, tile, tile_overlap, window_size, scale)
-        output = output[..., : h_old * scale, : w_old * scale]
-        output = output.data.squeeze().float().cpu().clamp_(0, 1).numpy()
-        if output.ndim == 3:
-            output = np.transpose(
-                output[[2, 1, 0], :, :], (1, 2, 0)
-            )  # CHW-RGB to HCW-BGR
-        output = (output * 255.0).round().astype(np.uint8)  # float32 to uint8
-        return Image.fromarray(output, "RGB")
-
-
-def inference(img, model, tile, tile_overlap, window_size, scale):
-    # test the image tile by tile
-    b, c, h, w = img.size()
-    tile = min(tile, h, w)
-    assert tile % window_size == 0, "tile size should be a multiple of window_size"
-    sf = scale
-
-    stride = tile - tile_overlap
-    h_idx_list = list(range(0, h - tile, stride)) + [h - tile]
-    w_idx_list = list(range(0, w - tile, stride)) + [w - tile]
-    E = torch.zeros(b, c, h * sf, w * sf, dtype=devices.dtype, device=device_swinir).type_as(img)
-    W = torch.zeros_like(E, dtype=devices.dtype, device=device_swinir)
-
-    with tqdm(total=len(h_idx_list) * len(w_idx_list), desc="SwinIR tiles") as pbar:
-        for h_idx in h_idx_list:
-            if state.interrupted or state.skipped:
-                break
-
-            for w_idx in w_idx_list:
-                if state.interrupted or state.skipped:
-                    break
-
-                in_patch = img[..., h_idx: h_idx + tile, w_idx: w_idx + tile]
-                out_patch = model(in_patch)
-                out_patch_mask = torch.ones_like(out_patch)
-
-                E[
-                ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf
-                ].add_(out_patch)
-                W[
-                ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf
-                ].add_(out_patch_mask)
-                pbar.update(1)
-    output = E.div_(W)
-
-    return output
+        model_descriptor = modelloader.load_spandrel_model(
+            filename,
+            device=self._get_device(),
+            prefer_half=(devices.dtype == torch.float16),
+            expected_architecture="SwinIR",
+        )
+        if getattr(shared.opts, 'SWIN_torch_compile', False):
+            try:
+                model_descriptor.model.compile()
+            except Exception:
+                logger.warning("Failed to compile SwinIR model, fallback to JIT", exc_info=True)
+        return model_descriptor
+
+    def _get_device(self):
+        return devices.get_device_for('swinir')
 
 
 def on_ui_settings():
@@ -185,8 +89,7 @@ def on_ui_settings():
 
     shared.opts.add_option("SWIN_tile", shared.OptionInfo(192, "Tile size for all SwinIR.", gr.Slider, {"minimum": 16, "maximum": 512, "step": 16}, section=('upscaling', "Upscaling")))
     shared.opts.add_option("SWIN_tile_overlap", shared.OptionInfo(8, "Tile overlap, in pixels for SwinIR. Low values = visible seam.", gr.Slider, {"minimum": 0, "maximum": 48, "step": 1}, section=('upscaling', "Upscaling")))
-    if int(torch.__version__.split('.')[0]) >= 2 and platform.system() != "Windows":    # torch.compile() require pytorch 2.0 or above, and not on Windows
-        shared.opts.add_option("SWIN_torch_compile", shared.OptionInfo(False, "Use torch.compile to accelerate SwinIR.", gr.Checkbox, {"interactive": True}, section=('upscaling', "Upscaling")).info("Takes longer on first run"))
+    shared.opts.add_option("SWIN_torch_compile", shared.OptionInfo(False, "Use torch.compile to accelerate SwinIR.", gr.Checkbox, {"interactive": True}, section=('upscaling', "Upscaling")).info("Takes longer on first run"))
 
 
 script_callbacks.on_ui_settings(on_ui_settings)
diff --git a/extensions-builtin/SwinIR/swinir_model_arch.py b/extensions-builtin/SwinIR/swinir_model_arch.py
deleted file mode 100644
index 93b9327473a..00000000000
--- a/extensions-builtin/SwinIR/swinir_model_arch.py
+++ /dev/null
@@ -1,867 +0,0 @@
-# -----------------------------------------------------------------------------------
-# SwinIR: Image Restoration Using Swin Transformer, https://arxiv.org/abs/2108.10257
-# Originally Written by Ze Liu, Modified by Jingyun Liang.
-# -----------------------------------------------------------------------------------
-
-import math
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-import torch.utils.checkpoint as checkpoint
-from timm.models.layers import DropPath, to_2tuple, trunc_normal_
-
-
-class Mlp(nn.Module):
-    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
-        super().__init__()
-        out_features = out_features or in_features
-        hidden_features = hidden_features or in_features
-        self.fc1 = nn.Linear(in_features, hidden_features)
-        self.act = act_layer()
-        self.fc2 = nn.Linear(hidden_features, out_features)
-        self.drop = nn.Dropout(drop)
-
-    def forward(self, x):
-        x = self.fc1(x)
-        x = self.act(x)
-        x = self.drop(x)
-        x = self.fc2(x)
-        x = self.drop(x)
-        return x
-
-
-def window_partition(x, window_size):
-    """
-    Args:
-        x: (B, H, W, C)
-        window_size (int): window size
-
-    Returns:
-        windows: (num_windows*B, window_size, window_size, C)
-    """
-    B, H, W, C = x.shape
-    x = x.view(B, H // window_size, window_size, W // window_size, window_size, C)
-    windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C)
-    return windows
-
-
-def window_reverse(windows, window_size, H, W):
-    """
-    Args:
-        windows: (num_windows*B, window_size, window_size, C)
-        window_size (int): Window size
-        H (int): Height of image
-        W (int): Width of image
-
-    Returns:
-        x: (B, H, W, C)
-    """
-    B = int(windows.shape[0] / (H * W / window_size / window_size))
-    x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1)
-    x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1)
-    return x
-
-
-class WindowAttention(nn.Module):
-    r""" Window based multi-head self attention (W-MSA) module with relative position bias.
-    It supports both of shifted and non-shifted window.
-
-    Args:
-        dim (int): Number of input channels.
-        window_size (tuple[int]): The height and width of the window.
-        num_heads (int): Number of attention heads.
-        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True
-        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set
-        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0
-        proj_drop (float, optional): Dropout ratio of output. Default: 0.0
-    """
-
-    def __init__(self, dim, window_size, num_heads, qkv_bias=True, qk_scale=None, attn_drop=0., proj_drop=0.):
-
-        super().__init__()
-        self.dim = dim
-        self.window_size = window_size  # Wh, Ww
-        self.num_heads = num_heads
-        head_dim = dim // num_heads
-        self.scale = qk_scale or head_dim ** -0.5
-
-        # define a parameter table of relative position bias
-        self.relative_position_bias_table = nn.Parameter(
-            torch.zeros((2 * window_size[0] - 1) * (2 * window_size[1] - 1), num_heads))  # 2*Wh-1 * 2*Ww-1, nH
-
-        # get pair-wise relative position index for each token inside the window
-        coords_h = torch.arange(self.window_size[0])
-        coords_w = torch.arange(self.window_size[1])
-        coords = torch.stack(torch.meshgrid([coords_h, coords_w]))  # 2, Wh, Ww
-        coords_flatten = torch.flatten(coords, 1)  # 2, Wh*Ww
-        relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :]  # 2, Wh*Ww, Wh*Ww
-        relative_coords = relative_coords.permute(1, 2, 0).contiguous()  # Wh*Ww, Wh*Ww, 2
-        relative_coords[:, :, 0] += self.window_size[0] - 1  # shift to start from 0
-        relative_coords[:, :, 1] += self.window_size[1] - 1
-        relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1
-        relative_position_index = relative_coords.sum(-1)  # Wh*Ww, Wh*Ww
-        self.register_buffer("relative_position_index", relative_position_index)
-
-        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
-        self.attn_drop = nn.Dropout(attn_drop)
-        self.proj = nn.Linear(dim, dim)
-
-        self.proj_drop = nn.Dropout(proj_drop)
-
-        trunc_normal_(self.relative_position_bias_table, std=.02)
-        self.softmax = nn.Softmax(dim=-1)
-
-    def forward(self, x, mask=None):
-        """
-        Args:
-            x: input features with shape of (num_windows*B, N, C)
-            mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None
-        """
-        B_, N, C = x.shape
-        qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
-        q, k, v = qkv[0], qkv[1], qkv[2]  # make torchscript happy (cannot use tensor as tuple)
-
-        q = q * self.scale
-        attn = (q @ k.transpose(-2, -1))
-
-        relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
-            self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1)  # Wh*Ww,Wh*Ww,nH
-        relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww
-        attn = attn + relative_position_bias.unsqueeze(0)
-
-        if mask is not None:
-            nW = mask.shape[0]
-            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0)
-            attn = attn.view(-1, self.num_heads, N, N)
-            attn = self.softmax(attn)
-        else:
-            attn = self.softmax(attn)
-
-        attn = self.attn_drop(attn)
-
-        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
-        x = self.proj(x)
-        x = self.proj_drop(x)
-        return x
-
-    def extra_repr(self) -> str:
-        return f'dim={self.dim}, window_size={self.window_size}, num_heads={self.num_heads}'
-
-    def flops(self, N):
-        # calculate flops for 1 window with token length of N
-        flops = 0
-        # qkv = self.qkv(x)
-        flops += N * self.dim * 3 * self.dim
-        # attn = (q @ k.transpose(-2, -1))
-        flops += self.num_heads * N * (self.dim // self.num_heads) * N
-        #  x = (attn @ v)
-        flops += self.num_heads * N * N * (self.dim // self.num_heads)
-        # x = self.proj(x)
-        flops += N * self.dim * self.dim
-        return flops
-
-
-class SwinTransformerBlock(nn.Module):
-    r""" Swin Transformer Block.
-
-    Args:
-        dim (int): Number of input channels.
-        input_resolution (tuple[int]): Input resolution.
-        num_heads (int): Number of attention heads.
-        window_size (int): Window size.
-        shift_size (int): Shift size for SW-MSA.
-        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
-        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
-        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
-        drop (float, optional): Dropout rate. Default: 0.0
-        attn_drop (float, optional): Attention dropout rate. Default: 0.0
-        drop_path (float, optional): Stochastic depth rate. Default: 0.0
-        act_layer (nn.Module, optional): Activation layer. Default: nn.GELU
-        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
-    """
-
-    def __init__(self, dim, input_resolution, num_heads, window_size=7, shift_size=0,
-                 mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0., drop_path=0.,
-                 act_layer=nn.GELU, norm_layer=nn.LayerNorm):
-        super().__init__()
-        self.dim = dim
-        self.input_resolution = input_resolution
-        self.num_heads = num_heads
-        self.window_size = window_size
-        self.shift_size = shift_size
-        self.mlp_ratio = mlp_ratio
-        if min(self.input_resolution) <= self.window_size:
-            # if window size is larger than input resolution, we don't partition windows
-            self.shift_size = 0
-            self.window_size = min(self.input_resolution)
-        assert 0 <= self.shift_size < self.window_size, "shift_size must in 0-window_size"
-
-        self.norm1 = norm_layer(dim)
-        self.attn = WindowAttention(
-            dim, window_size=to_2tuple(self.window_size), num_heads=num_heads,
-            qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
-
-        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
-        self.norm2 = norm_layer(dim)
-        mlp_hidden_dim = int(dim * mlp_ratio)
-        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)
-
-        if self.shift_size > 0:
-            attn_mask = self.calculate_mask(self.input_resolution)
-        else:
-            attn_mask = None
-
-        self.register_buffer("attn_mask", attn_mask)
-
-    def calculate_mask(self, x_size):
-        # calculate attention mask for SW-MSA
-        H, W = x_size
-        img_mask = torch.zeros((1, H, W, 1))  # 1 H W 1
-        h_slices = (slice(0, -self.window_size),
-                    slice(-self.window_size, -self.shift_size),
-                    slice(-self.shift_size, None))
-        w_slices = (slice(0, -self.window_size),
-                    slice(-self.window_size, -self.shift_size),
-                    slice(-self.shift_size, None))
-        cnt = 0
-        for h in h_slices:
-            for w in w_slices:
-                img_mask[:, h, w, :] = cnt
-                cnt += 1
-
-        mask_windows = window_partition(img_mask, self.window_size)  # nW, window_size, window_size, 1
-        mask_windows = mask_windows.view(-1, self.window_size * self.window_size)
-        attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
-        attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))
-
-        return attn_mask
-
-    def forward(self, x, x_size):
-        H, W = x_size
-        B, L, C = x.shape
-        # assert L == H * W, "input feature has wrong size"
-
-        shortcut = x
-        x = self.norm1(x)
-        x = x.view(B, H, W, C)
-
-        # cyclic shift
-        if self.shift_size > 0:
-            shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
-        else:
-            shifted_x = x
-
-        # partition windows
-        x_windows = window_partition(shifted_x, self.window_size)  # nW*B, window_size, window_size, C
-        x_windows = x_windows.view(-1, self.window_size * self.window_size, C)  # nW*B, window_size*window_size, C
-
-        # W-MSA/SW-MSA (to be compatible for testing on images whose shapes are the multiple of window size
-        if self.input_resolution == x_size:
-            attn_windows = self.attn(x_windows, mask=self.attn_mask)  # nW*B, window_size*window_size, C
-        else:
-            attn_windows = self.attn(x_windows, mask=self.calculate_mask(x_size).to(x.device))
-
-        # merge windows
-        attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C)
-        shifted_x = window_reverse(attn_windows, self.window_size, H, W)  # B H' W' C
-
-        # reverse cyclic shift
-        if self.shift_size > 0:
-            x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
-        else:
-            x = shifted_x
-        x = x.view(B, H * W, C)
-
-        # FFN
-        x = shortcut + self.drop_path(x)
-        x = x + self.drop_path(self.mlp(self.norm2(x)))
-
-        return x
-
-    def extra_repr(self) -> str:
-        return f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, " \
-               f"window_size={self.window_size}, shift_size={self.shift_size}, mlp_ratio={self.mlp_ratio}"
-
-    def flops(self):
-        flops = 0
-        H, W = self.input_resolution
-        # norm1
-        flops += self.dim * H * W
-        # W-MSA/SW-MSA
-        nW = H * W / self.window_size / self.window_size
-        flops += nW * self.attn.flops(self.window_size * self.window_size)
-        # mlp
-        flops += 2 * H * W * self.dim * self.dim * self.mlp_ratio
-        # norm2
-        flops += self.dim * H * W
-        return flops
-
-
-class PatchMerging(nn.Module):
-    r""" Patch Merging Layer.
-
-    Args:
-        input_resolution (tuple[int]): Resolution of input feature.
-        dim (int): Number of input channels.
-        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
-    """
-
-    def __init__(self, input_resolution, dim, norm_layer=nn.LayerNorm):
-        super().__init__()
-        self.input_resolution = input_resolution
-        self.dim = dim
-        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
-        self.norm = norm_layer(4 * dim)
-
-    def forward(self, x):
-        """
-        x: B, H*W, C
-        """
-        H, W = self.input_resolution
-        B, L, C = x.shape
-        assert L == H * W, "input feature has wrong size"
-        assert H % 2 == 0 and W % 2 == 0, f"x size ({H}*{W}) are not even."
-
-        x = x.view(B, H, W, C)
-
-        x0 = x[:, 0::2, 0::2, :]  # B H/2 W/2 C
-        x1 = x[:, 1::2, 0::2, :]  # B H/2 W/2 C
-        x2 = x[:, 0::2, 1::2, :]  # B H/2 W/2 C
-        x3 = x[:, 1::2, 1::2, :]  # B H/2 W/2 C
-        x = torch.cat([x0, x1, x2, x3], -1)  # B H/2 W/2 4*C
-        x = x.view(B, -1, 4 * C)  # B H/2*W/2 4*C
-
-        x = self.norm(x)
-        x = self.reduction(x)
-
-        return x
-
-    def extra_repr(self) -> str:
-        return f"input_resolution={self.input_resolution}, dim={self.dim}"
-
-    def flops(self):
-        H, W = self.input_resolution
-        flops = H * W * self.dim
-        flops += (H // 2) * (W // 2) * 4 * self.dim * 2 * self.dim
-        return flops
-
-
-class BasicLayer(nn.Module):
-    """ A basic Swin Transformer layer for one stage.
-
-    Args:
-        dim (int): Number of input channels.
-        input_resolution (tuple[int]): Input resolution.
-        depth (int): Number of blocks.
-        num_heads (int): Number of attention heads.
-        window_size (int): Local window size.
-        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
-        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
-        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
-        drop (float, optional): Dropout rate. Default: 0.0
-        attn_drop (float, optional): Attention dropout rate. Default: 0.0
-        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
-        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
-        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
-        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.
-    """
-
-    def __init__(self, dim, input_resolution, depth, num_heads, window_size,
-                 mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0.,
-                 drop_path=0., norm_layer=nn.LayerNorm, downsample=None, use_checkpoint=False):
-
-        super().__init__()
-        self.dim = dim
-        self.input_resolution = input_resolution
-        self.depth = depth
-        self.use_checkpoint = use_checkpoint
-
-        # build blocks
-        self.blocks = nn.ModuleList([
-            SwinTransformerBlock(dim=dim, input_resolution=input_resolution,
-                                 num_heads=num_heads, window_size=window_size,
-                                 shift_size=0 if (i % 2 == 0) else window_size // 2,
-                                 mlp_ratio=mlp_ratio,
-                                 qkv_bias=qkv_bias, qk_scale=qk_scale,
-                                 drop=drop, attn_drop=attn_drop,
-                                 drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,
-                                 norm_layer=norm_layer)
-            for i in range(depth)])
-
-        # patch merging layer
-        if downsample is not None:
-            self.downsample = downsample(input_resolution, dim=dim, norm_layer=norm_layer)
-        else:
-            self.downsample = None
-
-    def forward(self, x, x_size):
-        for blk in self.blocks:
-            if self.use_checkpoint:
-                x = checkpoint.checkpoint(blk, x, x_size)
-            else:
-                x = blk(x, x_size)
-        if self.downsample is not None:
-            x = self.downsample(x)
-        return x
-
-    def extra_repr(self) -> str:
-        return f"dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}"
-
-    def flops(self):
-        flops = 0
-        for blk in self.blocks:
-            flops += blk.flops()
-        if self.downsample is not None:
-            flops += self.downsample.flops()
-        return flops
-
-
-class RSTB(nn.Module):
-    """Residual Swin Transformer Block (RSTB).
-
-    Args:
-        dim (int): Number of input channels.
-        input_resolution (tuple[int]): Input resolution.
-        depth (int): Number of blocks.
-        num_heads (int): Number of attention heads.
-        window_size (int): Local window size.
-        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
-        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
-        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
-        drop (float, optional): Dropout rate. Default: 0.0
-        attn_drop (float, optional): Attention dropout rate. Default: 0.0
-        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
-        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
-        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
-        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.
-        img_size: Input image size.
-        patch_size: Patch size.
-        resi_connection: The convolutional block before residual connection.
-    """
-
-    def __init__(self, dim, input_resolution, depth, num_heads, window_size,
-                 mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0.,
-                 drop_path=0., norm_layer=nn.LayerNorm, downsample=None, use_checkpoint=False,
-                 img_size=224, patch_size=4, resi_connection='1conv'):
-        super(RSTB, self).__init__()
-
-        self.dim = dim
-        self.input_resolution = input_resolution
-
-        self.residual_group = BasicLayer(dim=dim,
-                                         input_resolution=input_resolution,
-                                         depth=depth,
-                                         num_heads=num_heads,
-                                         window_size=window_size,
-                                         mlp_ratio=mlp_ratio,
-                                         qkv_bias=qkv_bias, qk_scale=qk_scale,
-                                         drop=drop, attn_drop=attn_drop,
-                                         drop_path=drop_path,
-                                         norm_layer=norm_layer,
-                                         downsample=downsample,
-                                         use_checkpoint=use_checkpoint)
-
-        if resi_connection == '1conv':
-            self.conv = nn.Conv2d(dim, dim, 3, 1, 1)
-        elif resi_connection == '3conv':
-            # to save parameters and memory
-            self.conv = nn.Sequential(nn.Conv2d(dim, dim // 4, 3, 1, 1), nn.LeakyReLU(negative_slope=0.2, inplace=True),
-                                      nn.Conv2d(dim // 4, dim // 4, 1, 1, 0),
-                                      nn.LeakyReLU(negative_slope=0.2, inplace=True),
-                                      nn.Conv2d(dim // 4, dim, 3, 1, 1))
-
-        self.patch_embed = PatchEmbed(
-            img_size=img_size, patch_size=patch_size, in_chans=0, embed_dim=dim,
-            norm_layer=None)
-
-        self.patch_unembed = PatchUnEmbed(
-            img_size=img_size, patch_size=patch_size, in_chans=0, embed_dim=dim,
-            norm_layer=None)
-
-    def forward(self, x, x_size):
-        return self.patch_embed(self.conv(self.patch_unembed(self.residual_group(x, x_size), x_size))) + x
-
-    def flops(self):
-        flops = 0
-        flops += self.residual_group.flops()
-        H, W = self.input_resolution
-        flops += H * W * self.dim * self.dim * 9
-        flops += self.patch_embed.flops()
-        flops += self.patch_unembed.flops()
-
-        return flops
-
-
-class PatchEmbed(nn.Module):
-    r""" Image to Patch Embedding
-
-    Args:
-        img_size (int): Image size.  Default: 224.
-        patch_size (int): Patch token size. Default: 4.
-        in_chans (int): Number of input image channels. Default: 3.
-        embed_dim (int): Number of linear projection output channels. Default: 96.
-        norm_layer (nn.Module, optional): Normalization layer. Default: None
-    """
-
-    def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None):
-        super().__init__()
-        img_size = to_2tuple(img_size)
-        patch_size = to_2tuple(patch_size)
-        patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]]
-        self.img_size = img_size
-        self.patch_size = patch_size
-        self.patches_resolution = patches_resolution
-        self.num_patches = patches_resolution[0] * patches_resolution[1]
-
-        self.in_chans = in_chans
-        self.embed_dim = embed_dim
-
-        if norm_layer is not None:
-            self.norm = norm_layer(embed_dim)
-        else:
-            self.norm = None
-
-    def forward(self, x):
-        x = x.flatten(2).transpose(1, 2)  # B Ph*Pw C
-        if self.norm is not None:
-            x = self.norm(x)
-        return x
-
-    def flops(self):
-        flops = 0
-        H, W = self.img_size
-        if self.norm is not None:
-            flops += H * W * self.embed_dim
-        return flops
-
-
-class PatchUnEmbed(nn.Module):
-    r""" Image to Patch Unembedding
-
-    Args:
-        img_size (int): Image size.  Default: 224.
-        patch_size (int): Patch token size. Default: 4.
-        in_chans (int): Number of input image channels. Default: 3.
-        embed_dim (int): Number of linear projection output channels. Default: 96.
-        norm_layer (nn.Module, optional): Normalization layer. Default: None
-    """
-
-    def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None):
-        super().__init__()
-        img_size = to_2tuple(img_size)
-        patch_size = to_2tuple(patch_size)
-        patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]]
-        self.img_size = img_size
-        self.patch_size = patch_size
-        self.patches_resolution = patches_resolution
-        self.num_patches = patches_resolution[0] * patches_resolution[1]
-
-        self.in_chans = in_chans
-        self.embed_dim = embed_dim
-
-    def forward(self, x, x_size):
-        B, HW, C = x.shape
-        x = x.transpose(1, 2).view(B, self.embed_dim, x_size[0], x_size[1])  # B Ph*Pw C
-        return x
-
-    def flops(self):
-        flops = 0
-        return flops
-
-
-class Upsample(nn.Sequential):
-    """Upsample module.
-
-    Args:
-        scale (int): Scale factor. Supported scales: 2^n and 3.
-        num_feat (int): Channel number of intermediate features.
-    """
-
-    def __init__(self, scale, num_feat):
-        m = []
-        if (scale & (scale - 1)) == 0:  # scale = 2^n
-            for _ in range(int(math.log(scale, 2))):
-                m.append(nn.Conv2d(num_feat, 4 * num_feat, 3, 1, 1))
-                m.append(nn.PixelShuffle(2))
-        elif scale == 3:
-            m.append(nn.Conv2d(num_feat, 9 * num_feat, 3, 1, 1))
-            m.append(nn.PixelShuffle(3))
-        else:
-            raise ValueError(f'scale {scale} is not supported. ' 'Supported scales: 2^n and 3.')
-        super(Upsample, self).__init__(*m)
-
-
-class UpsampleOneStep(nn.Sequential):
-    """UpsampleOneStep module (the difference with Upsample is that it always only has 1conv + 1pixelshuffle)
-       Used in lightweight SR to save parameters.
-
-    Args:
-        scale (int): Scale factor. Supported scales: 2^n and 3.
-        num_feat (int): Channel number of intermediate features.
-
-    """
-
-    def __init__(self, scale, num_feat, num_out_ch, input_resolution=None):
-        self.num_feat = num_feat
-        self.input_resolution = input_resolution
-        m = []
-        m.append(nn.Conv2d(num_feat, (scale ** 2) * num_out_ch, 3, 1, 1))
-        m.append(nn.PixelShuffle(scale))
-        super(UpsampleOneStep, self).__init__(*m)
-
-    def flops(self):
-        H, W = self.input_resolution
-        flops = H * W * self.num_feat * 3 * 9
-        return flops
-
-
-class SwinIR(nn.Module):
-    r""" SwinIR
-        A PyTorch impl of : `SwinIR: Image Restoration Using Swin Transformer`, based on Swin Transformer.
-
-    Args:
-        img_size (int | tuple(int)): Input image size. Default 64
-        patch_size (int | tuple(int)): Patch size. Default: 1
-        in_chans (int): Number of input image channels. Default: 3
-        embed_dim (int): Patch embedding dimension. Default: 96
-        depths (tuple(int)): Depth of each Swin Transformer layer.
-        num_heads (tuple(int)): Number of attention heads in different layers.
-        window_size (int): Window size. Default: 7
-        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4
-        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: True
-        qk_scale (float): Override default qk scale of head_dim ** -0.5 if set. Default: None
-        drop_rate (float): Dropout rate. Default: 0
-        attn_drop_rate (float): Attention dropout rate. Default: 0
-        drop_path_rate (float): Stochastic depth rate. Default: 0.1
-        norm_layer (nn.Module): Normalization layer. Default: nn.LayerNorm.
-        ape (bool): If True, add absolute position embedding to the patch embedding. Default: False
-        patch_norm (bool): If True, add normalization after patch embedding. Default: True
-        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False
-        upscale: Upscale factor. 2/3/4/8 for image SR, 1 for denoising and compress artifact reduction
-        img_range: Image range. 1. or 255.
-        upsampler: The reconstruction reconstruction module. 'pixelshuffle'/'pixelshuffledirect'/'nearest+conv'/None
-        resi_connection: The convolutional block before residual connection. '1conv'/'3conv'
-    """
-
-    def __init__(self, img_size=64, patch_size=1, in_chans=3,
-                 embed_dim=96, depths=(6, 6, 6, 6), num_heads=(6, 6, 6, 6),
-                 window_size=7, mlp_ratio=4., qkv_bias=True, qk_scale=None,
-                 drop_rate=0., attn_drop_rate=0., drop_path_rate=0.1,
-                 norm_layer=nn.LayerNorm, ape=False, patch_norm=True,
-                 use_checkpoint=False, upscale=2, img_range=1., upsampler='', resi_connection='1conv',
-                 **kwargs):
-        super(SwinIR, self).__init__()
-        num_in_ch = in_chans
-        num_out_ch = in_chans
-        num_feat = 64
-        self.img_range = img_range
-        if in_chans == 3:
-            rgb_mean = (0.4488, 0.4371, 0.4040)
-            self.mean = torch.Tensor(rgb_mean).view(1, 3, 1, 1)
-        else:
-            self.mean = torch.zeros(1, 1, 1, 1)
-        self.upscale = upscale
-        self.upsampler = upsampler
-        self.window_size = window_size
-
-        #####################################################################################################
-        ################################### 1, shallow feature extraction ###################################
-        self.conv_first = nn.Conv2d(num_in_ch, embed_dim, 3, 1, 1)
-
-        #####################################################################################################
-        ################################### 2, deep feature extraction ######################################
-        self.num_layers = len(depths)
-        self.embed_dim = embed_dim
-        self.ape = ape
-        self.patch_norm = patch_norm
-        self.num_features = embed_dim
-        self.mlp_ratio = mlp_ratio
-
-        # split image into non-overlapping patches
-        self.patch_embed = PatchEmbed(
-            img_size=img_size, patch_size=patch_size, in_chans=embed_dim, embed_dim=embed_dim,
-            norm_layer=norm_layer if self.patch_norm else None)
-        num_patches = self.patch_embed.num_patches
-        patches_resolution = self.patch_embed.patches_resolution
-        self.patches_resolution = patches_resolution
-
-        # merge non-overlapping patches into image
-        self.patch_unembed = PatchUnEmbed(
-            img_size=img_size, patch_size=patch_size, in_chans=embed_dim, embed_dim=embed_dim,
-            norm_layer=norm_layer if self.patch_norm else None)
-
-        # absolute position embedding
-        if self.ape:
-            self.absolute_pos_embed = nn.Parameter(torch.zeros(1, num_patches, embed_dim))
-            trunc_normal_(self.absolute_pos_embed, std=.02)
-
-        self.pos_drop = nn.Dropout(p=drop_rate)
-
-        # stochastic depth
-        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))]  # stochastic depth decay rule
-
-        # build Residual Swin Transformer blocks (RSTB)
-        self.layers = nn.ModuleList()
-        for i_layer in range(self.num_layers):
-            layer = RSTB(dim=embed_dim,
-                         input_resolution=(patches_resolution[0],
-                                           patches_resolution[1]),
-                         depth=depths[i_layer],
-                         num_heads=num_heads[i_layer],
-                         window_size=window_size,
-                         mlp_ratio=self.mlp_ratio,
-                         qkv_bias=qkv_bias, qk_scale=qk_scale,
-                         drop=drop_rate, attn_drop=attn_drop_rate,
-                         drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])],  # no impact on SR results
-                         norm_layer=norm_layer,
-                         downsample=None,
-                         use_checkpoint=use_checkpoint,
-                         img_size=img_size,
-                         patch_size=patch_size,
-                         resi_connection=resi_connection
-
-                         )
-            self.layers.append(layer)
-        self.norm = norm_layer(self.num_features)
-
-        # build the last conv layer in deep feature extraction
-        if resi_connection == '1conv':
-            self.conv_after_body = nn.Conv2d(embed_dim, embed_dim, 3, 1, 1)
-        elif resi_connection == '3conv':
-            # to save parameters and memory
-            self.conv_after_body = nn.Sequential(nn.Conv2d(embed_dim, embed_dim // 4, 3, 1, 1),
-                                                 nn.LeakyReLU(negative_slope=0.2, inplace=True),
-                                                 nn.Conv2d(embed_dim // 4, embed_dim // 4, 1, 1, 0),
-                                                 nn.LeakyReLU(negative_slope=0.2, inplace=True),
-                                                 nn.Conv2d(embed_dim // 4, embed_dim, 3, 1, 1))
-
-        #####################################################################################################
-        ################################ 3, high quality image reconstruction ################################
-        if self.upsampler == 'pixelshuffle':
-            # for classical SR
-            self.conv_before_upsample = nn.Sequential(nn.Conv2d(embed_dim, num_feat, 3, 1, 1),
-                                                      nn.LeakyReLU(inplace=True))
-            self.upsample = Upsample(upscale, num_feat)
-            self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1)
-        elif self.upsampler == 'pixelshuffledirect':
-            # for lightweight SR (to save parameters)
-            self.upsample = UpsampleOneStep(upscale, embed_dim, num_out_ch,
-                                            (patches_resolution[0], patches_resolution[1]))
-        elif self.upsampler == 'nearest+conv':
-            # for real-world SR (less artifacts)
-            self.conv_before_upsample = nn.Sequential(nn.Conv2d(embed_dim, num_feat, 3, 1, 1),
-                                                      nn.LeakyReLU(inplace=True))
-            self.conv_up1 = nn.Conv2d(num_feat, num_feat, 3, 1, 1)
-            if self.upscale == 4:
-                self.conv_up2 = nn.Conv2d(num_feat, num_feat, 3, 1, 1)
-            self.conv_hr = nn.Conv2d(num_feat, num_feat, 3, 1, 1)
-            self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1)
-            self.lrelu = nn.LeakyReLU(negative_slope=0.2, inplace=True)
-        else:
-            # for image denoising and JPEG compression artifact reduction
-            self.conv_last = nn.Conv2d(embed_dim, num_out_ch, 3, 1, 1)
-
-        self.apply(self._init_weights)
-
-    def _init_weights(self, m):
-        if isinstance(m, nn.Linear):
-            trunc_normal_(m.weight, std=.02)
-            if isinstance(m, nn.Linear) and m.bias is not None:
-                nn.init.constant_(m.bias, 0)
-        elif isinstance(m, nn.LayerNorm):
-            nn.init.constant_(m.bias, 0)
-            nn.init.constant_(m.weight, 1.0)
-
-    @torch.jit.ignore
-    def no_weight_decay(self):
-        return {'absolute_pos_embed'}
-
-    @torch.jit.ignore
-    def no_weight_decay_keywords(self):
-        return {'relative_position_bias_table'}
-
-    def check_image_size(self, x):
-        _, _, h, w = x.size()
-        mod_pad_h = (self.window_size - h % self.window_size) % self.window_size
-        mod_pad_w = (self.window_size - w % self.window_size) % self.window_size
-        x = F.pad(x, (0, mod_pad_w, 0, mod_pad_h), 'reflect')
-        return x
-
-    def forward_features(self, x):
-        x_size = (x.shape[2], x.shape[3])
-        x = self.patch_embed(x)
-        if self.ape:
-            x = x + self.absolute_pos_embed
-        x = self.pos_drop(x)
-
-        for layer in self.layers:
-            x = layer(x, x_size)
-
-        x = self.norm(x)  # B L C
-        x = self.patch_unembed(x, x_size)
-
-        return x
-
-    def forward(self, x):
-        H, W = x.shape[2:]
-        x = self.check_image_size(x)
-
-        self.mean = self.mean.type_as(x)
-        x = (x - self.mean) * self.img_range
-
-        if self.upsampler == 'pixelshuffle':
-            # for classical SR
-            x = self.conv_first(x)
-            x = self.conv_after_body(self.forward_features(x)) + x
-            x = self.conv_before_upsample(x)
-            x = self.conv_last(self.upsample(x))
-        elif self.upsampler == 'pixelshuffledirect':
-            # for lightweight SR
-            x = self.conv_first(x)
-            x = self.conv_after_body(self.forward_features(x)) + x
-            x = self.upsample(x)
-        elif self.upsampler == 'nearest+conv':
-            # for real-world SR
-            x = self.conv_first(x)
-            x = self.conv_after_body(self.forward_features(x)) + x
-            x = self.conv_before_upsample(x)
-            x = self.lrelu(self.conv_up1(torch.nn.functional.interpolate(x, scale_factor=2, mode='nearest')))
-            if self.upscale == 4:
-                x = self.lrelu(self.conv_up2(torch.nn.functional.interpolate(x, scale_factor=2, mode='nearest')))
-            x = self.conv_last(self.lrelu(self.conv_hr(x)))
-        else:
-            # for image denoising and JPEG compression artifact reduction
-            x_first = self.conv_first(x)
-            res = self.conv_after_body(self.forward_features(x_first)) + x_first
-            x = x + self.conv_last(res)
-
-        x = x / self.img_range + self.mean
-
-        return x[:, :, :H*self.upscale, :W*self.upscale]
-
-    def flops(self):
-        flops = 0
-        H, W = self.patches_resolution
-        flops += H * W * 3 * self.embed_dim * 9
-        flops += self.patch_embed.flops()
-        for layer in self.layers:
-            flops += layer.flops()
-        flops += H * W * 3 * self.embed_dim * self.embed_dim
-        flops += self.upsample.flops()
-        return flops
-
-
-if __name__ == '__main__':
-    upscale = 4
-    window_size = 8
-    height = (1024 // upscale // window_size + 1) * window_size
-    width = (720 // upscale // window_size + 1) * window_size
-    model = SwinIR(upscale=2, img_size=(height, width),
-                   window_size=window_size, img_range=1., depths=[6, 6, 6, 6],
-                   embed_dim=60, num_heads=[6, 6, 6, 6], mlp_ratio=2, upsampler='pixelshuffledirect')
-    print(model)
-    print(height, width, model.flops() / 1e9)
-
-    x = torch.randn((1, 3, height, width))
-    x = model(x)
-    print(x.shape)
diff --git a/extensions-builtin/SwinIR/swinir_model_arch_v2.py b/extensions-builtin/SwinIR/swinir_model_arch_v2.py
deleted file mode 100644
index dad22cca29e..00000000000
--- a/extensions-builtin/SwinIR/swinir_model_arch_v2.py
+++ /dev/null
@@ -1,1017 +0,0 @@
-# -----------------------------------------------------------------------------------
-# Swin2SR: Swin2SR: SwinV2 Transformer for Compressed Image Super-Resolution and Restoration, https://arxiv.org/abs/
-# Written by Conde and Choi et al.
-# -----------------------------------------------------------------------------------
-
-import math
-import numpy as np
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-import torch.utils.checkpoint as checkpoint
-from timm.models.layers import DropPath, to_2tuple, trunc_normal_
-
-
-class Mlp(nn.Module):
-    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
-        super().__init__()
-        out_features = out_features or in_features
-        hidden_features = hidden_features or in_features
-        self.fc1 = nn.Linear(in_features, hidden_features)
-        self.act = act_layer()
-        self.fc2 = nn.Linear(hidden_features, out_features)
-        self.drop = nn.Dropout(drop)
-
-    def forward(self, x):
-        x = self.fc1(x)
-        x = self.act(x)
-        x = self.drop(x)
-        x = self.fc2(x)
-        x = self.drop(x)
-        return x
-
-
-def window_partition(x, window_size):
-    """
-    Args:
-        x: (B, H, W, C)
-        window_size (int): window size
-    Returns:
-        windows: (num_windows*B, window_size, window_size, C)
-    """
-    B, H, W, C = x.shape
-    x = x.view(B, H // window_size, window_size, W // window_size, window_size, C)
-    windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C)
-    return windows
-
-
-def window_reverse(windows, window_size, H, W):
-    """
-    Args:
-        windows: (num_windows*B, window_size, window_size, C)
-        window_size (int): Window size
-        H (int): Height of image
-        W (int): Width of image
-    Returns:
-        x: (B, H, W, C)
-    """
-    B = int(windows.shape[0] / (H * W / window_size / window_size))
-    x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1)
-    x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1)
-    return x
-
-class WindowAttention(nn.Module):
-    r""" Window based multi-head self attention (W-MSA) module with relative position bias.
-    It supports both of shifted and non-shifted window.
-    Args:
-        dim (int): Number of input channels.
-        window_size (tuple[int]): The height and width of the window.
-        num_heads (int): Number of attention heads.
-        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True
-        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0
-        proj_drop (float, optional): Dropout ratio of output. Default: 0.0
-        pretrained_window_size (tuple[int]): The height and width of the window in pre-training.
-    """
-
-    def __init__(self, dim, window_size, num_heads, qkv_bias=True, attn_drop=0., proj_drop=0.,
-                 pretrained_window_size=(0, 0)):
-
-        super().__init__()
-        self.dim = dim
-        self.window_size = window_size  # Wh, Ww
-        self.pretrained_window_size = pretrained_window_size
-        self.num_heads = num_heads
-
-        self.logit_scale = nn.Parameter(torch.log(10 * torch.ones((num_heads, 1, 1))), requires_grad=True)
-
-        # mlp to generate continuous relative position bias
-        self.cpb_mlp = nn.Sequential(nn.Linear(2, 512, bias=True),
-                                     nn.ReLU(inplace=True),
-                                     nn.Linear(512, num_heads, bias=False))
-
-        # get relative_coords_table
-        relative_coords_h = torch.arange(-(self.window_size[0] - 1), self.window_size[0], dtype=torch.float32)
-        relative_coords_w = torch.arange(-(self.window_size[1] - 1), self.window_size[1], dtype=torch.float32)
-        relative_coords_table = torch.stack(
-            torch.meshgrid([relative_coords_h,
-                            relative_coords_w])).permute(1, 2, 0).contiguous().unsqueeze(0)  # 1, 2*Wh-1, 2*Ww-1, 2
-        if pretrained_window_size[0] > 0:
-            relative_coords_table[:, :, :, 0] /= (pretrained_window_size[0] - 1)
-            relative_coords_table[:, :, :, 1] /= (pretrained_window_size[1] - 1)
-        else:
-            relative_coords_table[:, :, :, 0] /= (self.window_size[0] - 1)
-            relative_coords_table[:, :, :, 1] /= (self.window_size[1] - 1)
-        relative_coords_table *= 8  # normalize to -8, 8
-        relative_coords_table = torch.sign(relative_coords_table) * torch.log2(
-            torch.abs(relative_coords_table) + 1.0) / np.log2(8)
-
-        self.register_buffer("relative_coords_table", relative_coords_table)
-
-        # get pair-wise relative position index for each token inside the window
-        coords_h = torch.arange(self.window_size[0])
-        coords_w = torch.arange(self.window_size[1])
-        coords = torch.stack(torch.meshgrid([coords_h, coords_w]))  # 2, Wh, Ww
-        coords_flatten = torch.flatten(coords, 1)  # 2, Wh*Ww
-        relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :]  # 2, Wh*Ww, Wh*Ww
-        relative_coords = relative_coords.permute(1, 2, 0).contiguous()  # Wh*Ww, Wh*Ww, 2
-        relative_coords[:, :, 0] += self.window_size[0] - 1  # shift to start from 0
-        relative_coords[:, :, 1] += self.window_size[1] - 1
-        relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1
-        relative_position_index = relative_coords.sum(-1)  # Wh*Ww, Wh*Ww
-        self.register_buffer("relative_position_index", relative_position_index)
-
-        self.qkv = nn.Linear(dim, dim * 3, bias=False)
-        if qkv_bias:
-            self.q_bias = nn.Parameter(torch.zeros(dim))
-            self.v_bias = nn.Parameter(torch.zeros(dim))
-        else:
-            self.q_bias = None
-            self.v_bias = None
-        self.attn_drop = nn.Dropout(attn_drop)
-        self.proj = nn.Linear(dim, dim)
-        self.proj_drop = nn.Dropout(proj_drop)
-        self.softmax = nn.Softmax(dim=-1)
-
-    def forward(self, x, mask=None):
-        """
-        Args:
-            x: input features with shape of (num_windows*B, N, C)
-            mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None
-        """
-        B_, N, C = x.shape
-        qkv_bias = None
-        if self.q_bias is not None:
-            qkv_bias = torch.cat((self.q_bias, torch.zeros_like(self.v_bias, requires_grad=False), self.v_bias))
-        qkv = F.linear(input=x, weight=self.qkv.weight, bias=qkv_bias)
-        qkv = qkv.reshape(B_, N, 3, self.num_heads, -1).permute(2, 0, 3, 1, 4)
-        q, k, v = qkv[0], qkv[1], qkv[2]  # make torchscript happy (cannot use tensor as tuple)
-
-        # cosine attention
-        attn = (F.normalize(q, dim=-1) @ F.normalize(k, dim=-1).transpose(-2, -1))
-        logit_scale = torch.clamp(self.logit_scale, max=torch.log(torch.tensor(1. / 0.01)).to(self.logit_scale.device)).exp()
-        attn = attn * logit_scale
-
-        relative_position_bias_table = self.cpb_mlp(self.relative_coords_table).view(-1, self.num_heads)
-        relative_position_bias = relative_position_bias_table[self.relative_position_index.view(-1)].view(
-            self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1)  # Wh*Ww,Wh*Ww,nH
-        relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww
-        relative_position_bias = 16 * torch.sigmoid(relative_position_bias)
-        attn = attn + relative_position_bias.unsqueeze(0)
-
-        if mask is not None:
-            nW = mask.shape[0]
-            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0)
-            attn = attn.view(-1, self.num_heads, N, N)
-            attn = self.softmax(attn)
-        else:
-            attn = self.softmax(attn)
-
-        attn = self.attn_drop(attn)
-
-        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
-        x = self.proj(x)
-        x = self.proj_drop(x)
-        return x
-
-    def extra_repr(self) -> str:
-        return f'dim={self.dim}, window_size={self.window_size}, ' \
-               f'pretrained_window_size={self.pretrained_window_size}, num_heads={self.num_heads}'
-
-    def flops(self, N):
-        # calculate flops for 1 window with token length of N
-        flops = 0
-        # qkv = self.qkv(x)
-        flops += N * self.dim * 3 * self.dim
-        # attn = (q @ k.transpose(-2, -1))
-        flops += self.num_heads * N * (self.dim // self.num_heads) * N
-        #  x = (attn @ v)
-        flops += self.num_heads * N * N * (self.dim // self.num_heads)
-        # x = self.proj(x)
-        flops += N * self.dim * self.dim
-        return flops
-
-class SwinTransformerBlock(nn.Module):
-    r""" Swin Transformer Block.
-    Args:
-        dim (int): Number of input channels.
-        input_resolution (tuple[int]): Input resulotion.
-        num_heads (int): Number of attention heads.
-        window_size (int): Window size.
-        shift_size (int): Shift size for SW-MSA.
-        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
-        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
-        drop (float, optional): Dropout rate. Default: 0.0
-        attn_drop (float, optional): Attention dropout rate. Default: 0.0
-        drop_path (float, optional): Stochastic depth rate. Default: 0.0
-        act_layer (nn.Module, optional): Activation layer. Default: nn.GELU
-        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
-        pretrained_window_size (int): Window size in pre-training.
-    """
-
-    def __init__(self, dim, input_resolution, num_heads, window_size=7, shift_size=0,
-                 mlp_ratio=4., qkv_bias=True, drop=0., attn_drop=0., drop_path=0.,
-                 act_layer=nn.GELU, norm_layer=nn.LayerNorm, pretrained_window_size=0):
-        super().__init__()
-        self.dim = dim
-        self.input_resolution = input_resolution
-        self.num_heads = num_heads
-        self.window_size = window_size
-        self.shift_size = shift_size
-        self.mlp_ratio = mlp_ratio
-        if min(self.input_resolution) <= self.window_size:
-            # if window size is larger than input resolution, we don't partition windows
-            self.shift_size = 0
-            self.window_size = min(self.input_resolution)
-        assert 0 <= self.shift_size < self.window_size, "shift_size must in 0-window_size"
-
-        self.norm1 = norm_layer(dim)
-        self.attn = WindowAttention(
-            dim, window_size=to_2tuple(self.window_size), num_heads=num_heads,
-            qkv_bias=qkv_bias, attn_drop=attn_drop, proj_drop=drop,
-            pretrained_window_size=to_2tuple(pretrained_window_size))
-
-        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
-        self.norm2 = norm_layer(dim)
-        mlp_hidden_dim = int(dim * mlp_ratio)
-        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)
-
-        if self.shift_size > 0:
-            attn_mask = self.calculate_mask(self.input_resolution)
-        else:
-            attn_mask = None
-
-        self.register_buffer("attn_mask", attn_mask)
-
-    def calculate_mask(self, x_size):
-        # calculate attention mask for SW-MSA
-        H, W = x_size
-        img_mask = torch.zeros((1, H, W, 1))  # 1 H W 1
-        h_slices = (slice(0, -self.window_size),
-                    slice(-self.window_size, -self.shift_size),
-                    slice(-self.shift_size, None))
-        w_slices = (slice(0, -self.window_size),
-                    slice(-self.window_size, -self.shift_size),
-                    slice(-self.shift_size, None))
-        cnt = 0
-        for h in h_slices:
-            for w in w_slices:
-                img_mask[:, h, w, :] = cnt
-                cnt += 1
-
-        mask_windows = window_partition(img_mask, self.window_size)  # nW, window_size, window_size, 1
-        mask_windows = mask_windows.view(-1, self.window_size * self.window_size)
-        attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
-        attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))
-
-        return attn_mask
-
-    def forward(self, x, x_size):
-        H, W = x_size
-        B, L, C = x.shape
-        #assert L == H * W, "input feature has wrong size"
-
-        shortcut = x
-        x = x.view(B, H, W, C)
-
-        # cyclic shift
-        if self.shift_size > 0:
-            shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
-        else:
-            shifted_x = x
-
-        # partition windows
-        x_windows = window_partition(shifted_x, self.window_size)  # nW*B, window_size, window_size, C
-        x_windows = x_windows.view(-1, self.window_size * self.window_size, C)  # nW*B, window_size*window_size, C
-
-        # W-MSA/SW-MSA (to be compatible for testing on images whose shapes are the multiple of window size
-        if self.input_resolution == x_size:
-            attn_windows = self.attn(x_windows, mask=self.attn_mask)  # nW*B, window_size*window_size, C
-        else:
-            attn_windows = self.attn(x_windows, mask=self.calculate_mask(x_size).to(x.device))
-
-        # merge windows
-        attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C)
-        shifted_x = window_reverse(attn_windows, self.window_size, H, W)  # B H' W' C
-
-        # reverse cyclic shift
-        if self.shift_size > 0:
-            x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
-        else:
-            x = shifted_x
-        x = x.view(B, H * W, C)
-        x = shortcut + self.drop_path(self.norm1(x))
-
-        # FFN
-        x = x + self.drop_path(self.norm2(self.mlp(x)))
-
-        return x
-
-    def extra_repr(self) -> str:
-        return f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, " \
-               f"window_size={self.window_size}, shift_size={self.shift_size}, mlp_ratio={self.mlp_ratio}"
-
-    def flops(self):
-        flops = 0
-        H, W = self.input_resolution
-        # norm1
-        flops += self.dim * H * W
-        # W-MSA/SW-MSA
-        nW = H * W / self.window_size / self.window_size
-        flops += nW * self.attn.flops(self.window_size * self.window_size)
-        # mlp
-        flops += 2 * H * W * self.dim * self.dim * self.mlp_ratio
-        # norm2
-        flops += self.dim * H * W
-        return flops
-
-class PatchMerging(nn.Module):
-    r""" Patch Merging Layer.
-    Args:
-        input_resolution (tuple[int]): Resolution of input feature.
-        dim (int): Number of input channels.
-        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
-    """
-
-    def __init__(self, input_resolution, dim, norm_layer=nn.LayerNorm):
-        super().__init__()
-        self.input_resolution = input_resolution
-        self.dim = dim
-        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
-        self.norm = norm_layer(2 * dim)
-
-    def forward(self, x):
-        """
-        x: B, H*W, C
-        """
-        H, W = self.input_resolution
-        B, L, C = x.shape
-        assert L == H * W, "input feature has wrong size"
-        assert H % 2 == 0 and W % 2 == 0, f"x size ({H}*{W}) are not even."
-
-        x = x.view(B, H, W, C)
-
-        x0 = x[:, 0::2, 0::2, :]  # B H/2 W/2 C
-        x1 = x[:, 1::2, 0::2, :]  # B H/2 W/2 C
-        x2 = x[:, 0::2, 1::2, :]  # B H/2 W/2 C
-        x3 = x[:, 1::2, 1::2, :]  # B H/2 W/2 C
-        x = torch.cat([x0, x1, x2, x3], -1)  # B H/2 W/2 4*C
-        x = x.view(B, -1, 4 * C)  # B H/2*W/2 4*C
-
-        x = self.reduction(x)
-        x = self.norm(x)
-
-        return x
-
-    def extra_repr(self) -> str:
-        return f"input_resolution={self.input_resolution}, dim={self.dim}"
-
-    def flops(self):
-        H, W = self.input_resolution
-        flops = (H // 2) * (W // 2) * 4 * self.dim * 2 * self.dim
-        flops += H * W * self.dim // 2
-        return flops
-
-class BasicLayer(nn.Module):
-    """ A basic Swin Transformer layer for one stage.
-    Args:
-        dim (int): Number of input channels.
-        input_resolution (tuple[int]): Input resolution.
-        depth (int): Number of blocks.
-        num_heads (int): Number of attention heads.
-        window_size (int): Local window size.
-        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
-        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
-        drop (float, optional): Dropout rate. Default: 0.0
-        attn_drop (float, optional): Attention dropout rate. Default: 0.0
-        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
-        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
-        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
-        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.
-        pretrained_window_size (int): Local window size in pre-training.
-    """
-
-    def __init__(self, dim, input_resolution, depth, num_heads, window_size,
-                 mlp_ratio=4., qkv_bias=True, drop=0., attn_drop=0.,
-                 drop_path=0., norm_layer=nn.LayerNorm, downsample=None, use_checkpoint=False,
-                 pretrained_window_size=0):
-
-        super().__init__()
-        self.dim = dim
-        self.input_resolution = input_resolution
-        self.depth = depth
-        self.use_checkpoint = use_checkpoint
-
-        # build blocks
-        self.blocks = nn.ModuleList([
-            SwinTransformerBlock(dim=dim, input_resolution=input_resolution,
-                                 num_heads=num_heads, window_size=window_size,
-                                 shift_size=0 if (i % 2 == 0) else window_size // 2,
-                                 mlp_ratio=mlp_ratio,
-                                 qkv_bias=qkv_bias,
-                                 drop=drop, attn_drop=attn_drop,
-                                 drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,
-                                 norm_layer=norm_layer,
-                                 pretrained_window_size=pretrained_window_size)
-            for i in range(depth)])
-
-        # patch merging layer
-        if downsample is not None:
-            self.downsample = downsample(input_resolution, dim=dim, norm_layer=norm_layer)
-        else:
-            self.downsample = None
-
-    def forward(self, x, x_size):
-        for blk in self.blocks:
-            if self.use_checkpoint:
-                x = checkpoint.checkpoint(blk, x, x_size)
-            else:
-                x = blk(x, x_size)
-        if self.downsample is not None:
-            x = self.downsample(x)
-        return x
-
-    def extra_repr(self) -> str:
-        return f"dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}"
-
-    def flops(self):
-        flops = 0
-        for blk in self.blocks:
-            flops += blk.flops()
-        if self.downsample is not None:
-            flops += self.downsample.flops()
-        return flops
-
-    def _init_respostnorm(self):
-        for blk in self.blocks:
-            nn.init.constant_(blk.norm1.bias, 0)
-            nn.init.constant_(blk.norm1.weight, 0)
-            nn.init.constant_(blk.norm2.bias, 0)
-            nn.init.constant_(blk.norm2.weight, 0)
-
-class PatchEmbed(nn.Module):
-    r""" Image to Patch Embedding
-    Args:
-        img_size (int): Image size.  Default: 224.
-        patch_size (int): Patch token size. Default: 4.
-        in_chans (int): Number of input image channels. Default: 3.
-        embed_dim (int): Number of linear projection output channels. Default: 96.
-        norm_layer (nn.Module, optional): Normalization layer. Default: None
-    """
-
-    def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None):
-        super().__init__()
-        img_size = to_2tuple(img_size)
-        patch_size = to_2tuple(patch_size)
-        patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]]
-        self.img_size = img_size
-        self.patch_size = patch_size
-        self.patches_resolution = patches_resolution
-        self.num_patches = patches_resolution[0] * patches_resolution[1]
-
-        self.in_chans = in_chans
-        self.embed_dim = embed_dim
-
-        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)
-        if norm_layer is not None:
-            self.norm = norm_layer(embed_dim)
-        else:
-            self.norm = None
-
-    def forward(self, x):
-        B, C, H, W = x.shape
-        # FIXME look at relaxing size constraints
-        # assert H == self.img_size[0] and W == self.img_size[1],
-        #     f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})."
-        x = self.proj(x).flatten(2).transpose(1, 2)  # B Ph*Pw C
-        if self.norm is not None:
-            x = self.norm(x)
-        return x
-
-    def flops(self):
-        Ho, Wo = self.patches_resolution
-        flops = Ho * Wo * self.embed_dim * self.in_chans * (self.patch_size[0] * self.patch_size[1])
-        if self.norm is not None:
-            flops += Ho * Wo * self.embed_dim
-        return flops
-
-class RSTB(nn.Module):
-    """Residual Swin Transformer Block (RSTB).
-
-    Args:
-        dim (int): Number of input channels.
-        input_resolution (tuple[int]): Input resolution.
-        depth (int): Number of blocks.
-        num_heads (int): Number of attention heads.
-        window_size (int): Local window size.
-        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
-        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
-        drop (float, optional): Dropout rate. Default: 0.0
-        attn_drop (float, optional): Attention dropout rate. Default: 0.0
-        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
-        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
-        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
-        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.
-        img_size: Input image size.
-        patch_size: Patch size.
-        resi_connection: The convolutional block before residual connection.
-    """
-
-    def __init__(self, dim, input_resolution, depth, num_heads, window_size,
-                 mlp_ratio=4., qkv_bias=True, drop=0., attn_drop=0.,
-                 drop_path=0., norm_layer=nn.LayerNorm, downsample=None, use_checkpoint=False,
-                 img_size=224, patch_size=4, resi_connection='1conv'):
-        super(RSTB, self).__init__()
-
-        self.dim = dim
-        self.input_resolution = input_resolution
-
-        self.residual_group = BasicLayer(dim=dim,
-                                         input_resolution=input_resolution,
-                                         depth=depth,
-                                         num_heads=num_heads,
-                                         window_size=window_size,
-                                         mlp_ratio=mlp_ratio,
-                                         qkv_bias=qkv_bias,
-                                         drop=drop, attn_drop=attn_drop,
-                                         drop_path=drop_path,
-                                         norm_layer=norm_layer,
-                                         downsample=downsample,
-                                         use_checkpoint=use_checkpoint)
-
-        if resi_connection == '1conv':
-            self.conv = nn.Conv2d(dim, dim, 3, 1, 1)
-        elif resi_connection == '3conv':
-            # to save parameters and memory
-            self.conv = nn.Sequential(nn.Conv2d(dim, dim // 4, 3, 1, 1), nn.LeakyReLU(negative_slope=0.2, inplace=True),
-                                      nn.Conv2d(dim // 4, dim // 4, 1, 1, 0),
-                                      nn.LeakyReLU(negative_slope=0.2, inplace=True),
-                                      nn.Conv2d(dim // 4, dim, 3, 1, 1))
-
-        self.patch_embed = PatchEmbed(
-            img_size=img_size, patch_size=patch_size, in_chans=dim, embed_dim=dim,
-            norm_layer=None)
-
-        self.patch_unembed = PatchUnEmbed(
-            img_size=img_size, patch_size=patch_size, in_chans=dim, embed_dim=dim,
-            norm_layer=None)
-
-    def forward(self, x, x_size):
-        return self.patch_embed(self.conv(self.patch_unembed(self.residual_group(x, x_size), x_size))) + x
-
-    def flops(self):
-        flops = 0
-        flops += self.residual_group.flops()
-        H, W = self.input_resolution
-        flops += H * W * self.dim * self.dim * 9
-        flops += self.patch_embed.flops()
-        flops += self.patch_unembed.flops()
-
-        return flops
-
-class PatchUnEmbed(nn.Module):
-    r""" Image to Patch Unembedding
-
-    Args:
-        img_size (int): Image size.  Default: 224.
-        patch_size (int): Patch token size. Default: 4.
-        in_chans (int): Number of input image channels. Default: 3.
-        embed_dim (int): Number of linear projection output channels. Default: 96.
-        norm_layer (nn.Module, optional): Normalization layer. Default: None
-    """
-
-    def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None):
-        super().__init__()
-        img_size = to_2tuple(img_size)
-        patch_size = to_2tuple(patch_size)
-        patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]]
-        self.img_size = img_size
-        self.patch_size = patch_size
-        self.patches_resolution = patches_resolution
-        self.num_patches = patches_resolution[0] * patches_resolution[1]
-
-        self.in_chans = in_chans
-        self.embed_dim = embed_dim
-
-    def forward(self, x, x_size):
-        B, HW, C = x.shape
-        x = x.transpose(1, 2).view(B, self.embed_dim, x_size[0], x_size[1])  # B Ph*Pw C
-        return x
-
-    def flops(self):
-        flops = 0
-        return flops
-
-
-class Upsample(nn.Sequential):
-    """Upsample module.
-
-    Args:
-        scale (int): Scale factor. Supported scales: 2^n and 3.
-        num_feat (int): Channel number of intermediate features.
-    """
-
-    def __init__(self, scale, num_feat):
-        m = []
-        if (scale & (scale - 1)) == 0:  # scale = 2^n
-            for _ in range(int(math.log(scale, 2))):
-                m.append(nn.Conv2d(num_feat, 4 * num_feat, 3, 1, 1))
-                m.append(nn.PixelShuffle(2))
-        elif scale == 3:
-            m.append(nn.Conv2d(num_feat, 9 * num_feat, 3, 1, 1))
-            m.append(nn.PixelShuffle(3))
-        else:
-            raise ValueError(f'scale {scale} is not supported. ' 'Supported scales: 2^n and 3.')
-        super(Upsample, self).__init__(*m)
-
-class Upsample_hf(nn.Sequential):
-    """Upsample module.
-
-    Args:
-        scale (int): Scale factor. Supported scales: 2^n and 3.
-        num_feat (int): Channel number of intermediate features.
-    """
-
-    def __init__(self, scale, num_feat):
-        m = []
-        if (scale & (scale - 1)) == 0:  # scale = 2^n
-            for _ in range(int(math.log(scale, 2))):
-                m.append(nn.Conv2d(num_feat, 4 * num_feat, 3, 1, 1))
-                m.append(nn.PixelShuffle(2))
-        elif scale == 3:
-            m.append(nn.Conv2d(num_feat, 9 * num_feat, 3, 1, 1))
-            m.append(nn.PixelShuffle(3))
-        else:
-            raise ValueError(f'scale {scale} is not supported. ' 'Supported scales: 2^n and 3.')
-        super(Upsample_hf, self).__init__(*m)
-
-
-class UpsampleOneStep(nn.Sequential):
-    """UpsampleOneStep module (the difference with Upsample is that it always only has 1conv + 1pixelshuffle)
-       Used in lightweight SR to save parameters.
-
-    Args:
-        scale (int): Scale factor. Supported scales: 2^n and 3.
-        num_feat (int): Channel number of intermediate features.
-
-    """
-
-    def __init__(self, scale, num_feat, num_out_ch, input_resolution=None):
-        self.num_feat = num_feat
-        self.input_resolution = input_resolution
-        m = []
-        m.append(nn.Conv2d(num_feat, (scale ** 2) * num_out_ch, 3, 1, 1))
-        m.append(nn.PixelShuffle(scale))
-        super(UpsampleOneStep, self).__init__(*m)
-
-    def flops(self):
-        H, W = self.input_resolution
-        flops = H * W * self.num_feat * 3 * 9
-        return flops
-
-
-
-class Swin2SR(nn.Module):
-    r""" Swin2SR
-        A PyTorch impl of : `Swin2SR: SwinV2 Transformer for Compressed Image Super-Resolution and Restoration`.
-
-    Args:
-        img_size (int | tuple(int)): Input image size. Default 64
-        patch_size (int | tuple(int)): Patch size. Default: 1
-        in_chans (int): Number of input image channels. Default: 3
-        embed_dim (int): Patch embedding dimension. Default: 96
-        depths (tuple(int)): Depth of each Swin Transformer layer.
-        num_heads (tuple(int)): Number of attention heads in different layers.
-        window_size (int): Window size. Default: 7
-        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4
-        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: True
-        drop_rate (float): Dropout rate. Default: 0
-        attn_drop_rate (float): Attention dropout rate. Default: 0
-        drop_path_rate (float): Stochastic depth rate. Default: 0.1
-        norm_layer (nn.Module): Normalization layer. Default: nn.LayerNorm.
-        ape (bool): If True, add absolute position embedding to the patch embedding. Default: False
-        patch_norm (bool): If True, add normalization after patch embedding. Default: True
-        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False
-        upscale: Upscale factor. 2/3/4/8 for image SR, 1 for denoising and compress artifact reduction
-        img_range: Image range. 1. or 255.
-        upsampler: The reconstruction reconstruction module. 'pixelshuffle'/'pixelshuffledirect'/'nearest+conv'/None
-        resi_connection: The convolutional block before residual connection. '1conv'/'3conv'
-    """
-
-    def __init__(self, img_size=64, patch_size=1, in_chans=3,
-                 embed_dim=96, depths=(6, 6, 6, 6), num_heads=(6, 6, 6, 6),
-                 window_size=7, mlp_ratio=4., qkv_bias=True,
-                 drop_rate=0., attn_drop_rate=0., drop_path_rate=0.1,
-                 norm_layer=nn.LayerNorm, ape=False, patch_norm=True,
-                 use_checkpoint=False, upscale=2, img_range=1., upsampler='', resi_connection='1conv',
-                 **kwargs):
-        super(Swin2SR, self).__init__()
-        num_in_ch = in_chans
-        num_out_ch = in_chans
-        num_feat = 64
-        self.img_range = img_range
-        if in_chans == 3:
-            rgb_mean = (0.4488, 0.4371, 0.4040)
-            self.mean = torch.Tensor(rgb_mean).view(1, 3, 1, 1)
-        else:
-            self.mean = torch.zeros(1, 1, 1, 1)
-        self.upscale = upscale
-        self.upsampler = upsampler
-        self.window_size = window_size
-
-        #####################################################################################################
-        ################################### 1, shallow feature extraction ###################################
-        self.conv_first = nn.Conv2d(num_in_ch, embed_dim, 3, 1, 1)
-
-        #####################################################################################################
-        ################################### 2, deep feature extraction ######################################
-        self.num_layers = len(depths)
-        self.embed_dim = embed_dim
-        self.ape = ape
-        self.patch_norm = patch_norm
-        self.num_features = embed_dim
-        self.mlp_ratio = mlp_ratio
-
-        # split image into non-overlapping patches
-        self.patch_embed = PatchEmbed(
-            img_size=img_size, patch_size=patch_size, in_chans=embed_dim, embed_dim=embed_dim,
-            norm_layer=norm_layer if self.patch_norm else None)
-        num_patches = self.patch_embed.num_patches
-        patches_resolution = self.patch_embed.patches_resolution
-        self.patches_resolution = patches_resolution
-
-        # merge non-overlapping patches into image
-        self.patch_unembed = PatchUnEmbed(
-            img_size=img_size, patch_size=patch_size, in_chans=embed_dim, embed_dim=embed_dim,
-            norm_layer=norm_layer if self.patch_norm else None)
-
-        # absolute position embedding
-        if self.ape:
-            self.absolute_pos_embed = nn.Parameter(torch.zeros(1, num_patches, embed_dim))
-            trunc_normal_(self.absolute_pos_embed, std=.02)
-
-        self.pos_drop = nn.Dropout(p=drop_rate)
-
-        # stochastic depth
-        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))]  # stochastic depth decay rule
-
-        # build Residual Swin Transformer blocks (RSTB)
-        self.layers = nn.ModuleList()
-        for i_layer in range(self.num_layers):
-            layer = RSTB(dim=embed_dim,
-                         input_resolution=(patches_resolution[0],
-                                           patches_resolution[1]),
-                         depth=depths[i_layer],
-                         num_heads=num_heads[i_layer],
-                         window_size=window_size,
-                         mlp_ratio=self.mlp_ratio,
-                         qkv_bias=qkv_bias,
-                         drop=drop_rate, attn_drop=attn_drop_rate,
-                         drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])],  # no impact on SR results
-                         norm_layer=norm_layer,
-                         downsample=None,
-                         use_checkpoint=use_checkpoint,
-                         img_size=img_size,
-                         patch_size=patch_size,
-                         resi_connection=resi_connection
-
-                         )
-            self.layers.append(layer)
-
-        if self.upsampler == 'pixelshuffle_hf':
-            self.layers_hf = nn.ModuleList()
-            for i_layer in range(self.num_layers):
-                layer = RSTB(dim=embed_dim,
-                             input_resolution=(patches_resolution[0],
-                                               patches_resolution[1]),
-                             depth=depths[i_layer],
-                             num_heads=num_heads[i_layer],
-                             window_size=window_size,
-                             mlp_ratio=self.mlp_ratio,
-                             qkv_bias=qkv_bias,
-                             drop=drop_rate, attn_drop=attn_drop_rate,
-                             drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])],  # no impact on SR results
-                             norm_layer=norm_layer,
-                             downsample=None,
-                             use_checkpoint=use_checkpoint,
-                             img_size=img_size,
-                             patch_size=patch_size,
-                             resi_connection=resi_connection
-
-                             )
-                self.layers_hf.append(layer)
-
-        self.norm = norm_layer(self.num_features)
-
-        # build the last conv layer in deep feature extraction
-        if resi_connection == '1conv':
-            self.conv_after_body = nn.Conv2d(embed_dim, embed_dim, 3, 1, 1)
-        elif resi_connection == '3conv':
-            # to save parameters and memory
-            self.conv_after_body = nn.Sequential(nn.Conv2d(embed_dim, embed_dim // 4, 3, 1, 1),
-                                                 nn.LeakyReLU(negative_slope=0.2, inplace=True),
-                                                 nn.Conv2d(embed_dim // 4, embed_dim // 4, 1, 1, 0),
-                                                 nn.LeakyReLU(negative_slope=0.2, inplace=True),
-                                                 nn.Conv2d(embed_dim // 4, embed_dim, 3, 1, 1))
-
-        #####################################################################################################
-        ################################ 3, high quality image reconstruction ################################
-        if self.upsampler == 'pixelshuffle':
-            # for classical SR
-            self.conv_before_upsample = nn.Sequential(nn.Conv2d(embed_dim, num_feat, 3, 1, 1),
-                                                      nn.LeakyReLU(inplace=True))
-            self.upsample = Upsample(upscale, num_feat)
-            self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1)
-        elif self.upsampler == 'pixelshuffle_aux':
-            self.conv_bicubic = nn.Conv2d(num_in_ch, num_feat, 3, 1, 1)
-            self.conv_before_upsample = nn.Sequential(
-                nn.Conv2d(embed_dim, num_feat, 3, 1, 1),
-                nn.LeakyReLU(inplace=True))
-            self.conv_aux = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1)
-            self.conv_after_aux = nn.Sequential(
-                nn.Conv2d(3, num_feat, 3, 1, 1),
-                nn.LeakyReLU(inplace=True))
-            self.upsample = Upsample(upscale, num_feat)
-            self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1)
-
-        elif self.upsampler == 'pixelshuffle_hf':
-            self.conv_before_upsample = nn.Sequential(nn.Conv2d(embed_dim, num_feat, 3, 1, 1),
-                                                      nn.LeakyReLU(inplace=True))
-            self.upsample = Upsample(upscale, num_feat)
-            self.upsample_hf = Upsample_hf(upscale, num_feat)
-            self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1)
-            self.conv_first_hf = nn.Sequential(nn.Conv2d(num_feat, embed_dim, 3, 1, 1),
-                                                      nn.LeakyReLU(inplace=True))
-            self.conv_after_body_hf = nn.Conv2d(embed_dim, embed_dim, 3, 1, 1)
-            self.conv_before_upsample_hf = nn.Sequential(
-                nn.Conv2d(embed_dim, num_feat, 3, 1, 1),
-                nn.LeakyReLU(inplace=True))
-            self.conv_last_hf = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1)
-
-        elif self.upsampler == 'pixelshuffledirect':
-            # for lightweight SR (to save parameters)
-            self.upsample = UpsampleOneStep(upscale, embed_dim, num_out_ch,
-                                            (patches_resolution[0], patches_resolution[1]))
-        elif self.upsampler == 'nearest+conv':
-            # for real-world SR (less artifacts)
-            assert self.upscale == 4, 'only support x4 now.'
-            self.conv_before_upsample = nn.Sequential(nn.Conv2d(embed_dim, num_feat, 3, 1, 1),
-                                                      nn.LeakyReLU(inplace=True))
-            self.conv_up1 = nn.Conv2d(num_feat, num_feat, 3, 1, 1)
-            self.conv_up2 = nn.Conv2d(num_feat, num_feat, 3, 1, 1)
-            self.conv_hr = nn.Conv2d(num_feat, num_feat, 3, 1, 1)
-            self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1)
-            self.lrelu = nn.LeakyReLU(negative_slope=0.2, inplace=True)
-        else:
-            # for image denoising and JPEG compression artifact reduction
-            self.conv_last = nn.Conv2d(embed_dim, num_out_ch, 3, 1, 1)
-
-        self.apply(self._init_weights)
-
-    def _init_weights(self, m):
-        if isinstance(m, nn.Linear):
-            trunc_normal_(m.weight, std=.02)
-            if isinstance(m, nn.Linear) and m.bias is not None:
-                nn.init.constant_(m.bias, 0)
-        elif isinstance(m, nn.LayerNorm):
-            nn.init.constant_(m.bias, 0)
-            nn.init.constant_(m.weight, 1.0)
-
-    @torch.jit.ignore
-    def no_weight_decay(self):
-        return {'absolute_pos_embed'}
-
-    @torch.jit.ignore
-    def no_weight_decay_keywords(self):
-        return {'relative_position_bias_table'}
-
-    def check_image_size(self, x):
-        _, _, h, w = x.size()
-        mod_pad_h = (self.window_size - h % self.window_size) % self.window_size
-        mod_pad_w = (self.window_size - w % self.window_size) % self.window_size
-        x = F.pad(x, (0, mod_pad_w, 0, mod_pad_h), 'reflect')
-        return x
-
-    def forward_features(self, x):
-        x_size = (x.shape[2], x.shape[3])
-        x = self.patch_embed(x)
-        if self.ape:
-            x = x + self.absolute_pos_embed
-        x = self.pos_drop(x)
-
-        for layer in self.layers:
-            x = layer(x, x_size)
-
-        x = self.norm(x)  # B L C
-        x = self.patch_unembed(x, x_size)
-
-        return x
-
-    def forward_features_hf(self, x):
-        x_size = (x.shape[2], x.shape[3])
-        x = self.patch_embed(x)
-        if self.ape:
-            x = x + self.absolute_pos_embed
-        x = self.pos_drop(x)
-
-        for layer in self.layers_hf:
-            x = layer(x, x_size)
-
-        x = self.norm(x)  # B L C
-        x = self.patch_unembed(x, x_size)
-
-        return x
-
-    def forward(self, x):
-        H, W = x.shape[2:]
-        x = self.check_image_size(x)
-
-        self.mean = self.mean.type_as(x)
-        x = (x - self.mean) * self.img_range
-
-        if self.upsampler == 'pixelshuffle':
-            # for classical SR
-            x = self.conv_first(x)
-            x = self.conv_after_body(self.forward_features(x)) + x
-            x = self.conv_before_upsample(x)
-            x = self.conv_last(self.upsample(x))
-        elif self.upsampler == 'pixelshuffle_aux':
-            bicubic = F.interpolate(x, size=(H * self.upscale, W * self.upscale), mode='bicubic', align_corners=False)
-            bicubic = self.conv_bicubic(bicubic)
-            x = self.conv_first(x)
-            x = self.conv_after_body(self.forward_features(x)) + x
-            x = self.conv_before_upsample(x)
-            aux = self.conv_aux(x) # b, 3, LR_H, LR_W
-            x = self.conv_after_aux(aux)
-            x = self.upsample(x)[:, :, :H * self.upscale, :W * self.upscale] + bicubic[:, :, :H * self.upscale, :W * self.upscale]
-            x = self.conv_last(x)
-            aux = aux / self.img_range + self.mean
-        elif self.upsampler == 'pixelshuffle_hf':
-            # for classical SR with HF
-            x = self.conv_first(x)
-            x = self.conv_after_body(self.forward_features(x)) + x
-            x_before = self.conv_before_upsample(x)
-            x_out = self.conv_last(self.upsample(x_before))
-
-            x_hf = self.conv_first_hf(x_before)
-            x_hf = self.conv_after_body_hf(self.forward_features_hf(x_hf)) + x_hf
-            x_hf = self.conv_before_upsample_hf(x_hf)
-            x_hf = self.conv_last_hf(self.upsample_hf(x_hf))
-            x = x_out + x_hf
-            x_hf = x_hf / self.img_range + self.mean
-
-        elif self.upsampler == 'pixelshuffledirect':
-            # for lightweight SR
-            x = self.conv_first(x)
-            x = self.conv_after_body(self.forward_features(x)) + x
-            x = self.upsample(x)
-        elif self.upsampler == 'nearest+conv':
-            # for real-world SR
-            x = self.conv_first(x)
-            x = self.conv_after_body(self.forward_features(x)) + x
-            x = self.conv_before_upsample(x)
-            x = self.lrelu(self.conv_up1(torch.nn.functional.interpolate(x, scale_factor=2, mode='nearest')))
-            x = self.lrelu(self.conv_up2(torch.nn.functional.interpolate(x, scale_factor=2, mode='nearest')))
-            x = self.conv_last(self.lrelu(self.conv_hr(x)))
-        else:
-            # for image denoising and JPEG compression artifact reduction
-            x_first = self.conv_first(x)
-            res = self.conv_after_body(self.forward_features(x_first)) + x_first
-            x = x + self.conv_last(res)
-
-        x = x / self.img_range + self.mean
-        if self.upsampler == "pixelshuffle_aux":
-            return x[:, :, :H*self.upscale, :W*self.upscale], aux
-
-        elif self.upsampler == "pixelshuffle_hf":
-            x_out = x_out / self.img_range + self.mean
-            return x_out[:, :, :H*self.upscale, :W*self.upscale], x[:, :, :H*self.upscale, :W*self.upscale], x_hf[:, :, :H*self.upscale, :W*self.upscale]
-
-        else:
-            return x[:, :, :H*self.upscale, :W*self.upscale]
-
-    def flops(self):
-        flops = 0
-        H, W = self.patches_resolution
-        flops += H * W * 3 * self.embed_dim * 9
-        flops += self.patch_embed.flops()
-        for layer in self.layers:
-            flops += layer.flops()
-        flops += H * W * 3 * self.embed_dim * self.embed_dim
-        flops += self.upsample.flops()
-        return flops
-
-
-if __name__ == '__main__':
-    upscale = 4
-    window_size = 8
-    height = (1024 // upscale // window_size + 1) * window_size
-    width = (720 // upscale // window_size + 1) * window_size
-    model = Swin2SR(upscale=2, img_size=(height, width),
-                   window_size=window_size, img_range=1., depths=[6, 6, 6, 6],
-                   embed_dim=60, num_heads=[6, 6, 6, 6], mlp_ratio=2, upsampler='pixelshuffledirect')
-    print(model)
-    print(height, width, model.flops() / 1e9)
-
-    x = torch.randn((1, 3, height, width))
-    x = model(x)
-    print(x.shape)
diff --git a/extensions-builtin/canvas-zoom-and-pan/javascript/zoom.js b/extensions-builtin/canvas-zoom-and-pan/javascript/zoom.js
index 45c7600ac5f..7807f7f6185 100644
--- a/extensions-builtin/canvas-zoom-and-pan/javascript/zoom.js
+++ b/extensions-builtin/canvas-zoom-and-pan/javascript/zoom.js
@@ -29,6 +29,7 @@ onUiLoaded(async() => {
     });
 
     function getActiveTab(elements, all = false) {
+        if (!elements.img2imgTabs) return null;
         const tabs = elements.img2imgTabs.querySelectorAll("button");
 
         if (all) return tabs;
@@ -43,6 +44,7 @@ onUiLoaded(async() => {
     // Get tab ID
     function getTabId(elements) {
         const activeTab = getActiveTab(elements);
+        if (!activeTab) return null;
         return tabNameToElementId[activeTab.innerText];
     }
 
@@ -218,6 +220,8 @@ onUiLoaded(async() => {
         canvas_hotkey_fullscreen: "KeyS",
         canvas_hotkey_move: "KeyF",
         canvas_hotkey_overlap: "KeyO",
+        canvas_hotkey_shrink_brush: "KeyQ",
+        canvas_hotkey_grow_brush: "KeyW",
         canvas_disabled_functions: [],
         canvas_show_tooltip: true,
         canvas_auto_expand: true,
@@ -227,6 +231,8 @@ onUiLoaded(async() => {
     const functionMap = {
         "Zoom": "canvas_hotkey_zoom",
         "Adjust brush size": "canvas_hotkey_adjust",
+        "Hotkey shrink brush": "canvas_hotkey_shrink_brush",
+        "Hotkey enlarge brush": "canvas_hotkey_grow_brush",
         "Moving canvas": "canvas_hotkey_move",
         "Fullscreen": "canvas_hotkey_fullscreen",
         "Reset Zoom": "canvas_hotkey_reset",
@@ -248,6 +254,7 @@ onUiLoaded(async() => {
     let isMoving = false;
     let mouseX, mouseY;
     let activeElement;
+    let interactedWithAltKey = false;
 
     const elements = Object.fromEntries(
         Object.keys(elementIDs).map(id => [
@@ -273,7 +280,7 @@ onUiLoaded(async() => {
         const targetElement = gradioApp().querySelector(elemId);
 
         if (!targetElement) {
-            console.log("Element not found");
+            console.log("Element not found", elemId);
             return;
         }
 
@@ -288,7 +295,7 @@ onUiLoaded(async() => {
 
         // Create tooltip
         function createTooltip() {
-            const toolTipElemnt =
+            const toolTipElement =
                 targetElement.querySelector(".image-container");
             const tooltip = document.createElement("div");
             tooltip.className = "canvas-tooltip";
@@ -351,7 +358,7 @@ onUiLoaded(async() => {
             tooltip.appendChild(tooltipContent);
 
             // Add a hint element to the target element
-            toolTipElemnt.appendChild(tooltip);
+            toolTipElement.appendChild(tooltip);
         }
 
         //Show tool tip if setting enable
@@ -361,9 +368,9 @@ onUiLoaded(async() => {
 
         // In the course of research, it was found that the tag img is very harmful when zooming and creates white canvases. This hack allows you to almost never think about this problem, it has no effect on webui.
         function fixCanvas() {
-            const activeTab = getActiveTab(elements).textContent.trim();
+            const activeTab = getActiveTab(elements)?.textContent.trim();
 
-            if (activeTab !== "img2img") {
+            if (activeTab && activeTab !== "img2img") {
                 const img = targetElement.querySelector(`${elemId} img`);
 
                 if (img && img.style.display !== "none") {
@@ -504,6 +511,10 @@ onUiLoaded(async() => {
             if (isModifierKey(e, hotkeysConfig.canvas_hotkey_zoom)) {
                 e.preventDefault();
 
+                if (hotkeysConfig.canvas_hotkey_zoom === "Alt") {
+                    interactedWithAltKey = true;
+                }
+
                 let zoomPosX, zoomPosY;
                 let delta = 0.2;
                 if (elemData[elemId].zoomLevel > 7) {
@@ -686,7 +697,9 @@ onUiLoaded(async() => {
             const hotkeyActions = {
                 [hotkeysConfig.canvas_hotkey_reset]: resetZoom,
                 [hotkeysConfig.canvas_hotkey_overlap]: toggleOverlap,
-                [hotkeysConfig.canvas_hotkey_fullscreen]: fitToScreen
+                [hotkeysConfig.canvas_hotkey_fullscreen]: fitToScreen,
+                [hotkeysConfig.canvas_hotkey_shrink_brush]: () => adjustBrushSize(elemId, 10),
+                [hotkeysConfig.canvas_hotkey_grow_brush]: () => adjustBrushSize(elemId, -10)
             };
 
             const action = hotkeyActions[event.code];
@@ -777,23 +790,29 @@ onUiLoaded(async() => {
         targetElement.addEventListener("mouseleave", handleMouseLeave);
 
         // Reset zoom when click on another tab
-        elements.img2imgTabs.addEventListener("click", resetZoom);
-        elements.img2imgTabs.addEventListener("click", () => {
-            // targetElement.style.width = "";
-            if (parseInt(targetElement.style.width) > 865) {
-                setTimeout(fitToElement, 0);
-            }
-        });
+        if (elements.img2imgTabs) {
+            elements.img2imgTabs.addEventListener("click", resetZoom);
+            elements.img2imgTabs.addEventListener("click", () => {
+                // targetElement.style.width = "";
+                if (parseInt(targetElement.style.width) > 865) {
+                    setTimeout(fitToElement, 0);
+                }
+            });
+        }
 
         targetElement.addEventListener("wheel", e => {
             // change zoom level
-            const operation = e.deltaY > 0 ? "-" : "+";
+            const operation = (e.deltaY || -e.wheelDelta) > 0 ? "-" : "+";
             changeZoomLevel(operation, e);
 
             // Handle brush size adjustment with ctrl key pressed
             if (isModifierKey(e, hotkeysConfig.canvas_hotkey_adjust)) {
                 e.preventDefault();
 
+                if (hotkeysConfig.canvas_hotkey_adjust === "Alt") {
+                    interactedWithAltKey = true;
+                }
+
                 // Increase or decrease brush size based on scroll direction
                 adjustBrushSize(elemId, e.deltaY);
             }
@@ -833,6 +852,20 @@ onUiLoaded(async() => {
         document.addEventListener("keydown", handleMoveKeyDown);
         document.addEventListener("keyup", handleMoveKeyUp);
 
+
+        // Prevent firefox from opening main menu when alt is used as a hotkey for zoom or brush size
+        function handleAltKeyUp(e) {
+            if (e.key !== "Alt" || !interactedWithAltKey) {
+                return;
+            }
+
+            e.preventDefault();
+            interactedWithAltKey = false;
+        }
+
+        document.addEventListener("keyup", handleAltKeyUp);
+
+
         // Detect zoom level and update the pan speed.
         function updatePanPosition(movementX, movementY) {
             let panSpeed = 2;
diff --git a/extensions-builtin/canvas-zoom-and-pan/scripts/hotkey_config.py b/extensions-builtin/canvas-zoom-and-pan/scripts/hotkey_config.py
index 2d8d2d1c014..17b27b2741b 100644
--- a/extensions-builtin/canvas-zoom-and-pan/scripts/hotkey_config.py
+++ b/extensions-builtin/canvas-zoom-and-pan/scripts/hotkey_config.py
@@ -4,12 +4,14 @@
 shared.options_templates.update(shared.options_section(('canvas_hotkey', "Canvas Hotkeys"), {
     "canvas_hotkey_zoom": shared.OptionInfo("Alt", "Zoom canvas", gr.Radio, {"choices": ["Shift","Ctrl", "Alt"]}).info("If you choose 'Shift' you cannot scroll horizontally, 'Alt' can cause a little trouble in firefox"),
     "canvas_hotkey_adjust": shared.OptionInfo("Ctrl", "Adjust brush size", gr.Radio, {"choices": ["Shift","Ctrl", "Alt"]}).info("If you choose 'Shift' you cannot scroll horizontally, 'Alt' can cause a little trouble in firefox"),
+    "canvas_hotkey_shrink_brush": shared.OptionInfo("Q", "Shrink the brush size"),
+    "canvas_hotkey_grow_brush": shared.OptionInfo("W", "Enlarge the brush size"),
     "canvas_hotkey_move": shared.OptionInfo("F", "Moving the canvas").info("To work correctly in firefox, turn off 'Automatically search the page text when typing' in the browser settings"),
     "canvas_hotkey_fullscreen": shared.OptionInfo("S", "Fullscreen Mode, maximizes the picture so that it fits into the screen and stretches it to its full width "),
-    "canvas_hotkey_reset": shared.OptionInfo("R", "Reset zoom and canvas positon"),
-    "canvas_hotkey_overlap": shared.OptionInfo("O", "Toggle overlap").info("Technical button, neededs for testing"),
+    "canvas_hotkey_reset": shared.OptionInfo("R", "Reset zoom and canvas position"),
+    "canvas_hotkey_overlap": shared.OptionInfo("O", "Toggle overlap").info("Technical button, needed for testing"),
     "canvas_show_tooltip": shared.OptionInfo(True, "Enable tooltip on the canvas"),
     "canvas_auto_expand": shared.OptionInfo(True, "Automatically expands an image that does not fit completely in the canvas area, similar to manually pressing the S and R buttons"),
     "canvas_blur_prompt": shared.OptionInfo(False, "Take the focus off the prompt when working with a canvas"),
-    "canvas_disabled_functions": shared.OptionInfo(["Overlap"], "Disable function that you don't use", gr.CheckboxGroup, {"choices": ["Zoom","Adjust brush size", "Moving canvas","Fullscreen","Reset Zoom","Overlap"]}),
+    "canvas_disabled_functions": shared.OptionInfo(["Overlap"], "Disable function that you don't use", gr.CheckboxGroup, {"choices": ["Zoom","Adjust brush size","Hotkey enlarge brush","Hotkey shrink brush","Moving canvas","Fullscreen","Reset Zoom","Overlap"]}),
 }))
diff --git a/extensions-builtin/extra-options-section/scripts/extra_options_section.py b/extensions-builtin/extra-options-section/scripts/extra_options_section.py
index 983f87ff033..a91bea4fa9d 100644
--- a/extensions-builtin/extra-options-section/scripts/extra_options_section.py
+++ b/extensions-builtin/extra-options-section/scripts/extra_options_section.py
@@ -1,7 +1,7 @@
 import math
 
 import gradio as gr
-from modules import scripts, shared, ui_components, ui_settings, generation_parameters_copypaste
+from modules import scripts, shared, ui_components, ui_settings, infotext_utils, errors
 from modules.ui_components import FormColumn
 
 
@@ -23,11 +23,12 @@ def ui(self, is_img2img):
         self.setting_names = []
         self.infotext_fields = []
         extra_options = shared.opts.extra_options_img2img if is_img2img else shared.opts.extra_options_txt2img
+        elem_id_tabname = "extra_options_" + ("img2img" if is_img2img else "txt2img")
 
-        mapping = {k: v for v, k in generation_parameters_copypaste.infotext_to_setting_name_mapping}
+        mapping = {k: v for v, k in infotext_utils.infotext_to_setting_name_mapping}
 
         with gr.Blocks() as interface:
-            with gr.Accordion("Options", open=False) if shared.opts.extra_options_accordion and extra_options else gr.Group():
+            with gr.Accordion("Options", open=False, elem_id=elem_id_tabname) if shared.opts.extra_options_accordion and extra_options else gr.Group(elem_id=elem_id_tabname):
 
                 row_count = math.ceil(len(extra_options) / shared.opts.extra_options_cols)
 
@@ -41,7 +42,11 @@ def ui(self, is_img2img):
                             setting_name = extra_options[index]
 
                             with FormColumn():
-                                comp = ui_settings.create_setting_component(setting_name)
+                                try:
+                                    comp = ui_settings.create_setting_component(setting_name)
+                                except KeyError:
+                                    errors.report(f"Can't add extra options for {setting_name} in ui")
+                                    continue
 
                             self.comps.append(comp)
                             self.setting_names.append(setting_name)
@@ -64,11 +69,14 @@ def before_process(self, p, *args):
                 p.override_settings[name] = value
 
 
-shared.options_templates.update(shared.options_section(('ui', "User interface"), {
-    "extra_options_txt2img": shared.OptionInfo([], "Options in main UI - txt2img", ui_components.DropdownMulti, lambda: {"choices": list(shared.opts.data_labels.keys())}).js("info", "settingsHintsShowQuicksettings").info("setting entries that also appear in txt2img interfaces").needs_reload_ui(),
-    "extra_options_img2img": shared.OptionInfo([], "Options in main UI - img2img", ui_components.DropdownMulti, lambda: {"choices": list(shared.opts.data_labels.keys())}).js("info", "settingsHintsShowQuicksettings").info("setting entries that also appear in img2img interfaces").needs_reload_ui(),
-    "extra_options_cols": shared.OptionInfo(1, "Options in main UI - number of columns", gr.Number, {"precision": 0}).needs_reload_ui(),
-    "extra_options_accordion": shared.OptionInfo(False, "Options in main UI - place into an accordion").needs_reload_ui()
+shared.options_templates.update(shared.options_section(('settings_in_ui', "Settings in UI", "ui"), {
+    "settings_in_ui": shared.OptionHTML("""
+This page allows you to add some settings to the main interface of txt2img and img2img tabs.
+"""),
+    "extra_options_txt2img": shared.OptionInfo([], "Settings for txt2img", ui_components.DropdownMulti, lambda: {"choices": list(shared.opts.data_labels.keys())}).js("info", "settingsHintsShowQuicksettings").info("setting entries that also appear in txt2img interfaces").needs_reload_ui(),
+    "extra_options_img2img": shared.OptionInfo([], "Settings for img2img", ui_components.DropdownMulti, lambda: {"choices": list(shared.opts.data_labels.keys())}).js("info", "settingsHintsShowQuicksettings").info("setting entries that also appear in img2img interfaces").needs_reload_ui(),
+    "extra_options_cols": shared.OptionInfo(1, "Number of columns for added settings", gr.Slider, {"step": 1, "minimum": 1, "maximum": 20}).info("displayed amount will depend on the actual browser window width").needs_reload_ui(),
+    "extra_options_accordion": shared.OptionInfo(False, "Place added settings into an accordion").needs_reload_ui()
 }))
 
 
diff --git a/extensions-builtin/hypertile/hypertile.py b/extensions-builtin/hypertile/hypertile.py
new file mode 100644
index 00000000000..0f40e2d3925
--- /dev/null
+++ b/extensions-builtin/hypertile/hypertile.py
@@ -0,0 +1,351 @@
+"""
+Hypertile module for splitting attention layers in SD-1.5 U-Net and SD-1.5 VAE
+Warn: The patch works well only if the input image has a width and height that are multiples of 128
+Original author: @tfernd Github: https://github.com/tfernd/HyperTile
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+from typing import Callable
+
+from functools import wraps, cache
+
+import math
+import torch.nn as nn
+import random
+
+from einops import rearrange
+
+
+@dataclass
+class HypertileParams:
+    depth = 0
+    layer_name = ""
+    tile_size: int = 0
+    swap_size: int = 0
+    aspect_ratio: float = 1.0
+    forward = None
+    enabled = False
+
+
+
+# TODO add SD-XL layers
+DEPTH_LAYERS = {
+    0: [
+        # SD 1.5 U-Net (diffusers)
+        "down_blocks.0.attentions.0.transformer_blocks.0.attn1",
+        "down_blocks.0.attentions.1.transformer_blocks.0.attn1",
+        "up_blocks.3.attentions.0.transformer_blocks.0.attn1",
+        "up_blocks.3.attentions.1.transformer_blocks.0.attn1",
+        "up_blocks.3.attentions.2.transformer_blocks.0.attn1",
+        # SD 1.5 U-Net (ldm)
+        "input_blocks.1.1.transformer_blocks.0.attn1",
+        "input_blocks.2.1.transformer_blocks.0.attn1",
+        "output_blocks.9.1.transformer_blocks.0.attn1",
+        "output_blocks.10.1.transformer_blocks.0.attn1",
+        "output_blocks.11.1.transformer_blocks.0.attn1",
+        # SD 1.5 VAE
+        "decoder.mid_block.attentions.0",
+        "decoder.mid.attn_1",
+    ],
+    1: [
+        # SD 1.5 U-Net (diffusers)
+        "down_blocks.1.attentions.0.transformer_blocks.0.attn1",
+        "down_blocks.1.attentions.1.transformer_blocks.0.attn1",
+        "up_blocks.2.attentions.0.transformer_blocks.0.attn1",
+        "up_blocks.2.attentions.1.transformer_blocks.0.attn1",
+        "up_blocks.2.attentions.2.transformer_blocks.0.attn1",
+        # SD 1.5 U-Net (ldm)
+        "input_blocks.4.1.transformer_blocks.0.attn1",
+        "input_blocks.5.1.transformer_blocks.0.attn1",
+        "output_blocks.6.1.transformer_blocks.0.attn1",
+        "output_blocks.7.1.transformer_blocks.0.attn1",
+        "output_blocks.8.1.transformer_blocks.0.attn1",
+    ],
+    2: [
+        # SD 1.5 U-Net (diffusers)
+        "down_blocks.2.attentions.0.transformer_blocks.0.attn1",
+        "down_blocks.2.attentions.1.transformer_blocks.0.attn1",
+        "up_blocks.1.attentions.0.transformer_blocks.0.attn1",
+        "up_blocks.1.attentions.1.transformer_blocks.0.attn1",
+        "up_blocks.1.attentions.2.transformer_blocks.0.attn1",
+        # SD 1.5 U-Net (ldm)
+        "input_blocks.7.1.transformer_blocks.0.attn1",
+        "input_blocks.8.1.transformer_blocks.0.attn1",
+        "output_blocks.3.1.transformer_blocks.0.attn1",
+        "output_blocks.4.1.transformer_blocks.0.attn1",
+        "output_blocks.5.1.transformer_blocks.0.attn1",
+    ],
+    3: [
+        # SD 1.5 U-Net (diffusers)
+        "mid_block.attentions.0.transformer_blocks.0.attn1",
+        # SD 1.5 U-Net (ldm)
+        "middle_block.1.transformer_blocks.0.attn1",
+    ],
+}
+# XL layers, thanks for GitHub@gel-crabs for the help
+DEPTH_LAYERS_XL = {
+    0: [
+        # SD 1.5 U-Net (diffusers)
+        "down_blocks.0.attentions.0.transformer_blocks.0.attn1",
+        "down_blocks.0.attentions.1.transformer_blocks.0.attn1",
+        "up_blocks.3.attentions.0.transformer_blocks.0.attn1",
+        "up_blocks.3.attentions.1.transformer_blocks.0.attn1",
+        "up_blocks.3.attentions.2.transformer_blocks.0.attn1",
+        # SD 1.5 U-Net (ldm)
+        "input_blocks.4.1.transformer_blocks.0.attn1",
+        "input_blocks.5.1.transformer_blocks.0.attn1",
+        "output_blocks.3.1.transformer_blocks.0.attn1",
+        "output_blocks.4.1.transformer_blocks.0.attn1",
+        "output_blocks.5.1.transformer_blocks.0.attn1",
+        # SD 1.5 VAE
+        "decoder.mid_block.attentions.0",
+        "decoder.mid.attn_1",
+    ],
+    1: [
+        # SD 1.5 U-Net (diffusers)
+        #"down_blocks.1.attentions.0.transformer_blocks.0.attn1",
+        #"down_blocks.1.attentions.1.transformer_blocks.0.attn1",
+        #"up_blocks.2.attentions.0.transformer_blocks.0.attn1",
+        #"up_blocks.2.attentions.1.transformer_blocks.0.attn1",
+        #"up_blocks.2.attentions.2.transformer_blocks.0.attn1",
+        # SD 1.5 U-Net (ldm)
+        "input_blocks.4.1.transformer_blocks.1.attn1",
+        "input_blocks.5.1.transformer_blocks.1.attn1",
+        "output_blocks.3.1.transformer_blocks.1.attn1",
+        "output_blocks.4.1.transformer_blocks.1.attn1",
+        "output_blocks.5.1.transformer_blocks.1.attn1",
+        "input_blocks.7.1.transformer_blocks.0.attn1",
+        "input_blocks.8.1.transformer_blocks.0.attn1",
+        "output_blocks.0.1.transformer_blocks.0.attn1",
+        "output_blocks.1.1.transformer_blocks.0.attn1",
+        "output_blocks.2.1.transformer_blocks.0.attn1",
+        "input_blocks.7.1.transformer_blocks.1.attn1",
+        "input_blocks.8.1.transformer_blocks.1.attn1",
+        "output_blocks.0.1.transformer_blocks.1.attn1",
+        "output_blocks.1.1.transformer_blocks.1.attn1",
+        "output_blocks.2.1.transformer_blocks.1.attn1",
+        "input_blocks.7.1.transformer_blocks.2.attn1",
+        "input_blocks.8.1.transformer_blocks.2.attn1",
+        "output_blocks.0.1.transformer_blocks.2.attn1",
+        "output_blocks.1.1.transformer_blocks.2.attn1",
+        "output_blocks.2.1.transformer_blocks.2.attn1",
+        "input_blocks.7.1.transformer_blocks.3.attn1",
+        "input_blocks.8.1.transformer_blocks.3.attn1",
+        "output_blocks.0.1.transformer_blocks.3.attn1",
+        "output_blocks.1.1.transformer_blocks.3.attn1",
+        "output_blocks.2.1.transformer_blocks.3.attn1",
+        "input_blocks.7.1.transformer_blocks.4.attn1",
+        "input_blocks.8.1.transformer_blocks.4.attn1",
+        "output_blocks.0.1.transformer_blocks.4.attn1",
+        "output_blocks.1.1.transformer_blocks.4.attn1",
+        "output_blocks.2.1.transformer_blocks.4.attn1",
+        "input_blocks.7.1.transformer_blocks.5.attn1",
+        "input_blocks.8.1.transformer_blocks.5.attn1",
+        "output_blocks.0.1.transformer_blocks.5.attn1",
+        "output_blocks.1.1.transformer_blocks.5.attn1",
+        "output_blocks.2.1.transformer_blocks.5.attn1",
+        "input_blocks.7.1.transformer_blocks.6.attn1",
+        "input_blocks.8.1.transformer_blocks.6.attn1",
+        "output_blocks.0.1.transformer_blocks.6.attn1",
+        "output_blocks.1.1.transformer_blocks.6.attn1",
+        "output_blocks.2.1.transformer_blocks.6.attn1",
+        "input_blocks.7.1.transformer_blocks.7.attn1",
+        "input_blocks.8.1.transformer_blocks.7.attn1",
+        "output_blocks.0.1.transformer_blocks.7.attn1",
+        "output_blocks.1.1.transformer_blocks.7.attn1",
+        "output_blocks.2.1.transformer_blocks.7.attn1",
+        "input_blocks.7.1.transformer_blocks.8.attn1",
+        "input_blocks.8.1.transformer_blocks.8.attn1",
+        "output_blocks.0.1.transformer_blocks.8.attn1",
+        "output_blocks.1.1.transformer_blocks.8.attn1",
+        "output_blocks.2.1.transformer_blocks.8.attn1",
+        "input_blocks.7.1.transformer_blocks.9.attn1",
+        "input_blocks.8.1.transformer_blocks.9.attn1",
+        "output_blocks.0.1.transformer_blocks.9.attn1",
+        "output_blocks.1.1.transformer_blocks.9.attn1",
+        "output_blocks.2.1.transformer_blocks.9.attn1",
+    ],
+    2: [
+        # SD 1.5 U-Net (diffusers)
+        "mid_block.attentions.0.transformer_blocks.0.attn1",
+        # SD 1.5 U-Net (ldm)
+        "middle_block.1.transformer_blocks.0.attn1",
+        "middle_block.1.transformer_blocks.1.attn1",
+        "middle_block.1.transformer_blocks.2.attn1",
+        "middle_block.1.transformer_blocks.3.attn1",
+        "middle_block.1.transformer_blocks.4.attn1",
+        "middle_block.1.transformer_blocks.5.attn1",
+        "middle_block.1.transformer_blocks.6.attn1",
+        "middle_block.1.transformer_blocks.7.attn1",
+        "middle_block.1.transformer_blocks.8.attn1",
+        "middle_block.1.transformer_blocks.9.attn1",
+    ],
+    3 : [] # TODO - separate layers for SD-XL
+}
+
+
+RNG_INSTANCE = random.Random()
+
+@cache
+def get_divisors(value: int, min_value: int, /, max_options: int = 1) -> list[int]:
+    """
+    Returns divisors of value that
+        x * min_value <= value
+    in big -> small order, amount of divisors is limited by max_options
+    """
+    max_options = max(1, max_options) # at least 1 option should be returned
+    min_value = min(min_value, value)
+    divisors = [i for i in range(min_value, value + 1) if value % i == 0] # divisors in small -> big order
+    ns = [value // i for i in divisors[:max_options]]  # has at least 1 element # big -> small order
+    return ns
+
+
+def random_divisor(value: int, min_value: int, /, max_options: int = 1) -> int:
+    """
+    Returns a random divisor of value that
+        x * min_value <= value
+    if max_options is 1, the behavior is deterministic
+    """
+    ns = get_divisors(value, min_value, max_options=max_options) # get cached divisors
+    idx = RNG_INSTANCE.randint(0, len(ns) - 1)
+
+    return ns[idx]
+
+
+def set_hypertile_seed(seed: int) -> None:
+    RNG_INSTANCE.seed(seed)
+
+
+@cache
+def largest_tile_size_available(width: int, height: int) -> int:
+    """
+    Calculates the largest tile size available for a given width and height
+    Tile size is always a power of 2
+    """
+    gcd = math.gcd(width, height)
+    largest_tile_size_available = 1
+    while gcd % (largest_tile_size_available * 2) == 0:
+        largest_tile_size_available *= 2
+    return largest_tile_size_available
+
+
+def iterative_closest_divisors(hw:int, aspect_ratio:float) -> tuple[int, int]:
+    """
+    Finds h and w such that h*w = hw and h/w = aspect_ratio
+    We check all possible divisors of hw and return the closest to the aspect ratio
+    """
+    divisors = [i for i in range(2, hw + 1) if hw % i == 0] # all divisors of hw
+    pairs = [(i, hw // i) for i in divisors] # all pairs of divisors of hw
+    ratios = [w/h for h, w in pairs] # all ratios of pairs of divisors of hw
+    closest_ratio = min(ratios, key=lambda x: abs(x - aspect_ratio)) # closest ratio to aspect_ratio
+    closest_pair = pairs[ratios.index(closest_ratio)] # closest pair of divisors to aspect_ratio
+    return closest_pair
+
+
+@cache
+def find_hw_candidates(hw:int, aspect_ratio:float) -> tuple[int, int]:
+    """
+    Finds h and w such that h*w = hw and h/w = aspect_ratio
+    """
+    h, w = round(math.sqrt(hw * aspect_ratio)), round(math.sqrt(hw / aspect_ratio))
+    # find h and w such that h*w = hw and h/w = aspect_ratio
+    if h * w != hw:
+        w_candidate = hw / h
+        # check if w is an integer
+        if not w_candidate.is_integer():
+            h_candidate = hw / w
+            # check if h is an integer
+            if not h_candidate.is_integer():
+                return iterative_closest_divisors(hw, aspect_ratio)
+            else:
+                h = int(h_candidate)
+        else:
+            w = int(w_candidate)
+    return h, w
+
+
+def self_attn_forward(params: HypertileParams, scale_depth=True) -> Callable:
+
+    @wraps(params.forward)
+    def wrapper(*args, **kwargs):
+        if not params.enabled:
+            return params.forward(*args, **kwargs)
+
+        latent_tile_size = max(128, params.tile_size) // 8
+        x = args[0]
+
+        # VAE
+        if x.ndim == 4:
+            b, c, h, w = x.shape
+
+            nh = random_divisor(h, latent_tile_size, params.swap_size)
+            nw = random_divisor(w, latent_tile_size, params.swap_size)
+
+            if nh * nw > 1:
+                x = rearrange(x, "b c (nh h) (nw w) -> (b nh nw) c h w", nh=nh, nw=nw)  # split into nh * nw tiles
+
+            out = params.forward(x, *args[1:], **kwargs)
+
+            if nh * nw > 1:
+                out = rearrange(out, "(b nh nw) c h w -> b c (nh h) (nw w)", nh=nh, nw=nw)
+
+        # U-Net
+        else:
+            hw: int = x.size(1)
+            h, w = find_hw_candidates(hw, params.aspect_ratio)
+            assert h * w == hw, f"Invalid aspect ratio {params.aspect_ratio} for input of shape {x.shape}, hw={hw}, h={h}, w={w}"
+
+            factor = 2 ** params.depth if scale_depth else 1
+            nh = random_divisor(h, latent_tile_size * factor, params.swap_size)
+            nw = random_divisor(w, latent_tile_size * factor, params.swap_size)
+
+            if nh * nw > 1:
+                x = rearrange(x, "b (nh h nw w) c -> (b nh nw) (h w) c", h=h // nh, w=w // nw, nh=nh, nw=nw)
+
+            out = params.forward(x, *args[1:], **kwargs)
+
+            if nh * nw > 1:
+                out = rearrange(out, "(b nh nw) hw c -> b nh nw hw c", nh=nh, nw=nw)
+                out = rearrange(out, "b nh nw (h w) c -> b (nh h nw w) c", h=h // nh, w=w // nw)
+
+        return out
+
+    return wrapper
+
+
+def hypertile_hook_model(model: nn.Module, width, height, *, enable=False, tile_size_max=128, swap_size=1, max_depth=3, is_sdxl=False):
+    hypertile_layers = getattr(model, "__webui_hypertile_layers", None)
+    if hypertile_layers is None:
+        if not enable:
+            return
+
+        hypertile_layers = {}
+        layers = DEPTH_LAYERS_XL if is_sdxl else DEPTH_LAYERS
+
+        for depth in range(4):
+            for layer_name, module in model.named_modules():
+                if any(layer_name.endswith(try_name) for try_name in layers[depth]):
+                    params = HypertileParams()
+                    module.__webui_hypertile_params = params
+                    params.forward = module.forward
+                    params.depth = depth
+                    params.layer_name = layer_name
+                    module.forward = self_attn_forward(params)
+
+                    hypertile_layers[layer_name] = 1
+
+        model.__webui_hypertile_layers = hypertile_layers
+
+    aspect_ratio = width / height
+    tile_size = min(largest_tile_size_available(width, height), tile_size_max)
+
+    for layer_name, module in model.named_modules():
+        if layer_name in hypertile_layers:
+            params = module.__webui_hypertile_params
+
+            params.tile_size = tile_size
+            params.swap_size = swap_size
+            params.aspect_ratio = aspect_ratio
+            params.enabled = enable and params.depth <= max_depth
diff --git a/extensions-builtin/hypertile/scripts/hypertile_script.py b/extensions-builtin/hypertile/scripts/hypertile_script.py
new file mode 100644
index 00000000000..59e7f9907e5
--- /dev/null
+++ b/extensions-builtin/hypertile/scripts/hypertile_script.py
@@ -0,0 +1,122 @@
+import hypertile
+from modules import scripts, script_callbacks, shared
+
+
+class ScriptHypertile(scripts.Script):
+    name = "Hypertile"
+
+    def title(self):
+        return self.name
+
+    def show(self, is_img2img):
+        return scripts.AlwaysVisible
+
+    def process(self, p, *args):
+        hypertile.set_hypertile_seed(p.all_seeds[0])
+
+        configure_hypertile(p.width, p.height, enable_unet=shared.opts.hypertile_enable_unet)
+
+        self.add_infotext(p)
+
+    def before_hr(self, p, *args):
+
+        enable = shared.opts.hypertile_enable_unet_secondpass or shared.opts.hypertile_enable_unet
+
+        # exclusive hypertile seed for the second pass
+        if enable:
+            hypertile.set_hypertile_seed(p.all_seeds[0])
+
+        configure_hypertile(p.hr_upscale_to_x, p.hr_upscale_to_y, enable_unet=enable)
+
+        if enable and not shared.opts.hypertile_enable_unet:
+            p.extra_generation_params["Hypertile U-Net second pass"] = True
+
+            self.add_infotext(p, add_unet_params=True)
+
+    def add_infotext(self, p, add_unet_params=False):
+        def option(name):
+            value = getattr(shared.opts, name)
+            default_value = shared.opts.get_default(name)
+            return None if value == default_value else value
+
+        if shared.opts.hypertile_enable_unet:
+            p.extra_generation_params["Hypertile U-Net"] = True
+
+        if shared.opts.hypertile_enable_unet or add_unet_params:
+            p.extra_generation_params["Hypertile U-Net max depth"] = option('hypertile_max_depth_unet')
+            p.extra_generation_params["Hypertile U-Net max tile size"] = option('hypertile_max_tile_unet')
+            p.extra_generation_params["Hypertile U-Net swap size"] = option('hypertile_swap_size_unet')
+
+        if shared.opts.hypertile_enable_vae:
+            p.extra_generation_params["Hypertile VAE"] = True
+            p.extra_generation_params["Hypertile VAE max depth"] = option('hypertile_max_depth_vae')
+            p.extra_generation_params["Hypertile VAE max tile size"] = option('hypertile_max_tile_vae')
+            p.extra_generation_params["Hypertile VAE swap size"] = option('hypertile_swap_size_vae')
+
+
+def configure_hypertile(width, height, enable_unet=True):
+    hypertile.hypertile_hook_model(
+        shared.sd_model.first_stage_model,
+        width,
+        height,
+        swap_size=shared.opts.hypertile_swap_size_vae,
+        max_depth=shared.opts.hypertile_max_depth_vae,
+        tile_size_max=shared.opts.hypertile_max_tile_vae,
+        enable=shared.opts.hypertile_enable_vae,
+    )
+
+    hypertile.hypertile_hook_model(
+        shared.sd_model.model,
+        width,
+        height,
+        swap_size=shared.opts.hypertile_swap_size_unet,
+        max_depth=shared.opts.hypertile_max_depth_unet,
+        tile_size_max=shared.opts.hypertile_max_tile_unet,
+        enable=enable_unet,
+        is_sdxl=shared.sd_model.is_sdxl
+    )
+
+
+def on_ui_settings():
+    import gradio as gr
+
+    options = {
+        "hypertile_explanation": shared.OptionHTML("""
+    <a href='https://github.com/tfernd/HyperTile'>Hypertile</a> optimizes the self-attention layer within U-Net and VAE models,
+    resulting in a reduction in computation time ranging from 1 to 4 times. The larger the generated image is, the greater the
+    benefit.
+    """),
+
+        "hypertile_enable_unet": shared.OptionInfo(False, "Enable Hypertile U-Net", infotext="Hypertile U-Net").info("enables hypertile for all modes, including hires fix second pass; noticeable change in details of the generated picture"),
+        "hypertile_enable_unet_secondpass": shared.OptionInfo(False, "Enable Hypertile U-Net for hires fix second pass", infotext="Hypertile U-Net second pass").info("enables hypertile just for hires fix second pass - regardless of whether the above setting is enabled"),
+        "hypertile_max_depth_unet": shared.OptionInfo(3, "Hypertile U-Net max depth", gr.Slider, {"minimum": 0, "maximum": 3, "step": 1}, infotext="Hypertile U-Net max depth").info("larger = more neural network layers affected; minor effect on performance"),
+        "hypertile_max_tile_unet": shared.OptionInfo(256, "Hypertile U-Net max tile size", gr.Slider, {"minimum": 0, "maximum": 512, "step": 16}, infotext="Hypertile U-Net max tile size").info("larger = worse performance"),
+        "hypertile_swap_size_unet": shared.OptionInfo(3, "Hypertile U-Net swap size", gr.Slider, {"minimum": 0, "maximum": 64, "step": 1}, infotext="Hypertile U-Net swap size"),
+        "hypertile_enable_vae": shared.OptionInfo(False, "Enable Hypertile VAE", infotext="Hypertile VAE").info("minimal change in the generated picture"),
+        "hypertile_max_depth_vae": shared.OptionInfo(3, "Hypertile VAE max depth", gr.Slider, {"minimum": 0, "maximum": 3, "step": 1}, infotext="Hypertile VAE max depth"),
+        "hypertile_max_tile_vae": shared.OptionInfo(128, "Hypertile VAE max tile size", gr.Slider, {"minimum": 0, "maximum": 512, "step": 16}, infotext="Hypertile VAE max tile size"),
+        "hypertile_swap_size_vae": shared.OptionInfo(3, "Hypertile VAE swap size ", gr.Slider, {"minimum": 0, "maximum": 64, "step": 1}, infotext="Hypertile VAE swap size"),
+    }
+
+    for name, opt in options.items():
+        opt.section = ('hypertile', "Hypertile")
+        shared.opts.add_option(name, opt)
+
+
+def add_axis_options():
+    xyz_grid = [x for x in scripts.scripts_data if x.script_class.__module__ == "xyz_grid.py"][0].module
+    xyz_grid.axis_options.extend([
+        xyz_grid.AxisOption("[Hypertile] Unet First pass Enabled", str, xyz_grid.apply_override('hypertile_enable_unet', boolean=True), choices=xyz_grid.boolean_choice(reverse=True)),
+        xyz_grid.AxisOption("[Hypertile] Unet Second pass Enabled", str, xyz_grid.apply_override('hypertile_enable_unet_secondpass', boolean=True), choices=xyz_grid.boolean_choice(reverse=True)),
+        xyz_grid.AxisOption("[Hypertile] Unet Max Depth", int, xyz_grid.apply_override("hypertile_max_depth_unet"), confirm=xyz_grid.confirm_range(0, 3, '[Hypertile] Unet Max Depth'), choices=lambda: [str(x) for x in range(4)]),
+        xyz_grid.AxisOption("[Hypertile] Unet Max Tile Size", int, xyz_grid.apply_override("hypertile_max_tile_unet"), confirm=xyz_grid.confirm_range(0, 512, '[Hypertile] Unet Max Tile Size')),
+        xyz_grid.AxisOption("[Hypertile] Unet Swap Size", int, xyz_grid.apply_override("hypertile_swap_size_unet"), confirm=xyz_grid.confirm_range(0, 64, '[Hypertile] Unet Swap Size')),
+        xyz_grid.AxisOption("[Hypertile] VAE Enabled", str, xyz_grid.apply_override('hypertile_enable_vae', boolean=True), choices=xyz_grid.boolean_choice(reverse=True)),
+        xyz_grid.AxisOption("[Hypertile] VAE Max Depth", int, xyz_grid.apply_override("hypertile_max_depth_vae"), confirm=xyz_grid.confirm_range(0, 3, '[Hypertile] VAE Max Depth'), choices=lambda: [str(x) for x in range(4)]),
+        xyz_grid.AxisOption("[Hypertile] VAE Max Tile Size", int, xyz_grid.apply_override("hypertile_max_tile_vae"), confirm=xyz_grid.confirm_range(0, 512, '[Hypertile] VAE Max Tile Size')),
+        xyz_grid.AxisOption("[Hypertile] VAE Swap Size", int, xyz_grid.apply_override("hypertile_swap_size_vae"), confirm=xyz_grid.confirm_range(0, 64, '[Hypertile] VAE Swap Size')),
+    ])
+
+
+script_callbacks.on_ui_settings(on_ui_settings)
+script_callbacks.on_before_ui(add_axis_options)
diff --git a/extensions-builtin/mobile/javascript/mobile.js b/extensions-builtin/mobile/javascript/mobile.js
index 652f07ac7ec..bff1acedff3 100644
--- a/extensions-builtin/mobile/javascript/mobile.js
+++ b/extensions-builtin/mobile/javascript/mobile.js
@@ -12,6 +12,8 @@ function isMobile() {
 }
 
 function reportWindowSize() {
+    if (gradioApp().querySelector('.toprow-compact-tools')) return; // not applicable for compact prompt layout
+
     var currentlyMobile = isMobile();
     if (currentlyMobile == isSetupForMobile) return;
     isSetupForMobile = currentlyMobile;
diff --git a/extensions-builtin/postprocessing-for-training/scripts/postprocessing_autosized_crop.py b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_autosized_crop.py
new file mode 100644
index 00000000000..7e674989814
--- /dev/null
+++ b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_autosized_crop.py
@@ -0,0 +1,64 @@
+from PIL import Image
+
+from modules import scripts_postprocessing, ui_components
+import gradio as gr
+
+
+def center_crop(image: Image, w: int, h: int):
+    iw, ih = image.size
+    if ih / h < iw / w:
+        sw = w * ih / h
+        box = (iw - sw) / 2, 0, iw - (iw - sw) / 2, ih
+    else:
+        sh = h * iw / w
+        box = 0, (ih - sh) / 2, iw, ih - (ih - sh) / 2
+    return image.resize((w, h), Image.Resampling.LANCZOS, box)
+
+
+def multicrop_pic(image: Image, mindim, maxdim, minarea, maxarea, objective, threshold):
+    iw, ih = image.size
+    err = lambda w, h: 1 - (lambda x: x if x < 1 else 1 / x)(iw / ih / (w / h))
+    wh = max(((w, h) for w in range(mindim, maxdim + 1, 64) for h in range(mindim, maxdim + 1, 64)
+              if minarea <= w * h <= maxarea and err(w, h) <= threshold),
+             key=lambda wh: (wh[0] * wh[1], -err(*wh))[::1 if objective == 'Maximize area' else -1],
+             default=None
+             )
+    return wh and center_crop(image, *wh)
+
+
+class ScriptPostprocessingAutosizedCrop(scripts_postprocessing.ScriptPostprocessing):
+    name = "Auto-sized crop"
+    order = 4020
+
+    def ui(self):
+        with ui_components.InputAccordion(False, label="Auto-sized crop") as enable:
+            gr.Markdown('Each image is center-cropped with an automatically chosen width and height.')
+            with gr.Row():
+                mindim = gr.Slider(minimum=64, maximum=2048, step=8, label="Dimension lower bound", value=384, elem_id="postprocess_multicrop_mindim")
+                maxdim = gr.Slider(minimum=64, maximum=2048, step=8, label="Dimension upper bound", value=768, elem_id="postprocess_multicrop_maxdim")
+            with gr.Row():
+                minarea = gr.Slider(minimum=64 * 64, maximum=2048 * 2048, step=1, label="Area lower bound", value=64 * 64, elem_id="postprocess_multicrop_minarea")
+                maxarea = gr.Slider(minimum=64 * 64, maximum=2048 * 2048, step=1, label="Area upper bound", value=640 * 640, elem_id="postprocess_multicrop_maxarea")
+            with gr.Row():
+                objective = gr.Radio(["Maximize area", "Minimize error"], value="Maximize area", label="Resizing objective", elem_id="postprocess_multicrop_objective")
+                threshold = gr.Slider(minimum=0, maximum=1, step=0.01, label="Error threshold", value=0.1, elem_id="postprocess_multicrop_threshold")
+
+        return {
+            "enable": enable,
+            "mindim": mindim,
+            "maxdim": maxdim,
+            "minarea": minarea,
+            "maxarea": maxarea,
+            "objective": objective,
+            "threshold": threshold,
+        }
+
+    def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, mindim, maxdim, minarea, maxarea, objective, threshold):
+        if not enable:
+            return
+
+        cropped = multicrop_pic(pp.image, mindim, maxdim, minarea, maxarea, objective, threshold)
+        if cropped is not None:
+            pp.image = cropped
+        else:
+            print(f"skipped {pp.image.width}x{pp.image.height} image (can't find suitable size within error threshold)")
diff --git a/extensions-builtin/postprocessing-for-training/scripts/postprocessing_caption.py b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_caption.py
new file mode 100644
index 00000000000..5592a89870e
--- /dev/null
+++ b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_caption.py
@@ -0,0 +1,30 @@
+from modules import scripts_postprocessing, ui_components, deepbooru, shared
+import gradio as gr
+
+
+class ScriptPostprocessingCeption(scripts_postprocessing.ScriptPostprocessing):
+    name = "Caption"
+    order = 4040
+
+    def ui(self):
+        with ui_components.InputAccordion(False, label="Caption") as enable:
+            option = gr.CheckboxGroup(value=["Deepbooru"], choices=["Deepbooru", "BLIP"], show_label=False)
+
+        return {
+            "enable": enable,
+            "option": option,
+        }
+
+    def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, option):
+        if not enable:
+            return
+
+        captions = [pp.caption]
+
+        if "Deepbooru" in option:
+            captions.append(deepbooru.model.tag(pp.image))
+
+        if "BLIP" in option:
+            captions.append(shared.interrogator.interrogate(pp.image.convert("RGB")))
+
+        pp.caption = ", ".join([x for x in captions if x])
diff --git a/extensions-builtin/postprocessing-for-training/scripts/postprocessing_create_flipped_copies.py b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_create_flipped_copies.py
new file mode 100644
index 00000000000..b673003b6ea
--- /dev/null
+++ b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_create_flipped_copies.py
@@ -0,0 +1,32 @@
+from PIL import ImageOps, Image
+
+from modules import scripts_postprocessing, ui_components
+import gradio as gr
+
+
+class ScriptPostprocessingCreateFlippedCopies(scripts_postprocessing.ScriptPostprocessing):
+    name = "Create flipped copies"
+    order = 4030
+
+    def ui(self):
+        with ui_components.InputAccordion(False, label="Create flipped copies") as enable:
+            with gr.Row():
+                option = gr.CheckboxGroup(value=["Horizontal"], choices=["Horizontal", "Vertical", "Both"], show_label=False)
+
+        return {
+            "enable": enable,
+            "option": option,
+        }
+
+    def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, option):
+        if not enable:
+            return
+
+        if "Horizontal" in option:
+            pp.extra_images.append(ImageOps.mirror(pp.image))
+
+        if "Vertical" in option:
+            pp.extra_images.append(pp.image.transpose(Image.Transpose.FLIP_TOP_BOTTOM))
+
+        if "Both" in option:
+            pp.extra_images.append(pp.image.transpose(Image.Transpose.FLIP_TOP_BOTTOM).transpose(Image.Transpose.FLIP_LEFT_RIGHT))
diff --git a/extensions-builtin/postprocessing-for-training/scripts/postprocessing_focal_crop.py b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_focal_crop.py
new file mode 100644
index 00000000000..cff1dbc5470
--- /dev/null
+++ b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_focal_crop.py
@@ -0,0 +1,54 @@
+
+from modules import scripts_postprocessing, ui_components, errors
+import gradio as gr
+
+from modules.textual_inversion import autocrop
+
+
+class ScriptPostprocessingFocalCrop(scripts_postprocessing.ScriptPostprocessing):
+    name = "Auto focal point crop"
+    order = 4010
+
+    def ui(self):
+        with ui_components.InputAccordion(False, label="Auto focal point crop") as enable:
+            face_weight = gr.Slider(label='Focal point face weight', value=0.9, minimum=0.0, maximum=1.0, step=0.05, elem_id="postprocess_focal_crop_face_weight")
+            entropy_weight = gr.Slider(label='Focal point entropy weight', value=0.15, minimum=0.0, maximum=1.0, step=0.05, elem_id="postprocess_focal_crop_entropy_weight")
+            edges_weight = gr.Slider(label='Focal point edges weight', value=0.5, minimum=0.0, maximum=1.0, step=0.05, elem_id="postprocess_focal_crop_edges_weight")
+            debug = gr.Checkbox(label='Create debug image', elem_id="train_process_focal_crop_debug")
+
+        return {
+            "enable": enable,
+            "face_weight": face_weight,
+            "entropy_weight": entropy_weight,
+            "edges_weight": edges_weight,
+            "debug": debug,
+        }
+
+    def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, face_weight, entropy_weight, edges_weight, debug):
+        if not enable:
+            return
+
+        if not pp.shared.target_width or not pp.shared.target_height:
+            return
+
+        dnn_model_path = None
+        try:
+            dnn_model_path = autocrop.download_and_cache_models()
+        except Exception:
+            errors.report("Unable to load face detection model for auto crop selection. Falling back to lower quality haar method.", exc_info=True)
+
+        autocrop_settings = autocrop.Settings(
+            crop_width=pp.shared.target_width,
+            crop_height=pp.shared.target_height,
+            face_points_weight=face_weight,
+            entropy_points_weight=entropy_weight,
+            corner_points_weight=edges_weight,
+            annotate_image=debug,
+            dnn_model_path=dnn_model_path,
+        )
+
+        result, *others = autocrop.crop_image(pp.image, autocrop_settings)
+
+        pp.image = result
+        pp.extra_images = [pp.create_copy(x, nametags=["focal-crop-debug"], disable_processing=True) for x in others]
+
diff --git a/extensions-builtin/postprocessing-for-training/scripts/postprocessing_split_oversized.py b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_split_oversized.py
new file mode 100644
index 00000000000..133e199b838
--- /dev/null
+++ b/extensions-builtin/postprocessing-for-training/scripts/postprocessing_split_oversized.py
@@ -0,0 +1,71 @@
+import math
+
+from modules import scripts_postprocessing, ui_components
+import gradio as gr
+
+
+def split_pic(image, inverse_xy, width, height, overlap_ratio):
+    if inverse_xy:
+        from_w, from_h = image.height, image.width
+        to_w, to_h = height, width
+    else:
+        from_w, from_h = image.width, image.height
+        to_w, to_h = width, height
+    h = from_h * to_w // from_w
+    if inverse_xy:
+        image = image.resize((h, to_w))
+    else:
+        image = image.resize((to_w, h))
+
+    split_count = math.ceil((h - to_h * overlap_ratio) / (to_h * (1.0 - overlap_ratio)))
+    y_step = (h - to_h) / (split_count - 1)
+    for i in range(split_count):
+        y = int(y_step * i)
+        if inverse_xy:
+            splitted = image.crop((y, 0, y + to_h, to_w))
+        else:
+            splitted = image.crop((0, y, to_w, y + to_h))
+        yield splitted
+
+
+class ScriptPostprocessingSplitOversized(scripts_postprocessing.ScriptPostprocessing):
+    name = "Split oversized images"
+    order = 4000
+
+    def ui(self):
+        with ui_components.InputAccordion(False, label="Split oversized images") as enable:
+            with gr.Row():
+                split_threshold = gr.Slider(label='Threshold', value=0.5, minimum=0.0, maximum=1.0, step=0.05, elem_id="postprocess_split_threshold")
+                overlap_ratio = gr.Slider(label='Overlap ratio', value=0.2, minimum=0.0, maximum=0.9, step=0.05, elem_id="postprocess_overlap_ratio")
+
+        return {
+            "enable": enable,
+            "split_threshold": split_threshold,
+            "overlap_ratio": overlap_ratio,
+        }
+
+    def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, split_threshold, overlap_ratio):
+        if not enable:
+            return
+
+        width = pp.shared.target_width
+        height = pp.shared.target_height
+
+        if not width or not height:
+            return
+
+        if pp.image.height > pp.image.width:
+            ratio = (pp.image.width * height) / (pp.image.height * width)
+            inverse_xy = False
+        else:
+            ratio = (pp.image.height * width) / (pp.image.width * height)
+            inverse_xy = True
+
+        if ratio >= 1.0 or ratio > split_threshold:
+            return
+
+        result, *others = split_pic(pp.image, inverse_xy, width, height, overlap_ratio)
+
+        pp.image = result
+        pp.extra_images = [pp.create_copy(x) for x in others]
+
diff --git a/extensions-builtin/soft-inpainting/scripts/soft_inpainting.py b/extensions-builtin/soft-inpainting/scripts/soft_inpainting.py
new file mode 100644
index 00000000000..0e629963af4
--- /dev/null
+++ b/extensions-builtin/soft-inpainting/scripts/soft_inpainting.py
@@ -0,0 +1,760 @@
+import numpy as np
+import gradio as gr
+import math
+from modules.ui_components import InputAccordion
+import modules.scripts as scripts
+from modules.torch_utils import float64
+
+
+class SoftInpaintingSettings:
+    def __init__(self,
+                 mask_blend_power,
+                 mask_blend_scale,
+                 inpaint_detail_preservation,
+                 composite_mask_influence,
+                 composite_difference_threshold,
+                 composite_difference_contrast):
+        self.mask_blend_power = mask_blend_power
+        self.mask_blend_scale = mask_blend_scale
+        self.inpaint_detail_preservation = inpaint_detail_preservation
+        self.composite_mask_influence = composite_mask_influence
+        self.composite_difference_threshold = composite_difference_threshold
+        self.composite_difference_contrast = composite_difference_contrast
+
+    def add_generation_params(self, dest):
+        dest[enabled_gen_param_label] = True
+        dest[gen_param_labels.mask_blend_power] = self.mask_blend_power
+        dest[gen_param_labels.mask_blend_scale] = self.mask_blend_scale
+        dest[gen_param_labels.inpaint_detail_preservation] = self.inpaint_detail_preservation
+        dest[gen_param_labels.composite_mask_influence] = self.composite_mask_influence
+        dest[gen_param_labels.composite_difference_threshold] = self.composite_difference_threshold
+        dest[gen_param_labels.composite_difference_contrast] = self.composite_difference_contrast
+
+
+# ------------------- Methods -------------------
+
+def processing_uses_inpainting(p):
+    # TODO: Figure out a better way to determine if inpainting is being used by p
+    if getattr(p, "image_mask", None) is not None:
+        return True
+
+    if getattr(p, "mask", None) is not None:
+        return True
+
+    if getattr(p, "nmask", None) is not None:
+        return True
+
+    return False
+
+
+def latent_blend(settings, a, b, t):
+    """
+    Interpolates two latent image representations according to the parameter t,
+    where the interpolated vectors' magnitudes are also interpolated separately.
+    The "detail_preservation" factor biases the magnitude interpolation towards
+    the larger of the two magnitudes.
+    """
+    import torch
+
+    # NOTE: We use inplace operations wherever possible.
+
+    if len(t.shape) == 3:
+        # [4][w][h] to [1][4][w][h]
+        t2 = t.unsqueeze(0)
+        # [4][w][h] to [1][1][w][h] - the [4] seem redundant.
+        t3 = t[0].unsqueeze(0).unsqueeze(0)
+    else:
+        t2 = t
+        t3 = t[:, 0][:, None]
+
+    one_minus_t2 = 1 - t2
+    one_minus_t3 = 1 - t3
+
+    # Linearly interpolate the image vectors.
+    a_scaled = a * one_minus_t2
+    b_scaled = b * t2
+    image_interp = a_scaled
+    image_interp.add_(b_scaled)
+    result_type = image_interp.dtype
+    del a_scaled, b_scaled, t2, one_minus_t2
+
+    # Calculate the magnitude of the interpolated vectors. (We will remove this magnitude.)
+    # 64-bit operations are used here to allow large exponents.
+    current_magnitude = torch.norm(image_interp, p=2, dim=1, keepdim=True).to(float64(image_interp)).add_(0.00001)
+
+    # Interpolate the powered magnitudes, then un-power them (bring them back to a power of 1).
+    a_magnitude = torch.norm(a, p=2, dim=1, keepdim=True).to(float64(a)).pow_(settings.inpaint_detail_preservation) * one_minus_t3
+    b_magnitude = torch.norm(b, p=2, dim=1, keepdim=True).to(float64(b)).pow_(settings.inpaint_detail_preservation) * t3
+    desired_magnitude = a_magnitude
+    desired_magnitude.add_(b_magnitude).pow_(1 / settings.inpaint_detail_preservation)
+    del a_magnitude, b_magnitude, t3, one_minus_t3
+
+    # Change the linearly interpolated image vectors' magnitudes to the value we want.
+    # This is the last 64-bit operation.
+    image_interp_scaling_factor = desired_magnitude
+    image_interp_scaling_factor.div_(current_magnitude)
+    image_interp_scaling_factor = image_interp_scaling_factor.to(result_type)
+    image_interp_scaled = image_interp
+    image_interp_scaled.mul_(image_interp_scaling_factor)
+    del current_magnitude
+    del desired_magnitude
+    del image_interp
+    del image_interp_scaling_factor
+    del result_type
+
+    return image_interp_scaled
+
+
+def get_modified_nmask(settings, nmask, sigma):
+    """
+    Converts a negative mask representing the transparency of the original latent vectors being overlaid
+    to a mask that is scaled according to the denoising strength for this step.
+
+    Where:
+        0 = fully opaque, infinite density, fully masked
+        1 = fully transparent, zero density, fully unmasked
+
+    We bring this transparency to a power, as this allows one to simulate N number of blending operations
+    where N can be any positive real value. Using this one can control the balance of influence between
+    the denoiser and the original latents according to the sigma value.
+
+    NOTE: "mask" is not used
+    """
+    import torch
+    return torch.pow(nmask, (sigma ** settings.mask_blend_power) * settings.mask_blend_scale)
+
+
+def apply_adaptive_masks(
+        settings: SoftInpaintingSettings,
+        nmask,
+        latent_orig,
+        latent_processed,
+        overlay_images,
+        width, height,
+        paste_to):
+    import torch
+    import modules.processing as proc
+    import modules.images as images
+    from PIL import Image, ImageOps, ImageFilter
+
+    # TODO: Bias the blending according to the latent mask, add adjustable parameter for bias control.
+    if len(nmask.shape) == 3:
+        latent_mask = nmask[0].float()
+    else:
+        latent_mask = nmask[:, 0].float()
+    # convert the original mask into a form we use to scale distances for thresholding
+    mask_scalar = 1 - (torch.clamp(latent_mask, min=0, max=1) ** (settings.mask_blend_scale / 2))
+    mask_scalar = (0.5 * (1 - settings.composite_mask_influence)
+                   + mask_scalar * settings.composite_mask_influence)
+    mask_scalar = mask_scalar / (1.00001 - mask_scalar)
+    mask_scalar = mask_scalar.cpu().numpy()
+
+    latent_distance = torch.norm(latent_processed - latent_orig, p=2, dim=1)
+
+    kernel, kernel_center = get_gaussian_kernel(stddev_radius=1.5, max_radius=2)
+
+    masks_for_overlay = []
+
+    for i, (distance_map, overlay_image) in enumerate(zip(latent_distance, overlay_images)):
+        converted_mask = distance_map.float().cpu().numpy()
+        converted_mask = weighted_histogram_filter(converted_mask, kernel, kernel_center,
+                                                   percentile_min=0.9, percentile_max=1, min_width=1)
+        converted_mask = weighted_histogram_filter(converted_mask, kernel, kernel_center,
+                                                   percentile_min=0.25, percentile_max=0.75, min_width=1)
+
+        # The distance at which opacity of original decreases to 50%
+        if len(mask_scalar.shape) == 3:
+            if mask_scalar.shape[0] > i:
+                half_weighted_distance = settings.composite_difference_threshold * mask_scalar[i]
+            else:
+                half_weighted_distance = settings.composite_difference_threshold * mask_scalar[0]
+        else:
+            half_weighted_distance = settings.composite_difference_threshold * mask_scalar
+
+        converted_mask = converted_mask / half_weighted_distance
+
+        converted_mask = 1 / (1 + converted_mask ** settings.composite_difference_contrast)
+        converted_mask = smootherstep(converted_mask)
+        converted_mask = 1 - converted_mask
+        converted_mask = 255. * converted_mask
+        converted_mask = converted_mask.astype(np.uint8)
+        converted_mask = Image.fromarray(converted_mask)
+        converted_mask = images.resize_image(2, converted_mask, width, height)
+        converted_mask = proc.create_binary_mask(converted_mask, round=False)
+
+        # Remove aliasing artifacts using a gaussian blur.
+        converted_mask = converted_mask.filter(ImageFilter.GaussianBlur(radius=4))
+
+        # Expand the mask to fit the whole image if needed.
+        if paste_to is not None:
+            converted_mask = proc.uncrop(converted_mask,
+                                         (overlay_image.width, overlay_image.height),
+                                         paste_to)
+
+        masks_for_overlay.append(converted_mask)
+
+        image_masked = Image.new('RGBa', (overlay_image.width, overlay_image.height))
+        image_masked.paste(overlay_image.convert("RGBA").convert("RGBa"),
+                           mask=ImageOps.invert(converted_mask.convert('L')))
+
+        overlay_images[i] = image_masked.convert('RGBA')
+
+    return masks_for_overlay
+
+
+def apply_masks(
+        settings,
+        nmask,
+        overlay_images,
+        width, height,
+        paste_to):
+    import torch
+    import modules.processing as proc
+    import modules.images as images
+    from PIL import Image, ImageOps, ImageFilter
+
+    converted_mask = nmask[0].float()
+    converted_mask = torch.clamp(converted_mask, min=0, max=1).pow_(settings.mask_blend_scale / 2)
+    converted_mask = 255. * converted_mask
+    converted_mask = converted_mask.cpu().numpy().astype(np.uint8)
+    converted_mask = Image.fromarray(converted_mask)
+    converted_mask = images.resize_image(2, converted_mask, width, height)
+    converted_mask = proc.create_binary_mask(converted_mask, round=False)
+
+    # Remove aliasing artifacts using a gaussian blur.
+    converted_mask = converted_mask.filter(ImageFilter.GaussianBlur(radius=4))
+
+    # Expand the mask to fit the whole image if needed.
+    if paste_to is not None:
+        converted_mask = proc.uncrop(converted_mask,
+                                     (width, height),
+                                     paste_to)
+
+    masks_for_overlay = []
+
+    for i, overlay_image in enumerate(overlay_images):
+        masks_for_overlay[i] = converted_mask
+
+        image_masked = Image.new('RGBa', (overlay_image.width, overlay_image.height))
+        image_masked.paste(overlay_image.convert("RGBA").convert("RGBa"),
+                           mask=ImageOps.invert(converted_mask.convert('L')))
+
+        overlay_images[i] = image_masked.convert('RGBA')
+
+    return masks_for_overlay
+
+
+def weighted_histogram_filter(img, kernel, kernel_center, percentile_min=0.0, percentile_max=1.0, min_width=1.0):
+    """
+    Generalization convolution filter capable of applying
+    weighted mean, median, maximum, and minimum filters
+    parametrically using an arbitrary kernel.
+
+    Args:
+        img (nparray):
+            The image, a 2-D array of floats, to which the filter is being applied.
+        kernel (nparray):
+            The kernel, a 2-D array of floats.
+        kernel_center (nparray):
+            The kernel center coordinate, a 1-D array with two elements.
+        percentile_min (float):
+            The lower bound of the histogram window used by the filter,
+            from 0 to 1.
+        percentile_max (float):
+            The upper bound of the histogram window used by the filter,
+            from 0 to 1.
+        min_width (float):
+            The minimum size of the histogram window bounds, in weight units.
+            Must be greater than 0.
+
+    Returns:
+        (nparray): A filtered copy of the input image "img", a 2-D array of floats.
+    """
+
+    # Converts an index tuple into a vector.
+    def vec(x):
+        return np.array(x)
+
+    kernel_min = -kernel_center
+    kernel_max = vec(kernel.shape) - kernel_center
+
+    def weighted_histogram_filter_single(idx):
+        idx = vec(idx)
+        min_index = np.maximum(0, idx + kernel_min)
+        max_index = np.minimum(vec(img.shape), idx + kernel_max)
+        window_shape = max_index - min_index
+
+        class WeightedElement:
+            """
+            An element of the histogram, its weight
+            and bounds.
+            """
+
+            def __init__(self, value, weight):
+                self.value: float = value
+                self.weight: float = weight
+                self.window_min: float = 0.0
+                self.window_max: float = 1.0
+
+        # Collect the values in the image as WeightedElements,
+        # weighted by their corresponding kernel values.
+        values = []
+        for window_tup in np.ndindex(tuple(window_shape)):
+            window_index = vec(window_tup)
+            image_index = window_index + min_index
+            centered_kernel_index = image_index - idx
+            kernel_index = centered_kernel_index + kernel_center
+            element = WeightedElement(img[tuple(image_index)], kernel[tuple(kernel_index)])
+            values.append(element)
+
+        def sort_key(x: WeightedElement):
+            return x.value
+
+        values.sort(key=sort_key)
+
+        # Calculate the height of the stack (sum)
+        # and each sample's range they occupy in the stack
+        sum = 0
+        for i in range(len(values)):
+            values[i].window_min = sum
+            sum += values[i].weight
+            values[i].window_max = sum
+
+        # Calculate what range of this stack ("window")
+        # we want to get the weighted average across.
+        window_min = sum * percentile_min
+        window_max = sum * percentile_max
+        window_width = window_max - window_min
+
+        # Ensure the window is within the stack and at least a certain size.
+        if window_width < min_width:
+            window_center = (window_min + window_max) / 2
+            window_min = window_center - min_width / 2
+            window_max = window_center + min_width / 2
+
+            if window_max > sum:
+                window_max = sum
+                window_min = sum - min_width
+
+            if window_min < 0:
+                window_min = 0
+                window_max = min_width
+
+        value = 0
+        value_weight = 0
+
+        # Get the weighted average of all the samples
+        # that overlap with the window, weighted
+        # by the size of their overlap.
+        for i in range(len(values)):
+            if window_min >= values[i].window_max:
+                continue
+            if window_max <= values[i].window_min:
+                break
+
+            s = max(window_min, values[i].window_min)
+            e = min(window_max, values[i].window_max)
+            w = e - s
+
+            value += values[i].value * w
+            value_weight += w
+
+        return value / value_weight if value_weight != 0 else 0
+
+    img_out = img.copy()
+
+    # Apply the kernel operation over each pixel.
+    for index in np.ndindex(img.shape):
+        img_out[index] = weighted_histogram_filter_single(index)
+
+    return img_out
+
+
+def smoothstep(x):
+    """
+    The smoothstep function, input should be clamped to 0-1 range.
+    Turns a diagonal line (f(x) = x) into a sigmoid-like curve.
+    """
+    return x * x * (3 - 2 * x)
+
+
+def smootherstep(x):
+    """
+    The smootherstep function, input should be clamped to 0-1 range.
+    Turns a diagonal line (f(x) = x) into a sigmoid-like curve.
+    """
+    return x * x * x * (x * (6 * x - 15) + 10)
+
+
+def get_gaussian_kernel(stddev_radius=1.0, max_radius=2):
+    """
+    Creates a Gaussian kernel with thresholded edges.
+
+    Args:
+        stddev_radius (float):
+            Standard deviation of the gaussian kernel, in pixels.
+        max_radius (int):
+            The size of the filter kernel. The number of pixels is (max_radius*2+1) ** 2.
+            The kernel is thresholded so that any values one pixel beyond this radius
+            is weighted at 0.
+
+    Returns:
+        (nparray, nparray): A kernel array (shape: (N, N)), its center coordinate (shape: (2))
+    """
+
+    # Evaluates a 0-1 normalized gaussian function for a given square distance from the mean.
+    def gaussian(sqr_mag):
+        return math.exp(-sqr_mag / (stddev_radius * stddev_radius))
+
+    # Helper function for converting a tuple to an array.
+    def vec(x):
+        return np.array(x)
+
+    """
+    Since a gaussian is unbounded, we need to limit ourselves
+    to a finite range.
+    We taper the ends off at the end of that range so they equal zero
+    while preserving the maximum value of 1 at the mean.
+    """
+    zero_radius = max_radius + 1.0
+    gauss_zero = gaussian(zero_radius * zero_radius)
+    gauss_kernel_scale = 1 / (1 - gauss_zero)
+
+    def gaussian_kernel_func(coordinate):
+        x = coordinate[0] ** 2.0 + coordinate[1] ** 2.0
+        x = gaussian(x)
+        x -= gauss_zero
+        x *= gauss_kernel_scale
+        x = max(0.0, x)
+        return x
+
+    size = max_radius * 2 + 1
+    kernel_center = max_radius
+    kernel = np.zeros((size, size))
+
+    for index in np.ndindex(kernel.shape):
+        kernel[index] = gaussian_kernel_func(vec(index) - kernel_center)
+
+    return kernel, kernel_center
+
+
+# ------------------- Constants -------------------
+
+
+default = SoftInpaintingSettings(1, 0.5, 4, 0, 0.5, 2)
+
+enabled_ui_label = "Soft inpainting"
+enabled_gen_param_label = "Soft inpainting enabled"
+enabled_el_id = "soft_inpainting_enabled"
+
+ui_labels = SoftInpaintingSettings(
+    "Schedule bias",
+    "Preservation strength",
+    "Transition contrast boost",
+    "Mask influence",
+    "Difference threshold",
+    "Difference contrast")
+
+ui_info = SoftInpaintingSettings(
+    "Shifts when preservation of original content occurs during denoising.",
+    "How strongly partially masked content should be preserved.",
+    "Amplifies the contrast that may be lost in partially masked regions.",
+    "How strongly the original mask should bias the difference threshold.",
+    "How much an image region can change before the original pixels are not blended in anymore.",
+    "How sharp the transition should be between blended and not blended.")
+
+gen_param_labels = SoftInpaintingSettings(
+    "Soft inpainting schedule bias",
+    "Soft inpainting preservation strength",
+    "Soft inpainting transition contrast boost",
+    "Soft inpainting mask influence",
+    "Soft inpainting difference threshold",
+    "Soft inpainting difference contrast")
+
+el_ids = SoftInpaintingSettings(
+    "mask_blend_power",
+    "mask_blend_scale",
+    "inpaint_detail_preservation",
+    "composite_mask_influence",
+    "composite_difference_threshold",
+    "composite_difference_contrast")
+
+
+# ------------------- Script -------------------
+
+
+class Script(scripts.Script):
+    def __init__(self):
+        self.section = "inpaint"
+        self.masks_for_overlay = None
+        self.overlay_images = None
+
+    def title(self):
+        return "Soft Inpainting"
+
+    def show(self, is_img2img):
+        return scripts.AlwaysVisible if is_img2img else False
+
+    def ui(self, is_img2img):
+        if not is_img2img:
+            return
+
+        with InputAccordion(False, label=enabled_ui_label, elem_id=enabled_el_id) as soft_inpainting_enabled:
+            with gr.Group():
+                gr.Markdown(
+                    """
+                    Soft inpainting allows you to **seamlessly blend original content with inpainted content** according to the mask opacity.
+                    **High _Mask blur_** values are recommended!
+                    """)
+
+                power = \
+                    gr.Slider(label=ui_labels.mask_blend_power,
+                              info=ui_info.mask_blend_power,
+                              minimum=0,
+                              maximum=8,
+                              step=0.1,
+                              value=default.mask_blend_power,
+                              elem_id=el_ids.mask_blend_power)
+                scale = \
+                    gr.Slider(label=ui_labels.mask_blend_scale,
+                              info=ui_info.mask_blend_scale,
+                              minimum=0,
+                              maximum=8,
+                              step=0.05,
+                              value=default.mask_blend_scale,
+                              elem_id=el_ids.mask_blend_scale)
+                detail = \
+                    gr.Slider(label=ui_labels.inpaint_detail_preservation,
+                              info=ui_info.inpaint_detail_preservation,
+                              minimum=1,
+                              maximum=32,
+                              step=0.5,
+                              value=default.inpaint_detail_preservation,
+                              elem_id=el_ids.inpaint_detail_preservation)
+
+                gr.Markdown(
+                    """
+                    ### Pixel Composite Settings
+                    """)
+
+                mask_inf = \
+                    gr.Slider(label=ui_labels.composite_mask_influence,
+                              info=ui_info.composite_mask_influence,
+                              minimum=0,
+                              maximum=1,
+                              step=0.05,
+                              value=default.composite_mask_influence,
+                              elem_id=el_ids.composite_mask_influence)
+
+                dif_thresh = \
+                    gr.Slider(label=ui_labels.composite_difference_threshold,
+                              info=ui_info.composite_difference_threshold,
+                              minimum=0,
+                              maximum=8,
+                              step=0.25,
+                              value=default.composite_difference_threshold,
+                              elem_id=el_ids.composite_difference_threshold)
+
+                dif_contr = \
+                    gr.Slider(label=ui_labels.composite_difference_contrast,
+                              info=ui_info.composite_difference_contrast,
+                              minimum=0,
+                              maximum=8,
+                              step=0.25,
+                              value=default.composite_difference_contrast,
+                              elem_id=el_ids.composite_difference_contrast)
+
+                with gr.Accordion("Help", open=False):
+                    gr.Markdown(
+                        f"""
+                        ### {ui_labels.mask_blend_power}
+
+                        The blending strength of original content is scaled proportionally with the decreasing noise level values at each step (sigmas).
+                        This ensures that the influence of the denoiser and original content preservation is roughly balanced at each step.
+                        This balance can be shifted using this parameter, controlling whether earlier or later steps have stronger preservation.
+
+                        - **Below 1**: Stronger preservation near the end (with low sigma)
+                        - **1**: Balanced (proportional to sigma)
+                        - **Above 1**: Stronger preservation in the beginning (with high sigma)
+                        """)
+                    gr.Markdown(
+                        f"""
+                        ### {ui_labels.mask_blend_scale}
+
+                        Skews whether partially masked image regions should be more likely to preserve the original content or favor inpainted content.
+                        This may need to be adjusted depending on the {ui_labels.mask_blend_power}, CFG Scale, prompt and Denoising strength.
+
+                        - **Low values**: Favors generated content.
+                        - **High values**: Favors original content.
+                        """)
+                    gr.Markdown(
+                        f"""
+                        ### {ui_labels.inpaint_detail_preservation}
+
+                        This parameter controls how the original latent vectors and denoised latent vectors are interpolated.
+                        With higher values, the magnitude of the resulting blended vector will be closer to the maximum of the two interpolated vectors.
+                        This can prevent the loss of contrast that occurs with linear interpolation.
+
+                        - **Low values**: Softer blending, details may fade.
+                        - **High values**: Stronger contrast, may over-saturate colors.
+                        """)
+
+                    gr.Markdown(
+                        """
+                        ## Pixel Composite Settings
+
+                        Masks are generated based on how much a part of the image changed after denoising.
+                        These masks are used to blend the original and final images together.
+                        If the difference is low, the original pixels are used instead of the pixels returned by the inpainting process.
+                        """)
+
+                    gr.Markdown(
+                        f"""
+                        ### {ui_labels.composite_mask_influence}
+
+                        This parameter controls how much the mask should bias this sensitivity to difference.
+
+                        - **0**: Ignore the mask, only consider differences in image content.
+                        - **1**: Follow the mask closely despite image content changes.
+                        """)
+
+                    gr.Markdown(
+                        f"""
+                        ### {ui_labels.composite_difference_threshold}
+
+                        This value represents the difference at which the original pixels will have less than 50% opacity.
+
+                        - **Low values**: Two images patches must be almost the same in order to retain original pixels.
+                        - **High values**: Two images patches can be very different and still retain original pixels.
+                        """)
+
+                    gr.Markdown(
+                        f"""
+                        ### {ui_labels.composite_difference_contrast}
+
+                        This value represents the contrast between the opacity of the original and inpainted content.
+
+                        - **Low values**: The blend will be more gradual and have longer transitions, but may cause ghosting.
+                        - **High values**: Ghosting will be less common, but transitions may be very sudden.
+                        """)
+
+        self.infotext_fields = [(soft_inpainting_enabled, enabled_gen_param_label),
+                                (power, gen_param_labels.mask_blend_power),
+                                (scale, gen_param_labels.mask_blend_scale),
+                                (detail, gen_param_labels.inpaint_detail_preservation),
+                                (mask_inf, gen_param_labels.composite_mask_influence),
+                                (dif_thresh, gen_param_labels.composite_difference_threshold),
+                                (dif_contr, gen_param_labels.composite_difference_contrast)]
+
+        self.paste_field_names = []
+        for _, field_name in self.infotext_fields:
+            self.paste_field_names.append(field_name)
+
+        return [soft_inpainting_enabled,
+                power,
+                scale,
+                detail,
+                mask_inf,
+                dif_thresh,
+                dif_contr]
+
+    def process(self, p, enabled, power, scale, detail_preservation, mask_inf, dif_thresh, dif_contr):
+        if not enabled:
+            return
+
+        if not processing_uses_inpainting(p):
+            return
+
+        # Shut off the rounding it normally does.
+        p.mask_round = False
+
+        settings = SoftInpaintingSettings(power, scale, detail_preservation, mask_inf, dif_thresh, dif_contr)
+
+        # p.extra_generation_params["Mask rounding"] = False
+        settings.add_generation_params(p.extra_generation_params)
+
+    def on_mask_blend(self, p, mba: scripts.MaskBlendArgs, enabled, power, scale, detail_preservation, mask_inf,
+                      dif_thresh, dif_contr):
+        if not enabled:
+            return
+
+        if not processing_uses_inpainting(p):
+            return
+
+        if mba.is_final_blend:
+            mba.blended_latent = mba.current_latent
+            return
+
+        settings = SoftInpaintingSettings(power, scale, detail_preservation, mask_inf, dif_thresh, dif_contr)
+
+        # todo: Why is sigma 2D? Both values are the same.
+        mba.blended_latent = latent_blend(settings,
+                                          mba.init_latent,
+                                          mba.current_latent,
+                                          get_modified_nmask(settings, mba.nmask, mba.sigma[0]))
+
+    def post_sample(self, p, ps: scripts.PostSampleArgs, enabled, power, scale, detail_preservation, mask_inf,
+                    dif_thresh, dif_contr):
+        if not enabled:
+            return
+
+        if not processing_uses_inpainting(p):
+            return
+
+        nmask = getattr(p, "nmask", None)
+        if nmask is None:
+            return
+
+        from modules import images
+        from modules.shared import opts
+
+        settings = SoftInpaintingSettings(power, scale, detail_preservation, mask_inf, dif_thresh, dif_contr)
+
+        # since the original code puts holes in the existing overlay images,
+        # we have to rebuild them.
+        self.overlay_images = []
+        for img in p.init_images:
+
+            image = images.flatten(img, opts.img2img_background_color)
+
+            if p.paste_to is None and p.resize_mode != 3:
+                image = images.resize_image(p.resize_mode, image, p.width, p.height)
+
+            self.overlay_images.append(image.convert('RGBA'))
+
+        if len(p.init_images) == 1:
+            self.overlay_images = self.overlay_images * p.batch_size
+
+        if getattr(ps.samples, 'already_decoded', False):
+            self.masks_for_overlay = apply_masks(settings=settings,
+                                                 nmask=nmask,
+                                                 overlay_images=self.overlay_images,
+                                                 width=p.width,
+                                                 height=p.height,
+                                                 paste_to=p.paste_to)
+        else:
+            self.masks_for_overlay = apply_adaptive_masks(settings=settings,
+                                                          nmask=nmask,
+                                                          latent_orig=p.init_latent,
+                                                          latent_processed=ps.samples,
+                                                          overlay_images=self.overlay_images,
+                                                          width=p.width,
+                                                          height=p.height,
+                                                          paste_to=p.paste_to)
+
+    def postprocess_maskoverlay(self, p, ppmo: scripts.PostProcessMaskOverlayArgs, enabled, power, scale,
+                                detail_preservation, mask_inf, dif_thresh, dif_contr):
+        if not enabled:
+            return
+
+        if not processing_uses_inpainting(p):
+            return
+
+        if self.masks_for_overlay is None:
+            return
+
+        if self.overlay_images is None:
+            return
+
+        ppmo.mask_for_overlay = self.masks_for_overlay[ppmo.index]
+        ppmo.overlay_image = self.overlay_images[ppmo.index]
diff --git a/html/extra-networks-card.html b/html/extra-networks-card.html
index 39674666f1e..f1d959a6733 100644
--- a/html/extra-networks-card.html
+++ b/html/extra-networks-card.html
@@ -1,14 +1,9 @@
-<div class='card' style={style} onclick={card_clicked} data-name="{name}" {sort_keys}>
+<div class="card" style="{style}" onclick="{card_clicked}" data-name="{name}" {sort_keys}>
 	{background_image}
-	<div class="button-row">
-		{metadata_button}
-		{edit_button}
-	</div>
-	<div class='actions'>
-		<div class='additional'>
-			<span style="display:none" class='search_term{search_only}'>{search_term}</span>
-		</div>
-		<span class='name'>{name}</span>
-		<span class='description'>{description}</span>
+	<div class="button-row">{copy_path_button}{metadata_button}{edit_button}</div>
+	<div class="actions">
+		<div class="additional">{search_terms}</div>
+		<span class="name">{name}</span>
+		<span class="description">{description}</span>
 	</div>
 </div>
diff --git a/html/extra-networks-copy-path-button.html b/html/extra-networks-copy-path-button.html
new file mode 100644
index 00000000000..50304b42d4b
--- /dev/null
+++ b/html/extra-networks-copy-path-button.html
@@ -0,0 +1,5 @@
+<div class="copy-path-button card-button"
+    title="Copy path to clipboard"
+    onclick="extraNetworksCopyCardPath(event)"
+    data-clipboard-text="{filename}">
+</div>
\ No newline at end of file
diff --git a/html/extra-networks-edit-item-button.html b/html/extra-networks-edit-item-button.html
new file mode 100644
index 00000000000..fd728600f6d
--- /dev/null
+++ b/html/extra-networks-edit-item-button.html
@@ -0,0 +1,4 @@
+<div class="edit-button card-button"
+    title="Edit metadata"
+    onclick="extraNetworksEditUserMetadata(event, '{tabname}', '{extra_networks_tabname}')">
+</div>
\ No newline at end of file
diff --git a/html/extra-networks-metadata-button.html b/html/extra-networks-metadata-button.html
new file mode 100644
index 00000000000..4ef013bc02a
--- /dev/null
+++ b/html/extra-networks-metadata-button.html
@@ -0,0 +1,4 @@
+<div class="metadata-button card-button"
+    title="Show internal metadata"
+    onclick="extraNetworksRequestMetadata(event, '{extra_networks_tabname}')">
+</div>
\ No newline at end of file
diff --git a/html/extra-networks-pane-dirs.html b/html/extra-networks-pane-dirs.html
new file mode 100644
index 00000000000..5ce04289a5a
--- /dev/null
+++ b/html/extra-networks-pane-dirs.html
@@ -0,0 +1,8 @@
+    <div class="extra-network-pane-content-dirs">
+        <div id='{tabname}_{extra_networks_tabname}_dirs' class='extra-network-dirs'>
+            {dirs_html}
+        </div>
+        <div id='{tabname}_{extra_networks_tabname}_cards' class='extra-network-cards'>
+            {items_html}
+        </div>
+    </div>
diff --git a/html/extra-networks-pane-tree.html b/html/extra-networks-pane-tree.html
new file mode 100644
index 00000000000..88561fcdc85
--- /dev/null
+++ b/html/extra-networks-pane-tree.html
@@ -0,0 +1,8 @@
+    <div class="extra-network-pane-content-tree resize-handle-row">
+        <div id='{tabname}_{extra_networks_tabname}_tree' class='extra-network-tree' style='flex-basis: {extra_networks_tree_view_default_width}px'>
+            {tree_html}
+        </div>
+        <div id='{tabname}_{extra_networks_tabname}_cards' class='extra-network-cards' style='flex-grow: 1;'>
+            {items_html}
+        </div>
+    </div>
\ No newline at end of file
diff --git a/html/extra-networks-pane.html b/html/extra-networks-pane.html
new file mode 100644
index 00000000000..9a67baea90f
--- /dev/null
+++ b/html/extra-networks-pane.html
@@ -0,0 +1,81 @@
+<div id='{tabname}_{extra_networks_tabname}_pane' class='extra-network-pane {tree_view_div_default_display_class}'>
+    <div class="extra-network-control" id="{tabname}_{extra_networks_tabname}_controls" style="display:none" >
+        <div class="extra-network-control--search">
+            <input
+                id="{tabname}_{extra_networks_tabname}_extra_search"
+                class="extra-network-control--search-text"
+                type="search"
+                placeholder="Search"
+            >
+        </div>
+
+        <small>Sort: </small>
+        <div 
+            id="{tabname}_{extra_networks_tabname}_extra_sort_path"
+            class="extra-network-control--sort{sort_path_active}"
+            data-sortkey="default"
+            title="Sort by path"
+            onclick="extraNetworksControlSortOnClick(event, '{tabname}', '{extra_networks_tabname}');"
+        >
+            <i class="extra-network-control--icon extra-network-control--sort-icon"></i>
+        </div>
+        <div
+            id="{tabname}_{extra_networks_tabname}_extra_sort_name"
+            class="extra-network-control--sort{sort_name_active}"
+            data-sortkey="name"
+            title="Sort by name"
+            onclick="extraNetworksControlSortOnClick(event, '{tabname}', '{extra_networks_tabname}');"
+        >
+            <i class="extra-network-control--icon extra-network-control--sort-icon"></i>
+        </div>
+        <div
+            id="{tabname}_{extra_networks_tabname}_extra_sort_date_created"
+            class="extra-network-control--sort{sort_date_created_active}"
+            data-sortkey="date_created"
+            title="Sort by date created"
+            onclick="extraNetworksControlSortOnClick(event, '{tabname}', '{extra_networks_tabname}');"
+        >
+            <i class="extra-network-control--icon extra-network-control--sort-icon"></i>
+        </div>
+        <div
+            id="{tabname}_{extra_networks_tabname}_extra_sort_date_modified"
+            class="extra-network-control--sort{sort_date_modified_active}"
+            data-sortkey="date_modified"
+            title="Sort by date modified"
+            onclick="extraNetworksControlSortOnClick(event, '{tabname}', '{extra_networks_tabname}');"
+        >
+            <i class="extra-network-control--icon extra-network-control--sort-icon"></i>
+        </div>
+
+        <small> </small>
+        <div
+            id="{tabname}_{extra_networks_tabname}_extra_sort_dir"
+            class="extra-network-control--sort-dir"
+            data-sortdir="{data_sortdir}"
+            title="Sort ascending"
+            onclick="extraNetworksControlSortDirOnClick(event, '{tabname}', '{extra_networks_tabname}');"
+        >
+            <i class="extra-network-control--icon extra-network-control--sort-dir-icon"></i>
+        </div>
+
+
+        <small> </small>
+        <div
+            id="{tabname}_{extra_networks_tabname}_extra_tree_view"
+            class="extra-network-control--tree-view {tree_view_btn_extra_class}"
+            title="Enable Tree View"
+            onclick="extraNetworksControlTreeViewOnClick(event, '{tabname}', '{extra_networks_tabname}');"
+        >
+            <i class="extra-network-control--icon extra-network-control--tree-view-icon"></i>
+        </div>
+        <div
+            id="{tabname}_{extra_networks_tabname}_extra_refresh"
+            class="extra-network-control--refresh"
+            title="Refresh page"
+            onclick="extraNetworksControlRefreshOnClick(event, '{tabname}', '{extra_networks_tabname}');"
+        >
+            <i class="extra-network-control--icon extra-network-control--refresh-icon"></i>
+        </div>
+    </div>
+    {pane_content}
+</div>
diff --git a/html/extra-networks-tree-button.html b/html/extra-networks-tree-button.html
new file mode 100644
index 00000000000..9dc2e2a40c8
--- /dev/null
+++ b/html/extra-networks-tree-button.html
@@ -0,0 +1,23 @@
+<span data-filterable-item-text hidden>{search_terms}</span>
+<div class="tree-list-content {subclass}"
+    type="button"
+    onclick="extraNetworksTreeOnClick(event, '{tabname}', '{extra_networks_tabname}');{onclick_extra}"
+    data-path="{data_path}"
+    data-hash="{data_hash}"
+>
+    <span class='tree-list-item-action tree-list-item-action--leading'>
+        {action_list_item_action_leading}
+    </span>
+    <span class="tree-list-item-visual tree-list-item-visual--leading">
+        {action_list_item_visual_leading}
+    </span>
+    <span class="tree-list-item-label tree-list-item-label--truncate">
+        {action_list_item_label}
+    </span>
+    <span class="tree-list-item-visual tree-list-item-visual--trailing">
+        {action_list_item_visual_trailing}
+    </span>
+    <span class="tree-list-item-action tree-list-item-action--trailing">
+        {action_list_item_action_trailing}
+    </span>
+</div>
\ No newline at end of file
diff --git a/html/licenses.html b/html/licenses.html
index ef6f2c0a42b..9f5d1e9dc5c 100644
--- a/html/licenses.html
+++ b/html/licenses.html
@@ -4,107 +4,6 @@
     #licenses pre { margin: 1em 0 2em 0;}
 </style>
 
-<h2><a href="https://github.com/sczhou/CodeFormer/blob/master/LICENSE">CodeFormer</a></h2>
-<small>Parts of CodeFormer code had to be copied to be compatible with GFPGAN.</small>
-<pre>
-S-Lab License 1.0
-
-Copyright 2022 S-Lab
-
-Redistribution and use for non-commercial purpose in source and
-binary forms, with or without modification, are permitted provided
-that the following conditions are met:
-
-1. Redistributions of source code must retain the above copyright
-   notice, this list of conditions and the following disclaimer.
-
-2. Redistributions in binary form must reproduce the above copyright
-   notice, this list of conditions and the following disclaimer in
-   the documentation and/or other materials provided with the
-   distribution.
-
-3. Neither the name of the copyright holder nor the names of its
-   contributors may be used to endorse or promote products derived
-   from this software without specific prior written permission.
-
-THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
-"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
-LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
-A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
-HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
-SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
-LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
-DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
-THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
-(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
-OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-
-In the event that redistribution and/or use for commercial purpose in
-source or binary forms, with or without modification is required,
-please contact the contributor(s) of the work.
-</pre>
-
-
-<h2><a href="https://github.com/victorca25/iNNfer/blob/main/LICENSE">ESRGAN</a></h2>
-<small>Code for architecture and reading models copied.</small>
-<pre>
-MIT License
-
-Copyright (c) 2021 victorca25
-
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
-</pre>
-
-<h2><a href="https://github.com/xinntao/Real-ESRGAN/blob/master/LICENSE">Real-ESRGAN</a></h2>
-<small>Some code is copied to support ESRGAN models.</small>
-<pre>
-BSD 3-Clause License
-
-Copyright (c) 2021, Xintao Wang
-All rights reserved.
-
-Redistribution and use in source and binary forms, with or without
-modification, are permitted provided that the following conditions are met:
-
-1. Redistributions of source code must retain the above copyright notice, this
-   list of conditions and the following disclaimer.
-
-2. Redistributions in binary form must reproduce the above copyright notice,
-   this list of conditions and the following disclaimer in the documentation
-   and/or other materials provided with the distribution.
-
-3. Neither the name of the copyright holder nor the names of its
-   contributors may be used to endorse or promote products derived from
-   this software without specific prior written permission.
-
-THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
-AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
-IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
-DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
-FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
-DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
-SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
-CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
-OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
-OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-</pre>
-
 <h2><a href="https://github.com/invoke-ai/InvokeAI/blob/main/LICENSE">InvokeAI</a></h2>
 <small>Some code for compatibility with OSX is taken from lstein's repository.</small>
 <pre>
@@ -183,213 +82,6 @@ <h2><a href="https://github.com/pharmapsychotic/clip-interrogator/blob/main/LICE
 SOFTWARE.
 </pre>
 
-<h2><a href="https://github.com/JingyunLiang/SwinIR/blob/main/LICENSE">SwinIR</a></h2>
-<small>Code added by contributors, most likely copied from this repository.</small>
-
-<pre>
-                                 Apache License
-                           Version 2.0, January 2004
-                        http://www.apache.org/licenses/
-
-   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
-
-   1. Definitions.
-
-      "License" shall mean the terms and conditions for use, reproduction,
-      and distribution as defined by Sections 1 through 9 of this document.
-
-      "Licensor" shall mean the copyright owner or entity authorized by
-      the copyright owner that is granting the License.
-
-      "Legal Entity" shall mean the union of the acting entity and all
-      other entities that control, are controlled by, or are under common
-      control with that entity. For the purposes of this definition,
-      "control" means (i) the power, direct or indirect, to cause the
-      direction or management of such entity, whether by contract or
-      otherwise, or (ii) ownership of fifty percent (50%) or more of the
-      outstanding shares, or (iii) beneficial ownership of such entity.
-
-      "You" (or "Your") shall mean an individual or Legal Entity
-      exercising permissions granted by this License.
-
-      "Source" form shall mean the preferred form for making modifications,
-      including but not limited to software source code, documentation
-      source, and configuration files.
-
-      "Object" form shall mean any form resulting from mechanical
-      transformation or translation of a Source form, including but
-      not limited to compiled object code, generated documentation,
-      and conversions to other media types.
-
-      "Work" shall mean the work of authorship, whether in Source or
-      Object form, made available under the License, as indicated by a
-      copyright notice that is included in or attached to the work
-      (an example is provided in the Appendix below).
-
-      "Derivative Works" shall mean any work, whether in Source or Object
-      form, that is based on (or derived from) the Work and for which the
-      editorial revisions, annotations, elaborations, or other modifications
-      represent, as a whole, an original work of authorship. For the purposes
-      of this License, Derivative Works shall not include works that remain
-      separable from, or merely link (or bind by name) to the interfaces of,
-      the Work and Derivative Works thereof.
-
-      "Contribution" shall mean any work of authorship, including
-      the original version of the Work and any modifications or additions
-      to that Work or Derivative Works thereof, that is intentionally
-      submitted to Licensor for inclusion in the Work by the copyright owner
-      or by an individual or Legal Entity authorized to submit on behalf of
-      the copyright owner. For the purposes of this definition, "submitted"
-      means any form of electronic, verbal, or written communication sent
-      to the Licensor or its representatives, including but not limited to
-      communication on electronic mailing lists, source code control systems,
-      and issue tracking systems that are managed by, or on behalf of, the
-      Licensor for the purpose of discussing and improving the Work, but
-      excluding communication that is conspicuously marked or otherwise
-      designated in writing by the copyright owner as "Not a Contribution."
-
-      "Contributor" shall mean Licensor and any individual or Legal Entity
-      on behalf of whom a Contribution has been received by Licensor and
-      subsequently incorporated within the Work.
-
-   2. Grant of Copyright License. Subject to the terms and conditions of
-      this License, each Contributor hereby grants to You a perpetual,
-      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      copyright license to reproduce, prepare Derivative Works of,
-      publicly display, publicly perform, sublicense, and distribute the
-      Work and such Derivative Works in Source or Object form.
-
-   3. Grant of Patent License. Subject to the terms and conditions of
-      this License, each Contributor hereby grants to You a perpetual,
-      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      (except as stated in this section) patent license to make, have made,
-      use, offer to sell, sell, import, and otherwise transfer the Work,
-      where such license applies only to those patent claims licensable
-      by such Contributor that are necessarily infringed by their
-      Contribution(s) alone or by combination of their Contribution(s)
-      with the Work to which such Contribution(s) was submitted. If You
-      institute patent litigation against any entity (including a
-      cross-claim or counterclaim in a lawsuit) alleging that the Work
-      or a Contribution incorporated within the Work constitutes direct
-      or contributory patent infringement, then any patent licenses
-      granted to You under this License for that Work shall terminate
-      as of the date such litigation is filed.
-
-   4. Redistribution. You may reproduce and distribute copies of the
-      Work or Derivative Works thereof in any medium, with or without
-      modifications, and in Source or Object form, provided that You
-      meet the following conditions:
-
-      (a) You must give any other recipients of the Work or
-          Derivative Works a copy of this License; and
-
-      (b) You must cause any modified files to carry prominent notices
-          stating that You changed the files; and
-
-      (c) You must retain, in the Source form of any Derivative Works
-          that You distribute, all copyright, patent, trademark, and
-          attribution notices from the Source form of the Work,
-          excluding those notices that do not pertain to any part of
-          the Derivative Works; and
-
-      (d) If the Work includes a "NOTICE" text file as part of its
-          distribution, then any Derivative Works that You distribute must
-          include a readable copy of the attribution notices contained
-          within such NOTICE file, excluding those notices that do not
-          pertain to any part of the Derivative Works, in at least one
-          of the following places: within a NOTICE text file distributed
-          as part of the Derivative Works; within the Source form or
-          documentation, if provided along with the Derivative Works; or,
-          within a display generated by the Derivative Works, if and
-          wherever such third-party notices normally appear. The contents
-          of the NOTICE file are for informational purposes only and
-          do not modify the License. You may add Your own attribution
-          notices within Derivative Works that You distribute, alongside
-          or as an addendum to the NOTICE text from the Work, provided
-          that such additional attribution notices cannot be construed
-          as modifying the License.
-
-      You may add Your own copyright statement to Your modifications and
-      may provide additional or different license terms and conditions
-      for use, reproduction, or distribution of Your modifications, or
-      for any such Derivative Works as a whole, provided Your use,
-      reproduction, and distribution of the Work otherwise complies with
-      the conditions stated in this License.
-
-   5. Submission of Contributions. Unless You explicitly state otherwise,
-      any Contribution intentionally submitted for inclusion in the Work
-      by You to the Licensor shall be under the terms and conditions of
-      this License, without any additional terms or conditions.
-      Notwithstanding the above, nothing herein shall supersede or modify
-      the terms of any separate license agreement you may have executed
-      with Licensor regarding such Contributions.
-
-   6. Trademarks. This License does not grant permission to use the trade
-      names, trademarks, service marks, or product names of the Licensor,
-      except as required for reasonable and customary use in describing the
-      origin of the Work and reproducing the content of the NOTICE file.
-
-   7. Disclaimer of Warranty. Unless required by applicable law or
-      agreed to in writing, Licensor provides the Work (and each
-      Contributor provides its Contributions) on an "AS IS" BASIS,
-      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
-      implied, including, without limitation, any warranties or conditions
-      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
-      PARTICULAR PURPOSE. You are solely responsible for determining the
-      appropriateness of using or redistributing the Work and assume any
-      risks associated with Your exercise of permissions under this License.
-
-   8. Limitation of Liability. In no event and under no legal theory,
-      whether in tort (including negligence), contract, or otherwise,
-      unless required by applicable law (such as deliberate and grossly
-      negligent acts) or agreed to in writing, shall any Contributor be
-      liable to You for damages, including any direct, indirect, special,
-      incidental, or consequential damages of any character arising as a
-      result of this License or out of the use or inability to use the
-      Work (including but not limited to damages for loss of goodwill,
-      work stoppage, computer failure or malfunction, or any and all
-      other commercial damages or losses), even if such Contributor
-      has been advised of the possibility of such damages.
-
-   9. Accepting Warranty or Additional Liability. While redistributing
-      the Work or Derivative Works thereof, You may choose to offer,
-      and charge a fee for, acceptance of support, warranty, indemnity,
-      or other liability obligations and/or rights consistent with this
-      License. However, in accepting such obligations, You may act only
-      on Your own behalf and on Your sole responsibility, not on behalf
-      of any other Contributor, and only if You agree to indemnify,
-      defend, and hold each Contributor harmless for any liability
-      incurred by, or claims asserted against, such Contributor by reason
-      of your accepting any such warranty or additional liability.
-
-   END OF TERMS AND CONDITIONS
-
-   APPENDIX: How to apply the Apache License to your work.
-
-      To apply the Apache License to your work, attach the following
-      boilerplate notice, with the fields enclosed by brackets "[]"
-      replaced with your own identifying information. (Don't include
-      the brackets!)  The text should be enclosed in the appropriate
-      comment syntax for the file format. We also recommend that a
-      file or class name and description of purpose be included on the
-      same "printed page" as the copyright notice for easier
-      identification within third-party archives.
-
-   Copyright [2021] [SwinIR Authors]
-
-   Licensed under the Apache License, Version 2.0 (the "License");
-   you may not use this file except in compliance with the License.
-   You may obtain a copy of the License at
-
-       http://www.apache.org/licenses/LICENSE-2.0
-
-   Unless required by applicable law or agreed to in writing, software
-   distributed under the License is distributed on an "AS IS" BASIS,
-   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-   See the License for the specific language governing permissions and
-   limitations under the License.
-</pre>
-
 <h2><a href="https://github.com/AminRezaei0x443/memory-efficient-attention/blob/main/LICENSE">Memory Efficient Attention</a></h2>
 <small>The sub-quadratic cross attention optimization uses modified code from the Memory Efficient Attention package that Alex Birch optimized for 3D tensors. This license is updated to reflect that.</small>
 <pre>
@@ -687,4 +379,4 @@ <h2><a href="https://github.com/madebyollin/taesd/blob/main/LICENSE">TAESD</a></
 LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 SOFTWARE.
-</pre>
\ No newline at end of file
+</pre>
diff --git a/javascript/aspectRatioOverlay.js b/javascript/aspectRatioOverlay.js
index 2cf2d571fc0..c8751fe494f 100644
--- a/javascript/aspectRatioOverlay.js
+++ b/javascript/aspectRatioOverlay.js
@@ -50,17 +50,17 @@ function dimensionChange(e, is_width, is_height) {
         var scaledx = targetElement.naturalWidth * viewportscale;
         var scaledy = targetElement.naturalHeight * viewportscale;
 
-        var cleintRectTop = (viewportOffset.top + window.scrollY);
-        var cleintRectLeft = (viewportOffset.left + window.scrollX);
-        var cleintRectCentreY = cleintRectTop + (targetElement.clientHeight / 2);
-        var cleintRectCentreX = cleintRectLeft + (targetElement.clientWidth / 2);
+        var clientRectTop = (viewportOffset.top + window.scrollY);
+        var clientRectLeft = (viewportOffset.left + window.scrollX);
+        var clientRectCentreY = clientRectTop + (targetElement.clientHeight / 2);
+        var clientRectCentreX = clientRectLeft + (targetElement.clientWidth / 2);
 
         var arscale = Math.min(scaledx / currentWidth, scaledy / currentHeight);
         var arscaledx = currentWidth * arscale;
         var arscaledy = currentHeight * arscale;
 
-        var arRectTop = cleintRectCentreY - (arscaledy / 2);
-        var arRectLeft = cleintRectCentreX - (arscaledx / 2);
+        var arRectTop = clientRectCentreY - (arscaledy / 2);
+        var arRectLeft = clientRectCentreX - (arscaledx / 2);
         var arRectWidth = arscaledx;
         var arRectHeight = arscaledy;
 
diff --git a/javascript/contextMenus.js b/javascript/contextMenus.js
index ccae242f2b6..e01fd67e80e 100644
--- a/javascript/contextMenus.js
+++ b/javascript/contextMenus.js
@@ -8,9 +8,6 @@ var contextMenuInit = function() {
     };
 
     function showContextMenu(event, element, menuEntries) {
-        let posx = event.clientX + document.body.scrollLeft + document.documentElement.scrollLeft;
-        let posy = event.clientY + document.body.scrollTop + document.documentElement.scrollTop;
-
         let oldMenu = gradioApp().querySelector('#context-menu');
         if (oldMenu) {
             oldMenu.remove();
@@ -23,10 +20,8 @@ var contextMenuInit = function() {
         contextMenu.style.background = baseStyle.background;
         contextMenu.style.color = baseStyle.color;
         contextMenu.style.fontFamily = baseStyle.fontFamily;
-        contextMenu.style.top = posy + 'px';
-        contextMenu.style.left = posx + 'px';
-
-
+        contextMenu.style.top = event.pageY + 'px';
+        contextMenu.style.left = event.pageX + 'px';
 
         const contextMenuList = document.createElement('ul');
         contextMenuList.className = 'context-menu-items';
@@ -43,21 +38,6 @@ var contextMenuInit = function() {
         });
 
         gradioApp().appendChild(contextMenu);
-
-        let menuWidth = contextMenu.offsetWidth + 4;
-        let menuHeight = contextMenu.offsetHeight + 4;
-
-        let windowWidth = window.innerWidth;
-        let windowHeight = window.innerHeight;
-
-        if ((windowWidth - posx) < menuWidth) {
-            contextMenu.style.left = windowWidth - menuWidth + "px";
-        }
-
-        if ((windowHeight - posy) < menuHeight) {
-            contextMenu.style.top = windowHeight - menuHeight + "px";
-        }
-
     }
 
     function appendContextMenuOption(targetElementSelector, entryName, entryFunction) {
@@ -107,16 +87,23 @@ var contextMenuInit = function() {
                 oldMenu.remove();
             }
         });
-        gradioApp().addEventListener("contextmenu", function(e) {
-            let oldMenu = gradioApp().querySelector('#context-menu');
-            if (oldMenu) {
-                oldMenu.remove();
-            }
-            menuSpecs.forEach(function(v, k) {
-                if (e.composedPath()[0].matches(k)) {
-                    showContextMenu(e, e.composedPath()[0], v);
-                    e.preventDefault();
+        ['contextmenu', 'touchstart'].forEach((eventType) => {
+            gradioApp().addEventListener(eventType, function(e) {
+                let ev = e;
+                if (eventType.startsWith('touch')) {
+                    if (e.touches.length !== 2) return;
+                    ev = e.touches[0];
+                }
+                let oldMenu = gradioApp().querySelector('#context-menu');
+                if (oldMenu) {
+                    oldMenu.remove();
                 }
+                menuSpecs.forEach(function(v, k) {
+                    if (e.composedPath()[0].matches(k)) {
+                        showContextMenu(ev, e.composedPath()[0], v);
+                        e.preventDefault();
+                    }
+                });
             });
         });
         eventListenerApplied = true;
diff --git a/javascript/dragdrop.js b/javascript/dragdrop.js
index 5803daea5ef..882562d7367 100644
--- a/javascript/dragdrop.js
+++ b/javascript/dragdrop.js
@@ -56,6 +56,15 @@ function eventHasFiles(e) {
     return false;
 }
 
+function isURL(url) {
+    try {
+        const _ = new URL(url);
+        return true;
+    } catch {
+        return false;
+    }
+}
+
 function dragDropTargetIsPrompt(target) {
     if (target?.placeholder && target?.placeholder.indexOf("Prompt") >= 0) return true;
     if (target?.parentNode?.parentNode?.className?.indexOf("prompt") > 0) return true;
@@ -74,22 +83,39 @@ window.document.addEventListener('dragover', e => {
     e.dataTransfer.dropEffect = 'copy';
 });
 
-window.document.addEventListener('drop', e => {
+window.document.addEventListener('drop', async e => {
     const target = e.composedPath()[0];
-    if (!eventHasFiles(e)) return;
+    const url = e.dataTransfer.getData('text/uri-list') || e.dataTransfer.getData('text/plain');
+    if (!eventHasFiles(e) && !isURL(url)) return;
 
     if (dragDropTargetIsPrompt(target)) {
         e.stopPropagation();
         e.preventDefault();
 
-        let prompt_target = get_tab_index('tabs') == 1 ? "img2img_prompt_image" : "txt2img_prompt_image";
+        const isImg2img = get_tab_index('tabs') == 1;
+        let prompt_image_target = isImg2img ? "img2img_prompt_image" : "txt2img_prompt_image";
 
-        const imgParent = gradioApp().getElementById(prompt_target);
+        const imgParent = gradioApp().getElementById(prompt_image_target);
         const files = e.dataTransfer.files;
         const fileInput = imgParent.querySelector('input[type="file"]');
-        if (fileInput) {
+        if (eventHasFiles(e) && fileInput) {
             fileInput.files = files;
             fileInput.dispatchEvent(new Event('change'));
+        } else if (url) {
+            try {
+                const request = await fetch(url);
+                if (!request.ok) {
+                    console.error('Error fetching URL:', url, request.status);
+                    return;
+                }
+                const data = new DataTransfer();
+                data.items.add(new File([await request.blob()], 'image.png'));
+                fileInput.files = data.files;
+                fileInput.dispatchEvent(new Event('change'));
+            } catch (error) {
+                console.error('Error fetching URL:', url, error);
+                return;
+            }
         }
     }
 
@@ -119,7 +145,7 @@ window.addEventListener('paste', e => {
     }
 
     const firstFreeImageField = visibleImageFields
-        .filter(el => el.querySelector('input[type=file]'))?.[0];
+        .filter(el => !el.querySelector('img'))?.[0];
 
     dropReplaceImage(
         firstFreeImageField ?
diff --git a/javascript/edit-attention.js b/javascript/edit-attention.js
index 8906c8922e1..b07ba97cb9b 100644
--- a/javascript/edit-attention.js
+++ b/javascript/edit-attention.js
@@ -18,37 +18,43 @@ function keyupEditAttention(event) {
         const before = text.substring(0, selectionStart);
         let beforeParen = before.lastIndexOf(OPEN);
         if (beforeParen == -1) return false;
-        let beforeParenClose = before.lastIndexOf(CLOSE);
-        while (beforeParenClose !== -1 && beforeParenClose > beforeParen) {
-            beforeParen = before.lastIndexOf(OPEN, beforeParen - 1);
-            beforeParenClose = before.lastIndexOf(CLOSE, beforeParenClose - 1);
-        }
+
+        let beforeClosingParen = before.lastIndexOf(CLOSE);
+        if (beforeClosingParen != -1 && beforeClosingParen > beforeParen) return false;
 
         // Find closing parenthesis around current cursor
         const after = text.substring(selectionStart);
         let afterParen = after.indexOf(CLOSE);
         if (afterParen == -1) return false;
-        let afterParenOpen = after.indexOf(OPEN);
-        while (afterParenOpen !== -1 && afterParen > afterParenOpen) {
-            afterParen = after.indexOf(CLOSE, afterParen + 1);
-            afterParenOpen = after.indexOf(OPEN, afterParenOpen + 1);
-        }
-        if (beforeParen === -1 || afterParen === -1) return false;
+
+        let afterOpeningParen = after.indexOf(OPEN);
+        if (afterOpeningParen != -1 && afterOpeningParen < afterParen) return false;
 
         // Set the selection to the text between the parenthesis
         const parenContent = text.substring(beforeParen + 1, selectionStart + afterParen);
-        const lastColon = parenContent.lastIndexOf(":");
-        selectionStart = beforeParen + 1;
-        selectionEnd = selectionStart + lastColon;
+        if (/.*:-?[\d.]+/s.test(parenContent)) {
+            const lastColon = parenContent.lastIndexOf(":");
+            selectionStart = beforeParen + 1;
+            selectionEnd = selectionStart + lastColon;
+        } else {
+            selectionStart = beforeParen + 1;
+            selectionEnd = selectionStart + parenContent.length;
+        }
+
         target.setSelectionRange(selectionStart, selectionEnd);
         return true;
     }
 
     function selectCurrentWord() {
         if (selectionStart !== selectionEnd) return false;
-        const delimiters = opts.keyedit_delimiters + " \r\n\t";
+        const whitespace_delimiters = {"Tab": "\t", "Carriage Return": "\r", "Line Feed": "\n"};
+        let delimiters = opts.keyedit_delimiters;
+
+        for (let i of opts.keyedit_delimiters_whitespace) {
+            delimiters += whitespace_delimiters[i];
+        }
 
-        // seek backward until to find beggining
+        // seek backward to find beginning
         while (!delimiters.includes(text[selectionStart - 1]) && selectionStart > 0) {
             selectionStart--;
         }
@@ -58,12 +64,20 @@ function keyupEditAttention(event) {
             selectionEnd++;
         }
 
+        // deselect surrounding whitespace
+        while (text[selectionStart] == " " && selectionStart < selectionEnd) {
+            selectionStart++;
+        }
+        while (text[selectionEnd - 1] == " " && selectionEnd > selectionStart) {
+            selectionEnd--;
+        }
+
         target.setSelectionRange(selectionStart, selectionEnd);
         return true;
     }
 
     // If the user hasn't selected anything, let's select their current parenthesis block or word
-    if (!selectCurrentParenthesisBlock('<', '>') && !selectCurrentParenthesisBlock('(', ')')) {
+    if (!selectCurrentParenthesisBlock('<', '>') && !selectCurrentParenthesisBlock('(', ')') && !selectCurrentParenthesisBlock('[', ']')) {
         selectCurrentWord();
     }
 
@@ -71,33 +85,54 @@ function keyupEditAttention(event) {
 
     var closeCharacter = ')';
     var delta = opts.keyedit_precision_attention;
+    var start = selectionStart > 0 ? text[selectionStart - 1] : "";
+    var end = text[selectionEnd];
 
-    if (selectionStart > 0 && text[selectionStart - 1] == '<') {
+    if (start == '<') {
         closeCharacter = '>';
         delta = opts.keyedit_precision_extra;
-    } else if (selectionStart == 0 || text[selectionStart - 1] != "(") {
+    } else if (start == '(' && end == ')' || start == '[' && end == ']') { // convert old-style (((emphasis)))
+        let numParen = 0;
+
+        while (text[selectionStart - numParen - 1] == start && text[selectionEnd + numParen] == end) {
+            numParen++;
+        }
 
+        if (start == "[") {
+            weight = (1 / 1.1) ** numParen;
+        } else {
+            weight = 1.1 ** numParen;
+        }
+
+        weight = Math.round(weight / opts.keyedit_precision_attention) * opts.keyedit_precision_attention;
+
+        text = text.slice(0, selectionStart - numParen) + "(" + text.slice(selectionStart, selectionEnd) + ":" + weight + ")" + text.slice(selectionEnd + numParen);
+        selectionStart -= numParen - 1;
+        selectionEnd -= numParen - 1;
+    } else if (start != '(') {
         // do not include spaces at the end
         while (selectionEnd > selectionStart && text[selectionEnd - 1] == ' ') {
-            selectionEnd -= 1;
+            selectionEnd--;
         }
+
         if (selectionStart == selectionEnd) {
             return;
         }
 
         text = text.slice(0, selectionStart) + "(" + text.slice(selectionStart, selectionEnd) + ":1.0)" + text.slice(selectionEnd);
 
-        selectionStart += 1;
-        selectionEnd += 1;
+        selectionStart++;
+        selectionEnd++;
     }
 
-    var end = text.slice(selectionEnd + 1).indexOf(closeCharacter) + 1;
-    var weight = parseFloat(text.slice(selectionEnd + 1, selectionEnd + 1 + end));
+    if (text[selectionEnd] != ':') return;
+    var weightLength = text.slice(selectionEnd + 1).indexOf(closeCharacter) + 1;
+    var weight = parseFloat(text.slice(selectionEnd + 1, selectionEnd + weightLength));
     if (isNaN(weight)) return;
 
     weight += isPlus ? delta : -delta;
     weight = parseFloat(weight.toPrecision(12));
-    if (String(weight).length == 1) weight += ".0";
+    if (Number.isInteger(weight)) weight += ".0";
 
     if (closeCharacter == ')' && weight == 1) {
         var endParenPos = text.substring(selectionEnd).indexOf(')');
@@ -105,7 +140,7 @@ function keyupEditAttention(event) {
         selectionStart--;
         selectionEnd--;
     } else {
-        text = text.slice(0, selectionEnd + 1) + weight + text.slice(selectionEnd + end);
+        text = text.slice(0, selectionEnd + 1) + weight + text.slice(selectionEnd + weightLength);
     }
 
     target.focus();
diff --git a/javascript/extensions.js b/javascript/extensions.js
index 312131b76eb..cc8ee220b17 100644
--- a/javascript/extensions.js
+++ b/javascript/extensions.js
@@ -2,8 +2,11 @@
 function extensions_apply(_disabled_list, _update_list, disable_all) {
     var disable = [];
     var update = [];
-
-    gradioApp().querySelectorAll('#extensions input[type="checkbox"]').forEach(function(x) {
+    const extensions_input = gradioApp().querySelectorAll('#extensions input[type="checkbox"]');
+    if (extensions_input.length == 0) {
+        throw Error("Extensions page not yet loaded.");
+    }
+    extensions_input.forEach(function(x) {
         if (x.name.startsWith("enable_") && !x.checked) {
             disable.push(x.name.substring(7));
         }
diff --git a/javascript/extraNetworks.js b/javascript/extraNetworks.js
index 493f31af28a..c5cced97399 100644
--- a/javascript/extraNetworks.js
+++ b/javascript/extraNetworks.js
@@ -16,143 +16,207 @@ function toggleCss(key, css, enable) {
 }
 
 function setupExtraNetworksForTab(tabname) {
-    gradioApp().querySelector('#' + tabname + '_extra_tabs').classList.add('extra-networks');
-
-    var tabs = gradioApp().querySelector('#' + tabname + '_extra_tabs > div');
-    var searchDiv = gradioApp().getElementById(tabname + '_extra_search');
-    var search = searchDiv.querySelector('textarea');
-    var sort = gradioApp().getElementById(tabname + '_extra_sort');
-    var sortOrder = gradioApp().getElementById(tabname + '_extra_sortorder');
-    var refresh = gradioApp().getElementById(tabname + '_extra_refresh');
-    var showDirsDiv = gradioApp().getElementById(tabname + '_extra_show_dirs');
-    var showDirs = gradioApp().querySelector('#' + tabname + '_extra_show_dirs input');
-
-    sort.dataset.sortkey = 'sortDefault';
-    tabs.appendChild(searchDiv);
-    tabs.appendChild(sort);
-    tabs.appendChild(sortOrder);
-    tabs.appendChild(refresh);
-    tabs.appendChild(showDirsDiv);
-
-    var applyFilter = function() {
-        var searchTerm = search.value.toLowerCase();
-
-        gradioApp().querySelectorAll('#' + tabname + '_extra_tabs div.card').forEach(function(elem) {
-            var searchOnly = elem.querySelector('.search_only');
-            var text = elem.querySelector('.name').textContent.toLowerCase() + " " + elem.querySelector('.search_term').textContent.toLowerCase();
-
-            var visible = text.indexOf(searchTerm) != -1;
-
-            if (searchOnly && searchTerm.length < 4) {
-                visible = false;
-            }
+    function registerPrompt(tabname, id) {
+        var textarea = gradioApp().querySelector("#" + id + " > label > textarea");
 
-            elem.style.display = visible ? "" : "none";
+        if (!activePromptTextarea[tabname]) {
+            activePromptTextarea[tabname] = textarea;
+        }
+
+        textarea.addEventListener("focus", function() {
+            activePromptTextarea[tabname] = textarea;
         });
-    };
+    }
 
-    var applySort = function() {
-        var reverse = sortOrder.classList.contains("sortReverse");
-        var sortKey = sort.querySelector("input").value.toLowerCase().replace("sort", "").replaceAll(" ", "_").replace(/_+$/, "").trim();
-        sortKey = sortKey ? "sort" + sortKey.charAt(0).toUpperCase() + sortKey.slice(1) : "";
-        var sortKeyStore = sortKey ? sortKey + (reverse ? "Reverse" : "") : "";
-        if (!sortKey || sortKeyStore == sort.dataset.sortkey) {
-            return;
+    var tabnav = gradioApp().querySelector('#' + tabname + '_extra_tabs > div.tab-nav');
+    var controlsDiv = document.createElement('DIV');
+    controlsDiv.classList.add('extra-networks-controls-div');
+    tabnav.appendChild(controlsDiv);
+    tabnav.insertBefore(controlsDiv, null);
+
+    var this_tab = gradioApp().querySelector('#' + tabname + '_extra_tabs');
+    this_tab.querySelectorAll(":scope > [id^='" + tabname + "_']").forEach(function(elem) {
+        // tabname_full = {tabname}_{extra_networks_tabname}
+        var tabname_full = elem.id;
+        var search = gradioApp().querySelector("#" + tabname_full + "_extra_search");
+        var sort_dir = gradioApp().querySelector("#" + tabname_full + "_extra_sort_dir");
+        var refresh = gradioApp().querySelector("#" + tabname_full + "_extra_refresh");
+        var currentSort = '';
+
+        // If any of the buttons above don't exist, we want to skip this iteration of the loop.
+        if (!search || !sort_dir || !refresh) {
+            return; // `return` is equivalent of `continue` but for forEach loops.
         }
 
-        sort.dataset.sortkey = sortKeyStore;
+        var applyFilter = function(force) {
+            var searchTerm = search.value.toLowerCase();
+            gradioApp().querySelectorAll('#' + tabname + '_extra_tabs div.card').forEach(function(elem) {
+                var searchOnly = elem.querySelector('.search_only');
+                var text = Array.prototype.map.call(elem.querySelectorAll('.search_terms, .description'), function(t) {
+                    return t.textContent.toLowerCase();
+                }).join(" ");
+
+                var visible = text.indexOf(searchTerm) != -1;
+                if (searchOnly && searchTerm.length < 4) {
+                    visible = false;
+                }
+                if (visible) {
+                    elem.classList.remove("hidden");
+                } else {
+                    elem.classList.add("hidden");
+                }
+            });
 
-        var cards = gradioApp().querySelectorAll('#' + tabname + '_extra_tabs div.card');
-        cards.forEach(function(card) {
-            card.originalParentElement = card.parentElement;
-        });
-        var sortedCards = Array.from(cards);
-        sortedCards.sort(function(cardA, cardB) {
-            var a = cardA.dataset[sortKey];
-            var b = cardB.dataset[sortKey];
-            if (!isNaN(a) && !isNaN(b)) {
-                return parseInt(a) - parseInt(b);
+            applySort(force);
+        };
+
+        var applySort = function(force) {
+            var cards = gradioApp().querySelectorAll('#' + tabname_full + ' div.card');
+            var parent = gradioApp().querySelector('#' + tabname_full + "_cards");
+            var reverse = sort_dir.dataset.sortdir == "Descending";
+            var activeSearchElem = gradioApp().querySelector('#' + tabname_full + "_controls .extra-network-control--sort.extra-network-control--enabled");
+            var sortKey = activeSearchElem ? activeSearchElem.dataset.sortkey : "default";
+            var sortKeyDataField = "sort" + sortKey.charAt(0).toUpperCase() + sortKey.slice(1);
+            var sortKeyStore = sortKey + "-" + sort_dir.dataset.sortdir + "-" + cards.length;
+
+            if (sortKeyStore == currentSort && !force) {
+                return;
             }
+            currentSort = sortKeyStore;
+
+            var sortedCards = Array.from(cards);
+            sortedCards.sort(function(cardA, cardB) {
+                var a = cardA.dataset[sortKeyDataField];
+                var b = cardB.dataset[sortKeyDataField];
+                if (!isNaN(a) && !isNaN(b)) {
+                    return parseInt(a) - parseInt(b);
+                }
 
-            return (a < b ? -1 : (a > b ? 1 : 0));
-        });
-        if (reverse) {
-            sortedCards.reverse();
-        }
-        cards.forEach(function(card) {
-            card.remove();
-        });
-        sortedCards.forEach(function(card) {
-            card.originalParentElement.appendChild(card);
+                return (a < b ? -1 : (a > b ? 1 : 0));
+            });
+
+            if (reverse) {
+                sortedCards.reverse();
+            }
+
+            parent.innerHTML = '';
+
+            var frag = document.createDocumentFragment();
+            sortedCards.forEach(function(card) {
+                frag.appendChild(card);
+            });
+            parent.appendChild(frag);
+        };
+
+        search.addEventListener("input", function() {
+            applyFilter();
         });
-    };
+        applySort();
+        applyFilter();
+        extraNetworksApplySort[tabname_full] = applySort;
+        extraNetworksApplyFilter[tabname_full] = applyFilter;
 
-    search.addEventListener("input", applyFilter);
-    applyFilter();
-    ["change", "blur", "click"].forEach(function(evt) {
-        sort.querySelector("input").addEventListener(evt, applySort);
+        var controls = gradioApp().querySelector("#" + tabname_full + "_controls");
+        controlsDiv.insertBefore(controls, null);
+
+        if (elem.style.display != "none") {
+            extraNetworksShowControlsForPage(tabname, tabname_full);
+        }
     });
-    sortOrder.addEventListener("click", function() {
-        sortOrder.classList.toggle("sortReverse");
-        applySort();
+
+    registerPrompt(tabname, tabname + "_prompt");
+    registerPrompt(tabname, tabname + "_neg_prompt");
+}
+
+function extraNetworksMovePromptToTab(tabname, id, showPrompt, showNegativePrompt) {
+    if (!gradioApp().querySelector('.toprow-compact-tools')) return; // only applicable for compact prompt layout
+
+    var promptContainer = gradioApp().getElementById(tabname + '_prompt_container');
+    var prompt = gradioApp().getElementById(tabname + '_prompt_row');
+    var negPrompt = gradioApp().getElementById(tabname + '_neg_prompt_row');
+    var elem = id ? gradioApp().getElementById(id) : null;
+
+    if (showNegativePrompt && elem) {
+        elem.insertBefore(negPrompt, elem.firstChild);
+    } else {
+        promptContainer.insertBefore(negPrompt, promptContainer.firstChild);
+    }
+
+    if (showPrompt && elem) {
+        elem.insertBefore(prompt, elem.firstChild);
+    } else {
+        promptContainer.insertBefore(prompt, promptContainer.firstChild);
+    }
+
+    if (elem) {
+        elem.classList.toggle('extra-page-prompts-active', showNegativePrompt || showPrompt);
+    }
+}
+
+
+function extraNetworksShowControlsForPage(tabname, tabname_full) {
+    gradioApp().querySelectorAll('#' + tabname + '_extra_tabs .extra-networks-controls-div > div').forEach(function(elem) {
+        var targetId = tabname_full + "_controls";
+        elem.style.display = elem.id == targetId ? "" : "none";
     });
+}
+
+
+function extraNetworksUnrelatedTabSelected(tabname) { // called from python when user selects an unrelated tab (generate)
+    extraNetworksMovePromptToTab(tabname, '', false, false);
+
+    extraNetworksShowControlsForPage(tabname, null);
+}
 
-    extraNetworksApplyFilter[tabname] = applyFilter;
+function extraNetworksTabSelected(tabname, id, showPrompt, showNegativePrompt, tabname_full) { // called from python when user selects an extra networks tab
+    extraNetworksMovePromptToTab(tabname, id, showPrompt, showNegativePrompt);
 
-    var showDirsUpdate = function() {
-        var css = '#' + tabname + '_extra_tabs .extra-network-subdirs { display: none; }';
-        toggleCss(tabname + '_extra_show_dirs_style', css, !showDirs.checked);
-        localSet('extra-networks-show-dirs', showDirs.checked ? 1 : 0);
+    extraNetworksShowControlsForPage(tabname, tabname_full);
+}
+
+function applyExtraNetworkFilter(tabname_full) {
+    var doFilter = function() {
+        var applyFunction = extraNetworksApplyFilter[tabname_full];
+
+        if (applyFunction) {
+            applyFunction(true);
+        }
     };
-    showDirs.checked = localGet('extra-networks-show-dirs', 1) == 1;
-    showDirs.addEventListener("change", showDirsUpdate);
-    showDirsUpdate();
+    setTimeout(doFilter, 1);
 }
 
-function applyExtraNetworkFilter(tabname) {
-    setTimeout(extraNetworksApplyFilter[tabname], 1);
+function applyExtraNetworkSort(tabname_full) {
+    var doSort = function() {
+        extraNetworksApplySort[tabname_full](true);
+    };
+    setTimeout(doSort, 1);
 }
 
 var extraNetworksApplyFilter = {};
+var extraNetworksApplySort = {};
 var activePromptTextarea = {};
 
 function setupExtraNetworks() {
     setupExtraNetworksForTab('txt2img');
     setupExtraNetworksForTab('img2img');
-
-    function registerPrompt(tabname, id) {
-        var textarea = gradioApp().querySelector("#" + id + " > label > textarea");
-
-        if (!activePromptTextarea[tabname]) {
-            activePromptTextarea[tabname] = textarea;
-        }
-
-        textarea.addEventListener("focus", function() {
-            activePromptTextarea[tabname] = textarea;
-        });
-    }
-
-    registerPrompt('txt2img', 'txt2img_prompt');
-    registerPrompt('txt2img', 'txt2img_neg_prompt');
-    registerPrompt('img2img', 'img2img_prompt');
-    registerPrompt('img2img', 'img2img_neg_prompt');
 }
 
-onUiLoaded(setupExtraNetworks);
+var re_extranet = /<([^:^>]+:[^:]+):[\d.]+>(.*)/;
+var re_extranet_g = /<([^:^>]+:[^:]+):[\d.]+>/g;
 
-var re_extranet = /<([^:]+:[^:]+):[\d.]+>(.*)/;
-var re_extranet_g = /\s+<([^:]+:[^:]+):[\d.]+>/g;
-
-function tryToRemoveExtraNetworkFromPrompt(textarea, text) {
-    var m = text.match(re_extranet);
+var re_extranet_neg = /\(([^:^>]+:[\d.]+)\)/;
+var re_extranet_g_neg = /\(([^:^>]+:[\d.]+)\)/g;
+function tryToRemoveExtraNetworkFromPrompt(textarea, text, isNeg) {
+    var m = text.match(isNeg ? re_extranet_neg : re_extranet);
     var replaced = false;
     var newTextareaText;
+    var extraTextBeforeNet = opts.extra_networks_add_text_separator;
     if (m) {
         var extraTextAfterNet = m[2];
         var partToSearch = m[1];
         var foundAtPosition = -1;
-        newTextareaText = textarea.value.replaceAll(re_extranet_g, function(found, net, pos) {
-            m = found.match(re_extranet);
+        newTextareaText = textarea.value.replaceAll(isNeg ? re_extranet_g_neg : re_extranet_g, function(found, net, pos) {
+            m = found.match(isNeg ? re_extranet_neg : re_extranet);
             if (m[1] == partToSearch) {
                 replaced = true;
                 foundAtPosition = pos;
@@ -160,18 +224,17 @@ function tryToRemoveExtraNetworkFromPrompt(textarea, text) {
             }
             return found;
         });
-
-        if (foundAtPosition >= 0 && newTextareaText.substr(foundAtPosition, extraTextAfterNet.length) == extraTextAfterNet) {
-            newTextareaText = newTextareaText.substr(0, foundAtPosition) + newTextareaText.substr(foundAtPosition + extraTextAfterNet.length);
+        if (foundAtPosition >= 0) {
+            if (extraTextAfterNet && newTextareaText.substr(foundAtPosition, extraTextAfterNet.length) == extraTextAfterNet) {
+                newTextareaText = newTextareaText.substr(0, foundAtPosition) + newTextareaText.substr(foundAtPosition + extraTextAfterNet.length);
+            }
+            if (newTextareaText.substr(foundAtPosition - extraTextBeforeNet.length, extraTextBeforeNet.length) == extraTextBeforeNet) {
+                newTextareaText = newTextareaText.substr(0, foundAtPosition - extraTextBeforeNet.length) + newTextareaText.substr(foundAtPosition);
+            }
         }
     } else {
-        newTextareaText = textarea.value.replaceAll(new RegExp(text, "g"), function(found) {
-            if (found == text) {
-                replaced = true;
-                return "";
-            }
-            return found;
-        });
+        newTextareaText = textarea.value.replaceAll(new RegExp(`((?:${extraTextBeforeNet})?${text})`, "g"), "");
+        replaced = (newTextareaText != textarea.value);
     }
 
     if (replaced) {
@@ -182,14 +245,22 @@ function tryToRemoveExtraNetworkFromPrompt(textarea, text) {
     return false;
 }
 
-function cardClicked(tabname, textToAdd, allowNegativePrompt) {
-    var textarea = allowNegativePrompt ? activePromptTextarea[tabname] : gradioApp().querySelector("#" + tabname + "_prompt > label > textarea");
-
-    if (!tryToRemoveExtraNetworkFromPrompt(textarea, textToAdd)) {
-        textarea.value = textarea.value + opts.extra_networks_add_text_separator + textToAdd;
+function updatePromptArea(text, textArea, isNeg) {
+    if (!tryToRemoveExtraNetworkFromPrompt(textArea, text, isNeg)) {
+        textArea.value = textArea.value + opts.extra_networks_add_text_separator + text;
     }
 
-    updateInput(textarea);
+    updateInput(textArea);
+}
+
+function cardClicked(tabname, textToAdd, textToAddNegative, allowNegativePrompt) {
+    if (textToAddNegative.length > 0) {
+        updatePromptArea(textToAdd, gradioApp().querySelector("#" + tabname + "_prompt > label > textarea"));
+        updatePromptArea(textToAddNegative, gradioApp().querySelector("#" + tabname + "_neg_prompt > label > textarea"), true);
+    } else {
+        var textarea = allowNegativePrompt ? activePromptTextarea[tabname] : gradioApp().querySelector("#" + tabname + "_prompt > label > textarea");
+        updatePromptArea(textToAdd, textarea);
+    }
 }
 
 function saveCardPreview(event, tabname, filename) {
@@ -205,8 +276,8 @@ function saveCardPreview(event, tabname, filename) {
     event.preventDefault();
 }
 
-function extraNetworksSearchButton(tabs_id, event) {
-    var searchTextarea = gradioApp().querySelector("#" + tabs_id + ' > label > textarea');
+function extraNetworksSearchButton(tabname, extra_networks_tabname, event) {
+    var searchTextarea = gradioApp().querySelector("#" + tabname + "_" + extra_networks_tabname + "_extra_search");
     var button = event.target;
     var text = button.classList.contains("search-all") ? "" : button.textContent.trim();
 
@@ -214,29 +285,207 @@ function extraNetworksSearchButton(tabs_id, event) {
     updateInput(searchTextarea);
 }
 
+function extraNetworksTreeProcessFileClick(event, btn, tabname, extra_networks_tabname) {
+    /**
+     * Processes `onclick` events when user clicks on files in tree.
+     *
+     * @param event                     The generated event.
+     * @param btn                       The clicked `tree-list-item` button.
+     * @param tabname                   The name of the active tab in the sd webui. Ex: txt2img, img2img, etc.
+     * @param extra_networks_tabname    The id of the active extraNetworks tab. Ex: lora, checkpoints, etc.
+     */
+    // NOTE: Currently unused.
+    return;
+}
+
+function extraNetworksTreeProcessDirectoryClick(event, btn, tabname, extra_networks_tabname) {
+    /**
+     * Processes `onclick` events when user clicks on directories in tree.
+     *
+     * Here is how the tree reacts to clicks for various states:
+     * unselected unopened directory: Directory is selected and expanded.
+     * unselected opened directory: Directory is selected.
+     * selected opened directory: Directory is collapsed and deselected.
+     * chevron is clicked: Directory is expanded or collapsed. Selected state unchanged.
+     *
+     * @param event                     The generated event.
+     * @param btn                       The clicked `tree-list-item` button.
+     * @param tabname                   The name of the active tab in the sd webui. Ex: txt2img, img2img, etc.
+     * @param extra_networks_tabname    The id of the active extraNetworks tab. Ex: lora, checkpoints, etc.
+     */
+    var ul = btn.nextElementSibling;
+    // This is the actual target that the user clicked on within the target button.
+    // We use this to detect if the chevron was clicked.
+    var true_targ = event.target;
+
+    function _expand_or_collapse(_ul, _btn) {
+        // Expands <ul> if it is collapsed, collapses otherwise. Updates button attributes.
+        if (_ul.hasAttribute("hidden")) {
+            _ul.removeAttribute("hidden");
+            _btn.dataset.expanded = "";
+        } else {
+            _ul.setAttribute("hidden", "");
+            delete _btn.dataset.expanded;
+        }
+    }
+
+    function _remove_selected_from_all() {
+        // Removes the `selected` attribute from all buttons.
+        var sels = document.querySelectorAll("div.tree-list-content");
+        [...sels].forEach(el => {
+            delete el.dataset.selected;
+        });
+    }
+
+    function _select_button(_btn) {
+        // Removes `data-selected` attribute from all buttons then adds to passed button.
+        _remove_selected_from_all();
+        _btn.dataset.selected = "";
+    }
+
+    function _update_search(_tabname, _extra_networks_tabname, _search_text) {
+        // Update search input with select button's path.
+        var search_input_elem = gradioApp().querySelector("#" + tabname + "_" + extra_networks_tabname + "_extra_search");
+        search_input_elem.value = _search_text;
+        updateInput(search_input_elem);
+    }
+
+
+    // If user clicks on the chevron, then we do not select the folder.
+    if (true_targ.matches(".tree-list-item-action--leading, .tree-list-item-action-chevron")) {
+        _expand_or_collapse(ul, btn);
+    } else {
+        // User clicked anywhere else on the button.
+        if ("selected" in btn.dataset && !(ul.hasAttribute("hidden"))) {
+            // If folder is select and open, collapse and deselect button.
+            _expand_or_collapse(ul, btn);
+            delete btn.dataset.selected;
+            _update_search(tabname, extra_networks_tabname, "");
+        } else if (!(!("selected" in btn.dataset) && !(ul.hasAttribute("hidden")))) {
+            // If folder is open and not selected, then we don't collapse; just select.
+            // NOTE: Double inversion sucks but it is the clearest way to show the branching here.
+            _expand_or_collapse(ul, btn);
+            _select_button(btn, tabname, extra_networks_tabname);
+            _update_search(tabname, extra_networks_tabname, btn.dataset.path);
+        } else {
+            // All other cases, just select the button.
+            _select_button(btn, tabname, extra_networks_tabname);
+            _update_search(tabname, extra_networks_tabname, btn.dataset.path);
+        }
+    }
+}
+
+function extraNetworksTreeOnClick(event, tabname, extra_networks_tabname) {
+    /**
+     * Handles `onclick` events for buttons within an `extra-network-tree .tree-list--tree`.
+     *
+     * Determines whether the clicked button in the tree is for a file entry or a directory
+     * then calls the appropriate function.
+     *
+     * @param event                     The generated event.
+     * @param tabname                   The name of the active tab in the sd webui. Ex: txt2img, img2img, etc.
+     * @param extra_networks_tabname    The id of the active extraNetworks tab. Ex: lora, checkpoints, etc.
+     */
+    var btn = event.currentTarget;
+    var par = btn.parentElement;
+    if (par.dataset.treeEntryType === "file") {
+        extraNetworksTreeProcessFileClick(event, btn, tabname, extra_networks_tabname);
+    } else {
+        extraNetworksTreeProcessDirectoryClick(event, btn, tabname, extra_networks_tabname);
+    }
+}
+
+function extraNetworksControlSortOnClick(event, tabname, extra_networks_tabname) {
+    /** Handles `onclick` events for Sort Mode buttons. */
+
+    var self = event.currentTarget;
+    var parent = event.currentTarget.parentElement;
+
+    parent.querySelectorAll('.extra-network-control--sort').forEach(function(x) {
+        x.classList.remove('extra-network-control--enabled');
+    });
+
+    self.classList.add('extra-network-control--enabled');
+
+    applyExtraNetworkSort(tabname + "_" + extra_networks_tabname);
+}
+
+function extraNetworksControlSortDirOnClick(event, tabname, extra_networks_tabname) {
+    /**
+     * Handles `onclick` events for the Sort Direction button.
+     *
+     * Modifies the data attributes of the Sort Direction button to cycle between
+     * ascending and descending sort directions.
+     *
+     * @param event                     The generated event.
+     * @param tabname                   The name of the active tab in the sd webui. Ex: txt2img, img2img, etc.
+     * @param extra_networks_tabname    The id of the active extraNetworks tab. Ex: lora, checkpoints, etc.
+     */
+    if (event.currentTarget.dataset.sortdir == "Ascending") {
+        event.currentTarget.dataset.sortdir = "Descending";
+        event.currentTarget.setAttribute("title", "Sort descending");
+    } else {
+        event.currentTarget.dataset.sortdir = "Ascending";
+        event.currentTarget.setAttribute("title", "Sort ascending");
+    }
+    applyExtraNetworkSort(tabname + "_" + extra_networks_tabname);
+}
+
+function extraNetworksControlTreeViewOnClick(event, tabname, extra_networks_tabname) {
+    /**
+     * Handles `onclick` events for the Tree View button.
+     *
+     * Toggles the tree view in the extra networks pane.
+     *
+     * @param event                     The generated event.
+     * @param tabname                   The name of the active tab in the sd webui. Ex: txt2img, img2img, etc.
+     * @param extra_networks_tabname    The id of the active extraNetworks tab. Ex: lora, checkpoints, etc.
+     */
+    var button = event.currentTarget;
+    button.classList.toggle("extra-network-control--enabled");
+    var show = !button.classList.contains("extra-network-control--enabled");
+
+    var pane = gradioApp().getElementById(tabname + "_" + extra_networks_tabname + "_pane");
+    pane.classList.toggle("extra-network-dirs-hidden", show);
+}
+
+function extraNetworksControlRefreshOnClick(event, tabname, extra_networks_tabname) {
+    /**
+     * Handles `onclick` events for the Refresh Page button.
+     *
+     * In order to actually call the python functions in `ui_extra_networks.py`
+     * to refresh the page, we created an empty gradio button in that file with an
+     * event handler that refreshes the page. So what this function here does
+     * is it manually raises a `click` event on that button.
+     *
+     * @param event                     The generated event.
+     * @param tabname                   The name of the active tab in the sd webui. Ex: txt2img, img2img, etc.
+     * @param extra_networks_tabname    The id of the active extraNetworks tab. Ex: lora, checkpoints, etc.
+     */
+    var btn_refresh_internal = gradioApp().getElementById(tabname + "_" + extra_networks_tabname + "_extra_refresh_internal");
+    btn_refresh_internal.dispatchEvent(new Event("click"));
+}
+
 var globalPopup = null;
 var globalPopupInner = null;
+
 function closePopup() {
     if (!globalPopup) return;
-
     globalPopup.style.display = "none";
 }
+
 function popup(contents) {
     if (!globalPopup) {
         globalPopup = document.createElement('div');
-        globalPopup.onclick = closePopup;
         globalPopup.classList.add('global-popup');
 
         var close = document.createElement('div');
         close.classList.add('global-popup-close');
-        close.onclick = closePopup;
+        close.addEventListener("click", closePopup);
         close.title = "Close";
         globalPopup.appendChild(close);
 
         globalPopupInner = document.createElement('div');
-        globalPopupInner.onclick = function(event) {
-            event.stopPropagation(); return false;
-        };
         globalPopupInner.classList.add('global-popup-inner');
         globalPopup.appendChild(globalPopupInner);
 
@@ -258,12 +507,76 @@ function popupId(id) {
     popup(storedPopupIds[id]);
 }
 
+function extraNetworksFlattenMetadata(obj) {
+    const result = {};
+
+    // Convert any stringified JSON objects to actual objects
+    for (const key of Object.keys(obj)) {
+        if (typeof obj[key] === 'string') {
+            try {
+                const parsed = JSON.parse(obj[key]);
+                if (parsed && typeof parsed === 'object') {
+                    obj[key] = parsed;
+                }
+            } catch (error) {
+                continue;
+            }
+        }
+    }
+
+    // Flatten the object
+    for (const key of Object.keys(obj)) {
+        if (typeof obj[key] === 'object' && obj[key] !== null) {
+            const nested = extraNetworksFlattenMetadata(obj[key]);
+            for (const nestedKey of Object.keys(nested)) {
+                result[`${key}/${nestedKey}`] = nested[nestedKey];
+            }
+        } else {
+            result[key] = obj[key];
+        }
+    }
+
+    // Special case for handling modelspec keys
+    for (const key of Object.keys(result)) {
+        if (key.startsWith("modelspec.")) {
+            result[key.replaceAll(".", "/")] = result[key];
+            delete result[key];
+        }
+    }
+
+    // Add empty keys to designate hierarchy
+    for (const key of Object.keys(result)) {
+        const parts = key.split("/");
+        for (let i = 1; i < parts.length; i++) {
+            const parent = parts.slice(0, i).join("/");
+            if (!result[parent]) {
+                result[parent] = "";
+            }
+        }
+    }
+
+    return result;
+}
+
 function extraNetworksShowMetadata(text) {
+    try {
+        let parsed = JSON.parse(text);
+        if (parsed && typeof parsed === 'object') {
+            parsed = extraNetworksFlattenMetadata(parsed);
+            const table = createVisualizationTable(parsed, 0);
+            popup(table);
+            return;
+        }
+    } catch (error) {
+        console.error(error);
+    }
+
     var elem = document.createElement('pre');
     elem.classList.add('popup-metadata');
     elem.textContent = text;
 
     popup(elem);
+    return;
 }
 
 function requestGet(url, data, handler, errorHandler) {
@@ -292,11 +605,18 @@ function requestGet(url, data, handler, errorHandler) {
     xhr.send(js);
 }
 
-function extraNetworksRequestMetadata(event, extraPage, cardName) {
+function extraNetworksCopyCardPath(event) {
+    navigator.clipboard.writeText(event.target.getAttribute("data-clipboard-text"));
+    event.stopPropagation();
+}
+
+function extraNetworksRequestMetadata(event, extraPage) {
     var showError = function() {
         extraNetworksShowMetadata("there was an error getting metadata");
     };
 
+    var cardName = event.target.parentElement.parentElement.getAttribute("data-name");
+
     requestGet("./sd_extra_networks/metadata", {page: extraPage, item: cardName}, function(data) {
         if (data && data.metadata) {
             extraNetworksShowMetadata(data.metadata);
@@ -310,7 +630,7 @@ function extraNetworksRequestMetadata(event, extraPage, cardName) {
 
 var extraPageUserMetadataEditors = {};
 
-function extraNetworksEditUserMetadata(event, tabname, extraPage, cardName) {
+function extraNetworksEditUserMetadata(event, tabname, extraPage) {
     var id = tabname + '_' + extraPage + '_edit_user_metadata';
 
     var editor = extraPageUserMetadataEditors[id];
@@ -322,6 +642,7 @@ function extraNetworksEditUserMetadata(event, tabname, extraPage, cardName) {
         extraPageUserMetadataEditors[id] = editor;
     }
 
+    var cardName = event.target.parentElement.parentElement.getAttribute("data-name");
     editor.nameTextarea.value = cardName;
     updateInput(editor.nameTextarea);
 
@@ -335,7 +656,7 @@ function extraNetworksEditUserMetadata(event, tabname, extraPage, cardName) {
 function extraNetworksRefreshSingleCard(page, tabname, name) {
     requestGet("./sd_extra_networks/get-single-card", {page: page, tabname: tabname, name: name}, function(data) {
         if (data && data.html) {
-            var card = gradioApp().querySelector('.card[data-name=' + JSON.stringify(name) + ']'); // likely using the wrong stringify function
+            var card = gradioApp().querySelector(`#${tabname}_${page.replace(" ", "_")}_cards > .card[data-name="${name}"]`);
 
             var newDiv = document.createElement('DIV');
             newDiv.innerHTML = data.html;
@@ -347,3 +668,45 @@ function extraNetworksRefreshSingleCard(page, tabname, name) {
         }
     });
 }
+
+window.addEventListener("keydown", function(event) {
+    if (event.key == "Escape") {
+        closePopup();
+    }
+});
+
+/**
+ * Setup custom loading for this script.
+ * We need to wait for all of our HTML to be generated in the extra networks tabs
+ * before we can actually run the `setupExtraNetworks` function.
+ * The `onUiLoaded` function actually runs before all of our extra network tabs are
+ * finished generating. Thus we needed this new method.
+ *
+ */
+
+var uiAfterScriptsCallbacks = [];
+var uiAfterScriptsTimeout = null;
+var executedAfterScripts = false;
+
+function scheduleAfterScriptsCallbacks() {
+    clearTimeout(uiAfterScriptsTimeout);
+    uiAfterScriptsTimeout = setTimeout(function() {
+        executeCallbacks(uiAfterScriptsCallbacks);
+    }, 200);
+}
+
+onUiLoaded(function() {
+    var mutationObserver = new MutationObserver(function(m) {
+        let existingSearchfields = gradioApp().querySelectorAll("[id$='_extra_search']").length;
+        let neededSearchfields = gradioApp().querySelectorAll("[id$='_extra_tabs'] > .tab-nav > button").length - 2;
+
+        if (!executedAfterScripts && existingSearchfields >= neededSearchfields) {
+            mutationObserver.disconnect();
+            executedAfterScripts = true;
+            scheduleAfterScriptsCallbacks();
+        }
+    });
+    mutationObserver.observe(gradioApp(), {childList: true, subtree: true});
+});
+
+uiAfterScriptsCallbacks.push(setupExtraNetworks);
diff --git a/javascript/imageviewer.js b/javascript/imageviewer.js
index c21d396eefd..9b23f4700b3 100644
--- a/javascript/imageviewer.js
+++ b/javascript/imageviewer.js
@@ -6,6 +6,8 @@ function closeModal() {
 function showModal(event) {
     const source = event.target || event.srcElement;
     const modalImage = gradioApp().getElementById("modalImage");
+    const modalToggleLivePreviewBtn = gradioApp().getElementById("modal_toggle_live_preview");
+    modalToggleLivePreviewBtn.innerHTML = opts.js_live_preview_in_modal_lightbox ? "&#x1F5C7;" : "&#x1F5C6;";
     const lb = gradioApp().getElementById("lightboxModal");
     modalImage.src = source.src;
     if (modalImage.style.display === 'none') {
@@ -33,8 +35,11 @@ function updateOnBackgroundChange() {
     const modalImage = gradioApp().getElementById("modalImage");
     if (modalImage && modalImage.offsetParent) {
         let currentButton = selected_gallery_button();
-
-        if (currentButton?.children?.length > 0 && modalImage.src != currentButton.children[0].src) {
+        let preview = gradioApp().querySelectorAll('.livePreview > img');
+        if (opts.js_live_preview_in_modal_lightbox && preview.length > 0) {
+            // show preview image if available
+            modalImage.src = preview[preview.length - 1].src;
+        } else if (currentButton?.children?.length > 0 && modalImage.src != currentButton.children[0].src) {
             modalImage.src = currentButton.children[0].src;
             if (modalImage.style.display === 'none') {
                 const modal = gradioApp().getElementById("lightboxModal");
@@ -48,14 +53,7 @@ function modalImageSwitch(offset) {
     var galleryButtons = all_gallery_buttons();
 
     if (galleryButtons.length > 1) {
-        var currentButton = selected_gallery_button();
-
-        var result = -1;
-        galleryButtons.forEach(function(v, i) {
-            if (v == currentButton) {
-                result = i;
-            }
-        });
+        var result = selected_gallery_index();
 
         if (result != -1) {
             var nextButton = galleryButtons[negmod((result + offset), galleryButtons.length)];
@@ -128,19 +126,15 @@ function setupImageForLightbox(e) {
     e.style.cursor = 'pointer';
     e.style.userSelect = 'none';
 
-    var isFirefox = navigator.userAgent.toLowerCase().indexOf('firefox') > -1;
-
-    // For Firefox, listening on click first switched to next image then shows the lightbox.
-    // If you know how to fix this without switching to mousedown event, please.
-    // For other browsers the event is click to make it possiblr to drag picture.
-    var event = isFirefox ? 'mousedown' : 'click';
-
-    e.addEventListener(event, function(evt) {
+    e.addEventListener('mousedown', function(evt) {
         if (evt.button == 1) {
             open(evt.target.src);
             evt.preventDefault();
             return;
         }
+    }, true);
+
+    e.addEventListener('click', function(evt) {
         if (!opts.js_modal_lightbox || evt.button != 0) return;
 
         modalZoomSet(gradioApp().getElementById('modalImage'), opts.js_modal_lightbox_initially_zoomed);
@@ -160,6 +154,13 @@ function modalZoomToggle(event) {
     event.stopPropagation();
 }
 
+function modalLivePreviewToggle(event) {
+    const modalToggleLivePreview = gradioApp().getElementById("modal_toggle_live_preview");
+    opts.js_live_preview_in_modal_lightbox = !opts.js_live_preview_in_modal_lightbox;
+    modalToggleLivePreview.innerHTML = opts.js_live_preview_in_modal_lightbox ? "&#x1F5C7;" : "&#x1F5C6;";
+    event.stopPropagation();
+}
+
 function modalTileImageToggle(event) {
     const modalImage = gradioApp().getElementById("modalImage");
     const modal = gradioApp().getElementById("lightboxModal");
@@ -217,6 +218,14 @@ document.addEventListener("DOMContentLoaded", function() {
     modalSave.title = "Save Image(s)";
     modalControls.appendChild(modalSave);
 
+    const modalToggleLivePreview = document.createElement('span');
+    modalToggleLivePreview.className = 'modalToggleLivePreview cursor';
+    modalToggleLivePreview.id = "modal_toggle_live_preview";
+    modalToggleLivePreview.innerHTML = "&#x1F5C6;";
+    modalToggleLivePreview.onclick = modalLivePreviewToggle;
+    modalToggleLivePreview.title = "Toggle live preview";
+    modalControls.appendChild(modalToggleLivePreview);
+
     const modalClose = document.createElement('span');
     modalClose.className = 'modalClose cursor';
     modalClose.innerHTML = '&times;';
diff --git a/javascript/inputAccordion.js b/javascript/inputAccordion.js
index f2839852ee7..7570309aa73 100644
--- a/javascript/inputAccordion.js
+++ b/javascript/inputAccordion.js
@@ -1,37 +1,68 @@
-var observerAccordionOpen = new MutationObserver(function(mutations) {
-    mutations.forEach(function(mutationRecord) {
-        var elem = mutationRecord.target;
-        var open = elem.classList.contains('open');
+function inputAccordionChecked(id, checked) {
+    var accordion = gradioApp().getElementById(id);
+    accordion.visibleCheckbox.checked = checked;
+    accordion.onVisibleCheckboxChange();
+}
 
-        var accordion = elem.parentNode;
-        accordion.classList.toggle('input-accordion-open', open);
+function setupAccordion(accordion) {
+    var labelWrap = accordion.querySelector('.label-wrap');
+    var gradioCheckbox = gradioApp().querySelector('#' + accordion.id + "-checkbox input");
+    var extra = gradioApp().querySelector('#' + accordion.id + "-extra");
+    var span = labelWrap.querySelector('span');
+    var linked = true;
 
-        var checkbox = gradioApp().querySelector('#' + accordion.id + "-checkbox input");
-        checkbox.checked = open;
-        updateInput(checkbox);
+    var isOpen = function() {
+        return labelWrap.classList.contains('open');
+    };
 
-        var extra = gradioApp().querySelector('#' + accordion.id + "-extra");
-        if (extra) {
-            extra.style.display = open ? "" : "none";
-        }
+    var observerAccordionOpen = new MutationObserver(function(mutations) {
+        mutations.forEach(function(mutationRecord) {
+            accordion.classList.toggle('input-accordion-open', isOpen());
+
+            if (linked) {
+                accordion.visibleCheckbox.checked = isOpen();
+                accordion.onVisibleCheckboxChange();
+            }
+        });
     });
-});
+    observerAccordionOpen.observe(labelWrap, {attributes: true, attributeFilter: ['class']});
 
-function inputAccordionChecked(id, checked) {
-    var label = gradioApp().querySelector('#' + id + " .label-wrap");
-    if (label.classList.contains('open') != checked) {
-        label.click();
+    if (extra) {
+        labelWrap.insertBefore(extra, labelWrap.lastElementChild);
     }
+
+    accordion.onChecked = function(checked) {
+        if (isOpen() != checked) {
+            labelWrap.click();
+        }
+    };
+
+    var visibleCheckbox = document.createElement('INPUT');
+    visibleCheckbox.type = 'checkbox';
+    visibleCheckbox.checked = isOpen();
+    visibleCheckbox.id = accordion.id + "-visible-checkbox";
+    visibleCheckbox.className = gradioCheckbox.className + " input-accordion-checkbox";
+    span.insertBefore(visibleCheckbox, span.firstChild);
+
+    accordion.visibleCheckbox = visibleCheckbox;
+    accordion.onVisibleCheckboxChange = function() {
+        if (linked && isOpen() != visibleCheckbox.checked) {
+            labelWrap.click();
+        }
+
+        gradioCheckbox.checked = visibleCheckbox.checked;
+        updateInput(gradioCheckbox);
+    };
+
+    visibleCheckbox.addEventListener('click', function(event) {
+        linked = false;
+        event.stopPropagation();
+    });
+    visibleCheckbox.addEventListener('input', accordion.onVisibleCheckboxChange);
 }
 
 onUiLoaded(function() {
     for (var accordion of gradioApp().querySelectorAll('.input-accordion')) {
-        var labelWrap = accordion.querySelector('.label-wrap');
-        observerAccordionOpen.observe(labelWrap, {attributes: true, attributeFilter: ['class']});
-
-        var extra = gradioApp().querySelector('#' + accordion.id + "-extra");
-        if (extra) {
-            labelWrap.insertBefore(extra, labelWrap.lastElementChild);
-        }
+        setupAccordion(accordion);
     }
 });
diff --git a/javascript/notification.js b/javascript/notification.js
index 6d79956125c..3ee972ae166 100644
--- a/javascript/notification.js
+++ b/javascript/notification.js
@@ -26,7 +26,11 @@ onAfterUiUpdate(function() {
     lastHeadImg = headImg;
 
     // play notification sound if available
-    gradioApp().querySelector('#audio_notification audio')?.play();
+    const notificationAudio = gradioApp().querySelector('#audio_notification audio');
+    if (notificationAudio) {
+        notificationAudio.volume = opts.notification_volume / 100.0 || 1.0;
+        notificationAudio.play();
+    }
 
     if (document.hasFocus()) return;
 
diff --git a/javascript/profilerVisualization.js b/javascript/profilerVisualization.js
index 9d8e5f42f32..9822f4b2a2a 100644
--- a/javascript/profilerVisualization.js
+++ b/javascript/profilerVisualization.js
@@ -33,120 +33,141 @@ function createRow(table, cellName, items) {
     return res;
 }
 
-function showProfile(path, cutoff = 0.05) {
-    requestGet(path, {}, function(data) {
-        var table = document.createElement('table');
-        table.className = 'popup-table';
-
-        data.records['total'] = data.total;
-        var keys = Object.keys(data.records).sort(function(a, b) {
-            return data.records[b] - data.records[a];
+function createVisualizationTable(data, cutoff = 0, sort = "") {
+    var table = document.createElement('table');
+    table.className = 'popup-table';
+
+    var keys = Object.keys(data);
+    if (sort === "number") {
+        keys = keys.sort(function(a, b) {
+            return data[b] - data[a];
         });
-        var items = keys.map(function(x) {
-            return {key: x, parts: x.split('/'), time: data.records[x]};
+    } else {
+        keys = keys.sort();
+    }
+    var items = keys.map(function(x) {
+        return {key: x, parts: x.split('/'), value: data[x]};
+    });
+    var maxLength = items.reduce(function(a, b) {
+        return Math.max(a, b.parts.length);
+    }, 0);
+
+    var cols = createRow(
+        table,
+        'th',
+        [
+            cutoff === 0 ? 'key' : 'record',
+            cutoff === 0 ? 'value' : 'seconds'
+        ]
+    );
+    cols[0].colSpan = maxLength;
+
+    function arraysEqual(a, b) {
+        return !(a < b || b < a);
+    }
+
+    var addLevel = function(level, parent, hide) {
+        var matching = items.filter(function(x) {
+            return x.parts[level] && !x.parts[level + 1] && arraysEqual(x.parts.slice(0, level), parent);
         });
-        var maxLength = items.reduce(function(a, b) {
-            return Math.max(a, b.parts.length);
-        }, 0);
-
-        var cols = createRow(table, 'th', ['record', 'seconds']);
-        cols[0].colSpan = maxLength;
-
-        function arraysEqual(a, b) {
-            return !(a < b || b < a);
+        if (sort === "number") {
+            matching = matching.sort(function(a, b) {
+                return b.value - a.value;
+            });
+        } else {
+            matching = matching.sort();
         }
+        var othersTime = 0;
+        var othersList = [];
+        var othersRows = [];
+        var childrenRows = [];
+        matching.forEach(function(x) {
+            var visible = (cutoff === 0 && !hide) || (x.value >= cutoff && !hide);
+
+            var cells = [];
+            for (var i = 0; i < maxLength; i++) {
+                cells.push(x.parts[i]);
+            }
+            cells.push(cutoff === 0 ? x.value : x.value.toFixed(3));
+            var cols = createRow(table, 'td', cells);
+            for (i = 0; i < level; i++) {
+                cols[i].className = 'muted';
+            }
 
-        var addLevel = function(level, parent, hide) {
-            var matching = items.filter(function(x) {
-                return x.parts[level] && !x.parts[level + 1] && arraysEqual(x.parts.slice(0, level), parent);
-            });
-            var sorted = matching.sort(function(a, b) {
-                return b.time - a.time;
-            });
-            var othersTime = 0;
-            var othersList = [];
-            var othersRows = [];
-            var childrenRows = [];
-            sorted.forEach(function(x) {
-                var visible = x.time >= cutoff && !hide;
-
-                var cells = [];
-                for (var i = 0; i < maxLength; i++) {
-                    cells.push(x.parts[i]);
-                }
-                cells.push(x.time.toFixed(3));
-                var cols = createRow(table, 'td', cells);
-                for (i = 0; i < level; i++) {
-                    cols[i].className = 'muted';
-                }
-
-                var tr = cols[0].parentNode;
-                if (!visible) {
-                    tr.classList.add("hidden");
-                }
-
-                if (x.time >= cutoff) {
-                    childrenRows.push(tr);
-                } else {
-                    othersTime += x.time;
-                    othersList.push(x.parts[level]);
-                    othersRows.push(tr);
-                }
-
-                var children = addLevel(level + 1, parent.concat([x.parts[level]]), true);
-                if (children.length > 0) {
-                    var cell = cols[level];
-                    var onclick = function() {
-                        cell.classList.remove("link");
-                        cell.removeEventListener("click", onclick);
-                        children.forEach(function(x) {
-                            x.classList.remove("hidden");
-                        });
-                    };
-                    cell.classList.add("link");
-                    cell.addEventListener("click", onclick);
-                }
-            });
+            var tr = cols[0].parentNode;
+            if (!visible) {
+                tr.classList.add("hidden");
+            }
 
-            if (othersTime > 0) {
-                var cells = [];
-                for (var i = 0; i < maxLength; i++) {
-                    cells.push(parent[i]);
-                }
-                cells.push(othersTime.toFixed(3));
-                cells[level] = 'others';
-                var cols = createRow(table, 'td', cells);
-                for (i = 0; i < level; i++) {
-                    cols[i].className = 'muted';
-                }
+            if (cutoff === 0 || x.value >= cutoff) {
+                childrenRows.push(tr);
+            } else {
+                othersTime += x.value;
+                othersList.push(x.parts[level]);
+                othersRows.push(tr);
+            }
 
+            var children = addLevel(level + 1, parent.concat([x.parts[level]]), true);
+            if (children.length > 0) {
                 var cell = cols[level];
-                var tr = cell.parentNode;
                 var onclick = function() {
-                    tr.classList.add("hidden");
                     cell.classList.remove("link");
                     cell.removeEventListener("click", onclick);
-                    othersRows.forEach(function(x) {
+                    children.forEach(function(x) {
                         x.classList.remove("hidden");
                     });
                 };
-
-                cell.title = othersList.join(", ");
                 cell.classList.add("link");
                 cell.addEventListener("click", onclick);
+            }
+        });
 
-                if (hide) {
-                    tr.classList.add("hidden");
-                }
+        if (othersTime > 0) {
+            var cells = [];
+            for (var i = 0; i < maxLength; i++) {
+                cells.push(parent[i]);
+            }
+            cells.push(othersTime.toFixed(3));
+            cells[level] = 'others';
+            var cols = createRow(table, 'td', cells);
+            for (i = 0; i < level; i++) {
+                cols[i].className = 'muted';
+            }
 
-                childrenRows.push(tr);
+            var cell = cols[level];
+            var tr = cell.parentNode;
+            var onclick = function() {
+                tr.classList.add("hidden");
+                cell.classList.remove("link");
+                cell.removeEventListener("click", onclick);
+                othersRows.forEach(function(x) {
+                    x.classList.remove("hidden");
+                });
+            };
+
+            cell.title = othersList.join(", ");
+            cell.classList.add("link");
+            cell.addEventListener("click", onclick);
+
+            if (hide) {
+                tr.classList.add("hidden");
             }
 
-            return childrenRows;
-        };
+            childrenRows.push(tr);
+        }
+
+        return childrenRows;
+    };
 
-        addLevel(0, []);
+    addLevel(0, []);
+
+    return table;
+}
 
+function showProfile(path, cutoff = 0.05) {
+    requestGet(path, {}, function(data) {
+        data.records['total'] = data.total;
+        const table = createVisualizationTable(data.records, cutoff, "number");
         popup(table);
     });
 }
diff --git a/javascript/progressbar.js b/javascript/progressbar.js
index 777614954b2..23dea64ceda 100644
--- a/javascript/progressbar.js
+++ b/javascript/progressbar.js
@@ -45,8 +45,15 @@ function formatTime(secs) {
     }
 }
 
+
+var originalAppTitle = undefined;
+
+onUiLoaded(function() {
+    originalAppTitle = document.title;
+});
+
 function setTitle(progress) {
-    var title = 'Stable Diffusion';
+    var title = originalAppTitle;
 
     if (opts.show_progress_in_title && progress) {
         title = '[' + progress.trim() + '] ' + title;
@@ -69,6 +76,26 @@ function requestProgress(id_task, progressbarContainer, gallery, atEnd, onProgre
     var dateStart = new Date();
     var wasEverActive = false;
     var parentProgressbar = progressbarContainer.parentNode;
+    var wakeLock = null;
+
+    var requestWakeLock = async function() {
+        if (!opts.prevent_screen_sleep_during_generation || wakeLock) return;
+        try {
+            wakeLock = await navigator.wakeLock.request('screen');
+        } catch (err) {
+            console.error('Wake Lock is not supported.');
+        }
+    };
+
+    var releaseWakeLock = async function() {
+        if (!opts.prevent_screen_sleep_during_generation || !wakeLock) return;
+        try {
+            await wakeLock.release();
+            wakeLock = null;
+        } catch (err) {
+            console.error('Wake Lock release failed', err);
+        }
+    };
 
     var divProgress = document.createElement('div');
     divProgress.className = 'progressDiv';
@@ -82,6 +109,7 @@ function requestProgress(id_task, progressbarContainer, gallery, atEnd, onProgre
     var livePreview = null;
 
     var removeProgressBar = function() {
+        releaseWakeLock();
         if (!divProgress) return;
 
         setTitle("");
@@ -93,6 +121,7 @@ function requestProgress(id_task, progressbarContainer, gallery, atEnd, onProgre
     };
 
     var funProgress = function(id_task) {
+        requestWakeLock();
         request("./internal/progress", {id_task: id_task, live_preview: false}, function(res) {
             if (res.completed) {
                 removeProgressBar();
diff --git a/javascript/resizeHandle.js b/javascript/resizeHandle.js
index 8c5c5169210..4aeb14b41f3 100644
--- a/javascript/resizeHandle.js
+++ b/javascript/resizeHandle.js
@@ -1,8 +1,8 @@
 (function() {
     const GRADIO_MIN_WIDTH = 320;
-    const GRID_TEMPLATE_COLUMNS = '1fr 16px 1fr';
     const PAD = 16;
     const DEBOUNCE_TIME = 100;
+    const DOUBLE_TAP_DELAY = 200; //ms
 
     const R = {
         tracking: false,
@@ -11,6 +11,7 @@
         leftCol: null,
         leftColStartWidth: null,
         screenX: null,
+        lastTapTime: null,
     };
 
     let resizeTimer;
@@ -21,30 +22,29 @@
     }
 
     function displayResizeHandle(parent) {
+        if (!parent.needHideOnMoblie) {
+            return true;
+        }
         if (window.innerWidth < GRADIO_MIN_WIDTH * 2 + PAD * 4) {
             parent.style.display = 'flex';
-            if (R.handle != null) {
-                R.handle.style.opacity = '0';
-            }
+            parent.resizeHandle.style.display = "none";
             return false;
         } else {
             parent.style.display = 'grid';
-            if (R.handle != null) {
-                R.handle.style.opacity = '100';
-            }
+            parent.resizeHandle.style.display = "block";
             return true;
         }
     }
 
     function afterResize(parent) {
-        if (displayResizeHandle(parent) && parent.style.gridTemplateColumns != GRID_TEMPLATE_COLUMNS) {
+        if (displayResizeHandle(parent) && parent.style.gridTemplateColumns != parent.style.originalGridTemplateColumns) {
             const oldParentWidth = R.parentWidth;
             const newParentWidth = parent.offsetWidth;
             const widthL = parseInt(parent.style.gridTemplateColumns.split(' ')[0]);
 
             const ratio = newParentWidth / oldParentWidth;
 
-            const newWidthL = Math.max(Math.floor(ratio * widthL), GRADIO_MIN_WIDTH);
+            const newWidthL = Math.max(Math.floor(ratio * widthL), parent.minLeftColWidth);
             setLeftColGridTemplate(parent, newWidthL);
 
             R.parentWidth = newParentWidth;
@@ -52,6 +52,14 @@
     }
 
     function setup(parent) {
+
+        function onDoubleClick(evt) {
+            evt.preventDefault();
+            evt.stopPropagation();
+
+            parent.style.gridTemplateColumns = parent.style.originalGridTemplateColumns;
+        }
+
         const leftCol = parent.firstElementChild;
         const rightCol = parent.lastElementChild;
 
@@ -59,63 +67,114 @@
 
         parent.style.display = 'grid';
         parent.style.gap = '0';
-        parent.style.gridTemplateColumns = GRID_TEMPLATE_COLUMNS;
+        let leftColTemplate = "";
+        if (parent.children[0].style.flexGrow) {
+            leftColTemplate = `${parent.children[0].style.flexGrow}fr`;
+            parent.minLeftColWidth = GRADIO_MIN_WIDTH;
+            parent.minRightColWidth = GRADIO_MIN_WIDTH;
+            parent.needHideOnMoblie = true;
+        } else {
+            leftColTemplate = parent.children[0].style.flexBasis;
+            parent.minLeftColWidth = parent.children[0].style.flexBasis.slice(0, -2) / 2;
+            parent.minRightColWidth = 0;
+            parent.needHideOnMoblie = false;
+        }
+
+        if (!leftColTemplate) {
+            leftColTemplate = '1fr';
+        }
+
+        const gridTemplateColumns = `${leftColTemplate} ${PAD}px ${parent.children[1].style.flexGrow}fr`;
+        parent.style.gridTemplateColumns = gridTemplateColumns;
+        parent.style.originalGridTemplateColumns = gridTemplateColumns;
 
         const resizeHandle = document.createElement('div');
         resizeHandle.classList.add('resize-handle');
         parent.insertBefore(resizeHandle, rightCol);
-
-        resizeHandle.addEventListener('mousedown', (evt) => {
-            if (evt.button !== 0) return;
-
-            evt.preventDefault();
-            evt.stopPropagation();
-
-            document.body.classList.add('resizing');
-
-            R.tracking = true;
-            R.parent = parent;
-            R.parentWidth = parent.offsetWidth;
-            R.handle = resizeHandle;
-            R.leftCol = leftCol;
-            R.leftColStartWidth = leftCol.offsetWidth;
-            R.screenX = evt.screenX;
+        parent.resizeHandle = resizeHandle;
+
+        ['mousedown', 'touchstart'].forEach((eventType) => {
+            resizeHandle.addEventListener(eventType, (evt) => {
+                if (eventType.startsWith('mouse')) {
+                    if (evt.button !== 0) return;
+                } else {
+                    if (evt.changedTouches.length !== 1) return;
+
+                    const currentTime = new Date().getTime();
+                    if (R.lastTapTime && currentTime - R.lastTapTime <= DOUBLE_TAP_DELAY) {
+                        onDoubleClick(evt);
+                        return;
+                    }
+
+                    R.lastTapTime = currentTime;
+                }
+
+                evt.preventDefault();
+                evt.stopPropagation();
+
+                document.body.classList.add('resizing');
+
+                R.tracking = true;
+                R.parent = parent;
+                R.parentWidth = parent.offsetWidth;
+                R.leftCol = leftCol;
+                R.leftColStartWidth = leftCol.offsetWidth;
+                if (eventType.startsWith('mouse')) {
+                    R.screenX = evt.screenX;
+                } else {
+                    R.screenX = evt.changedTouches[0].screenX;
+                }
+            });
         });
 
-        resizeHandle.addEventListener('dblclick', (evt) => {
-            evt.preventDefault();
-            evt.stopPropagation();
-
-            parent.style.gridTemplateColumns = GRID_TEMPLATE_COLUMNS;
-        });
+        resizeHandle.addEventListener('dblclick', onDoubleClick);
 
         afterResize(parent);
     }
 
-    window.addEventListener('mousemove', (evt) => {
-        if (evt.button !== 0) return;
-
-        if (R.tracking) {
-            evt.preventDefault();
-            evt.stopPropagation();
+    ['mousemove', 'touchmove'].forEach((eventType) => {
+        window.addEventListener(eventType, (evt) => {
+            if (eventType.startsWith('mouse')) {
+                if (evt.button !== 0) return;
+            } else {
+                if (evt.changedTouches.length !== 1) return;
+            }
 
-            const delta = R.screenX - evt.screenX;
-            const leftColWidth = Math.max(Math.min(R.leftColStartWidth - delta, R.parent.offsetWidth - GRADIO_MIN_WIDTH - PAD), GRADIO_MIN_WIDTH);
-            setLeftColGridTemplate(R.parent, leftColWidth);
-        }
+            if (R.tracking) {
+                if (eventType.startsWith('mouse')) {
+                    evt.preventDefault();
+                }
+                evt.stopPropagation();
+
+                let delta = 0;
+                if (eventType.startsWith('mouse')) {
+                    delta = R.screenX - evt.screenX;
+                } else {
+                    delta = R.screenX - evt.changedTouches[0].screenX;
+                }
+                const leftColWidth = Math.max(Math.min(R.leftColStartWidth - delta, R.parent.offsetWidth - R.parent.minRightColWidth - PAD), R.parent.minLeftColWidth);
+                setLeftColGridTemplate(R.parent, leftColWidth);
+            }
+        });
     });
 
-    window.addEventListener('mouseup', (evt) => {
-        if (evt.button !== 0) return;
+    ['mouseup', 'touchend'].forEach((eventType) => {
+        window.addEventListener(eventType, (evt) => {
+            if (eventType.startsWith('mouse')) {
+                if (evt.button !== 0) return;
+            } else {
+                if (evt.changedTouches.length !== 1) return;
+            }
 
-        if (R.tracking) {
-            evt.preventDefault();
-            evt.stopPropagation();
+            if (R.tracking) {
+                evt.preventDefault();
+                evt.stopPropagation();
 
-            R.tracking = false;
+                R.tracking = false;
 
-            document.body.classList.remove('resizing');
-        }
+                document.body.classList.remove('resizing');
+            }
+        });
     });
 
 
@@ -132,10 +191,15 @@
     setupResizeHandle = setup;
 })();
 
-onUiLoaded(function() {
+
+function setupAllResizeHandles() {
     for (var elem of gradioApp().querySelectorAll('.resize-handle-row')) {
-        if (!elem.querySelector('.resize-handle')) {
+        if (!elem.querySelector('.resize-handle') && !elem.children[0].classList.contains("hidden")) {
             setupResizeHandle(elem);
         }
     }
-});
+}
+
+
+onUiLoaded(setupAllResizeHandles);
+
diff --git a/javascript/settings.js b/javascript/settings.js
new file mode 100644
index 00000000000..b2d981c2144
--- /dev/null
+++ b/javascript/settings.js
@@ -0,0 +1,71 @@
+let settingsExcludeTabsFromShowAll = {
+    settings_tab_defaults: 1,
+    settings_tab_sysinfo: 1,
+    settings_tab_actions: 1,
+    settings_tab_licenses: 1,
+};
+
+function settingsShowAllTabs() {
+    gradioApp().querySelectorAll('#settings > div').forEach(function(elem) {
+        if (settingsExcludeTabsFromShowAll[elem.id]) return;
+
+        elem.style.display = "block";
+    });
+}
+
+function settingsShowOneTab() {
+    gradioApp().querySelector('#settings_show_one_page').click();
+}
+
+onUiLoaded(function() {
+    var edit = gradioApp().querySelector('#settings_search');
+    var editTextarea = gradioApp().querySelector('#settings_search > label > input');
+    var buttonShowAllPages = gradioApp().getElementById('settings_show_all_pages');
+    var settings_tabs = gradioApp().querySelector('#settings div');
+
+    onEdit('settingsSearch', editTextarea, 250, function() {
+        var searchText = (editTextarea.value || "").trim().toLowerCase();
+
+        gradioApp().querySelectorAll('#settings > div[id^=settings_] div[id^=column_settings_] > *').forEach(function(elem) {
+            var visible = elem.textContent.trim().toLowerCase().indexOf(searchText) != -1;
+            elem.style.display = visible ? "" : "none";
+        });
+
+        if (searchText != "") {
+            settingsShowAllTabs();
+        } else {
+            settingsShowOneTab();
+        }
+    });
+
+    settings_tabs.insertBefore(edit, settings_tabs.firstChild);
+    settings_tabs.appendChild(buttonShowAllPages);
+
+
+    buttonShowAllPages.addEventListener("click", settingsShowAllTabs);
+});
+
+
+onOptionsChanged(function() {
+    if (gradioApp().querySelector('#settings .settings-category')) return;
+
+    var sectionMap = {};
+    gradioApp().querySelectorAll('#settings > div > button').forEach(function(x) {
+        sectionMap[x.textContent.trim()] = x;
+    });
+
+    opts._categories.forEach(function(x) {
+        var section = localization[x[0]] ?? x[0];
+        var category = localization[x[1]] ?? x[1];
+
+        var span = document.createElement('SPAN');
+        span.textContent = category;
+        span.className = 'settings-category';
+
+        var sectionElem = sectionMap[section];
+        if (!sectionElem) return;
+
+        sectionElem.parentElement.insertBefore(span, sectionElem);
+    });
+});
+
diff --git a/javascript/token-counters.js b/javascript/token-counters.js
index 9d81a723b01..eeea7a5d26c 100644
--- a/javascript/token-counters.js
+++ b/javascript/token-counters.js
@@ -1,10 +1,9 @@
-let promptTokenCountDebounceTime = 800;
-let promptTokenCountTimeouts = {};
-var promptTokenCountUpdateFunctions = {};
+let promptTokenCountUpdateFunctions = {};
 
 function update_txt2img_tokens(...args) {
     // Called from Gradio
     update_token_counter("txt2img_token_button");
+    update_token_counter("txt2img_negative_token_button");
     if (args.length == 2) {
         return args[0];
     }
@@ -14,6 +13,7 @@ function update_txt2img_tokens(...args) {
 function update_img2img_tokens(...args) {
     // Called from Gradio
     update_token_counter("img2img_token_button");
+    update_token_counter("img2img_negative_token_button");
     if (args.length == 2) {
         return args[0];
     }
@@ -21,16 +21,7 @@ function update_img2img_tokens(...args) {
 }
 
 function update_token_counter(button_id) {
-    if (opts.disable_token_counters) {
-        return;
-    }
-    if (promptTokenCountTimeouts[button_id]) {
-        clearTimeout(promptTokenCountTimeouts[button_id]);
-    }
-    promptTokenCountTimeouts[button_id] = setTimeout(
-        () => gradioApp().getElementById(button_id)?.click(),
-        promptTokenCountDebounceTime,
-    );
+    promptTokenCountUpdateFunctions[button_id]?.();
 }
 
 
@@ -57,11 +48,6 @@ function setupTokenCounting(id, id_counter, id_button) {
     var counter = gradioApp().getElementById(id_counter);
     var textarea = gradioApp().querySelector(`#${id} > label > textarea`);
 
-    if (opts.disable_token_counters) {
-        counter.style.display = "none";
-        return;
-    }
-
     if (counter.parentElement == prompt.parentElement) {
         return;
     }
@@ -69,15 +55,33 @@ function setupTokenCounting(id, id_counter, id_button) {
     prompt.parentElement.insertBefore(counter, prompt);
     prompt.parentElement.style.position = "relative";
 
-    promptTokenCountUpdateFunctions[id] = function() {
-        update_token_counter(id_button);
-    };
-    textarea.addEventListener("input", promptTokenCountUpdateFunctions[id]);
+    var func = onEdit(id, textarea, 800, function() {
+        if (counter.classList.contains("token-counter-visible")) {
+            gradioApp().getElementById(id_button)?.click();
+        }
+    });
+    promptTokenCountUpdateFunctions[id] = func;
+    promptTokenCountUpdateFunctions[id_button] = func;
 }
 
-function setupTokenCounters() {
-    setupTokenCounting('txt2img_prompt', 'txt2img_token_counter', 'txt2img_token_button');
-    setupTokenCounting('txt2img_neg_prompt', 'txt2img_negative_token_counter', 'txt2img_negative_token_button');
-    setupTokenCounting('img2img_prompt', 'img2img_token_counter', 'img2img_token_button');
-    setupTokenCounting('img2img_neg_prompt', 'img2img_negative_token_counter', 'img2img_negative_token_button');
+function toggleTokenCountingVisibility(id, id_counter, id_button) {
+    var counter = gradioApp().getElementById(id_counter);
+
+    counter.style.display = opts.disable_token_counters ? "none" : "block";
+    counter.classList.toggle("token-counter-visible", !opts.disable_token_counters);
+}
+
+function runCodeForTokenCounters(fun) {
+    fun('txt2img_prompt', 'txt2img_token_counter', 'txt2img_token_button');
+    fun('txt2img_neg_prompt', 'txt2img_negative_token_counter', 'txt2img_negative_token_button');
+    fun('img2img_prompt', 'img2img_token_counter', 'img2img_token_button');
+    fun('img2img_neg_prompt', 'img2img_negative_token_counter', 'img2img_negative_token_button');
 }
+
+onUiLoaded(function() {
+    runCodeForTokenCounters(setupTokenCounting);
+});
+
+onOptionsChanged(function() {
+    runCodeForTokenCounters(toggleTokenCountingVisibility);
+});
diff --git a/javascript/ui.js b/javascript/ui.js
index bedcbf3e211..20309634fb6 100644
--- a/javascript/ui.js
+++ b/javascript/ui.js
@@ -26,6 +26,14 @@ function selected_gallery_index() {
     return all_gallery_buttons().findIndex(elem => elem.classList.contains('selected'));
 }
 
+function gallery_container_buttons(gallery_container) {
+    return gradioApp().querySelectorAll(`#${gallery_container} .thumbnail-item.thumbnail-small`);
+}
+
+function selected_gallery_index_id(gallery_container) {
+    return Array.from(gallery_container_buttons(gallery_container)).findIndex(elem => elem.classList.contains('selected'));
+}
+
 function extract_image_from_gallery(gallery) {
     if (gallery.length == 0) {
         return [null];
@@ -119,16 +127,24 @@ function create_submit_args(args) {
     return res;
 }
 
+function setSubmitButtonsVisibility(tabname, showInterrupt, showSkip, showInterrupting) {
+    gradioApp().getElementById(tabname + '_interrupt').style.display = showInterrupt ? "block" : "none";
+    gradioApp().getElementById(tabname + '_skip').style.display = showSkip ? "block" : "none";
+    gradioApp().getElementById(tabname + '_interrupting').style.display = showInterrupting ? "block" : "none";
+}
+
 function showSubmitButtons(tabname, show) {
-    gradioApp().getElementById(tabname + '_interrupt').style.display = show ? "none" : "block";
-    gradioApp().getElementById(tabname + '_skip').style.display = show ? "none" : "block";
+    setSubmitButtonsVisibility(tabname, !show, !show, false);
+}
+
+function showSubmitInterruptingPlaceholder(tabname) {
+    setSubmitButtonsVisibility(tabname, false, true, true);
 }
 
 function showRestoreProgressButton(tabname, show) {
     var button = gradioApp().getElementById(tabname + "_restore_progress");
     if (!button) return;
-
-    button.style.display = show ? "flex" : "none";
+    button.style.setProperty('display', show ? 'flex' : 'none', 'important');
 }
 
 function submit() {
@@ -150,6 +166,14 @@ function submit() {
     return res;
 }
 
+function submit_txt2img_upscale() {
+    var res = submit(...arguments);
+
+    res[2] = selected_gallery_index();
+
+    return res;
+}
+
 function submit_img2img() {
     showSubmitButtons('img2img', false);
 
@@ -170,11 +194,29 @@ function submit_img2img() {
     return res;
 }
 
+function submit_extras() {
+    showSubmitButtons('extras', false);
+
+    var id = randomId();
+
+    requestProgress(id, gradioApp().getElementById('extras_gallery_container'), gradioApp().getElementById('extras_gallery'), function() {
+        showSubmitButtons('extras', true);
+    });
+
+    var res = create_submit_args(arguments);
+
+    res[0] = id;
+
+    console.log(res);
+    return res;
+}
+
 function restoreProgressTxt2img() {
     showRestoreProgressButton("txt2img", false);
     var id = localGet("txt2img_task_id");
 
     if (id) {
+        showSubmitInterruptingPlaceholder('txt2img');
         requestProgress(id, gradioApp().getElementById('txt2img_gallery_container'), gradioApp().getElementById('txt2img_gallery'), function() {
             showSubmitButtons('txt2img', true);
         }, null, 0);
@@ -189,6 +231,7 @@ function restoreProgressImg2img() {
     var id = localGet("img2img_task_id");
 
     if (id) {
+        showSubmitInterruptingPlaceholder('img2img');
         requestProgress(id, gradioApp().getElementById('img2img_gallery_container'), gradioApp().getElementById('img2img_gallery'), function() {
             showSubmitButtons('img2img', true);
         }, null, 0);
@@ -198,9 +241,33 @@ function restoreProgressImg2img() {
 }
 
 
+/**
+ * Configure the width and height elements on `tabname` to accept
+ * pasting of resolutions in the form of "width x height".
+ */
+function setupResolutionPasting(tabname) {
+    var width = gradioApp().querySelector(`#${tabname}_width input[type=number]`);
+    var height = gradioApp().querySelector(`#${tabname}_height input[type=number]`);
+    for (const el of [width, height]) {
+        el.addEventListener('paste', function(event) {
+            var pasteData = event.clipboardData.getData('text/plain');
+            var parsed = pasteData.match(/^\s*(\d+)\D+(\d+)\s*$/);
+            if (parsed) {
+                width.value = parsed[1];
+                height.value = parsed[2];
+                updateInput(width);
+                updateInput(height);
+                event.preventDefault();
+            }
+        });
+    }
+}
+
 onUiLoaded(function() {
     showRestoreProgressButton('txt2img', localGet("txt2img_task_id"));
     showRestoreProgressButton('img2img', localGet("img2img_task_id"));
+    setupResolutionPasting('txt2img');
+    setupResolutionPasting('img2img');
 });
 
 
@@ -240,6 +307,7 @@ onAfterUiUpdate(function() {
     var jsdata = textarea.value;
     opts = JSON.parse(jsdata);
 
+    executeCallbacks(optionsAvailableCallbacks); /*global optionsAvailableCallbacks*/
     executeCallbacks(optionsChangedCallbacks); /*global optionsChangedCallbacks*/
 
     Object.defineProperty(textarea, 'value', {
@@ -261,23 +329,6 @@ onAfterUiUpdate(function() {
     });
 
     json_elem.parentElement.style.display = "none";
-
-    setupTokenCounters();
-
-    var show_all_pages = gradioApp().getElementById('settings_show_all_pages');
-    var settings_tabs = gradioApp().querySelector('#settings div');
-    if (show_all_pages && settings_tabs) {
-        settings_tabs.appendChild(show_all_pages);
-        show_all_pages.onclick = function() {
-            gradioApp().querySelectorAll('#settings > div').forEach(function(elem) {
-                if (elem.id == "settings_tab_licenses") {
-                    return;
-                }
-
-                elem.style.display = "block";
-            });
-        };
-    }
 });
 
 onOptionsChanged(function() {
@@ -295,8 +346,8 @@ onOptionsChanged(function() {
 let txt2img_textarea, img2img_textarea = undefined;
 
 function restart_reload() {
+    document.body.style.backgroundColor = "var(--background-fill-primary)";
     document.body.innerHTML = '<h1 style="font-family:monospace;margin-top:20%;color:lightgray;text-align:center;">Reloading...</h1>';
-
     var requestPing = function() {
         requestGet("./internal/ping", {}, function(data) {
             location.reload();
@@ -366,3 +417,20 @@ function switchWidthHeight(tabname) {
     updateInput(height);
     return [];
 }
+
+
+var onEditTimers = {};
+
+// calls func after afterMs milliseconds has passed since the input elem has been edited by user
+function onEdit(editId, elem, afterMs, func) {
+    var edited = function() {
+        var existingTimer = onEditTimers[editId];
+        if (existingTimer) clearTimeout(existingTimer);
+
+        onEditTimers[editId] = setTimeout(func, afterMs);
+    };
+
+    elem.addEventListener("input", edited);
+
+    return edited;
+}
diff --git a/modules/api/api.py b/modules/api/api.py
index e6edffe7144..97ec7514ea1 100644
--- a/modules/api/api.py
+++ b/modules/api/api.py
@@ -17,23 +17,21 @@
 from secrets import compare_digest
 
 import modules.shared as shared
-from modules import sd_samplers, deepbooru, sd_hijack, images, scripts, ui, postprocessing, errors, restart, shared_items
+from modules import sd_samplers, deepbooru, sd_hijack, images, scripts, ui, postprocessing, errors, restart, shared_items, script_callbacks, infotext_utils, sd_models, sd_schedulers
 from modules.api import models
 from modules.shared import opts
 from modules.processing import StableDiffusionProcessingTxt2Img, StableDiffusionProcessingImg2Img, process_images
 from modules.textual_inversion.textual_inversion import create_embedding, train_embedding
-from modules.textual_inversion.preprocess import preprocess
 from modules.hypernetworks.hypernetwork import create_hypernetwork, train_hypernetwork
-from PIL import PngImagePlugin,Image
-from modules.sd_models import unload_model_weights, reload_model_weights, checkpoint_aliases
+from PIL import PngImagePlugin
 from modules.sd_models_config import find_checkpoint_config_near_filename
 from modules.realesrgan_model import get_realesrgan_models
 from modules import devices
-from typing import Dict, List, Any
+from typing import Any
 import piexif
 import piexif.helper
 from contextlib import closing
-
+from modules.progress import create_task_id, add_task_to_queue, start_task, finish_task, current_task
 
 def script_name_to_index(name, scripts):
     try:
@@ -45,7 +43,7 @@ def script_name_to_index(name, scripts):
 def validate_sampler_name(name):
     config = sd_samplers.all_samplers_map.get(name, None)
     if config is None:
-        raise HTTPException(status_code=404, detail="Sampler not found")
+        raise HTTPException(status_code=400, detail="Sampler not found")
 
     return name
 
@@ -87,7 +85,7 @@ def decode_base64_to_image(encoding):
         headers = {'user-agent': opts.api_useragent} if opts.api_useragent else {}
         response = requests.get(encoding, timeout=30, headers=headers)
         try:
-            image = Image.open(BytesIO(response.content))
+            image = images.read(BytesIO(response.content))
             return image
         except Exception as e:
             raise HTTPException(status_code=500, detail="Invalid image url") from e
@@ -95,7 +93,7 @@ def decode_base64_to_image(encoding):
     if encoding.startswith("data:image/"):
         encoding = encoding.split(";")[1].split(",")[1]
     try:
-        image = Image.open(BytesIO(base64.b64decode(encoding)))
+        image = images.read(BytesIO(base64.b64decode(encoding)))
         return image
     except Exception as e:
         raise HTTPException(status_code=500, detail="Invalid encoded image") from e
@@ -103,7 +101,8 @@ def decode_base64_to_image(encoding):
 
 def encode_pil_to_base64(image):
     with io.BytesIO() as output_bytes:
-
+        if isinstance(image, str):
+            return image
         if opts.samples_format.lower() == 'png':
             use_metadata = False
             metadata = PngImagePlugin.PngInfo()
@@ -114,7 +113,7 @@ def encode_pil_to_base64(image):
             image.save(output_bytes, format="PNG", pnginfo=(metadata if use_metadata else None), quality=opts.jpeg_quality)
 
         elif opts.samples_format.lower() in ("jpg", "jpeg", "webp"):
-            if image.mode == "RGBA":
+            if image.mode in ("RGBA", "P"):
                 image = image.convert("RGB")
             parameters = image.info.get('parameters', None)
             exif_bytes = piexif.dump({
@@ -221,28 +220,30 @@ def __init__(self, app: FastAPI, queue_lock: Lock):
         self.add_api_route("/sdapi/v1/options", self.get_config, methods=["GET"], response_model=models.OptionsModel)
         self.add_api_route("/sdapi/v1/options", self.set_config, methods=["POST"])
         self.add_api_route("/sdapi/v1/cmd-flags", self.get_cmd_flags, methods=["GET"], response_model=models.FlagsModel)
-        self.add_api_route("/sdapi/v1/samplers", self.get_samplers, methods=["GET"], response_model=List[models.SamplerItem])
-        self.add_api_route("/sdapi/v1/upscalers", self.get_upscalers, methods=["GET"], response_model=List[models.UpscalerItem])
-        self.add_api_route("/sdapi/v1/latent-upscale-modes", self.get_latent_upscale_modes, methods=["GET"], response_model=List[models.LatentUpscalerModeItem])
-        self.add_api_route("/sdapi/v1/sd-models", self.get_sd_models, methods=["GET"], response_model=List[models.SDModelItem])
-        self.add_api_route("/sdapi/v1/sd-vae", self.get_sd_vaes, methods=["GET"], response_model=List[models.SDVaeItem])
-        self.add_api_route("/sdapi/v1/hypernetworks", self.get_hypernetworks, methods=["GET"], response_model=List[models.HypernetworkItem])
-        self.add_api_route("/sdapi/v1/face-restorers", self.get_face_restorers, methods=["GET"], response_model=List[models.FaceRestorerItem])
-        self.add_api_route("/sdapi/v1/realesrgan-models", self.get_realesrgan_models, methods=["GET"], response_model=List[models.RealesrganItem])
-        self.add_api_route("/sdapi/v1/prompt-styles", self.get_prompt_styles, methods=["GET"], response_model=List[models.PromptStyleItem])
+        self.add_api_route("/sdapi/v1/samplers", self.get_samplers, methods=["GET"], response_model=list[models.SamplerItem])
+        self.add_api_route("/sdapi/v1/schedulers", self.get_schedulers, methods=["GET"], response_model=list[models.SchedulerItem])
+        self.add_api_route("/sdapi/v1/upscalers", self.get_upscalers, methods=["GET"], response_model=list[models.UpscalerItem])
+        self.add_api_route("/sdapi/v1/latent-upscale-modes", self.get_latent_upscale_modes, methods=["GET"], response_model=list[models.LatentUpscalerModeItem])
+        self.add_api_route("/sdapi/v1/sd-models", self.get_sd_models, methods=["GET"], response_model=list[models.SDModelItem])
+        self.add_api_route("/sdapi/v1/sd-vae", self.get_sd_vaes, methods=["GET"], response_model=list[models.SDVaeItem])
+        self.add_api_route("/sdapi/v1/hypernetworks", self.get_hypernetworks, methods=["GET"], response_model=list[models.HypernetworkItem])
+        self.add_api_route("/sdapi/v1/face-restorers", self.get_face_restorers, methods=["GET"], response_model=list[models.FaceRestorerItem])
+        self.add_api_route("/sdapi/v1/realesrgan-models", self.get_realesrgan_models, methods=["GET"], response_model=list[models.RealesrganItem])
+        self.add_api_route("/sdapi/v1/prompt-styles", self.get_prompt_styles, methods=["GET"], response_model=list[models.PromptStyleItem])
         self.add_api_route("/sdapi/v1/embeddings", self.get_embeddings, methods=["GET"], response_model=models.EmbeddingsResponse)
+        self.add_api_route("/sdapi/v1/refresh-embeddings", self.refresh_embeddings, methods=["POST"])
         self.add_api_route("/sdapi/v1/refresh-checkpoints", self.refresh_checkpoints, methods=["POST"])
         self.add_api_route("/sdapi/v1/refresh-vae", self.refresh_vae, methods=["POST"])
         self.add_api_route("/sdapi/v1/create/embedding", self.create_embedding, methods=["POST"], response_model=models.CreateResponse)
         self.add_api_route("/sdapi/v1/create/hypernetwork", self.create_hypernetwork, methods=["POST"], response_model=models.CreateResponse)
-        self.add_api_route("/sdapi/v1/preprocess", self.preprocess, methods=["POST"], response_model=models.PreprocessResponse)
         self.add_api_route("/sdapi/v1/train/embedding", self.train_embedding, methods=["POST"], response_model=models.TrainResponse)
         self.add_api_route("/sdapi/v1/train/hypernetwork", self.train_hypernetwork, methods=["POST"], response_model=models.TrainResponse)
         self.add_api_route("/sdapi/v1/memory", self.get_memory, methods=["GET"], response_model=models.MemoryResponse)
         self.add_api_route("/sdapi/v1/unload-checkpoint", self.unloadapi, methods=["POST"])
         self.add_api_route("/sdapi/v1/reload-checkpoint", self.reloadapi, methods=["POST"])
         self.add_api_route("/sdapi/v1/scripts", self.get_scripts_list, methods=["GET"], response_model=models.ScriptsList)
-        self.add_api_route("/sdapi/v1/script-info", self.get_script_info, methods=["GET"], response_model=List[models.ScriptInfo])
+        self.add_api_route("/sdapi/v1/script-info", self.get_script_info, methods=["GET"], response_model=list[models.ScriptInfo])
+        self.add_api_route("/sdapi/v1/extensions", self.get_extensions_list, methods=["GET"], response_model=list[models.ExtensionItem])
 
         if shared.cmd_opts.api_server_stop:
             self.add_api_route("/sdapi/v1/server-kill", self.kill_webui, methods=["POST"])
@@ -252,6 +253,24 @@ def __init__(self, app: FastAPI, queue_lock: Lock):
         self.default_script_arg_txt2img = []
         self.default_script_arg_img2img = []
 
+        txt2img_script_runner = scripts.scripts_txt2img
+        img2img_script_runner = scripts.scripts_img2img
+
+        if not txt2img_script_runner.scripts or not img2img_script_runner.scripts:
+            ui.create_ui()
+
+        if not txt2img_script_runner.scripts:
+            txt2img_script_runner.initialize_scripts(False)
+        if not self.default_script_arg_txt2img:
+            self.default_script_arg_txt2img = self.init_default_script_args(txt2img_script_runner)
+
+        if not img2img_script_runner.scripts:
+            img2img_script_runner.initialize_scripts(True)
+        if not self.default_script_arg_img2img:
+            self.default_script_arg_img2img = self.init_default_script_args(img2img_script_runner)
+
+
+
     def add_api_route(self, path: str, endpoint, **kwargs):
         if shared.cmd_opts.api_auth:
             return self.app.add_api_route(path, endpoint, dependencies=[Depends(self.auth)], **kwargs)
@@ -313,8 +332,13 @@ def init_default_script_args(self, script_runner):
                     script_args[script.args_from:script.args_to] = ui_default_values
         return script_args
 
-    def init_script_args(self, request, default_script_args, selectable_scripts, selectable_idx, script_runner):
+    def init_script_args(self, request, default_script_args, selectable_scripts, selectable_idx, script_runner, *, input_script_args=None):
         script_args = default_script_args.copy()
+
+        if input_script_args is not None:
+            for index, value in input_script_args.items():
+                script_args[index] = value
+
         # position 0 in script_arg is the idx+1 of the selectable script that is going to be run when using scripts.scripts_*2img.run()
         if selectable_scripts:
             script_args[selectable_scripts.args_from:selectable_scripts.args_to] = request.script_args
@@ -336,33 +360,110 @@ def init_script_args(self, request, default_script_args, selectable_scripts, sel
                         script_args[alwayson_script.args_from + idx] = request.alwayson_scripts[alwayson_script_name]["args"][idx]
         return script_args
 
+    def apply_infotext(self, request, tabname, *, script_runner=None, mentioned_script_args=None):
+        """Processes `infotext` field from the `request`, and sets other fields of the `request` according to what's in infotext.
+
+        If request already has a field set, and that field is encountered in infotext too, the value from infotext is ignored.
+
+        Additionally, fills `mentioned_script_args` dict with index: value pairs for script arguments read from infotext.
+        """
+
+        if not request.infotext:
+            return {}
+
+        possible_fields = infotext_utils.paste_fields[tabname]["fields"]
+        set_fields = request.model_dump(exclude_unset=True) if hasattr(request, "request") else request.dict(exclude_unset=True)  # pydantic v1/v2 have different names for this
+        params = infotext_utils.parse_generation_parameters(request.infotext)
+
+        def get_field_value(field, params):
+            value = field.function(params) if field.function else params.get(field.label)
+            if value is None:
+                return None
+
+            if field.api in request.__fields__:
+                target_type = request.__fields__[field.api].type_
+            else:
+                target_type = type(field.component.value)
+
+            if target_type == type(None):
+                return None
+
+            if isinstance(value, dict) and value.get('__type__') == 'generic_update':  # this is a gradio.update rather than a value
+                value = value.get('value')
+
+            if value is not None and not isinstance(value, target_type):
+                value = target_type(value)
+
+            return value
+
+        for field in possible_fields:
+            if not field.api:
+                continue
+
+            if field.api in set_fields:
+                continue
+
+            value = get_field_value(field, params)
+            if value is not None:
+                setattr(request, field.api, value)
+
+        if request.override_settings is None:
+            request.override_settings = {}
+
+        overridden_settings = infotext_utils.get_override_settings(params)
+        for _, setting_name, value in overridden_settings:
+            if setting_name not in request.override_settings:
+                request.override_settings[setting_name] = value
+
+        if script_runner is not None and mentioned_script_args is not None:
+            indexes = {v: i for i, v in enumerate(script_runner.inputs)}
+            script_fields = ((field, indexes[field.component]) for field in possible_fields if field.component in indexes)
+
+            for field, index in script_fields:
+                value = get_field_value(field, params)
+
+                if value is None:
+                    continue
+
+                mentioned_script_args[index] = value
+
+        return params
+
     def text2imgapi(self, txt2imgreq: models.StableDiffusionTxt2ImgProcessingAPI):
+        task_id = txt2imgreq.force_task_id or create_task_id("txt2img")
+
         script_runner = scripts.scripts_txt2img
-        if not script_runner.scripts:
-            script_runner.initialize_scripts(False)
-            ui.create_ui()
-        if not self.default_script_arg_txt2img:
-            self.default_script_arg_txt2img = self.init_default_script_args(script_runner)
+
+        infotext_script_args = {}
+        self.apply_infotext(txt2imgreq, "txt2img", script_runner=script_runner, mentioned_script_args=infotext_script_args)
+
         selectable_scripts, selectable_script_idx = self.get_selectable_script(txt2imgreq.script_name, script_runner)
+        sampler, scheduler = sd_samplers.get_sampler_and_scheduler(txt2imgreq.sampler_name or txt2imgreq.sampler_index, txt2imgreq.scheduler)
 
         populate = txt2imgreq.copy(update={  # Override __init__ params
-            "sampler_name": validate_sampler_name(txt2imgreq.sampler_name or txt2imgreq.sampler_index),
+            "sampler_name": validate_sampler_name(sampler),
             "do_not_save_samples": not txt2imgreq.save_images,
             "do_not_save_grid": not txt2imgreq.save_images,
         })
         if populate.sampler_name:
             populate.sampler_index = None  # prevent a warning later on
 
+        if not populate.scheduler and scheduler != "Automatic":
+            populate.scheduler = scheduler
+
         args = vars(populate)
         args.pop('script_name', None)
         args.pop('script_args', None) # will refeed them to the pipeline directly after initializing them
         args.pop('alwayson_scripts', None)
+        args.pop('infotext', None)
 
-        script_args = self.init_script_args(txt2imgreq, self.default_script_arg_txt2img, selectable_scripts, selectable_script_idx, script_runner)
+        script_args = self.init_script_args(txt2imgreq, self.default_script_arg_txt2img, selectable_scripts, selectable_script_idx, script_runner, input_script_args=infotext_script_args)
 
         send_images = args.pop('send_images', True)
         args.pop('save_images', None)
 
+        add_task_to_queue(task_id)
+
         with self.queue_lock:
             with closing(StableDiffusionProcessingTxt2Img(sd_model=shared.sd_model, **args)) as p:
                 p.is_api = True
@@ -372,12 +473,14 @@ def text2imgapi(self, txt2imgreq: models.StableDiffusionTxt2ImgProcessingAPI):
 
                 try:
                     shared.state.begin(job="scripts_txt2img")
+                    start_task(task_id)
                     if selectable_scripts is not None:
                         p.script_args = script_args
                         processed = scripts.scripts_txt2img.run(p, *p.script_args) # Need to pass args as list here
                     else:
                         p.script_args = tuple(script_args) # Need to pass args as tuple here
                         processed = process_images(p)
+                    finish_task(task_id)
                 finally:
                     shared.state.end()
                     shared.total_tqdm.clear()
@@ -387,6 +490,8 @@ def text2imgapi(self, txt2imgreq: models.StableDiffusionTxt2ImgProcessingAPI):
         return models.TextToImageResponse(images=b64images, parameters=vars(txt2imgreq), info=processed.js())
 
     def img2imgapi(self, img2imgreq: models.StableDiffusionImg2ImgProcessingAPI):
+        task_id = img2imgreq.force_task_id or create_task_id("img2img")
+
         init_images = img2imgreq.init_images
         if init_images is None:
             raise HTTPException(status_code=404, detail="Init image not found")
@@ -396,15 +501,15 @@ def img2imgapi(self, img2imgreq: models.StableDiffusionImg2ImgProcessingAPI):
             mask = decode_base64_to_image(mask)
 
         script_runner = scripts.scripts_img2img
-        if not script_runner.scripts:
-            script_runner.initialize_scripts(True)
-            ui.create_ui()
-        if not self.default_script_arg_img2img:
-            self.default_script_arg_img2img = self.init_default_script_args(script_runner)
+
+        infotext_script_args = {}
+        self.apply_infotext(img2imgreq, "img2img", script_runner=script_runner, mentioned_script_args=infotext_script_args)
+
         selectable_scripts, selectable_script_idx = self.get_selectable_script(img2imgreq.script_name, script_runner)
+        sampler, scheduler = sd_samplers.get_sampler_and_scheduler(img2imgreq.sampler_name or img2imgreq.sampler_index, img2imgreq.scheduler)
 
         populate = img2imgreq.copy(update={  # Override __init__ params
-            "sampler_name": validate_sampler_name(img2imgreq.sampler_name or img2imgreq.sampler_index),
+            "sampler_name": validate_sampler_name(sampler),
             "do_not_save_samples": not img2imgreq.save_images,
             "do_not_save_grid": not img2imgreq.save_images,
             "mask": mask,
@@ -412,17 +517,23 @@ def img2imgapi(self, img2imgreq: models.StableDiffusionImg2ImgProcessingAPI):
         if populate.sampler_name:
             populate.sampler_index = None  # prevent a warning later on
 
+        if not populate.scheduler and scheduler != "Automatic":
+            populate.scheduler = scheduler
+
         args = vars(populate)
         args.pop('include_init_images', None)  # this is meant to be done by "exclude": True in model, but it's for a reason that I cannot determine.
         args.pop('script_name', None)
         args.pop('script_args', None)  # will refeed them to the pipeline directly after initializing them
         args.pop('alwayson_scripts', None)
+        args.pop('infotext', None)
 
-        script_args = self.init_script_args(img2imgreq, self.default_script_arg_img2img, selectable_scripts, selectable_script_idx, script_runner)
+        script_args = self.init_script_args(img2imgreq, self.default_script_arg_img2img, selectable_scripts, selectable_script_idx, script_runner, input_script_args=infotext_script_args)
 
         send_images = args.pop('send_images', True)
         args.pop('save_images', None)
 
+        add_task_to_queue(task_id)
+
         with self.queue_lock:
             with closing(StableDiffusionProcessingImg2Img(sd_model=shared.sd_model, **args)) as p:
                 p.init_images = [decode_base64_to_image(x) for x in init_images]
@@ -433,12 +544,14 @@ def img2imgapi(self, img2imgreq: models.StableDiffusionImg2ImgProcessingAPI):
 
                 try:
                     shared.state.begin(job="scripts_img2img")
+                    start_task(task_id)
                     if selectable_scripts is not None:
                         p.script_args = script_args
                         processed = scripts.scripts_img2img.run(p, *p.script_args) # Need to pass args as list here
                     else:
                         p.script_args = tuple(script_args) # Need to pass args as tuple here
                         processed = process_images(p)
+                    finish_task(task_id)
                 finally:
                     shared.state.end()
                     shared.total_tqdm.clear()
@@ -473,9 +586,6 @@ def extras_batch_images_api(self, req: models.ExtrasBatchImagesRequest):
         return models.ExtrasBatchImagesResponse(images=list(map(encode_pil_to_base64, result[0])), html_info=result[1])
 
     def pnginfoapi(self, req: models.PNGInfoRequest):
-        if(not req.image.strip()):
-            return models.PNGInfoResponse(info="")
-
         image = decode_base64_to_image(req.image.strip())
         if image is None:
             return models.PNGInfoResponse(info="")
@@ -484,9 +594,10 @@ def pnginfoapi(self, req: models.PNGInfoRequest):
         if geninfo is None:
             geninfo = ""
 
-        items = {**{'parameters': geninfo}, **items}
+        params = infotext_utils.parse_generation_parameters(geninfo)
+        script_callbacks.infotext_pasted_callback(geninfo, params)
 
-        return models.PNGInfoResponse(info=geninfo, items=items)
+        return models.PNGInfoResponse(info=geninfo, items=items, parameters=params)
 
     def progressapi(self, req: models.ProgressRequest = Depends()):
         # copy from check_progress_call of ui.py
@@ -514,7 +625,7 @@ def progressapi(self, req: models.ProgressRequest = Depends()):
         if shared.state.current_image and not req.skip_current_image:
             current_image = encode_pil_to_base64(shared.state.current_image)
 
-        return models.ProgressResponse(progress=progress, eta_relative=eta_relative, state=shared.state.dict(), current_image=current_image, textinfo=shared.state.textinfo)
+        return models.ProgressResponse(progress=progress, eta_relative=eta_relative, state=shared.state.dict(), current_image=current_image, textinfo=shared.state.textinfo, current_task=current_task)
 
     def interrogateapi(self, interrogatereq: models.InterrogateRequest):
         image_b64 = interrogatereq.image
@@ -541,12 +652,12 @@ def interruptapi(self):
         return {}
 
     def unloadapi(self):
-        unload_model_weights()
+        sd_models.unload_model_weights()
 
         return {}
 
     def reloadapi(self):
-        reload_model_weights()
+        sd_models.send_model_to_device(shared.sd_model)
 
         return {}
 
@@ -564,9 +675,9 @@ def get_config(self):
 
         return options
 
-    def set_config(self, req: Dict[str, Any]):
+    def set_config(self, req: dict[str, Any]):
         checkpoint_name = req.get("sd_model_checkpoint", None)
-        if checkpoint_name is not None and checkpoint_name not in checkpoint_aliases:
+        if checkpoint_name is not None and checkpoint_name not in sd_models.checkpoint_aliases:
             raise RuntimeError(f"model {checkpoint_name!r} not found")
 
         for k, v in req.items():
@@ -581,6 +692,17 @@ def get_cmd_flags(self):
     def get_samplers(self):
         return [{"name": sampler[0], "aliases":sampler[2], "options":sampler[3]} for sampler in sd_samplers.all_samplers]
 
+    def get_schedulers(self):
+        return [
+            {
+                "name": scheduler.name,
+                "label": scheduler.label,
+                "aliases": scheduler.aliases,
+                "default_rho": scheduler.default_rho,
+                "need_inner_model": scheduler.need_inner_model,
+            }
+            for scheduler in sd_schedulers.schedulers]
+
     def get_upscalers(self):
         return [
             {
@@ -646,6 +768,10 @@ def convert_embeddings(embeddings):
             "skipped": convert_embeddings(db.skipped_embeddings),
         }
 
+    def refresh_embeddings(self):
+        with self.queue_lock:
+            sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings(force_reload=True)
+
     def refresh_checkpoints(self):
         with self.queue_lock:
             shared.refresh_checkpoints()
@@ -676,19 +802,6 @@ def create_hypernetwork(self, args: dict):
         finally:
             shared.state.end()
 
-    def preprocess(self, args: dict):
-        try:
-            shared.state.begin(job="preprocess")
-            preprocess(**args) # quick operation unless blip/booru interrogation is enabled
-            shared.state.end()
-            return models.PreprocessResponse(info='preprocess complete')
-        except KeyError as e:
-            return models.PreprocessResponse(info=f"preprocess error: invalid token: {e}")
-        except Exception as e:
-            return models.PreprocessResponse(info=f"preprocess error: {e}")
-        finally:
-            shared.state.end()
-
     def train_embedding(self, args: dict):
         try:
             shared.state.begin(job="train_embedding")
@@ -770,9 +883,36 @@ def get_memory(self):
             cuda = {'error': f'{err}'}
         return models.MemoryResponse(ram=ram, cuda=cuda)
 
+    def get_extensions_list(self):
+        from modules import extensions
+        extensions.list_extensions()
+        ext_list = []
+        for ext in extensions.extensions:
+            ext: extensions.Extension
+            ext.read_info_from_repo()
+            if ext.remote is not None:
+                ext_list.append({
+                    "name": ext.name,
+                    "remote": ext.remote,
+                    "branch": ext.branch,
+                    "commit_hash":ext.commit_hash,
+                    "commit_date":ext.commit_date,
+                    "version":ext.version,
+                    "enabled":ext.enabled
+                })
+        return ext_list
+
     def launch(self, server_name, port, root_path):
         self.app.include_router(self.router)
-        uvicorn.run(self.app, host=server_name, port=port, timeout_keep_alive=shared.cmd_opts.timeout_keep_alive, root_path=root_path)
+        uvicorn.run(
+            self.app,
+            host=server_name,
+            port=port,
+            timeout_keep_alive=shared.cmd_opts.timeout_keep_alive,
+            root_path=root_path,
+            ssl_keyfile=shared.cmd_opts.tls_keyfile,
+            ssl_certfile=shared.cmd_opts.tls_certfile
+        )
 
     def kill_webui(self):
         restart.stop_program()
diff --git a/modules/api/models.py b/modules/api/models.py
index 6a574771c33..79c4cbbbbbc 100644
--- a/modules/api/models.py
+++ b/modules/api/models.py
@@ -1,12 +1,10 @@
 import inspect
 
 from pydantic import BaseModel, Field, create_model
-from typing import Any, Optional
-from typing_extensions import Literal
+from typing import Any, Optional, Literal
 from inflection import underscore
 from modules.processing import StableDiffusionProcessingTxt2Img, StableDiffusionProcessingImg2Img
 from modules.shared import sd_upscalers, opts, parser
-from typing import Dict, List
 
 API_NOT_ALLOWED = [
     "self",
@@ -109,6 +107,8 @@ def generate_model(self):
         {"key": "send_images", "type": bool, "default": True},
         {"key": "save_images", "type": bool, "default": False},
         {"key": "alwayson_scripts", "type": dict, "default": {}},
+        {"key": "force_task_id", "type": str, "default": None},
+        {"key": "infotext", "type": str, "default": None},
     ]
 ).generate_model()
 
@@ -126,16 +126,18 @@ def generate_model(self):
         {"key": "send_images", "type": bool, "default": True},
         {"key": "save_images", "type": bool, "default": False},
         {"key": "alwayson_scripts", "type": dict, "default": {}},
+        {"key": "force_task_id", "type": str, "default": None},
+        {"key": "infotext", "type": str, "default": None},
     ]
 ).generate_model()
 
 class TextToImageResponse(BaseModel):
-    images: List[str] = Field(default=None, title="Image", description="The generated image in base64 format.")
+    images: list[str] = Field(default=None, title="Image", description="The generated image in base64 format.")
     parameters: dict
     info: str
 
 class ImageToImageResponse(BaseModel):
-    images: List[str] = Field(default=None, title="Image", description="The generated image in base64 format.")
+    images: list[str] = Field(default=None, title="Image", description="The generated image in base64 format.")
     parameters: dict
     info: str
 
@@ -145,7 +147,7 @@ class ExtrasBaseRequest(BaseModel):
     gfpgan_visibility: float = Field(default=0, title="GFPGAN Visibility", ge=0, le=1, allow_inf_nan=False, description="Sets the visibility of GFPGAN, values should be between 0 and 1.")
     codeformer_visibility: float = Field(default=0, title="CodeFormer Visibility", ge=0, le=1, allow_inf_nan=False, description="Sets the visibility of CodeFormer, values should be between 0 and 1.")
     codeformer_weight: float = Field(default=0, title="CodeFormer Weight", ge=0, le=1, allow_inf_nan=False, description="Sets the weight of CodeFormer, values should be between 0 and 1.")
-    upscaling_resize: float = Field(default=2, title="Upscaling Factor", ge=1, le=8, description="By how much to upscale the image, only used when resize_mode=0.")
+    upscaling_resize: float = Field(default=2, title="Upscaling Factor", gt=0, description="By how much to upscale the image, only used when resize_mode=0.")
     upscaling_resize_w: int = Field(default=512, title="Target Width", ge=1, description="Target width for the upscaler to hit. Only used when resize_mode=1.")
     upscaling_resize_h: int = Field(default=512, title="Target Height", ge=1, description="Target height for the upscaler to hit. Only used when resize_mode=1.")
     upscaling_crop: bool = Field(default=True, title="Crop to fit", description="Should the upscaler crop the image to fit in the chosen size?")
@@ -168,17 +170,18 @@ class FileData(BaseModel):
     name: str = Field(title="File name")
 
 class ExtrasBatchImagesRequest(ExtrasBaseRequest):
-    imageList: List[FileData] = Field(title="Images", description="List of images to work on. Must be Base64 strings")
+    imageList: list[FileData] = Field(title="Images", description="List of images to work on. Must be Base64 strings")
 
 class ExtrasBatchImagesResponse(ExtraBaseResponse):
-    images: List[str] = Field(title="Images", description="The generated images in base64 format.")
+    images: list[str] = Field(title="Images", description="The generated images in base64 format.")
 
 class PNGInfoRequest(BaseModel):
     image: str = Field(title="Image", description="The base64 encoded PNG image")
 
 class PNGInfoResponse(BaseModel):
     info: str = Field(title="Image info", description="A string with the parameters used to generate the image")
-    items: dict = Field(title="Items", description="An object containing all the info the image had")
+    items: dict = Field(title="Items", description="A dictionary containing all the other fields the image had")
+    parameters: dict = Field(title="Parameters", description="A dictionary with parsed generation info fields")
 
 class ProgressRequest(BaseModel):
     skip_current_image: bool = Field(default=False, title="Skip current image", description="Skip current image serialization")
@@ -203,9 +206,6 @@ class TrainResponse(BaseModel):
 class CreateResponse(BaseModel):
     info: str = Field(title="Create info", description="Response string from create embedding or hypernetwork task.")
 
-class PreprocessResponse(BaseModel):
-    info: str = Field(title="Preprocess info", description="Response string from preprocessing task.")
-
 fields = {}
 for key, metadata in opts.data_labels.items():
     value = opts.data.get(key)
@@ -232,8 +232,15 @@ class PreprocessResponse(BaseModel):
 
 class SamplerItem(BaseModel):
     name: str = Field(title="Name")
-    aliases: List[str] = Field(title="Aliases")
-    options: Dict[str, str] = Field(title="Options")
+    aliases: list[str] = Field(title="Aliases")
+    options: dict[str, str] = Field(title="Options")
+
+class SchedulerItem(BaseModel):
+    name: str = Field(title="Name")
+    label: str = Field(title="Label")
+    aliases: Optional[list[str]] = Field(title="Aliases")
+    default_rho: Optional[float] = Field(title="Default Rho")
+    need_inner_model: Optional[bool] = Field(title="Needs Inner Model")
 
 class UpscalerItem(BaseModel):
     name: str = Field(title="Name")
@@ -284,8 +291,8 @@ class EmbeddingItem(BaseModel):
     vectors: int = Field(title="Vectors", description="The number of vectors in the embedding")
 
 class EmbeddingsResponse(BaseModel):
-    loaded: Dict[str, EmbeddingItem] = Field(title="Loaded", description="Embeddings loaded for the current model")
-    skipped: Dict[str, EmbeddingItem] = Field(title="Skipped", description="Embeddings skipped for the current model (likely due to architecture incompatibility)")
+    loaded: dict[str, EmbeddingItem] = Field(title="Loaded", description="Embeddings loaded for the current model")
+    skipped: dict[str, EmbeddingItem] = Field(title="Skipped", description="Embeddings skipped for the current model (likely due to architecture incompatibility)")
 
 class MemoryResponse(BaseModel):
     ram: dict = Field(title="RAM", description="System memory stats")
@@ -303,11 +310,20 @@ class ScriptArg(BaseModel):
     minimum: Optional[Any] = Field(default=None, title="Minimum", description="Minimum allowed value for the argumentin UI")
     maximum: Optional[Any] = Field(default=None, title="Minimum", description="Maximum allowed value for the argumentin UI")
     step: Optional[Any] = Field(default=None, title="Minimum", description="Step for changing value of the argumentin UI")
-    choices: Optional[List[str]] = Field(default=None, title="Choices", description="Possible values for the argument")
+    choices: Optional[list[str]] = Field(default=None, title="Choices", description="Possible values for the argument")
 
 
 class ScriptInfo(BaseModel):
     name: str = Field(default=None, title="Name", description="Script name")
     is_alwayson: bool = Field(default=None, title="IsAlwayson", description="Flag specifying whether this script is an alwayson script")
     is_img2img: bool = Field(default=None, title="IsImg2img", description="Flag specifying whether this script is an img2img script")
-    args: List[ScriptArg] = Field(title="Arguments", description="List of script's arguments")
+    args: list[ScriptArg] = Field(title="Arguments", description="List of script's arguments")
+
+class ExtensionItem(BaseModel):
+    name: str = Field(title="Name", description="Extension name")
+    remote: str = Field(title="Remote", description="Extension Repository URL")
+    branch: str = Field(title="Branch", description="Extension Repository Branch")
+    commit_hash: str = Field(title="Commit Hash", description="Extension Repository Commit Hash")
+    version: str = Field(title="Version", description="Extension Version")
+    commit_date: str = Field(title="Commit Date", description="Extension Repository Commit Date")
+    enabled: bool = Field(title="Enabled", description="Flag specifying whether this extension is enabled")
diff --git a/modules/cache.py b/modules/cache.py
index ff26a2132d9..f4e5f702b42 100644
--- a/modules/cache.py
+++ b/modules/cache.py
@@ -2,48 +2,55 @@
 import os
 import os.path
 import threading
-import time
+
+import diskcache
+import tqdm
 
 from modules.paths import data_path, script_path
 
 cache_filename = os.environ.get('SD_WEBUI_CACHE_FILE', os.path.join(data_path, "cache.json"))
-cache_data = None
+cache_dir = os.environ.get('SD_WEBUI_CACHE_DIR', os.path.join(data_path, "cache"))
+caches = {}
 cache_lock = threading.Lock()
 
-dump_cache_after = None
-dump_cache_thread = None
-
 
 def dump_cache():
-    """
-    Marks cache for writing to disk. 5 seconds after no one else flags the cache for writing, it is written.
-    """
+    """old function for dumping cache to disk; does nothing since diskcache."""
 
-    global dump_cache_after
-    global dump_cache_thread
+    pass
 
-    def thread_func():
-        global dump_cache_after
-        global dump_cache_thread
 
-        while dump_cache_after is not None and time.time() < dump_cache_after:
-            time.sleep(1)
+def make_cache(subsection: str) -> diskcache.Cache:
+    return diskcache.Cache(
+        os.path.join(cache_dir, subsection),
+        size_limit=2**32,  # 4 GB, culling oldest first
+        disk_min_file_size=2**18,  # keep up to 256KB in Sqlite
+    )
 
-        with cache_lock:
-            cache_filename_tmp = cache_filename + "-"
-            with open(cache_filename_tmp, "w", encoding="utf8") as file:
-                json.dump(cache_data, file, indent=4)
 
-            os.replace(cache_filename_tmp, cache_filename)
+def convert_old_cached_data():
+    try:
+        with open(cache_filename, "r", encoding="utf8") as file:
+            data = json.load(file)
+    except FileNotFoundError:
+        return
+    except Exception:
+        os.replace(cache_filename, os.path.join(script_path, "tmp", "cache.json"))
+        print('[ERROR] issue occurred while trying to read cache.json; old cache has been moved to tmp/cache.json')
+        return
 
-            dump_cache_after = None
-            dump_cache_thread = None
+    total_count = sum(len(keyvalues) for keyvalues in data.values())
 
-    with cache_lock:
-        dump_cache_after = time.time() + 5
-        if dump_cache_thread is None:
-            dump_cache_thread = threading.Thread(name='cache-writer', target=thread_func)
-            dump_cache_thread.start()
+    with tqdm.tqdm(total=total_count, desc="converting cache") as progress:
+        for subsection, keyvalues in data.items():
+            cache_obj = caches.get(subsection)
+            if cache_obj is None:
+                cache_obj = make_cache(subsection)
+                caches[subsection] = cache_obj
+
+            for key, value in keyvalues.items():
+                cache_obj[key] = value
+                progress.update(1)
 
 
 def cache(subsection):
@@ -54,29 +61,21 @@ def cache(subsection):
         subsection (str): The subsection identifier for the cache.
 
     Returns:
-        dict: The cache data for the specified subsection.
+        diskcache.Cache: The cache data for the specified subsection.
     """
 
-    global cache_data
-
-    if cache_data is None:
+    cache_obj = caches.get(subsection)
+    if not cache_obj:
         with cache_lock:
-            if cache_data is None:
-                if not os.path.isfile(cache_filename):
-                    cache_data = {}
-                else:
-                    try:
-                        with open(cache_filename, "r", encoding="utf8") as file:
-                            cache_data = json.load(file)
-                    except Exception:
-                        os.replace(cache_filename, os.path.join(script_path, "tmp", "cache.json"))
-                        print('[ERROR] issue occurred while trying to read cache.json, move current cache to tmp/cache.json and create new cache')
-                        cache_data = {}
-
-    s = cache_data.get(subsection, {})
-    cache_data[subsection] = s
-
-    return s
+            if not os.path.exists(cache_dir) and os.path.isfile(cache_filename):
+                convert_old_cached_data()
+
+            cache_obj = caches.get(subsection)
+            if not cache_obj:
+                cache_obj = make_cache(subsection)
+                caches[subsection] = cache_obj
+
+    return cache_obj
 
 
 def cached_data_for_file(subsection, title, filename, func):
diff --git a/modules/call_queue.py b/modules/call_queue.py
index ddf0d57383c..555c35312dd 100644
--- a/modules/call_queue.py
+++ b/modules/call_queue.py
@@ -1,8 +1,9 @@
+import os.path
 from functools import wraps
 import html
 import time
 
-from modules import shared, progress, errors, devices, fifo_lock
+from modules import shared, progress, errors, devices, fifo_lock, profiling
 
 queue_lock = fifo_lock.FIFOLock()
 
@@ -46,6 +47,22 @@ def f(*args, **kwargs):
 
 
 def wrap_gradio_call(func, extra_outputs=None, add_stats=False):
+    @wraps(func)
+    def f(*args, **kwargs):
+        try:
+            res = func(*args, **kwargs)
+        finally:
+            shared.state.skipped = False
+            shared.state.interrupted = False
+            shared.state.stopping_generation = False
+            shared.state.job_count = 0
+            shared.state.job = ""
+        return res
+
+    return wrap_gradio_call_no_job(f, extra_outputs, add_stats)
+
+
+def wrap_gradio_call_no_job(func, extra_outputs=None, add_stats=False):
     @wraps(func)
     def f(*args, extra_outputs_array=extra_outputs, **kwargs):
         run_memmon = shared.opts.memmon_poll_rate > 0 and not shared.mem_mon.disabled and add_stats
@@ -65,9 +82,6 @@ def f(*args, extra_outputs_array=extra_outputs, **kwargs):
                 arg_str += f" (Argument list truncated at {max_debug_str_len}/{len(arg_str)} characters)"
             errors.report(f"{message}\n{arg_str}", exc_info=True)
 
-            shared.state.job = ""
-            shared.state.job_count = 0
-
             if extra_outputs_array is None:
                 extra_outputs_array = [None, '']
 
@@ -76,10 +90,6 @@ def f(*args, extra_outputs_array=extra_outputs, **kwargs):
 
         devices.torch_gc()
 
-        shared.state.skipped = False
-        shared.state.interrupted = False
-        shared.state.job_count = 0
-
         if not add_stats:
             return tuple(res)
 
@@ -99,8 +109,8 @@ def f(*args, extra_outputs_array=extra_outputs, **kwargs):
             sys_pct = sys_peak/max(sys_total, 1) * 100
 
             toltip_a = "Active: peak amount of video memory used during generation (excluding cached data)"
-            toltip_r = "Reserved: total amout of video memory allocated by the Torch library "
-            toltip_sys = "System: peak amout of video memory allocated by all running programs, out of total capacity"
+            toltip_r = "Reserved: total amount of video memory allocated by the Torch library "
+            toltip_sys = "System: peak amount of video memory allocated by all running programs, out of total capacity"
 
             text_a = f"<abbr title='{toltip_a}'>A</abbr>: <span class='measurement'>{active_peak/1024:.2f} GB</span>"
             text_r = f"<abbr title='{toltip_r}'>R</abbr>: <span class='measurement'>{reserved_peak/1024:.2f} GB</span>"
@@ -110,9 +120,15 @@ def f(*args, extra_outputs_array=extra_outputs, **kwargs):
         else:
             vram_html = ''
 
+        if shared.opts.profiling_enable and os.path.exists(shared.opts.profiling_filename):
+            profiling_html = f"<p class='profile'> [ <a href='{profiling.webpath()}' download>Profile</a> ] </p>"
+        else:
+            profiling_html = ''
+
         # last item is always HTML
-        res[-1] += f"<div class='performance'><p class='time'>Time taken: <wbr><span class='measurement'>{elapsed_text}</span></p>{vram_html}</div>"
+        res[-1] += f"<div class='performance'><p class='time'>Time taken: <wbr><span class='measurement'>{elapsed_text}</span></p>{vram_html}{profiling_html}</div>"
 
         return tuple(res)
 
     return f
+
diff --git a/modules/cmd_args.py b/modules/cmd_args.py
index aab62286e24..d71982b2c12 100644
--- a/modules/cmd_args.py
+++ b/modules/cmd_args.py
@@ -1,7 +1,7 @@
 import argparse
 import json
 import os
-from modules.paths_internal import models_path, script_path, data_path, extensions_dir, extensions_builtin_dir, sd_default_config, sd_model_file  # noqa: F401
+from modules.paths_internal import normalized_filepath, models_path, script_path, data_path, extensions_dir, extensions_builtin_dir, sd_default_config, sd_model_file  # noqa: F401
 
 parser = argparse.ArgumentParser()
 
@@ -19,21 +19,22 @@
 parser.add_argument("--dump-sysinfo", action='store_true', help="launch.py argument: dump limited sysinfo file (without information about extensions, options) to disk and quit")
 parser.add_argument("--loglevel", type=str, help="log level; one of: CRITICAL, ERROR, WARNING, INFO, DEBUG", default=None)
 parser.add_argument("--do-not-download-clip", action='store_true', help="do not download CLIP model even if it's not included in the checkpoint")
-parser.add_argument("--data-dir", type=str, default=os.path.dirname(os.path.dirname(os.path.realpath(__file__))), help="base path where all user data is stored")
-parser.add_argument("--config", type=str, default=sd_default_config, help="path to config which constructs model",)
-parser.add_argument("--ckpt", type=str, default=sd_model_file, help="path to checkpoint of stable diffusion model; if specified, this checkpoint will be added to the list of checkpoints and loaded",)
-parser.add_argument("--ckpt-dir", type=str, default=None, help="Path to directory with stable diffusion checkpoints")
-parser.add_argument("--vae-dir", type=str, default=None, help="Path to directory with VAE files")
-parser.add_argument("--gfpgan-dir", type=str, help="GFPGAN directory", default=('./src/gfpgan' if os.path.exists('./src/gfpgan') else './GFPGAN'))
-parser.add_argument("--gfpgan-model", type=str, help="GFPGAN model file name", default=None)
+parser.add_argument("--data-dir", type=normalized_filepath, default=os.path.dirname(os.path.dirname(os.path.realpath(__file__))), help="base path where all user data is stored")
+parser.add_argument("--models-dir", type=normalized_filepath, default=None, help="base path where models are stored; overrides --data-dir")
+parser.add_argument("--config", type=normalized_filepath, default=sd_default_config, help="path to config which constructs model",)
+parser.add_argument("--ckpt", type=normalized_filepath, default=sd_model_file, help="path to checkpoint of stable diffusion model; if specified, this checkpoint will be added to the list of checkpoints and loaded",)
+parser.add_argument("--ckpt-dir", type=normalized_filepath, default=None, help="Path to directory with stable diffusion checkpoints")
+parser.add_argument("--vae-dir", type=normalized_filepath, default=None, help="Path to directory with VAE files")
+parser.add_argument("--gfpgan-dir", type=normalized_filepath, help="GFPGAN directory", default=('./src/gfpgan' if os.path.exists('./src/gfpgan') else './GFPGAN'))
+parser.add_argument("--gfpgan-model", type=normalized_filepath, help="GFPGAN model file name", default=None)
 parser.add_argument("--no-half", action='store_true', help="do not switch the model to 16-bit floats")
 parser.add_argument("--no-half-vae", action='store_true', help="do not switch the VAE model to 16-bit floats")
 parser.add_argument("--no-progressbar-hiding", action='store_true', help="do not hide progressbar in gradio UI (we hide it because it slows down ML if you have hardware acceleration in browser)")
-parser.add_argument("--max-batch-count", type=int, default=16, help="maximum batch count value for the UI")
-parser.add_argument("--embeddings-dir", type=str, default=os.path.join(data_path, 'embeddings'), help="embeddings directory for textual inversion (default: embeddings)")
-parser.add_argument("--textual-inversion-templates-dir", type=str, default=os.path.join(script_path, 'textual_inversion_templates'), help="directory with textual inversion templates")
-parser.add_argument("--hypernetwork-dir", type=str, default=os.path.join(models_path, 'hypernetworks'), help="hypernetwork directory")
-parser.add_argument("--localizations-dir", type=str, default=os.path.join(script_path, 'localizations'), help="localizations directory")
+parser.add_argument("--max-batch-count", type=int, default=16, help="does not do anything")
+parser.add_argument("--embeddings-dir", type=normalized_filepath, default=os.path.join(data_path, 'embeddings'), help="embeddings directory for textual inversion (default: embeddings)")
+parser.add_argument("--textual-inversion-templates-dir", type=normalized_filepath, default=os.path.join(script_path, 'textual_inversion_templates'), help="directory with textual inversion templates")
+parser.add_argument("--hypernetwork-dir", type=normalized_filepath, default=os.path.join(models_path, 'hypernetworks'), help="hypernetwork directory")
+parser.add_argument("--localizations-dir", type=normalized_filepath, default=os.path.join(script_path, 'localizations'), help="localizations directory")
 parser.add_argument("--allow-code", action='store_true', help="allow custom script execution from webui")
 parser.add_argument("--medvram", action='store_true', help="enable stable diffusion model optimizations for sacrificing a little speed for low VRM usage")
 parser.add_argument("--medvram-sdxl", action='store_true', help="enable --medvram optimization just for SDXL models")
@@ -41,19 +42,20 @@
 parser.add_argument("--lowram", action='store_true', help="load stable diffusion checkpoint weights to VRAM instead of RAM")
 parser.add_argument("--always-batch-cond-uncond", action='store_true', help="does not do anything")
 parser.add_argument("--unload-gfpgan", action='store_true', help="does not do anything.")
-parser.add_argument("--precision", type=str, help="evaluate at this precision", choices=["full", "autocast"], default="autocast")
+parser.add_argument("--precision", type=str, help="evaluate at this precision", choices=["full", "half", "autocast"], default="autocast")
 parser.add_argument("--upcast-sampling", action='store_true', help="upcast sampling. No effect with --no-half. Usually produces similar results to --no-half with better performance while using less memory.")
 parser.add_argument("--share", action='store_true', help="use share=True for gradio and make the UI accessible through their site")
 parser.add_argument("--ngrok", type=str, help="ngrok authtoken, alternative to gradio --share", default=None)
 parser.add_argument("--ngrok-region", type=str, help="does not do anything.", default="")
 parser.add_argument("--ngrok-options", type=json.loads, help='The options to pass to ngrok in JSON format, e.g.: \'{"authtoken_from_env":true, "basic_auth":"user:password", "oauth_provider":"google", "oauth_allow_emails":"user@asdf.com"}\'', default=dict())
 parser.add_argument("--enable-insecure-extension-access", action='store_true', help="enable extensions tab regardless of other options")
-parser.add_argument("--codeformer-models-path", type=str, help="Path to directory with codeformer model file(s).", default=os.path.join(models_path, 'Codeformer'))
-parser.add_argument("--gfpgan-models-path", type=str, help="Path to directory with GFPGAN model file(s).", default=os.path.join(models_path, 'GFPGAN'))
-parser.add_argument("--esrgan-models-path", type=str, help="Path to directory with ESRGAN model file(s).", default=os.path.join(models_path, 'ESRGAN'))
-parser.add_argument("--bsrgan-models-path", type=str, help="Path to directory with BSRGAN model file(s).", default=os.path.join(models_path, 'BSRGAN'))
-parser.add_argument("--realesrgan-models-path", type=str, help="Path to directory with RealESRGAN model file(s).", default=os.path.join(models_path, 'RealESRGAN'))
-parser.add_argument("--clip-models-path", type=str, help="Path to directory with CLIP model file(s).", default=None)
+parser.add_argument("--codeformer-models-path", type=normalized_filepath, help="Path to directory with codeformer model file(s).", default=os.path.join(models_path, 'Codeformer'))
+parser.add_argument("--gfpgan-models-path", type=normalized_filepath, help="Path to directory with GFPGAN model file(s).", default=os.path.join(models_path, 'GFPGAN'))
+parser.add_argument("--esrgan-models-path", type=normalized_filepath, help="Path to directory with ESRGAN model file(s).", default=os.path.join(models_path, 'ESRGAN'))
+parser.add_argument("--bsrgan-models-path", type=normalized_filepath, help="Path to directory with BSRGAN model file(s).", default=os.path.join(models_path, 'BSRGAN'))
+parser.add_argument("--realesrgan-models-path", type=normalized_filepath, help="Path to directory with RealESRGAN model file(s).", default=os.path.join(models_path, 'RealESRGAN'))
+parser.add_argument("--dat-models-path", type=normalized_filepath, help="Path to directory with DAT model file(s).", default=os.path.join(models_path, 'DAT'))
+parser.add_argument("--clip-models-path", type=normalized_filepath, help="Path to directory with CLIP model file(s).", default=None)
 parser.add_argument("--xformers", action='store_true', help="enable xformers for cross attention layers")
 parser.add_argument("--force-enable-xformers", action='store_true', help="enable xformers for cross attention layers regardless of whether the checking code thinks you can run it; do not make bug reports if this fails to work")
 parser.add_argument("--xformers-flash-attention", action='store_true', help="enable xformers with Flash Attention to improve reproducibility (supported for SD2.x or variant only)")
@@ -70,28 +72,31 @@
 parser.add_argument("--disable-opt-split-attention", action='store_true', help="prefer no cross-attention layer optimization for automatic choice of optimization")
 parser.add_argument("--disable-nan-check", action='store_true', help="do not check if produced images/latent spaces have nans; useful for running without a checkpoint in CI")
 parser.add_argument("--use-cpu", nargs='+', help="use CPU as torch device for specified modules", default=[], type=str.lower)
+parser.add_argument("--use-ipex", action="store_true", help="use Intel XPU as torch device")
 parser.add_argument("--disable-model-loading-ram-optimization", action='store_true', help="disable an optimization that reduces RAM use when loading a model")
 parser.add_argument("--listen", action='store_true', help="launch gradio with 0.0.0.0 as server name, allowing to respond to network requests")
 parser.add_argument("--port", type=int, help="launch gradio with given server port, you need root/admin rights for ports < 1024, defaults to 7860 if available", default=None)
 parser.add_argument("--show-negative-prompt", action='store_true', help="does not do anything", default=False)
 parser.add_argument("--ui-config-file", type=str, help="filename to use for ui configuration", default=os.path.join(data_path, 'ui-config.json'))
 parser.add_argument("--hide-ui-dir-config", action='store_true', help="hide directory configuration from webui", default=False)
-parser.add_argument("--freeze-settings", action='store_true', help="disable editing settings", default=False)
+parser.add_argument("--freeze-settings", action='store_true', help="disable editing of all settings globally", default=False)
+parser.add_argument("--freeze-settings-in-sections", type=str, help='disable editing settings in specific sections of the settings page by specifying a comma-delimited list such like "saving-images,upscaling". The list of setting names can be found in the modules/shared_options.py file', default=None)
+parser.add_argument("--freeze-specific-settings", type=str, help='disable editing of individual settings by specifying a comma-delimited list like "samples_save,samples_format". The list of setting names can be found in the config.json file', default=None)
 parser.add_argument("--ui-settings-file", type=str, help="filename to use for ui settings", default=os.path.join(data_path, 'config.json'))
 parser.add_argument("--gradio-debug",  action='store_true', help="launch gradio with --debug option")
 parser.add_argument("--gradio-auth", type=str, help='set gradio authentication like "username:password"; or comma-delimit multiple like "u1:p1,u2:p2,u3:p3"', default=None)
-parser.add_argument("--gradio-auth-path", type=str, help='set gradio authentication file path ex. "/path/to/auth/file" same auth format as --gradio-auth', default=None)
+parser.add_argument("--gradio-auth-path", type=normalized_filepath, help='set gradio authentication file path ex. "/path/to/auth/file" same auth format as --gradio-auth', default=None)
 parser.add_argument("--gradio-img2img-tool", type=str, help='does not do anything')
 parser.add_argument("--gradio-inpaint-tool", type=str, help="does not do anything")
 parser.add_argument("--gradio-allowed-path", action='append', help="add path to gradio's allowed_paths, make it possible to serve files from it", default=[data_path])
 parser.add_argument("--opt-channelslast", action='store_true', help="change memory type for stable diffusion to channels last")
-parser.add_argument("--styles-file", type=str, help="filename to use for styles", default=os.path.join(data_path, 'styles.csv'))
+parser.add_argument("--styles-file", type=str, action='append', help="path or wildcard path of styles files, allow multiple entries.", default=[])
 parser.add_argument("--autolaunch", action='store_true', help="open the webui URL in the system's default browser upon launch", default=False)
 parser.add_argument("--theme", type=str, help="launches the UI with light or dark theme", default=None)
 parser.add_argument("--use-textbox-seed", action='store_true', help="use textbox for seeds in UI (no up/down, but possible to input long seeds)", default=False)
 parser.add_argument("--disable-console-progressbars", action='store_true', help="do not output progressbars to console", default=False)
-parser.add_argument("--enable-console-prompts", action='store_true', help="print prompts to console when generating with txt2img and img2img", default=False)
-parser.add_argument('--vae-path', type=str, help='Checkpoint to use as VAE; setting this argument disables all settings related to VAE', default=None)
+parser.add_argument("--enable-console-prompts", action='store_true', help="does not do anything", default=False)  # Legacy compatibility, use as default value shared.opts.enable_console_prompts
+parser.add_argument('--vae-path', type=normalized_filepath, help='Checkpoint to use as VAE; setting this argument disables all settings related to VAE', default=None)
 parser.add_argument("--disable-safe-unpickle", action='store_true', help="disable checking pytorch models for malicious code", default=False)
 parser.add_argument("--api", action='store_true', help="use api=True to launch the API together with the webui (use --nowebui instead for only the API)")
 parser.add_argument("--api-auth", type=str, help='Set authentication for API like "username:password"; or comma-delimit multiple like "u1:p1,u2:p2,u3:p3"', default=None)
@@ -107,13 +112,17 @@
 parser.add_argument("--disable-tls-verify", action="store_false", help="When passed, enables the use of self-signed certificates.", default=None)
 parser.add_argument("--server-name", type=str, help="Sets hostname of server", default=None)
 parser.add_argument("--gradio-queue", action='store_true', help="does not do anything", default=True)
-parser.add_argument("--no-gradio-queue", action='store_true', help="Disables gradio queue; causes the webpage to use http requests instead of websockets; was the defaul in earlier versions")
+parser.add_argument("--no-gradio-queue", action='store_true', help="Disables gradio queue; causes the webpage to use http requests instead of websockets; was the default in earlier versions")
 parser.add_argument("--skip-version-check", action='store_true', help="Do not check versions of torch and xformers")
 parser.add_argument("--no-hashing", action='store_true', help="disable sha256 hashing of checkpoints to help loading performance", default=False)
 parser.add_argument("--no-download-sd-model", action='store_true', help="don't download SD1.5 model even if no model is found in --ckpt-dir", default=False)
 parser.add_argument('--subpath', type=str, help='customize the subpath for gradio, use with reverse proxy')
-parser.add_argument('--add-stop-route', action='store_true', help='add /_stop route to stop server')
+parser.add_argument('--add-stop-route', action='store_true', help='does not do anything')
 parser.add_argument('--api-server-stop', action='store_true', help='enable server stop/restart/kill via api')
 parser.add_argument('--timeout-keep-alive', type=int, default=30, help='set timeout_keep_alive for uvicorn')
 parser.add_argument("--disable-all-extensions", action='store_true', help="prevent all extensions from running regardless of any other settings", default=False)
-parser.add_argument("--disable-extra-extensions", action='store_true', help=" prevent all extensions except built-in from running regardless of any other settings", default=False)
+parser.add_argument("--disable-extra-extensions", action='store_true', help="prevent all extensions except built-in from running regardless of any other settings", default=False)
+parser.add_argument("--skip-load-model-at-start", action='store_true', help="if load a model at web start, only take effect when --nowebui")
+parser.add_argument("--unix-filenames-sanitization", action='store_true', help="allow any symbols except '/' in filenames. May conflict with your browser and file system")
+parser.add_argument("--filenames-max-length", type=int, default=128, help='maximal length of filenames of saved images. If you override it, it can conflict with your file system')
+parser.add_argument("--no-prompt-history", action='store_true', help="disable read prompt from last generation feature; settings this argument will not create '--data_path/params.txt' file")
diff --git a/modules/codeformer/codeformer_arch.py b/modules/codeformer/codeformer_arch.py
deleted file mode 100644
index 12db6814268..00000000000
--- a/modules/codeformer/codeformer_arch.py
+++ /dev/null
@@ -1,276 +0,0 @@
-# this file is copied from CodeFormer repository. Please see comment in modules/codeformer_model.py
-
-import math
-import torch
-from torch import nn, Tensor
-import torch.nn.functional as F
-from typing import Optional
-
-from modules.codeformer.vqgan_arch import VQAutoEncoder, ResBlock
-from basicsr.utils.registry import ARCH_REGISTRY
-
-def calc_mean_std(feat, eps=1e-5):
-    """Calculate mean and std for adaptive_instance_normalization.
-
-    Args:
-        feat (Tensor): 4D tensor.
-        eps (float): A small value added to the variance to avoid
-            divide-by-zero. Default: 1e-5.
-    """
-    size = feat.size()
-    assert len(size) == 4, 'The input feature should be 4D tensor.'
-    b, c = size[:2]
-    feat_var = feat.view(b, c, -1).var(dim=2) + eps
-    feat_std = feat_var.sqrt().view(b, c, 1, 1)
-    feat_mean = feat.view(b, c, -1).mean(dim=2).view(b, c, 1, 1)
-    return feat_mean, feat_std
-
-
-def adaptive_instance_normalization(content_feat, style_feat):
-    """Adaptive instance normalization.
-
-    Adjust the reference features to have the similar color and illuminations
-    as those in the degradate features.
-
-    Args:
-        content_feat (Tensor): The reference feature.
-        style_feat (Tensor): The degradate features.
-    """
-    size = content_feat.size()
-    style_mean, style_std = calc_mean_std(style_feat)
-    content_mean, content_std = calc_mean_std(content_feat)
-    normalized_feat = (content_feat - content_mean.expand(size)) / content_std.expand(size)
-    return normalized_feat * style_std.expand(size) + style_mean.expand(size)
-
-
-class PositionEmbeddingSine(nn.Module):
-    """
-    This is a more standard version of the position embedding, very similar to the one
-    used by the Attention is all you need paper, generalized to work on images.
-    """
-
-    def __init__(self, num_pos_feats=64, temperature=10000, normalize=False, scale=None):
-        super().__init__()
-        self.num_pos_feats = num_pos_feats
-        self.temperature = temperature
-        self.normalize = normalize
-        if scale is not None and normalize is False:
-            raise ValueError("normalize should be True if scale is passed")
-        if scale is None:
-            scale = 2 * math.pi
-        self.scale = scale
-
-    def forward(self, x, mask=None):
-        if mask is None:
-            mask = torch.zeros((x.size(0), x.size(2), x.size(3)), device=x.device, dtype=torch.bool)
-        not_mask = ~mask
-        y_embed = not_mask.cumsum(1, dtype=torch.float32)
-        x_embed = not_mask.cumsum(2, dtype=torch.float32)
-        if self.normalize:
-            eps = 1e-6
-            y_embed = y_embed / (y_embed[:, -1:, :] + eps) * self.scale
-            x_embed = x_embed / (x_embed[:, :, -1:] + eps) * self.scale
-
-        dim_t = torch.arange(self.num_pos_feats, dtype=torch.float32, device=x.device)
-        dim_t = self.temperature ** (2 * (dim_t // 2) / self.num_pos_feats)
-
-        pos_x = x_embed[:, :, :, None] / dim_t
-        pos_y = y_embed[:, :, :, None] / dim_t
-        pos_x = torch.stack(
-            (pos_x[:, :, :, 0::2].sin(), pos_x[:, :, :, 1::2].cos()), dim=4
-        ).flatten(3)
-        pos_y = torch.stack(
-            (pos_y[:, :, :, 0::2].sin(), pos_y[:, :, :, 1::2].cos()), dim=4
-        ).flatten(3)
-        pos = torch.cat((pos_y, pos_x), dim=3).permute(0, 3, 1, 2)
-        return pos
-
-def _get_activation_fn(activation):
-    """Return an activation function given a string"""
-    if activation == "relu":
-        return F.relu
-    if activation == "gelu":
-        return F.gelu
-    if activation == "glu":
-        return F.glu
-    raise RuntimeError(F"activation should be relu/gelu, not {activation}.")
-
-
-class TransformerSALayer(nn.Module):
-    def __init__(self, embed_dim, nhead=8, dim_mlp=2048, dropout=0.0, activation="gelu"):
-        super().__init__()
-        self.self_attn = nn.MultiheadAttention(embed_dim, nhead, dropout=dropout)
-        # Implementation of Feedforward model - MLP
-        self.linear1 = nn.Linear(embed_dim, dim_mlp)
-        self.dropout = nn.Dropout(dropout)
-        self.linear2 = nn.Linear(dim_mlp, embed_dim)
-
-        self.norm1 = nn.LayerNorm(embed_dim)
-        self.norm2 = nn.LayerNorm(embed_dim)
-        self.dropout1 = nn.Dropout(dropout)
-        self.dropout2 = nn.Dropout(dropout)
-
-        self.activation = _get_activation_fn(activation)
-
-    def with_pos_embed(self, tensor, pos: Optional[Tensor]):
-        return tensor if pos is None else tensor + pos
-
-    def forward(self, tgt,
-                tgt_mask: Optional[Tensor] = None,
-                tgt_key_padding_mask: Optional[Tensor] = None,
-                query_pos: Optional[Tensor] = None):
-
-        # self attention
-        tgt2 = self.norm1(tgt)
-        q = k = self.with_pos_embed(tgt2, query_pos)
-        tgt2 = self.self_attn(q, k, value=tgt2, attn_mask=tgt_mask,
-                              key_padding_mask=tgt_key_padding_mask)[0]
-        tgt = tgt + self.dropout1(tgt2)
-
-        # ffn
-        tgt2 = self.norm2(tgt)
-        tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt2))))
-        tgt = tgt + self.dropout2(tgt2)
-        return tgt
-
-class Fuse_sft_block(nn.Module):
-    def __init__(self, in_ch, out_ch):
-        super().__init__()
-        self.encode_enc = ResBlock(2*in_ch, out_ch)
-
-        self.scale = nn.Sequential(
-                    nn.Conv2d(in_ch, out_ch, kernel_size=3, padding=1),
-                    nn.LeakyReLU(0.2, True),
-                    nn.Conv2d(out_ch, out_ch, kernel_size=3, padding=1))
-
-        self.shift = nn.Sequential(
-                    nn.Conv2d(in_ch, out_ch, kernel_size=3, padding=1),
-                    nn.LeakyReLU(0.2, True),
-                    nn.Conv2d(out_ch, out_ch, kernel_size=3, padding=1))
-
-    def forward(self, enc_feat, dec_feat, w=1):
-        enc_feat = self.encode_enc(torch.cat([enc_feat, dec_feat], dim=1))
-        scale = self.scale(enc_feat)
-        shift = self.shift(enc_feat)
-        residual = w * (dec_feat * scale + shift)
-        out = dec_feat + residual
-        return out
-
-
-@ARCH_REGISTRY.register()
-class CodeFormer(VQAutoEncoder):
-    def __init__(self, dim_embd=512, n_head=8, n_layers=9,
-                codebook_size=1024, latent_size=256,
-                connect_list=('32', '64', '128', '256'),
-                fix_modules=('quantize', 'generator')):
-        super(CodeFormer, self).__init__(512, 64, [1, 2, 2, 4, 4, 8], 'nearest',2, [16], codebook_size)
-
-        if fix_modules is not None:
-            for module in fix_modules:
-                for param in getattr(self, module).parameters():
-                    param.requires_grad = False
-
-        self.connect_list = connect_list
-        self.n_layers = n_layers
-        self.dim_embd = dim_embd
-        self.dim_mlp = dim_embd*2
-
-        self.position_emb = nn.Parameter(torch.zeros(latent_size, self.dim_embd))
-        self.feat_emb = nn.Linear(256, self.dim_embd)
-
-        # transformer
-        self.ft_layers = nn.Sequential(*[TransformerSALayer(embed_dim=dim_embd, nhead=n_head, dim_mlp=self.dim_mlp, dropout=0.0)
-                                    for _ in range(self.n_layers)])
-
-        # logits_predict head
-        self.idx_pred_layer = nn.Sequential(
-            nn.LayerNorm(dim_embd),
-            nn.Linear(dim_embd, codebook_size, bias=False))
-
-        self.channels = {
-            '16': 512,
-            '32': 256,
-            '64': 256,
-            '128': 128,
-            '256': 128,
-            '512': 64,
-        }
-
-        # after second residual block for > 16, before attn layer for ==16
-        self.fuse_encoder_block = {'512':2, '256':5, '128':8, '64':11, '32':14, '16':18}
-        # after first residual block for > 16, before attn layer for ==16
-        self.fuse_generator_block = {'16':6, '32': 9, '64':12, '128':15, '256':18, '512':21}
-
-        # fuse_convs_dict
-        self.fuse_convs_dict = nn.ModuleDict()
-        for f_size in self.connect_list:
-            in_ch = self.channels[f_size]
-            self.fuse_convs_dict[f_size] = Fuse_sft_block(in_ch, in_ch)
-
-    def _init_weights(self, module):
-        if isinstance(module, (nn.Linear, nn.Embedding)):
-            module.weight.data.normal_(mean=0.0, std=0.02)
-            if isinstance(module, nn.Linear) and module.bias is not None:
-                module.bias.data.zero_()
-        elif isinstance(module, nn.LayerNorm):
-            module.bias.data.zero_()
-            module.weight.data.fill_(1.0)
-
-    def forward(self, x, w=0, detach_16=True, code_only=False, adain=False):
-        # ################### Encoder #####################
-        enc_feat_dict = {}
-        out_list = [self.fuse_encoder_block[f_size] for f_size in self.connect_list]
-        for i, block in enumerate(self.encoder.blocks):
-            x = block(x)
-            if i in out_list:
-                enc_feat_dict[str(x.shape[-1])] = x.clone()
-
-        lq_feat = x
-        # ################# Transformer ###################
-        # quant_feat, codebook_loss, quant_stats = self.quantize(lq_feat)
-        pos_emb = self.position_emb.unsqueeze(1).repeat(1,x.shape[0],1)
-        # BCHW -> BC(HW) -> (HW)BC
-        feat_emb = self.feat_emb(lq_feat.flatten(2).permute(2,0,1))
-        query_emb = feat_emb
-        # Transformer encoder
-        for layer in self.ft_layers:
-            query_emb = layer(query_emb, query_pos=pos_emb)
-
-        # output logits
-        logits = self.idx_pred_layer(query_emb) # (hw)bn
-        logits = logits.permute(1,0,2) # (hw)bn -> b(hw)n
-
-        if code_only: # for training stage II
-          # logits doesn't need softmax before cross_entropy loss
-            return logits, lq_feat
-
-        # ################# Quantization ###################
-        # if self.training:
-        #     quant_feat = torch.einsum('btn,nc->btc', [soft_one_hot, self.quantize.embedding.weight])
-        #     # b(hw)c -> bc(hw) -> bchw
-        #     quant_feat = quant_feat.permute(0,2,1).view(lq_feat.shape)
-        # ------------
-        soft_one_hot = F.softmax(logits, dim=2)
-        _, top_idx = torch.topk(soft_one_hot, 1, dim=2)
-        quant_feat = self.quantize.get_codebook_feat(top_idx, shape=[x.shape[0],16,16,256])
-        # preserve gradients
-        # quant_feat = lq_feat + (quant_feat - lq_feat).detach()
-
-        if detach_16:
-            quant_feat = quant_feat.detach() # for training stage III
-        if adain:
-            quant_feat = adaptive_instance_normalization(quant_feat, lq_feat)
-
-        # ################## Generator ####################
-        x = quant_feat
-        fuse_list = [self.fuse_generator_block[f_size] for f_size in self.connect_list]
-
-        for i, block in enumerate(self.generator.blocks):
-            x = block(x)
-            if i in fuse_list: # fuse after i-th block
-                f_size = str(x.shape[-1])
-                if w>0:
-                    x = self.fuse_convs_dict[f_size](enc_feat_dict[f_size].detach(), x, w)
-        out = x
-        # logits doesn't need softmax before cross_entropy loss
-        return out, logits, lq_feat
diff --git a/modules/codeformer/vqgan_arch.py b/modules/codeformer/vqgan_arch.py
deleted file mode 100644
index 09ee6660dc5..00000000000
--- a/modules/codeformer/vqgan_arch.py
+++ /dev/null
@@ -1,435 +0,0 @@
-# this file is copied from CodeFormer repository. Please see comment in modules/codeformer_model.py
-
-'''
-VQGAN code, adapted from the original created by the Unleashing Transformers authors:
-https://github.com/samb-t/unleashing-transformers/blob/master/models/vqgan.py
-
-'''
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-from basicsr.utils import get_root_logger
-from basicsr.utils.registry import ARCH_REGISTRY
-
-def normalize(in_channels):
-    return torch.nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)
-
-
-@torch.jit.script
-def swish(x):
-    return x*torch.sigmoid(x)
-
-
-#  Define VQVAE classes
-class VectorQuantizer(nn.Module):
-    def __init__(self, codebook_size, emb_dim, beta):
-        super(VectorQuantizer, self).__init__()
-        self.codebook_size = codebook_size  # number of embeddings
-        self.emb_dim = emb_dim  # dimension of embedding
-        self.beta = beta  # commitment cost used in loss term, beta * ||z_e(x)-sg[e]||^2
-        self.embedding = nn.Embedding(self.codebook_size, self.emb_dim)
-        self.embedding.weight.data.uniform_(-1.0 / self.codebook_size, 1.0 / self.codebook_size)
-
-    def forward(self, z):
-        # reshape z -> (batch, height, width, channel) and flatten
-        z = z.permute(0, 2, 3, 1).contiguous()
-        z_flattened = z.view(-1, self.emb_dim)
-
-        # distances from z to embeddings e_j (z - e)^2 = z^2 + e^2 - 2 e * z
-        d = (z_flattened ** 2).sum(dim=1, keepdim=True) + (self.embedding.weight**2).sum(1) - \
-            2 * torch.matmul(z_flattened, self.embedding.weight.t())
-
-        mean_distance = torch.mean(d)
-        # find closest encodings
-        # min_encoding_indices = torch.argmin(d, dim=1).unsqueeze(1)
-        min_encoding_scores, min_encoding_indices = torch.topk(d, 1, dim=1, largest=False)
-        # [0-1], higher score, higher confidence
-        min_encoding_scores = torch.exp(-min_encoding_scores/10)
-
-        min_encodings = torch.zeros(min_encoding_indices.shape[0], self.codebook_size).to(z)
-        min_encodings.scatter_(1, min_encoding_indices, 1)
-
-        # get quantized latent vectors
-        z_q = torch.matmul(min_encodings, self.embedding.weight).view(z.shape)
-        # compute loss for embedding
-        loss = torch.mean((z_q.detach()-z)**2) + self.beta * torch.mean((z_q - z.detach()) ** 2)
-        # preserve gradients
-        z_q = z + (z_q - z).detach()
-
-        # perplexity
-        e_mean = torch.mean(min_encodings, dim=0)
-        perplexity = torch.exp(-torch.sum(e_mean * torch.log(e_mean + 1e-10)))
-        # reshape back to match original input shape
-        z_q = z_q.permute(0, 3, 1, 2).contiguous()
-
-        return z_q, loss, {
-            "perplexity": perplexity,
-            "min_encodings": min_encodings,
-            "min_encoding_indices": min_encoding_indices,
-            "min_encoding_scores": min_encoding_scores,
-            "mean_distance": mean_distance
-            }
-
-    def get_codebook_feat(self, indices, shape):
-        # input indices: batch*token_num -> (batch*token_num)*1
-        # shape: batch, height, width, channel
-        indices = indices.view(-1,1)
-        min_encodings = torch.zeros(indices.shape[0], self.codebook_size).to(indices)
-        min_encodings.scatter_(1, indices, 1)
-        # get quantized latent vectors
-        z_q = torch.matmul(min_encodings.float(), self.embedding.weight)
-
-        if shape is not None:  # reshape back to match original input shape
-            z_q = z_q.view(shape).permute(0, 3, 1, 2).contiguous()
-
-        return z_q
-
-
-class GumbelQuantizer(nn.Module):
-    def __init__(self, codebook_size, emb_dim, num_hiddens, straight_through=False, kl_weight=5e-4, temp_init=1.0):
-        super().__init__()
-        self.codebook_size = codebook_size  # number of embeddings
-        self.emb_dim = emb_dim  # dimension of embedding
-        self.straight_through = straight_through
-        self.temperature = temp_init
-        self.kl_weight = kl_weight
-        self.proj = nn.Conv2d(num_hiddens, codebook_size, 1)  # projects last encoder layer to quantized logits
-        self.embed = nn.Embedding(codebook_size, emb_dim)
-
-    def forward(self, z):
-        hard = self.straight_through if self.training else True
-
-        logits = self.proj(z)
-
-        soft_one_hot = F.gumbel_softmax(logits, tau=self.temperature, dim=1, hard=hard)
-
-        z_q = torch.einsum("b n h w, n d -> b d h w", soft_one_hot, self.embed.weight)
-
-        # + kl divergence to the prior loss
-        qy = F.softmax(logits, dim=1)
-        diff = self.kl_weight * torch.sum(qy * torch.log(qy * self.codebook_size + 1e-10), dim=1).mean()
-        min_encoding_indices = soft_one_hot.argmax(dim=1)
-
-        return z_q, diff, {
-            "min_encoding_indices": min_encoding_indices
-        }
-
-
-class Downsample(nn.Module):
-    def __init__(self, in_channels):
-        super().__init__()
-        self.conv = torch.nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=2, padding=0)
-
-    def forward(self, x):
-        pad = (0, 1, 0, 1)
-        x = torch.nn.functional.pad(x, pad, mode="constant", value=0)
-        x = self.conv(x)
-        return x
-
-
-class Upsample(nn.Module):
-    def __init__(self, in_channels):
-        super().__init__()
-        self.conv = nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=1, padding=1)
-
-    def forward(self, x):
-        x = F.interpolate(x, scale_factor=2.0, mode="nearest")
-        x = self.conv(x)
-
-        return x
-
-
-class ResBlock(nn.Module):
-    def __init__(self, in_channels, out_channels=None):
-        super(ResBlock, self).__init__()
-        self.in_channels = in_channels
-        self.out_channels = in_channels if out_channels is None else out_channels
-        self.norm1 = normalize(in_channels)
-        self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1, padding=1)
-        self.norm2 = normalize(out_channels)
-        self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=1, padding=1)
-        if self.in_channels != self.out_channels:
-            self.conv_out = nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, padding=0)
-
-    def forward(self, x_in):
-        x = x_in
-        x = self.norm1(x)
-        x = swish(x)
-        x = self.conv1(x)
-        x = self.norm2(x)
-        x = swish(x)
-        x = self.conv2(x)
-        if self.in_channels != self.out_channels:
-            x_in = self.conv_out(x_in)
-
-        return x + x_in
-
-
-class AttnBlock(nn.Module):
-    def __init__(self, in_channels):
-        super().__init__()
-        self.in_channels = in_channels
-
-        self.norm = normalize(in_channels)
-        self.q = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=1,
-            stride=1,
-            padding=0
-        )
-        self.k = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=1,
-            stride=1,
-            padding=0
-        )
-        self.v = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=1,
-            stride=1,
-            padding=0
-        )
-        self.proj_out = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=1,
-            stride=1,
-            padding=0
-        )
-
-    def forward(self, x):
-        h_ = x
-        h_ = self.norm(h_)
-        q = self.q(h_)
-        k = self.k(h_)
-        v = self.v(h_)
-
-        # compute attention
-        b, c, h, w = q.shape
-        q = q.reshape(b, c, h*w)
-        q = q.permute(0, 2, 1)
-        k = k.reshape(b, c, h*w)
-        w_ = torch.bmm(q, k)
-        w_ = w_ * (int(c)**(-0.5))
-        w_ = F.softmax(w_, dim=2)
-
-        # attend to values
-        v = v.reshape(b, c, h*w)
-        w_ = w_.permute(0, 2, 1)
-        h_ = torch.bmm(v, w_)
-        h_ = h_.reshape(b, c, h, w)
-
-        h_ = self.proj_out(h_)
-
-        return x+h_
-
-
-class Encoder(nn.Module):
-    def __init__(self, in_channels, nf, emb_dim, ch_mult, num_res_blocks, resolution, attn_resolutions):
-        super().__init__()
-        self.nf = nf
-        self.num_resolutions = len(ch_mult)
-        self.num_res_blocks = num_res_blocks
-        self.resolution = resolution
-        self.attn_resolutions = attn_resolutions
-
-        curr_res = self.resolution
-        in_ch_mult = (1,)+tuple(ch_mult)
-
-        blocks = []
-        # initial convultion
-        blocks.append(nn.Conv2d(in_channels, nf, kernel_size=3, stride=1, padding=1))
-
-        # residual and downsampling blocks, with attention on smaller res (16x16)
-        for i in range(self.num_resolutions):
-            block_in_ch = nf * in_ch_mult[i]
-            block_out_ch = nf * ch_mult[i]
-            for _ in range(self.num_res_blocks):
-                blocks.append(ResBlock(block_in_ch, block_out_ch))
-                block_in_ch = block_out_ch
-                if curr_res in attn_resolutions:
-                    blocks.append(AttnBlock(block_in_ch))
-
-            if i != self.num_resolutions - 1:
-                blocks.append(Downsample(block_in_ch))
-                curr_res = curr_res // 2
-
-        # non-local attention block
-        blocks.append(ResBlock(block_in_ch, block_in_ch))
-        blocks.append(AttnBlock(block_in_ch))
-        blocks.append(ResBlock(block_in_ch, block_in_ch))
-
-        # normalise and convert to latent size
-        blocks.append(normalize(block_in_ch))
-        blocks.append(nn.Conv2d(block_in_ch, emb_dim, kernel_size=3, stride=1, padding=1))
-        self.blocks = nn.ModuleList(blocks)
-
-    def forward(self, x):
-        for block in self.blocks:
-            x = block(x)
-
-        return x
-
-
-class Generator(nn.Module):
-    def __init__(self, nf, emb_dim, ch_mult, res_blocks, img_size, attn_resolutions):
-        super().__init__()
-        self.nf = nf
-        self.ch_mult = ch_mult
-        self.num_resolutions = len(self.ch_mult)
-        self.num_res_blocks = res_blocks
-        self.resolution = img_size
-        self.attn_resolutions = attn_resolutions
-        self.in_channels = emb_dim
-        self.out_channels = 3
-        block_in_ch = self.nf * self.ch_mult[-1]
-        curr_res = self.resolution // 2 ** (self.num_resolutions-1)
-
-        blocks = []
-        # initial conv
-        blocks.append(nn.Conv2d(self.in_channels, block_in_ch, kernel_size=3, stride=1, padding=1))
-
-        # non-local attention block
-        blocks.append(ResBlock(block_in_ch, block_in_ch))
-        blocks.append(AttnBlock(block_in_ch))
-        blocks.append(ResBlock(block_in_ch, block_in_ch))
-
-        for i in reversed(range(self.num_resolutions)):
-            block_out_ch = self.nf * self.ch_mult[i]
-
-            for _ in range(self.num_res_blocks):
-                blocks.append(ResBlock(block_in_ch, block_out_ch))
-                block_in_ch = block_out_ch
-
-                if curr_res in self.attn_resolutions:
-                    blocks.append(AttnBlock(block_in_ch))
-
-            if i != 0:
-                blocks.append(Upsample(block_in_ch))
-                curr_res = curr_res * 2
-
-        blocks.append(normalize(block_in_ch))
-        blocks.append(nn.Conv2d(block_in_ch, self.out_channels, kernel_size=3, stride=1, padding=1))
-
-        self.blocks = nn.ModuleList(blocks)
-
-
-    def forward(self, x):
-        for block in self.blocks:
-            x = block(x)
-
-        return x
-
-
-@ARCH_REGISTRY.register()
-class VQAutoEncoder(nn.Module):
-    def __init__(self, img_size, nf, ch_mult, quantizer="nearest", res_blocks=2, attn_resolutions=None, codebook_size=1024, emb_dim=256,
-                beta=0.25, gumbel_straight_through=False, gumbel_kl_weight=1e-8, model_path=None):
-        super().__init__()
-        logger = get_root_logger()
-        self.in_channels = 3
-        self.nf = nf
-        self.n_blocks = res_blocks
-        self.codebook_size = codebook_size
-        self.embed_dim = emb_dim
-        self.ch_mult = ch_mult
-        self.resolution = img_size
-        self.attn_resolutions = attn_resolutions or [16]
-        self.quantizer_type = quantizer
-        self.encoder = Encoder(
-            self.in_channels,
-            self.nf,
-            self.embed_dim,
-            self.ch_mult,
-            self.n_blocks,
-            self.resolution,
-            self.attn_resolutions
-        )
-        if self.quantizer_type == "nearest":
-            self.beta = beta #0.25
-            self.quantize = VectorQuantizer(self.codebook_size, self.embed_dim, self.beta)
-        elif self.quantizer_type == "gumbel":
-            self.gumbel_num_hiddens = emb_dim
-            self.straight_through = gumbel_straight_through
-            self.kl_weight = gumbel_kl_weight
-            self.quantize = GumbelQuantizer(
-                self.codebook_size,
-                self.embed_dim,
-                self.gumbel_num_hiddens,
-                self.straight_through,
-                self.kl_weight
-            )
-        self.generator = Generator(
-            self.nf,
-            self.embed_dim,
-            self.ch_mult,
-            self.n_blocks,
-            self.resolution,
-            self.attn_resolutions
-        )
-
-        if model_path is not None:
-            chkpt = torch.load(model_path, map_location='cpu')
-            if 'params_ema' in chkpt:
-                self.load_state_dict(torch.load(model_path, map_location='cpu')['params_ema'])
-                logger.info(f'vqgan is loaded from: {model_path} [params_ema]')
-            elif 'params' in chkpt:
-                self.load_state_dict(torch.load(model_path, map_location='cpu')['params'])
-                logger.info(f'vqgan is loaded from: {model_path} [params]')
-            else:
-                raise ValueError('Wrong params!')
-
-
-    def forward(self, x):
-        x = self.encoder(x)
-        quant, codebook_loss, quant_stats = self.quantize(x)
-        x = self.generator(quant)
-        return x, codebook_loss, quant_stats
-
-
-
-# patch based discriminator
-@ARCH_REGISTRY.register()
-class VQGANDiscriminator(nn.Module):
-    def __init__(self, nc=3, ndf=64, n_layers=4, model_path=None):
-        super().__init__()
-
-        layers = [nn.Conv2d(nc, ndf, kernel_size=4, stride=2, padding=1), nn.LeakyReLU(0.2, True)]
-        ndf_mult = 1
-        ndf_mult_prev = 1
-        for n in range(1, n_layers):  # gradually increase the number of filters
-            ndf_mult_prev = ndf_mult
-            ndf_mult = min(2 ** n, 8)
-            layers += [
-                nn.Conv2d(ndf * ndf_mult_prev, ndf * ndf_mult, kernel_size=4, stride=2, padding=1, bias=False),
-                nn.BatchNorm2d(ndf * ndf_mult),
-                nn.LeakyReLU(0.2, True)
-            ]
-
-        ndf_mult_prev = ndf_mult
-        ndf_mult = min(2 ** n_layers, 8)
-
-        layers += [
-            nn.Conv2d(ndf * ndf_mult_prev, ndf * ndf_mult, kernel_size=4, stride=1, padding=1, bias=False),
-            nn.BatchNorm2d(ndf * ndf_mult),
-            nn.LeakyReLU(0.2, True)
-        ]
-
-        layers += [
-            nn.Conv2d(ndf * ndf_mult, 1, kernel_size=4, stride=1, padding=1)]  # output 1 channel prediction map
-        self.main = nn.Sequential(*layers)
-
-        if model_path is not None:
-            chkpt = torch.load(model_path, map_location='cpu')
-            if 'params_d' in chkpt:
-                self.load_state_dict(torch.load(model_path, map_location='cpu')['params_d'])
-            elif 'params' in chkpt:
-                self.load_state_dict(torch.load(model_path, map_location='cpu')['params'])
-            else:
-                raise ValueError('Wrong params!')
-
-    def forward(self, x):
-        return self.main(x)
diff --git a/modules/codeformer_model.py b/modules/codeformer_model.py
index da42b5e9932..0b353353be2 100644
--- a/modules/codeformer_model.py
+++ b/modules/codeformer_model.py
@@ -1,132 +1,64 @@
-import os
+from __future__ import annotations
 
-import cv2
-import torch
-
-import modules.face_restoration
-import modules.shared
-from modules import shared, devices, modelloader, errors
-from modules.paths import models_path
-
-# codeformer people made a choice to include modified basicsr library to their project which makes
-# it utterly impossible to use it alongside with other libraries that also use basicsr, like GFPGAN.
-# I am making a choice to include some files from codeformer to work around this issue.
-model_dir = "Codeformer"
-model_path = os.path.join(models_path, model_dir)
-model_url = 'https://github.com/sczhou/CodeFormer/releases/download/v0.1.0/codeformer.pth'
-
-codeformer = None
-
-
-def setup_model(dirname):
-    os.makedirs(model_path, exist_ok=True)
-
-    path = modules.paths.paths.get("CodeFormer", None)
-    if path is None:
-        return
-
-    try:
-        from torchvision.transforms.functional import normalize
-        from modules.codeformer.codeformer_arch import CodeFormer
-        from basicsr.utils import img2tensor, tensor2img
-        from facelib.utils.face_restoration_helper import FaceRestoreHelper
-        from facelib.detection.retinaface import retinaface
-
-        net_class = CodeFormer
-
-        class FaceRestorerCodeFormer(modules.face_restoration.FaceRestoration):
-            def name(self):
-                return "CodeFormer"
-
-            def __init__(self, dirname):
-                self.net = None
-                self.face_helper = None
-                self.cmd_dir = dirname
+import logging
 
-            def create_models(self):
-
-                if self.net is not None and self.face_helper is not None:
-                    self.net.to(devices.device_codeformer)
-                    return self.net, self.face_helper
-                model_paths = modelloader.load_models(model_path, model_url, self.cmd_dir, download_name='codeformer-v0.1.0.pth', ext_filter=['.pth'])
-                if len(model_paths) != 0:
-                    ckpt_path = model_paths[0]
-                else:
-                    print("Unable to load codeformer model.")
-                    return None, None
-                net = net_class(dim_embd=512, codebook_size=1024, n_head=8, n_layers=9, connect_list=['32', '64', '128', '256']).to(devices.device_codeformer)
-                checkpoint = torch.load(ckpt_path)['params_ema']
-                net.load_state_dict(checkpoint)
-                net.eval()
-
-                if hasattr(retinaface, 'device'):
-                    retinaface.device = devices.device_codeformer
-                face_helper = FaceRestoreHelper(1, face_size=512, crop_ratio=(1, 1), det_model='retinaface_resnet50', save_ext='png', use_parse=True, device=devices.device_codeformer)
-
-                self.net = net
-                self.face_helper = face_helper
-
-                return net, face_helper
-
-            def send_model_to(self, device):
-                self.net.to(device)
-                self.face_helper.face_det.to(device)
-                self.face_helper.face_parse.to(device)
-
-            def restore(self, np_image, w=None):
-                np_image = np_image[:, :, ::-1]
-
-                original_resolution = np_image.shape[0:2]
+import torch
 
-                self.create_models()
-                if self.net is None or self.face_helper is None:
-                    return np_image
+from modules import (
+    devices,
+    errors,
+    face_restoration,
+    face_restoration_utils,
+    modelloader,
+    shared,
+)
 
-                self.send_model_to(devices.device_codeformer)
+logger = logging.getLogger(__name__)
 
-                self.face_helper.clean_all()
-                self.face_helper.read_image(np_image)
-                self.face_helper.get_face_landmarks_5(only_center_face=False, resize=640, eye_dist_threshold=5)
-                self.face_helper.align_warp_face()
+model_url = 'https://github.com/sczhou/CodeFormer/releases/download/v0.1.0/codeformer.pth'
+model_download_name = 'codeformer-v0.1.0.pth'
 
-                for cropped_face in self.face_helper.cropped_faces:
-                    cropped_face_t = img2tensor(cropped_face / 255., bgr2rgb=True, float32=True)
-                    normalize(cropped_face_t, (0.5, 0.5, 0.5), (0.5, 0.5, 0.5), inplace=True)
-                    cropped_face_t = cropped_face_t.unsqueeze(0).to(devices.device_codeformer)
+# used by e.g. postprocessing_codeformer.py
+codeformer: face_restoration.FaceRestoration | None = None
 
-                    try:
-                        with torch.no_grad():
-                            output = self.net(cropped_face_t, w=w if w is not None else shared.opts.code_former_weight, adain=True)[0]
-                            restored_face = tensor2img(output, rgb2bgr=True, min_max=(-1, 1))
-                        del output
-                        devices.torch_gc()
-                    except Exception:
-                        errors.report('Failed inference for CodeFormer', exc_info=True)
-                        restored_face = tensor2img(cropped_face_t, rgb2bgr=True, min_max=(-1, 1))
 
-                    restored_face = restored_face.astype('uint8')
-                    self.face_helper.add_restored_face(restored_face)
+class FaceRestorerCodeFormer(face_restoration_utils.CommonFaceRestoration):
+    def name(self):
+        return "CodeFormer"
 
-                self.face_helper.get_inverse_affine(None)
+    def load_net(self) -> torch.Module:
+        for model_path in modelloader.load_models(
+            model_path=self.model_path,
+            model_url=model_url,
+            command_path=self.model_path,
+            download_name=model_download_name,
+            ext_filter=['.pth'],
+        ):
+            return modelloader.load_spandrel_model(
+                model_path,
+                device=devices.device_codeformer,
+                expected_architecture='CodeFormer',
+            ).model
+        raise ValueError("No codeformer model found")
 
-                restored_img = self.face_helper.paste_faces_to_input_image()
-                restored_img = restored_img[:, :, ::-1]
+    def get_device(self):
+        return devices.device_codeformer
 
-                if original_resolution != restored_img.shape[0:2]:
-                    restored_img = cv2.resize(restored_img, (0, 0), fx=original_resolution[1]/restored_img.shape[1], fy=original_resolution[0]/restored_img.shape[0], interpolation=cv2.INTER_LINEAR)
+    def restore(self, np_image, w: float | None = None):
+        if w is None:
+            w = getattr(shared.opts, "code_former_weight", 0.5)
 
-                self.face_helper.clean_all()
+        def restore_face(cropped_face_t):
+            assert self.net is not None
+            return self.net(cropped_face_t, weight=w, adain=True)[0]
 
-                if shared.opts.face_restoration_unload:
-                    self.send_model_to(devices.cpu)
+        return self.restore_with_helper(np_image, restore_face)
 
-                return restored_img
 
-        global codeformer
+def setup_model(dirname: str) -> None:
+    global codeformer
+    try:
         codeformer = FaceRestorerCodeFormer(dirname)
         shared.face_restorers.append(codeformer)
-
     except Exception:
         errors.report("Error setting up CodeFormer", exc_info=True)
-
-   # sys.path = stored_sys_path
diff --git a/modules/config_states.py b/modules/config_states.py
index b766aef11d8..651793c7f6f 100644
--- a/modules/config_states.py
+++ b/modules/config_states.py
@@ -4,7 +4,6 @@
 
 import os
 import json
-import time
 import tqdm
 
 from datetime import datetime
@@ -38,7 +37,7 @@ def list_config_states():
     config_states = sorted(config_states, key=lambda cs: cs["created_at"], reverse=True)
 
     for cs in config_states:
-        timestamp = time.asctime(time.gmtime(cs["created_at"]))
+        timestamp = datetime.fromtimestamp(cs["created_at"]).strftime('%Y-%m-%d %H:%M:%S')
         name = cs.get("name", "Config")
         full_name = f"{name}: {timestamp}"
         all_config_states[full_name] = cs
diff --git a/modules/dat_model.py b/modules/dat_model.py
new file mode 100644
index 00000000000..495d5f4937d
--- /dev/null
+++ b/modules/dat_model.py
@@ -0,0 +1,79 @@
+import os
+
+from modules import modelloader, errors
+from modules.shared import cmd_opts, opts
+from modules.upscaler import Upscaler, UpscalerData
+from modules.upscaler_utils import upscale_with_model
+
+
+class UpscalerDAT(Upscaler):
+    def __init__(self, user_path):
+        self.name = "DAT"
+        self.user_path = user_path
+        self.scalers = []
+        super().__init__()
+
+        for file in self.find_models(ext_filter=[".pt", ".pth"]):
+            name = modelloader.friendly_name(file)
+            scaler_data = UpscalerData(name, file, upscaler=self, scale=None)
+            self.scalers.append(scaler_data)
+
+        for model in get_dat_models(self):
+            if model.name in opts.dat_enabled_models:
+                self.scalers.append(model)
+
+    def do_upscale(self, img, path):
+        try:
+            info = self.load_model(path)
+        except Exception:
+            errors.report(f"Unable to load DAT model {path}", exc_info=True)
+            return img
+
+        model_descriptor = modelloader.load_spandrel_model(
+            info.local_data_path,
+            device=self.device,
+            prefer_half=(not cmd_opts.no_half and not cmd_opts.upcast_sampling),
+            expected_architecture="DAT",
+        )
+        return upscale_with_model(
+            model_descriptor,
+            img,
+            tile_size=opts.DAT_tile,
+            tile_overlap=opts.DAT_tile_overlap,
+        )
+
+    def load_model(self, path):
+        for scaler in self.scalers:
+            if scaler.data_path == path:
+                if scaler.local_data_path.startswith("http"):
+                    scaler.local_data_path = modelloader.load_file_from_url(
+                        scaler.data_path,
+                        model_dir=self.model_download_path,
+                    )
+                if not os.path.exists(scaler.local_data_path):
+                    raise FileNotFoundError(f"DAT data missing: {scaler.local_data_path}")
+                return scaler
+        raise ValueError(f"Unable to find model info: {path}")
+
+
+def get_dat_models(scaler):
+    return [
+        UpscalerData(
+            name="DAT x2",
+            path="https://github.com/n0kovo/dat_upscaler_models/raw/main/DAT/DAT_x2.pth",
+            scale=2,
+            upscaler=scaler,
+        ),
+        UpscalerData(
+            name="DAT x3",
+            path="https://github.com/n0kovo/dat_upscaler_models/raw/main/DAT/DAT_x3.pth",
+            scale=3,
+            upscaler=scaler,
+        ),
+        UpscalerData(
+            name="DAT x4",
+            path="https://github.com/n0kovo/dat_upscaler_models/raw/main/DAT/DAT_x4.pth",
+            scale=4,
+            upscaler=scaler,
+        ),
+    ]
diff --git a/modules/deepbooru.py b/modules/deepbooru.py
index 547e1b4c67a..fb043feb296 100644
--- a/modules/deepbooru.py
+++ b/modules/deepbooru.py
@@ -57,7 +57,7 @@ def tag_multi(self, pil_image, force_disable_ranks=False):
         a = np.expand_dims(np.array(pic, dtype=np.float32), 0) / 255
 
         with torch.no_grad(), devices.autocast():
-            x = torch.from_numpy(a).to(devices.device)
+            x = torch.from_numpy(a).to(devices.device, devices.dtype)
             y = self.model(x)[0].detach().cpu().numpy()
 
         probability_dict = {}
diff --git a/modules/devices.py b/modules/devices.py
index c01f06024b4..ee679141ad7 100644
--- a/modules/devices.py
+++ b/modules/devices.py
@@ -3,11 +3,18 @@
 from functools import lru_cache
 
 import torch
-from modules import errors, shared
+from modules import errors, shared, npu_specific
 
 if sys.platform == "darwin":
     from modules import mac_specific
 
+if shared.cmd_opts.use_ipex:
+    from modules import xpu_specific
+
+
+def has_xpu() -> bool:
+    return shared.cmd_opts.use_ipex and xpu_specific.has_xpu
+
 
 def has_mps() -> bool:
     if sys.platform != "darwin":
@@ -16,6 +23,23 @@ def has_mps() -> bool:
         return mac_specific.has_mps
 
 
+def cuda_no_autocast(device_id=None) -> bool:
+    if device_id is None:
+        device_id = get_cuda_device_id()
+    return (
+        torch.cuda.get_device_capability(device_id) == (7, 5)
+        and torch.cuda.get_device_name(device_id).startswith("NVIDIA GeForce GTX 16")
+    )
+
+
+def get_cuda_device_id():
+    return (
+        int(shared.cmd_opts.device_id)
+        if shared.cmd_opts.device_id is not None and shared.cmd_opts.device_id.isdigit()
+        else 0
+    ) or torch.cuda.current_device()
+
+
 def get_cuda_device_string():
     if shared.cmd_opts.device_id is not None:
         return f"cuda:{shared.cmd_opts.device_id}"
@@ -30,6 +54,12 @@ def get_optimal_device_name():
     if has_mps():
         return "mps"
 
+    if has_xpu():
+        return xpu_specific.get_xpu_device_string()
+
+    if npu_specific.has_npu:
+        return npu_specific.get_npu_device_string()
+
     return "cpu"
 
 
@@ -38,7 +68,7 @@ def get_optimal_device():
 
 
 def get_device_for(task):
-    if task in shared.cmd_opts.use_cpu:
+    if task in shared.cmd_opts.use_cpu or "all" in shared.cmd_opts.use_cpu:
         return cpu
 
     return get_optimal_device()
@@ -54,13 +84,26 @@ def torch_gc():
     if has_mps():
         mac_specific.torch_mps_gc()
 
+    if has_xpu():
+        xpu_specific.torch_xpu_gc()
+
+    if npu_specific.has_npu:
+        torch_npu_set_device()
+        npu_specific.torch_npu_gc()
+
+
+def torch_npu_set_device():
+    # Work around due to bug in torch_npu, revert me after fixed, @see https://gitee.com/ascend/pytorch/issues/I8KECW?from=project-issue
+    if npu_specific.has_npu:
+        torch.npu.set_device(0)
+
 
 def enable_tf32():
     if torch.cuda.is_available():
 
         # enabling benchmark option seems to enable a range of cards to do fp16 when they otherwise can't
         # see https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/4407
-        if any(torch.cuda.get_device_capability(devid) == (7, 5) for devid in range(0, torch.cuda.device_count())):
+        if cuda_no_autocast():
             torch.backends.cudnn.benchmark = True
 
         torch.backends.cuda.matmul.allow_tf32 = True
@@ -70,6 +113,10 @@ def enable_tf32():
 errors.run(enable_tf32, "Enabling TF32")
 
 cpu: torch.device = torch.device("cpu")
+fp8: bool = False
+# Force fp16 for all models in inference. No casting during inference.
+# This flag is controlled by "--precision half" command line arg.
+force_fp16: bool = False
 device: torch.device = None
 device_interrogate: torch.device = None
 device_gfpgan: torch.device = None
@@ -78,10 +125,13 @@ def enable_tf32():
 dtype: torch.dtype = torch.float16
 dtype_vae: torch.dtype = torch.float16
 dtype_unet: torch.dtype = torch.float16
+dtype_inference: torch.dtype = torch.float16
 unet_needs_upcast = False
 
 
 def cond_cast_unet(input):
+    if force_fp16:
+        return input.to(torch.float16)
     return input.to(dtype_unet) if unet_needs_upcast else input
 
 
@@ -90,15 +140,94 @@ def cond_cast_float(input):
 
 
 nv_rng = None
+patch_module_list = [
+    torch.nn.Linear,
+    torch.nn.Conv2d,
+    torch.nn.MultiheadAttention,
+    torch.nn.GroupNorm,
+    torch.nn.LayerNorm,
+]
+
+
+def manual_cast_forward(target_dtype):
+    def forward_wrapper(self, *args, **kwargs):
+        if any(
+            isinstance(arg, torch.Tensor) and arg.dtype != target_dtype
+            for arg in args
+        ):
+            args = [arg.to(target_dtype) if isinstance(arg, torch.Tensor) else arg for arg in args]
+            kwargs = {k: v.to(target_dtype) if isinstance(v, torch.Tensor) else v for k, v in kwargs.items()}
+
+        org_dtype = target_dtype
+        for param in self.parameters():
+            if param.dtype != target_dtype:
+                org_dtype = param.dtype
+                break
+
+        if org_dtype != target_dtype:
+            self.to(target_dtype)
+        result = self.org_forward(*args, **kwargs)
+        if org_dtype != target_dtype:
+            self.to(org_dtype)
+
+        if target_dtype != dtype_inference:
+            if isinstance(result, tuple):
+                result = tuple(
+                    i.to(dtype_inference)
+                    if isinstance(i, torch.Tensor)
+                    else i
+                    for i in result
+                )
+            elif isinstance(result, torch.Tensor):
+                result = result.to(dtype_inference)
+        return result
+    return forward_wrapper
+
+
+@contextlib.contextmanager
+def manual_cast(target_dtype):
+    applied = False
+    for module_type in patch_module_list:
+        if hasattr(module_type, "org_forward"):
+            continue
+        applied = True
+        org_forward = module_type.forward
+        if module_type == torch.nn.MultiheadAttention:
+            module_type.forward = manual_cast_forward(torch.float32)
+        else:
+            module_type.forward = manual_cast_forward(target_dtype)
+        module_type.org_forward = org_forward
+    try:
+        yield None
+    finally:
+        if applied:
+            for module_type in patch_module_list:
+                if hasattr(module_type, "org_forward"):
+                    module_type.forward = module_type.org_forward
+                    delattr(module_type, "org_forward")
 
 
 def autocast(disable=False):
     if disable:
         return contextlib.nullcontext()
 
-    if dtype == torch.float32 or shared.cmd_opts.precision == "full":
+    if force_fp16:
+        # No casting during inference if force_fp16 is enabled.
+        # All tensor dtype conversion happens before inference.
+        return contextlib.nullcontext()
+
+    if fp8 and device==cpu:
+        return torch.autocast("cpu", dtype=torch.bfloat16, enabled=True)
+
+    if fp8 and dtype_inference == torch.float32:
+        return manual_cast(dtype)
+
+    if dtype == torch.float32 or dtype_inference == torch.float32:
         return contextlib.nullcontext()
 
+    if has_xpu() or has_mps() or cuda_no_autocast():
+        return manual_cast(dtype)
+
     return torch.autocast("cuda")
 
 
@@ -114,22 +243,22 @@ def test_for_nans(x, where):
     if shared.cmd_opts.disable_nan_check:
         return
 
-    if not torch.all(torch.isnan(x)).item():
+    if not torch.isnan(x[(0, ) * len(x.shape)]):
         return
 
     if where == "unet":
-        message = "A tensor with all NaNs was produced in Unet."
+        message = "A tensor with NaNs was produced in Unet."
 
         if not shared.cmd_opts.no_half:
             message += " This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the \"Upcast cross attention layer to float32\" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this."
 
     elif where == "vae":
-        message = "A tensor with all NaNs was produced in VAE."
+        message = "A tensor with NaNs was produced in VAE."
 
         if not shared.cmd_opts.no_half and not shared.cmd_opts.no_half_vae:
             message += " This could be because there's not enough precision to represent the picture. Try adding --no-half-vae commandline argument to fix this."
     else:
-        message = "A tensor with all NaNs was produced."
+        message = "A tensor with NaNs was produced."
 
     message += " Use --disable-nan-check commandline argument to disable this check."
 
@@ -139,8 +268,8 @@ def test_for_nans(x, where):
 @lru_cache
 def first_time_calculation():
     """
-    just do any calculation with pytorch layers - the first time this is done it allocaltes about 700MB of memory and
-    spends about 2.7 seconds doing that, at least wih NVidia.
+    just do any calculation with pytorch layers - the first time this is done it allocates about 700MB of memory and
+    spends about 2.7 seconds doing that, at least with NVidia.
     """
 
     x = torch.zeros((1, 1)).to(device, dtype)
@@ -151,3 +280,16 @@ def first_time_calculation():
     conv2d = torch.nn.Conv2d(1, 1, (3, 3)).to(device, dtype)
     conv2d(x)
 
+
+def force_model_fp16():
+    """
+    ldm and sgm has modules.diffusionmodules.util.GroupNorm32.forward, which
+    force conversion of input to float32. If force_fp16 is enabled, we need to
+    prevent this casting.
+    """
+    assert force_fp16
+    import sgm.modules.diffusionmodules.util as sgm_util
+    import ldm.modules.diffusionmodules.util as ldm_util
+    sgm_util.GroupNorm32 = torch.nn.GroupNorm
+    ldm_util.GroupNorm32 = torch.nn.GroupNorm
+    print("ldm/sgm GroupNorm32 replaced with normal torch.nn.GroupNorm due to `--precision half`.")
diff --git a/modules/errors.py b/modules/errors.py
index 8c339464d46..48aa13a1728 100644
--- a/modules/errors.py
+++ b/modules/errors.py
@@ -6,6 +6,21 @@
 exception_records = []
 
 
+def format_traceback(tb):
+    return [[f"{x.filename}, line {x.lineno}, {x.name}", x.line] for x in traceback.extract_tb(tb)]
+
+
+def format_exception(e, tb):
+    return {"exception": str(e), "traceback": format_traceback(tb)}
+
+
+def get_exceptions():
+    try:
+        return list(reversed(exception_records))
+    except Exception as e:
+        return str(e)
+
+
 def record_exception():
     _, e, tb = sys.exc_info()
     if e is None:
@@ -14,8 +29,7 @@ def record_exception():
     if exception_records and exception_records[-1] == e:
         return
 
-    from modules import sysinfo
-    exception_records.append(sysinfo.format_exception(e, tb))
+    exception_records.append(format_exception(e, tb))
 
     if len(exception_records) > 5:
         exception_records.pop(0)
@@ -93,8 +107,8 @@ def check_versions():
     import torch
     import gradio
 
-    expected_torch_version = "2.0.0"
-    expected_xformers_version = "0.0.20"
+    expected_torch_version = "2.1.2"
+    expected_xformers_version = "0.0.23.post1"
     expected_gradio_version = "3.41.2"
 
     if version.parse(torch.__version__) < version.parse(expected_torch_version):
diff --git a/modules/esrgan_model.py b/modules/esrgan_model.py
index 02a1727d280..70041ab0234 100644
--- a/modules/esrgan_model.py
+++ b/modules/esrgan_model.py
@@ -1,121 +1,7 @@
-import sys
-
-import numpy as np
-import torch
-from PIL import Image
-
-import modules.esrgan_model_arch as arch
-from modules import modelloader, images, devices
+from modules import modelloader, devices, errors
 from modules.shared import opts
 from modules.upscaler import Upscaler, UpscalerData
-
-
-def mod2normal(state_dict):
-    # this code is copied from https://github.com/victorca25/iNNfer
-    if 'conv_first.weight' in state_dict:
-        crt_net = {}
-        items = list(state_dict)
-
-        crt_net['model.0.weight'] = state_dict['conv_first.weight']
-        crt_net['model.0.bias'] = state_dict['conv_first.bias']
-
-        for k in items.copy():
-            if 'RDB' in k:
-                ori_k = k.replace('RRDB_trunk.', 'model.1.sub.')
-                if '.weight' in k:
-                    ori_k = ori_k.replace('.weight', '.0.weight')
-                elif '.bias' in k:
-                    ori_k = ori_k.replace('.bias', '.0.bias')
-                crt_net[ori_k] = state_dict[k]
-                items.remove(k)
-
-        crt_net['model.1.sub.23.weight'] = state_dict['trunk_conv.weight']
-        crt_net['model.1.sub.23.bias'] = state_dict['trunk_conv.bias']
-        crt_net['model.3.weight'] = state_dict['upconv1.weight']
-        crt_net['model.3.bias'] = state_dict['upconv1.bias']
-        crt_net['model.6.weight'] = state_dict['upconv2.weight']
-        crt_net['model.6.bias'] = state_dict['upconv2.bias']
-        crt_net['model.8.weight'] = state_dict['HRconv.weight']
-        crt_net['model.8.bias'] = state_dict['HRconv.bias']
-        crt_net['model.10.weight'] = state_dict['conv_last.weight']
-        crt_net['model.10.bias'] = state_dict['conv_last.bias']
-        state_dict = crt_net
-    return state_dict
-
-
-def resrgan2normal(state_dict, nb=23):
-    # this code is copied from https://github.com/victorca25/iNNfer
-    if "conv_first.weight" in state_dict and "body.0.rdb1.conv1.weight" in state_dict:
-        re8x = 0
-        crt_net = {}
-        items = list(state_dict)
-
-        crt_net['model.0.weight'] = state_dict['conv_first.weight']
-        crt_net['model.0.bias'] = state_dict['conv_first.bias']
-
-        for k in items.copy():
-            if "rdb" in k:
-                ori_k = k.replace('body.', 'model.1.sub.')
-                ori_k = ori_k.replace('.rdb', '.RDB')
-                if '.weight' in k:
-                    ori_k = ori_k.replace('.weight', '.0.weight')
-                elif '.bias' in k:
-                    ori_k = ori_k.replace('.bias', '.0.bias')
-                crt_net[ori_k] = state_dict[k]
-                items.remove(k)
-
-        crt_net[f'model.1.sub.{nb}.weight'] = state_dict['conv_body.weight']
-        crt_net[f'model.1.sub.{nb}.bias'] = state_dict['conv_body.bias']
-        crt_net['model.3.weight'] = state_dict['conv_up1.weight']
-        crt_net['model.3.bias'] = state_dict['conv_up1.bias']
-        crt_net['model.6.weight'] = state_dict['conv_up2.weight']
-        crt_net['model.6.bias'] = state_dict['conv_up2.bias']
-
-        if 'conv_up3.weight' in state_dict:
-            # modification supporting: https://github.com/ai-forever/Real-ESRGAN/blob/main/RealESRGAN/rrdbnet_arch.py
-            re8x = 3
-            crt_net['model.9.weight'] = state_dict['conv_up3.weight']
-            crt_net['model.9.bias'] = state_dict['conv_up3.bias']
-
-        crt_net[f'model.{8+re8x}.weight'] = state_dict['conv_hr.weight']
-        crt_net[f'model.{8+re8x}.bias'] = state_dict['conv_hr.bias']
-        crt_net[f'model.{10+re8x}.weight'] = state_dict['conv_last.weight']
-        crt_net[f'model.{10+re8x}.bias'] = state_dict['conv_last.bias']
-
-        state_dict = crt_net
-    return state_dict
-
-
-def infer_params(state_dict):
-    # this code is copied from https://github.com/victorca25/iNNfer
-    scale2x = 0
-    scalemin = 6
-    n_uplayer = 0
-    plus = False
-
-    for block in list(state_dict):
-        parts = block.split(".")
-        n_parts = len(parts)
-        if n_parts == 5 and parts[2] == "sub":
-            nb = int(parts[3])
-        elif n_parts == 3:
-            part_num = int(parts[1])
-            if (part_num > scalemin
-                and parts[0] == "model"
-                and parts[2] == "weight"):
-                scale2x += 1
-            if part_num > n_uplayer:
-                n_uplayer = part_num
-                out_nc = state_dict[block].shape[0]
-        if not plus and "conv1x1" in block:
-            plus = True
-
-    nf = state_dict["model.0.weight"].shape[0]
-    in_nc = state_dict["model.0.weight"].shape[1]
-    out_nc = out_nc
-    scale = 2 ** scale2x
-
-    return in_nc, out_nc, nf, nb, plus, scale
+from modules.upscaler_utils import upscale_with_model
 
 
 class UpscalerESRGAN(Upscaler):
@@ -143,12 +29,11 @@ def __init__(self, dirname):
     def do_upscale(self, img, selected_model):
         try:
             model = self.load_model(selected_model)
-        except Exception as e:
-            print(f"Unable to load ESRGAN model {selected_model}: {e}", file=sys.stderr)
+        except Exception:
+            errors.report(f"Unable to load ESRGAN model {selected_model}", exc_info=True)
             return img
         model.to(devices.device_esrgan)
-        img = esrgan_upscale(model, img)
-        return img
+        return esrgan_upscale(model, img)
 
     def load_model(self, path: str):
         if path.startswith("http"):
@@ -161,69 +46,17 @@ def load_model(self, path: str):
         else:
             filename = path
 
-        state_dict = torch.load(filename, map_location='cpu' if devices.device_esrgan.type == 'mps' else None)
-
-        if "params_ema" in state_dict:
-            state_dict = state_dict["params_ema"]
-        elif "params" in state_dict:
-            state_dict = state_dict["params"]
-            num_conv = 16 if "realesr-animevideov3" in filename else 32
-            model = arch.SRVGGNetCompact(num_in_ch=3, num_out_ch=3, num_feat=64, num_conv=num_conv, upscale=4, act_type='prelu')
-            model.load_state_dict(state_dict)
-            model.eval()
-            return model
-
-        if "body.0.rdb1.conv1.weight" in state_dict and "conv_first.weight" in state_dict:
-            nb = 6 if "RealESRGAN_x4plus_anime_6B" in filename else 23
-            state_dict = resrgan2normal(state_dict, nb)
-        elif "conv_first.weight" in state_dict:
-            state_dict = mod2normal(state_dict)
-        elif "model.0.weight" not in state_dict:
-            raise Exception("The file is not a recognized ESRGAN model.")
-
-        in_nc, out_nc, nf, nb, plus, mscale = infer_params(state_dict)
-
-        model = arch.RRDBNet(in_nc=in_nc, out_nc=out_nc, nf=nf, nb=nb, upscale=mscale, plus=plus)
-        model.load_state_dict(state_dict)
-        model.eval()
-
-        return model
-
-
-def upscale_without_tiling(model, img):
-    img = np.array(img)
-    img = img[:, :, ::-1]
-    img = np.ascontiguousarray(np.transpose(img, (2, 0, 1))) / 255
-    img = torch.from_numpy(img).float()
-    img = img.unsqueeze(0).to(devices.device_esrgan)
-    with torch.no_grad():
-        output = model(img)
-    output = output.squeeze().float().cpu().clamp_(0, 1).numpy()
-    output = 255. * np.moveaxis(output, 0, 2)
-    output = output.astype(np.uint8)
-    output = output[:, :, ::-1]
-    return Image.fromarray(output, 'RGB')
+        return modelloader.load_spandrel_model(
+            filename,
+            device=('cpu' if devices.device_esrgan.type == 'mps' else None),
+            expected_architecture='ESRGAN',
+        )
 
 
 def esrgan_upscale(model, img):
-    if opts.ESRGAN_tile == 0:
-        return upscale_without_tiling(model, img)
-
-    grid = images.split_grid(img, opts.ESRGAN_tile, opts.ESRGAN_tile, opts.ESRGAN_tile_overlap)
-    newtiles = []
-    scale_factor = 1
-
-    for y, h, row in grid.tiles:
-        newrow = []
-        for tiledata in row:
-            x, w, tile = tiledata
-
-            output = upscale_without_tiling(model, tile)
-            scale_factor = output.width // tile.width
-
-            newrow.append([x * scale_factor, w * scale_factor, output])
-        newtiles.append([y * scale_factor, h * scale_factor, newrow])
-
-    newgrid = images.Grid(newtiles, grid.tile_w * scale_factor, grid.tile_h * scale_factor, grid.image_w * scale_factor, grid.image_h * scale_factor, grid.overlap * scale_factor)
-    output = images.combine_grid(newgrid)
-    return output
+    return upscale_with_model(
+        model,
+        img,
+        tile_size=opts.ESRGAN_tile,
+        tile_overlap=opts.ESRGAN_tile_overlap,
+    )
diff --git a/modules/esrgan_model_arch.py b/modules/esrgan_model_arch.py
deleted file mode 100644
index 2b9888bafbc..00000000000
--- a/modules/esrgan_model_arch.py
+++ /dev/null
@@ -1,465 +0,0 @@
-# this file is adapted from https://github.com/victorca25/iNNfer
-
-from collections import OrderedDict
-import math
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-
-
-####################
-# RRDBNet Generator
-####################
-
-class RRDBNet(nn.Module):
-    def __init__(self, in_nc, out_nc, nf, nb, nr=3, gc=32, upscale=4, norm_type=None,
-            act_type='leakyrelu', mode='CNA', upsample_mode='upconv', convtype='Conv2D',
-            finalact=None, gaussian_noise=False, plus=False):
-        super(RRDBNet, self).__init__()
-        n_upscale = int(math.log(upscale, 2))
-        if upscale == 3:
-            n_upscale = 1
-
-        self.resrgan_scale = 0
-        if in_nc % 16 == 0:
-            self.resrgan_scale = 1
-        elif in_nc != 4 and in_nc % 4 == 0:
-            self.resrgan_scale = 2
-
-        fea_conv = conv_block(in_nc, nf, kernel_size=3, norm_type=None, act_type=None, convtype=convtype)
-        rb_blocks = [RRDB(nf, nr, kernel_size=3, gc=32, stride=1, bias=1, pad_type='zero',
-            norm_type=norm_type, act_type=act_type, mode='CNA', convtype=convtype,
-            gaussian_noise=gaussian_noise, plus=plus) for _ in range(nb)]
-        LR_conv = conv_block(nf, nf, kernel_size=3, norm_type=norm_type, act_type=None, mode=mode, convtype=convtype)
-
-        if upsample_mode == 'upconv':
-            upsample_block = upconv_block
-        elif upsample_mode == 'pixelshuffle':
-            upsample_block = pixelshuffle_block
-        else:
-            raise NotImplementedError(f'upsample mode [{upsample_mode}] is not found')
-        if upscale == 3:
-            upsampler = upsample_block(nf, nf, 3, act_type=act_type, convtype=convtype)
-        else:
-            upsampler = [upsample_block(nf, nf, act_type=act_type, convtype=convtype) for _ in range(n_upscale)]
-        HR_conv0 = conv_block(nf, nf, kernel_size=3, norm_type=None, act_type=act_type, convtype=convtype)
-        HR_conv1 = conv_block(nf, out_nc, kernel_size=3, norm_type=None, act_type=None, convtype=convtype)
-
-        outact = act(finalact) if finalact else None
-
-        self.model = sequential(fea_conv, ShortcutBlock(sequential(*rb_blocks, LR_conv)),
-            *upsampler, HR_conv0, HR_conv1, outact)
-
-    def forward(self, x, outm=None):
-        if self.resrgan_scale == 1:
-            feat = pixel_unshuffle(x, scale=4)
-        elif self.resrgan_scale == 2:
-            feat = pixel_unshuffle(x, scale=2)
-        else:
-            feat = x
-
-        return self.model(feat)
-
-
-class RRDB(nn.Module):
-    """
-    Residual in Residual Dense Block
-    (ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks)
-    """
-
-    def __init__(self, nf, nr=3, kernel_size=3, gc=32, stride=1, bias=1, pad_type='zero',
-            norm_type=None, act_type='leakyrelu', mode='CNA', convtype='Conv2D',
-            spectral_norm=False, gaussian_noise=False, plus=False):
-        super(RRDB, self).__init__()
-        # This is for backwards compatibility with existing models
-        if nr == 3:
-            self.RDB1 = ResidualDenseBlock_5C(nf, kernel_size, gc, stride, bias, pad_type,
-                    norm_type, act_type, mode, convtype, spectral_norm=spectral_norm,
-                    gaussian_noise=gaussian_noise, plus=plus)
-            self.RDB2 = ResidualDenseBlock_5C(nf, kernel_size, gc, stride, bias, pad_type,
-                    norm_type, act_type, mode, convtype, spectral_norm=spectral_norm,
-                    gaussian_noise=gaussian_noise, plus=plus)
-            self.RDB3 = ResidualDenseBlock_5C(nf, kernel_size, gc, stride, bias, pad_type,
-                    norm_type, act_type, mode, convtype, spectral_norm=spectral_norm,
-                    gaussian_noise=gaussian_noise, plus=plus)
-        else:
-            RDB_list = [ResidualDenseBlock_5C(nf, kernel_size, gc, stride, bias, pad_type,
-                                              norm_type, act_type, mode, convtype, spectral_norm=spectral_norm,
-                                              gaussian_noise=gaussian_noise, plus=plus) for _ in range(nr)]
-            self.RDBs = nn.Sequential(*RDB_list)
-
-    def forward(self, x):
-        if hasattr(self, 'RDB1'):
-            out = self.RDB1(x)
-            out = self.RDB2(out)
-            out = self.RDB3(out)
-        else:
-            out = self.RDBs(x)
-        return out * 0.2 + x
-
-
-class ResidualDenseBlock_5C(nn.Module):
-    """
-    Residual Dense Block
-    The core module of paper: (Residual Dense Network for Image Super-Resolution, CVPR 18)
-    Modified options that can be used:
-        - "Partial Convolution based Padding" arXiv:1811.11718
-        - "Spectral normalization" arXiv:1802.05957
-        - "ICASSP 2020 - ESRGAN+ : Further Improving ESRGAN" N. C.
-            {Rakotonirina} and A. {Rasoanaivo}
-    """
-
-    def __init__(self, nf=64, kernel_size=3, gc=32, stride=1, bias=1, pad_type='zero',
-            norm_type=None, act_type='leakyrelu', mode='CNA', convtype='Conv2D',
-            spectral_norm=False, gaussian_noise=False, plus=False):
-        super(ResidualDenseBlock_5C, self).__init__()
-
-        self.noise = GaussianNoise() if gaussian_noise else None
-        self.conv1x1 = conv1x1(nf, gc) if plus else None
-
-        self.conv1 = conv_block(nf, gc, kernel_size, stride, bias=bias, pad_type=pad_type,
-            norm_type=norm_type, act_type=act_type, mode=mode, convtype=convtype,
-            spectral_norm=spectral_norm)
-        self.conv2 = conv_block(nf+gc, gc, kernel_size, stride, bias=bias, pad_type=pad_type,
-            norm_type=norm_type, act_type=act_type, mode=mode, convtype=convtype,
-            spectral_norm=spectral_norm)
-        self.conv3 = conv_block(nf+2*gc, gc, kernel_size, stride, bias=bias, pad_type=pad_type,
-            norm_type=norm_type, act_type=act_type, mode=mode, convtype=convtype,
-            spectral_norm=spectral_norm)
-        self.conv4 = conv_block(nf+3*gc, gc, kernel_size, stride, bias=bias, pad_type=pad_type,
-            norm_type=norm_type, act_type=act_type, mode=mode, convtype=convtype,
-            spectral_norm=spectral_norm)
-        if mode == 'CNA':
-            last_act = None
-        else:
-            last_act = act_type
-        self.conv5 = conv_block(nf+4*gc, nf, 3, stride, bias=bias, pad_type=pad_type,
-            norm_type=norm_type, act_type=last_act, mode=mode, convtype=convtype,
-            spectral_norm=spectral_norm)
-
-    def forward(self, x):
-        x1 = self.conv1(x)
-        x2 = self.conv2(torch.cat((x, x1), 1))
-        if self.conv1x1:
-            x2 = x2 + self.conv1x1(x)
-        x3 = self.conv3(torch.cat((x, x1, x2), 1))
-        x4 = self.conv4(torch.cat((x, x1, x2, x3), 1))
-        if self.conv1x1:
-            x4 = x4 + x2
-        x5 = self.conv5(torch.cat((x, x1, x2, x3, x4), 1))
-        if self.noise:
-            return self.noise(x5.mul(0.2) + x)
-        else:
-            return x5 * 0.2 + x
-
-
-####################
-# ESRGANplus
-####################
-
-class GaussianNoise(nn.Module):
-    def __init__(self, sigma=0.1, is_relative_detach=False):
-        super().__init__()
-        self.sigma = sigma
-        self.is_relative_detach = is_relative_detach
-        self.noise = torch.tensor(0, dtype=torch.float)
-
-    def forward(self, x):
-        if self.training and self.sigma != 0:
-            self.noise = self.noise.to(x.device)
-            scale = self.sigma * x.detach() if self.is_relative_detach else self.sigma * x
-            sampled_noise = self.noise.repeat(*x.size()).normal_() * scale
-            x = x + sampled_noise
-        return x
-
-def conv1x1(in_planes, out_planes, stride=1):
-    return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride, bias=False)
-
-
-####################
-# SRVGGNetCompact
-####################
-
-class SRVGGNetCompact(nn.Module):
-    """A compact VGG-style network structure for super-resolution.
-    This class is copied from https://github.com/xinntao/Real-ESRGAN
-    """
-
-    def __init__(self, num_in_ch=3, num_out_ch=3, num_feat=64, num_conv=16, upscale=4, act_type='prelu'):
-        super(SRVGGNetCompact, self).__init__()
-        self.num_in_ch = num_in_ch
-        self.num_out_ch = num_out_ch
-        self.num_feat = num_feat
-        self.num_conv = num_conv
-        self.upscale = upscale
-        self.act_type = act_type
-
-        self.body = nn.ModuleList()
-        # the first conv
-        self.body.append(nn.Conv2d(num_in_ch, num_feat, 3, 1, 1))
-        # the first activation
-        if act_type == 'relu':
-            activation = nn.ReLU(inplace=True)
-        elif act_type == 'prelu':
-            activation = nn.PReLU(num_parameters=num_feat)
-        elif act_type == 'leakyrelu':
-            activation = nn.LeakyReLU(negative_slope=0.1, inplace=True)
-        self.body.append(activation)
-
-        # the body structure
-        for _ in range(num_conv):
-            self.body.append(nn.Conv2d(num_feat, num_feat, 3, 1, 1))
-            # activation
-            if act_type == 'relu':
-                activation = nn.ReLU(inplace=True)
-            elif act_type == 'prelu':
-                activation = nn.PReLU(num_parameters=num_feat)
-            elif act_type == 'leakyrelu':
-                activation = nn.LeakyReLU(negative_slope=0.1, inplace=True)
-            self.body.append(activation)
-
-        # the last conv
-        self.body.append(nn.Conv2d(num_feat, num_out_ch * upscale * upscale, 3, 1, 1))
-        # upsample
-        self.upsampler = nn.PixelShuffle(upscale)
-
-    def forward(self, x):
-        out = x
-        for i in range(0, len(self.body)):
-            out = self.body[i](out)
-
-        out = self.upsampler(out)
-        # add the nearest upsampled image, so that the network learns the residual
-        base = F.interpolate(x, scale_factor=self.upscale, mode='nearest')
-        out += base
-        return out
-
-
-####################
-# Upsampler
-####################
-
-class Upsample(nn.Module):
-    r"""Upsamples a given multi-channel 1D (temporal), 2D (spatial) or 3D (volumetric) data.
-    The input data is assumed to be of the form
-    `minibatch x channels x [optional depth] x [optional height] x width`.
-    """
-
-    def __init__(self, size=None, scale_factor=None, mode="nearest", align_corners=None):
-        super(Upsample, self).__init__()
-        if isinstance(scale_factor, tuple):
-            self.scale_factor = tuple(float(factor) for factor in scale_factor)
-        else:
-            self.scale_factor = float(scale_factor) if scale_factor else None
-        self.mode = mode
-        self.size = size
-        self.align_corners = align_corners
-
-    def forward(self, x):
-        return nn.functional.interpolate(x, size=self.size, scale_factor=self.scale_factor, mode=self.mode, align_corners=self.align_corners)
-
-    def extra_repr(self):
-        if self.scale_factor is not None:
-            info = f'scale_factor={self.scale_factor}'
-        else:
-            info = f'size={self.size}'
-        info += f', mode={self.mode}'
-        return info
-
-
-def pixel_unshuffle(x, scale):
-    """ Pixel unshuffle.
-    Args:
-        x (Tensor): Input feature with shape (b, c, hh, hw).
-        scale (int): Downsample ratio.
-    Returns:
-        Tensor: the pixel unshuffled feature.
-    """
-    b, c, hh, hw = x.size()
-    out_channel = c * (scale**2)
-    assert hh % scale == 0 and hw % scale == 0
-    h = hh // scale
-    w = hw // scale
-    x_view = x.view(b, c, h, scale, w, scale)
-    return x_view.permute(0, 1, 3, 5, 2, 4).reshape(b, out_channel, h, w)
-
-
-def pixelshuffle_block(in_nc, out_nc, upscale_factor=2, kernel_size=3, stride=1, bias=True,
-                        pad_type='zero', norm_type=None, act_type='relu', convtype='Conv2D'):
-    """
-    Pixel shuffle layer
-    (Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional
-    Neural Network, CVPR17)
-    """
-    conv = conv_block(in_nc, out_nc * (upscale_factor ** 2), kernel_size, stride, bias=bias,
-                        pad_type=pad_type, norm_type=None, act_type=None, convtype=convtype)
-    pixel_shuffle = nn.PixelShuffle(upscale_factor)
-
-    n = norm(norm_type, out_nc) if norm_type else None
-    a = act(act_type) if act_type else None
-    return sequential(conv, pixel_shuffle, n, a)
-
-
-def upconv_block(in_nc, out_nc, upscale_factor=2, kernel_size=3, stride=1, bias=True,
-                pad_type='zero', norm_type=None, act_type='relu', mode='nearest', convtype='Conv2D'):
-    """ Upconv layer """
-    upscale_factor = (1, upscale_factor, upscale_factor) if convtype == 'Conv3D' else upscale_factor
-    upsample = Upsample(scale_factor=upscale_factor, mode=mode)
-    conv = conv_block(in_nc, out_nc, kernel_size, stride, bias=bias,
-                        pad_type=pad_type, norm_type=norm_type, act_type=act_type, convtype=convtype)
-    return sequential(upsample, conv)
-
-
-
-
-
-
-
-
-####################
-# Basic blocks
-####################
-
-
-def make_layer(basic_block, num_basic_block, **kwarg):
-    """Make layers by stacking the same blocks.
-    Args:
-        basic_block (nn.module): nn.module class for basic block. (block)
-        num_basic_block (int): number of blocks. (n_layers)
-    Returns:
-        nn.Sequential: Stacked blocks in nn.Sequential.
-    """
-    layers = []
-    for _ in range(num_basic_block):
-        layers.append(basic_block(**kwarg))
-    return nn.Sequential(*layers)
-
-
-def act(act_type, inplace=True, neg_slope=0.2, n_prelu=1, beta=1.0):
-    """ activation helper """
-    act_type = act_type.lower()
-    if act_type == 'relu':
-        layer = nn.ReLU(inplace)
-    elif act_type in ('leakyrelu', 'lrelu'):
-        layer = nn.LeakyReLU(neg_slope, inplace)
-    elif act_type == 'prelu':
-        layer = nn.PReLU(num_parameters=n_prelu, init=neg_slope)
-    elif act_type == 'tanh':  # [-1, 1] range output
-        layer = nn.Tanh()
-    elif act_type == 'sigmoid':  # [0, 1] range output
-        layer = nn.Sigmoid()
-    else:
-        raise NotImplementedError(f'activation layer [{act_type}] is not found')
-    return layer
-
-
-class Identity(nn.Module):
-    def __init__(self, *kwargs):
-        super(Identity, self).__init__()
-
-    def forward(self, x, *kwargs):
-        return x
-
-
-def norm(norm_type, nc):
-    """ Return a normalization layer """
-    norm_type = norm_type.lower()
-    if norm_type == 'batch':
-        layer = nn.BatchNorm2d(nc, affine=True)
-    elif norm_type == 'instance':
-        layer = nn.InstanceNorm2d(nc, affine=False)
-    elif norm_type == 'none':
-        def norm_layer(x): return Identity()
-    else:
-        raise NotImplementedError(f'normalization layer [{norm_type}] is not found')
-    return layer
-
-
-def pad(pad_type, padding):
-    """ padding layer helper """
-    pad_type = pad_type.lower()
-    if padding == 0:
-        return None
-    if pad_type == 'reflect':
-        layer = nn.ReflectionPad2d(padding)
-    elif pad_type == 'replicate':
-        layer = nn.ReplicationPad2d(padding)
-    elif pad_type == 'zero':
-        layer = nn.ZeroPad2d(padding)
-    else:
-        raise NotImplementedError(f'padding layer [{pad_type}] is not implemented')
-    return layer
-
-
-def get_valid_padding(kernel_size, dilation):
-    kernel_size = kernel_size + (kernel_size - 1) * (dilation - 1)
-    padding = (kernel_size - 1) // 2
-    return padding
-
-
-class ShortcutBlock(nn.Module):
-    """ Elementwise sum the output of a submodule to its input """
-    def __init__(self, submodule):
-        super(ShortcutBlock, self).__init__()
-        self.sub = submodule
-
-    def forward(self, x):
-        output = x + self.sub(x)
-        return output
-
-    def __repr__(self):
-        return 'Identity + \n|' + self.sub.__repr__().replace('\n', '\n|')
-
-
-def sequential(*args):
-    """ Flatten Sequential. It unwraps nn.Sequential. """
-    if len(args) == 1:
-        if isinstance(args[0], OrderedDict):
-            raise NotImplementedError('sequential does not support OrderedDict input.')
-        return args[0]  # No sequential is needed.
-    modules = []
-    for module in args:
-        if isinstance(module, nn.Sequential):
-            for submodule in module.children():
-                modules.append(submodule)
-        elif isinstance(module, nn.Module):
-            modules.append(module)
-    return nn.Sequential(*modules)
-
-
-def conv_block(in_nc, out_nc, kernel_size, stride=1, dilation=1, groups=1, bias=True,
-               pad_type='zero', norm_type=None, act_type='relu', mode='CNA', convtype='Conv2D',
-               spectral_norm=False):
-    """ Conv layer with padding, normalization, activation """
-    assert mode in ['CNA', 'NAC', 'CNAC'], f'Wrong conv mode [{mode}]'
-    padding = get_valid_padding(kernel_size, dilation)
-    p = pad(pad_type, padding) if pad_type and pad_type != 'zero' else None
-    padding = padding if pad_type == 'zero' else 0
-
-    if convtype=='PartialConv2D':
-        from torchvision.ops import PartialConv2d  # this is definitely not going to work, but PartialConv2d doesn't work anyway and this shuts up static analyzer
-        c = PartialConv2d(in_nc, out_nc, kernel_size=kernel_size, stride=stride, padding=padding,
-               dilation=dilation, bias=bias, groups=groups)
-    elif convtype=='DeformConv2D':
-        from torchvision.ops import DeformConv2d  # not tested
-        c = DeformConv2d(in_nc, out_nc, kernel_size=kernel_size, stride=stride, padding=padding,
-               dilation=dilation, bias=bias, groups=groups)
-    elif convtype=='Conv3D':
-        c = nn.Conv3d(in_nc, out_nc, kernel_size=kernel_size, stride=stride, padding=padding,
-                dilation=dilation, bias=bias, groups=groups)
-    else:
-        c = nn.Conv2d(in_nc, out_nc, kernel_size=kernel_size, stride=stride, padding=padding,
-                dilation=dilation, bias=bias, groups=groups)
-
-    if spectral_norm:
-        c = nn.utils.spectral_norm(c)
-
-    a = act(act_type) if act_type else None
-    if 'CNA' in mode:
-        n = norm(norm_type, out_nc) if norm_type else None
-        return sequential(p, c, n, a)
-    elif mode == 'NAC':
-        if norm_type is None and act_type is not None:
-            a = act(act_type, inplace=False)
-        n = norm(norm_type, in_nc) if norm_type else None
-        return sequential(n, a, p, c)
diff --git a/modules/extensions.py b/modules/extensions.py
index bf9a1878f5d..24de766eb90 100644
--- a/modules/extensions.py
+++ b/modules/extensions.py
@@ -1,11 +1,19 @@
+from __future__ import annotations
+
+import configparser
+import dataclasses
 import os
 import threading
+import re
 
 from modules import shared, errors, cache, scripts
 from modules.gitpython_hack import Repo
 from modules.paths_internal import extensions_dir, extensions_builtin_dir, script_path  # noqa: F401
 
-extensions = []
+extensions: list[Extension] = []
+extension_paths: dict[str, Extension] = {}
+loaded_extensions: dict[str, Exception] = {}
+
 
 os.makedirs(extensions_dir, exist_ok=True)
 
@@ -19,11 +27,87 @@ def active():
         return [x for x in extensions if x.enabled]
 
 
+@dataclasses.dataclass
+class CallbackOrderInfo:
+    name: str
+    before: list
+    after: list
+
+
+class ExtensionMetadata:
+    filename = "metadata.ini"
+    config: configparser.ConfigParser
+    canonical_name: str
+    requires: list
+
+    def __init__(self, path, canonical_name):
+        self.config = configparser.ConfigParser()
+
+        filepath = os.path.join(path, self.filename)
+        # `self.config.read()` will quietly swallow OSErrors (which FileNotFoundError is),
+        # so no need to check whether the file exists beforehand.
+        try:
+            self.config.read(filepath)
+        except Exception:
+            errors.report(f"Error reading {self.filename} for extension {canonical_name}.", exc_info=True)
+
+        self.canonical_name = self.config.get("Extension", "Name", fallback=canonical_name)
+        self.canonical_name = canonical_name.lower().strip()
+
+        self.requires = None
+
+    def get_script_requirements(self, field, section, extra_section=None):
+        """reads a list of requirements from the config; field is the name of the field in the ini file,
+        like Requires or Before, and section is the name of the [section] in the ini file; additionally,
+        reads more requirements from [extra_section] if specified."""
+
+        x = self.config.get(section, field, fallback='')
+
+        if extra_section:
+            x = x + ', ' + self.config.get(extra_section, field, fallback='')
+
+        listed_requirements = self.parse_list(x.lower())
+        res = []
+
+        for requirement in listed_requirements:
+            loaded_requirements = (x for x in requirement.split("|") if x in loaded_extensions)
+            relevant_requirement = next(loaded_requirements, requirement)
+            res.append(relevant_requirement)
+
+        return res
+
+    def parse_list(self, text):
+        """converts a line from config ("ext1 ext2, ext3  ") into a python list (["ext1", "ext2", "ext3"])"""
+
+        if not text:
+            return []
+
+        # both "," and " " are accepted as separator
+        return [x for x in re.split(r"[,\s]+", text.strip()) if x]
+
+    def list_callback_order_instructions(self):
+        for section in self.config.sections():
+            if not section.startswith("callbacks/"):
+                continue
+
+            callback_name = section[10:]
+
+            if not callback_name.startswith(self.canonical_name):
+                errors.report(f"Callback order section for extension {self.canonical_name} is referencing the wrong extension: {section}")
+                continue
+
+            before = self.parse_list(self.config.get(section, 'Before', fallback=''))
+            after = self.parse_list(self.config.get(section, 'After', fallback=''))
+
+            yield CallbackOrderInfo(callback_name, before, after)
+
+
 class Extension:
     lock = threading.Lock()
     cached_fields = ['remote', 'commit_date', 'branch', 'commit_hash', 'version']
+    metadata: ExtensionMetadata
 
-    def __init__(self, name, path, enabled=True, is_builtin=False):
+    def __init__(self, name, path, enabled=True, is_builtin=False, metadata=None):
         self.name = name
         self.path = path
         self.enabled = enabled
@@ -36,6 +120,8 @@ def __init__(self, name, path, enabled=True, is_builtin=False):
         self.branch = None
         self.remote = None
         self.have_info_from_repo = False
+        self.metadata = metadata if metadata else ExtensionMetadata(self.path, name.lower())
+        self.canonical_name = metadata.canonical_name
 
     def to_dict(self):
         return {x: getattr(self, x) for x in self.cached_fields}
@@ -56,6 +142,7 @@ def read_from_repo():
                 self.do_read_info_from_repo()
 
                 return self.to_dict()
+
         try:
             d = cache.cached_data_for_file('extensions-git', self.name, os.path.join(self.path, ".git"), read_from_repo)
             self.from_dict(d)
@@ -104,14 +191,17 @@ def list_files(self, subdir, extension):
 
     def check_updates(self):
         repo = Repo(self.path)
+        branch_name = f'{repo.remote().name}/{self.branch}'
         for fetch in repo.remote().fetch(dry_run=True):
+            if self.branch and fetch.name != branch_name:
+                continue
             if fetch.flags != fetch.HEAD_UPTODATE:
                 self.can_update = True
                 self.status = "new commits"
                 return
 
         try:
-            origin = repo.rev_parse('origin')
+            origin = repo.rev_parse(branch_name)
             if repo.head.commit != origin:
                 self.can_update = True
                 self.status = "behind HEAD"
@@ -124,8 +214,10 @@ def check_updates(self):
         self.can_update = False
         self.status = "latest"
 
-    def fetch_and_reset_hard(self, commit='origin'):
+    def fetch_and_reset_hard(self, commit=None):
         repo = Repo(self.path)
+        if commit is None:
+            commit = f'{repo.remote().name}/{self.branch}'
         # Fix: `error: Your local changes to the following files would be overwritten by merge`,
         # because WSL2 Docker set 755 file permissions instead of 644, this results to the error.
         repo.git.fetch(all=True)
@@ -135,9 +227,8 @@ def fetch_and_reset_hard(self, commit='origin'):
 
 def list_extensions():
     extensions.clear()
-
-    if not os.path.isdir(extensions_dir):
-        return
+    extension_paths.clear()
+    loaded_extensions.clear()
 
     if shared.cmd_opts.disable_all_extensions:
         print("*** \"--disable-all-extensions\" arg was used, will not load any extensions ***")
@@ -148,18 +239,61 @@ def list_extensions():
     elif shared.opts.disable_all_extensions == "extra":
         print("*** \"Disable all extensions\" option was set, will only load built-in extensions ***")
 
-    extension_paths = []
-    for dirname in [extensions_dir, extensions_builtin_dir]:
+
+    # scan through extensions directory and load metadata
+    for dirname in [extensions_builtin_dir, extensions_dir]:
         if not os.path.isdir(dirname):
-            return
+            continue
 
         for extension_dirname in sorted(os.listdir(dirname)):
             path = os.path.join(dirname, extension_dirname)
             if not os.path.isdir(path):
                 continue
 
-            extension_paths.append((extension_dirname, path, dirname == extensions_builtin_dir))
+            canonical_name = extension_dirname
+            metadata = ExtensionMetadata(path, canonical_name)
+
+            # check for duplicated canonical names
+            already_loaded_extension = loaded_extensions.get(metadata.canonical_name)
+            if already_loaded_extension is not None:
+                errors.report(f'Duplicate canonical name "{canonical_name}" found in extensions "{extension_dirname}" and "{already_loaded_extension.name}". Former will be discarded.', exc_info=False)
+                continue
+
+            is_builtin = dirname == extensions_builtin_dir
+            extension = Extension(name=extension_dirname, path=path, enabled=extension_dirname not in shared.opts.disabled_extensions, is_builtin=is_builtin, metadata=metadata)
+            extensions.append(extension)
+            extension_paths[extension.path] = extension
+            loaded_extensions[canonical_name] = extension
+
+    for extension in extensions:
+        extension.metadata.requires = extension.metadata.get_script_requirements("Requires", "Extension")
+
+    # check for requirements
+    for extension in extensions:
+        if not extension.enabled:
+            continue
+
+        for req in extension.metadata.requires:
+            required_extension = loaded_extensions.get(req)
+            if required_extension is None:
+                errors.report(f'Extension "{extension.name}" requires "{req}" which is not installed.', exc_info=False)
+                continue
+
+            if not required_extension.enabled:
+                errors.report(f'Extension "{extension.name}" requires "{required_extension.name}" which is disabled.', exc_info=False)
+                continue
+
+
+def find_extension(filename):
+    parentdir = os.path.dirname(os.path.realpath(filename))
+
+    while parentdir != filename:
+        extension = extension_paths.get(parentdir)
+        if extension is not None:
+            return extension
+
+        filename = parentdir
+        parentdir = os.path.dirname(filename)
+
+    return None
 
-    for dirname, path, is_builtin in extension_paths:
-        extension = Extension(name=dirname, path=path, enabled=dirname not in shared.opts.disabled_extensions, is_builtin=is_builtin)
-        extensions.append(extension)
diff --git a/modules/extra_networks.py b/modules/extra_networks.py
index b9533677887..ae8d42d9b38 100644
--- a/modules/extra_networks.py
+++ b/modules/extra_networks.py
@@ -60,7 +60,7 @@ def activate(self, p, params_list):
         Where name matches the name of this ExtraNetwork object, and arg1:arg2:arg3 are any natural number of text arguments
         separated by colon.
 
-        Even if the user does not mention this ExtraNetwork in his prompt, the call will stil be made, with empty params_list -
+        Even if the user does not mention this ExtraNetwork in his prompt, the call will still be made, with empty params_list -
         in this case, all effects of this extra networks should be disabled.
 
         Can be called multiple times before deactivate() - each new call should override the previous call completely.
@@ -206,7 +206,7 @@ def parse_prompts(prompts):
     return res, extra_data
 
 
-def get_user_metadata(filename):
+def get_user_metadata(filename, lister=None):
     if filename is None:
         return {}
 
@@ -215,7 +215,8 @@ def get_user_metadata(filename):
 
     metadata = {}
     try:
-        if os.path.isfile(metadata_filename):
+        exists = lister.exists(metadata_filename) if lister else os.path.exists(metadata_filename)
+        if exists:
             with open(metadata_filename, "r", encoding="utf8") as file:
                 metadata = json.load(file)
     except Exception as e:
diff --git a/modules/face_restoration_utils.py b/modules/face_restoration_utils.py
new file mode 100644
index 00000000000..1cbac236480
--- /dev/null
+++ b/modules/face_restoration_utils.py
@@ -0,0 +1,180 @@
+from __future__ import annotations
+
+import logging
+import os
+from functools import cached_property
+from typing import TYPE_CHECKING, Callable
+
+import cv2
+import numpy as np
+import torch
+
+from modules import devices, errors, face_restoration, shared
+
+if TYPE_CHECKING:
+    from facexlib.utils.face_restoration_helper import FaceRestoreHelper
+
+logger = logging.getLogger(__name__)
+
+
+def bgr_image_to_rgb_tensor(img: np.ndarray) -> torch.Tensor:
+    """Convert a BGR NumPy image in [0..1] range to a PyTorch RGB float32 tensor."""
+    assert img.shape[2] == 3, "image must be RGB"
+    if img.dtype == "float64":
+        img = img.astype("float32")
+    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+    return torch.from_numpy(img.transpose(2, 0, 1)).float()
+
+
+def rgb_tensor_to_bgr_image(tensor: torch.Tensor, *, min_max=(0.0, 1.0)) -> np.ndarray:
+    """
+    Convert a PyTorch RGB tensor in range `min_max` to a BGR NumPy image in [0..1] range.
+    """
+    tensor = tensor.squeeze(0).float().detach().cpu().clamp_(*min_max)
+    tensor = (tensor - min_max[0]) / (min_max[1] - min_max[0])
+    assert tensor.dim() == 3, "tensor must be RGB"
+    img_np = tensor.numpy().transpose(1, 2, 0)
+    if img_np.shape[2] == 1:  # gray image, no RGB/BGR required
+        return np.squeeze(img_np, axis=2)
+    return cv2.cvtColor(img_np, cv2.COLOR_BGR2RGB)
+
+
+def create_face_helper(device) -> FaceRestoreHelper:
+    from facexlib.detection import retinaface
+    from facexlib.utils.face_restoration_helper import FaceRestoreHelper
+    if hasattr(retinaface, 'device'):
+        retinaface.device = device
+    return FaceRestoreHelper(
+        upscale_factor=1,
+        face_size=512,
+        crop_ratio=(1, 1),
+        det_model='retinaface_resnet50',
+        save_ext='png',
+        use_parse=True,
+        device=device,
+    )
+
+
+def restore_with_face_helper(
+    np_image: np.ndarray,
+    face_helper: FaceRestoreHelper,
+    restore_face: Callable[[torch.Tensor], torch.Tensor],
+) -> np.ndarray:
+    """
+    Find faces in the image using face_helper, restore them using restore_face, and paste them back into the image.
+
+    `restore_face` should take a cropped face image and return a restored face image.
+    """
+    from torchvision.transforms.functional import normalize
+    np_image = np_image[:, :, ::-1]
+    original_resolution = np_image.shape[0:2]
+
+    try:
+        logger.debug("Detecting faces...")
+        face_helper.clean_all()
+        face_helper.read_image(np_image)
+        face_helper.get_face_landmarks_5(only_center_face=False, resize=640, eye_dist_threshold=5)
+        face_helper.align_warp_face()
+        logger.debug("Found %d faces, restoring", len(face_helper.cropped_faces))
+        for cropped_face in face_helper.cropped_faces:
+            cropped_face_t = bgr_image_to_rgb_tensor(cropped_face / 255.0)
+            normalize(cropped_face_t, (0.5, 0.5, 0.5), (0.5, 0.5, 0.5), inplace=True)
+            cropped_face_t = cropped_face_t.unsqueeze(0).to(devices.device_codeformer)
+
+            try:
+                with torch.no_grad():
+                    cropped_face_t = restore_face(cropped_face_t)
+                devices.torch_gc()
+            except Exception:
+                errors.report('Failed face-restoration inference', exc_info=True)
+
+            restored_face = rgb_tensor_to_bgr_image(cropped_face_t, min_max=(-1, 1))
+            restored_face = (restored_face * 255.0).astype('uint8')
+            face_helper.add_restored_face(restored_face)
+
+        logger.debug("Merging restored faces into image")
+        face_helper.get_inverse_affine(None)
+        img = face_helper.paste_faces_to_input_image()
+        img = img[:, :, ::-1]
+        if original_resolution != img.shape[0:2]:
+            img = cv2.resize(
+                img,
+                (0, 0),
+                fx=original_resolution[1] / img.shape[1],
+                fy=original_resolution[0] / img.shape[0],
+                interpolation=cv2.INTER_LINEAR,
+            )
+        logger.debug("Face restoration complete")
+    finally:
+        face_helper.clean_all()
+    return img
+
+
+class CommonFaceRestoration(face_restoration.FaceRestoration):
+    net: torch.Module | None
+    model_url: str
+    model_download_name: str
+
+    def __init__(self, model_path: str):
+        super().__init__()
+        self.net = None
+        self.model_path = model_path
+        os.makedirs(model_path, exist_ok=True)
+
+    @cached_property
+    def face_helper(self) -> FaceRestoreHelper:
+        return create_face_helper(self.get_device())
+
+    def send_model_to(self, device):
+        if self.net:
+            logger.debug("Sending %s to %s", self.net, device)
+            self.net.to(device)
+        if self.face_helper:
+            logger.debug("Sending face helper to %s", device)
+            self.face_helper.face_det.to(device)
+            self.face_helper.face_parse.to(device)
+
+    def get_device(self):
+        raise NotImplementedError("get_device must be implemented by subclasses")
+
+    def load_net(self) -> torch.Module:
+        raise NotImplementedError("load_net must be implemented by subclasses")
+
+    def restore_with_helper(
+        self,
+        np_image: np.ndarray,
+        restore_face: Callable[[torch.Tensor], torch.Tensor],
+    ) -> np.ndarray:
+        try:
+            if self.net is None:
+                self.net = self.load_net()
+        except Exception:
+            logger.warning("Unable to load face-restoration model", exc_info=True)
+            return np_image
+
+        try:
+            self.send_model_to(self.get_device())
+            return restore_with_face_helper(np_image, self.face_helper, restore_face)
+        finally:
+            if shared.opts.face_restoration_unload:
+                self.send_model_to(devices.cpu)
+
+
+def patch_facexlib(dirname: str) -> None:
+    import facexlib.detection
+    import facexlib.parsing
+
+    det_facex_load_file_from_url = facexlib.detection.load_file_from_url
+    par_facex_load_file_from_url = facexlib.parsing.load_file_from_url
+
+    def update_kwargs(kwargs):
+        return dict(kwargs, save_dir=dirname, model_dir=None)
+
+    def facex_load_file_from_url(**kwargs):
+        return det_facex_load_file_from_url(**update_kwargs(kwargs))
+
+    def facex_load_file_from_url2(**kwargs):
+        return par_facex_load_file_from_url(**update_kwargs(kwargs))
+
+    facexlib.detection.load_file_from_url = facex_load_file_from_url
+    facexlib.parsing.load_file_from_url = facex_load_file_from_url2
diff --git a/modules/gfpgan_model.py b/modules/gfpgan_model.py
index 8e0f13bdc7d..01ef899e4a6 100644
--- a/modules/gfpgan_model.py
+++ b/modules/gfpgan_model.py
@@ -1,110 +1,69 @@
+from __future__ import annotations
+
+import logging
 import os
 
-import facexlib
-import gfpgan
+import torch
 
-import modules.face_restoration
-from modules import paths, shared, devices, modelloader, errors
+from modules import (
+    devices,
+    errors,
+    face_restoration,
+    face_restoration_utils,
+    modelloader,
+    shared,
+)
 
-model_dir = "GFPGAN"
-user_path = None
-model_path = os.path.join(paths.models_path, model_dir)
+logger = logging.getLogger(__name__)
 model_url = "https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.4.pth"
-have_gfpgan = False
-loaded_gfpgan_model = None
-
+model_download_name = "GFPGANv1.4.pth"
+gfpgan_face_restorer: face_restoration.FaceRestoration | None = None
 
-def gfpgann():
-    global loaded_gfpgan_model
-    global model_path
-    if loaded_gfpgan_model is not None:
-        loaded_gfpgan_model.gfpgan.to(devices.device_gfpgan)
-        return loaded_gfpgan_model
 
-    if gfpgan_constructor is None:
-        return None
+class FaceRestorerGFPGAN(face_restoration_utils.CommonFaceRestoration):
+    def name(self):
+        return "GFPGAN"
 
-    models = modelloader.load_models(model_path, model_url, user_path, ext_filter="GFPGAN")
-    if len(models) == 1 and models[0].startswith("http"):
-        model_file = models[0]
-    elif len(models) != 0:
-        latest_file = max(models, key=os.path.getctime)
-        model_file = latest_file
-    else:
-        print("Unable to load gfpgan model!")
-        return None
-    if hasattr(facexlib.detection.retinaface, 'device'):
-        facexlib.detection.retinaface.device = devices.device_gfpgan
-    model = gfpgan_constructor(model_path=model_file, upscale=1, arch='clean', channel_multiplier=2, bg_upsampler=None, device=devices.device_gfpgan)
-    loaded_gfpgan_model = model
+    def get_device(self):
+        return devices.device_gfpgan
 
-    return model
+    def load_net(self) -> torch.Module:
+        for model_path in modelloader.load_models(
+            model_path=self.model_path,
+            model_url=model_url,
+            command_path=self.model_path,
+            download_name=model_download_name,
+            ext_filter=['.pth'],
+        ):
+            if 'GFPGAN' in os.path.basename(model_path):
+                return modelloader.load_spandrel_model(
+                    model_path,
+                    device=self.get_device(),
+                    expected_architecture='GFPGAN',
+                ).model
+        raise ValueError("No GFPGAN model found")
 
+    def restore(self, np_image):
+        def restore_face(cropped_face_t):
+            assert self.net is not None
+            return self.net(cropped_face_t, return_rgb=False)[0]
 
-def send_model_to(model, device):
-    model.gfpgan.to(device)
-    model.face_helper.face_det.to(device)
-    model.face_helper.face_parse.to(device)
+        return self.restore_with_helper(np_image, restore_face)
 
 
 def gfpgan_fix_faces(np_image):
-    model = gfpgann()
-    if model is None:
-        return np_image
-
-    send_model_to(model, devices.device_gfpgan)
-
-    np_image_bgr = np_image[:, :, ::-1]
-    cropped_faces, restored_faces, gfpgan_output_bgr = model.enhance(np_image_bgr, has_aligned=False, only_center_face=False, paste_back=True)
-    np_image = gfpgan_output_bgr[:, :, ::-1]
-
-    model.face_helper.clean_all()
-
-    if shared.opts.face_restoration_unload:
-        send_model_to(model, devices.cpu)
-
+    if gfpgan_face_restorer:
+        return gfpgan_face_restorer.restore(np_image)
+    logger.warning("GFPGAN face restorer not set up")
     return np_image
 
 
-gfpgan_constructor = None
-
+def setup_model(dirname: str) -> None:
+    global gfpgan_face_restorer
 
-def setup_model(dirname):
     try:
-        os.makedirs(model_path, exist_ok=True)
-        from gfpgan import GFPGANer
-        from facexlib import detection, parsing  # noqa: F401
-        global user_path
-        global have_gfpgan
-        global gfpgan_constructor
-
-        load_file_from_url_orig = gfpgan.utils.load_file_from_url
-        facex_load_file_from_url_orig = facexlib.detection.load_file_from_url
-        facex_load_file_from_url_orig2 = facexlib.parsing.load_file_from_url
-
-        def my_load_file_from_url(**kwargs):
-            return load_file_from_url_orig(**dict(kwargs, model_dir=model_path))
-
-        def facex_load_file_from_url(**kwargs):
-            return facex_load_file_from_url_orig(**dict(kwargs, save_dir=model_path, model_dir=None))
-
-        def facex_load_file_from_url2(**kwargs):
-            return facex_load_file_from_url_orig2(**dict(kwargs, save_dir=model_path, model_dir=None))
-
-        gfpgan.utils.load_file_from_url = my_load_file_from_url
-        facexlib.detection.load_file_from_url = facex_load_file_from_url
-        facexlib.parsing.load_file_from_url = facex_load_file_from_url2
-        user_path = dirname
-        have_gfpgan = True
-        gfpgan_constructor = GFPGANer
-
-        class FaceRestorerGFPGAN(modules.face_restoration.FaceRestoration):
-            def name(self):
-                return "GFPGAN"
-
-            def restore(self, np_image):
-                return gfpgan_fix_faces(np_image)
-
-        shared.face_restorers.append(FaceRestorerGFPGAN())
+        face_restoration_utils.patch_facexlib(dirname)
+        gfpgan_face_restorer = FaceRestorerGFPGAN(model_path=dirname)
+        shared.face_restorers.append(gfpgan_face_restorer)
     except Exception:
         errors.report("Error setting up GFPGAN", exc_info=True)
diff --git a/modules/gitpython_hack.py b/modules/gitpython_hack.py
index e537c1df93e..b55f0640e5e 100644
--- a/modules/gitpython_hack.py
+++ b/modules/gitpython_hack.py
@@ -23,7 +23,7 @@ def get_object_header(self, ref: str | bytes) -> tuple[str, str, int]:
         )
         return self._parse_object_header(ret)
 
-    def stream_object_data(self, ref: str) -> tuple[str, str, int, "Git.CatFileContentStream"]:
+    def stream_object_data(self, ref: str) -> tuple[str, str, int, Git.CatFileContentStream]:
         # Not really streaming, per se; this buffers the entire object in memory.
         # Shouldn't be a problem for our use case, since we're only using this for
         # object headers (commit objects).
diff --git a/modules/gradio_extensons.py b/modules/gradio_extensons.py
index e6b6835adcc..7d88dc984bb 100644
--- a/modules/gradio_extensons.py
+++ b/modules/gradio_extensons.py
@@ -47,10 +47,20 @@ def Block_get_config(self):
 
 
 def BlockContext_init(self, *args, **kwargs):
+    if scripts.scripts_current is not None:
+        scripts.scripts_current.before_component(self, **kwargs)
+
+    scripts.script_callbacks.before_component_callback(self, **kwargs)
+
     res = original_BlockContext_init(self, *args, **kwargs)
 
     add_classes_to_gradio_component(self)
 
+    scripts.script_callbacks.after_component_callback(self, **kwargs)
+
+    if scripts.scripts_current is not None:
+        scripts.scripts_current.after_component(self, **kwargs)
+
     return res
 
 
diff --git a/modules/hashes.py b/modules/hashes.py
index b7a33b427c5..d22e5fadc47 100644
--- a/modules/hashes.py
+++ b/modules/hashes.py
@@ -21,7 +21,10 @@ def calculate_sha256(filename):
 
 def sha256_from_cache(filename, title, use_addnet_hash=False):
     hashes = cache("hashes-addnet") if use_addnet_hash else cache("hashes")
-    ondisk_mtime = os.path.getmtime(filename)
+    try:
+        ondisk_mtime = os.path.getmtime(filename)
+    except FileNotFoundError:
+        return None
 
     if title not in hashes:
         return None
diff --git a/modules/hat_model.py b/modules/hat_model.py
new file mode 100644
index 00000000000..7f2abb41660
--- /dev/null
+++ b/modules/hat_model.py
@@ -0,0 +1,43 @@
+import os
+import sys
+
+from modules import modelloader, devices
+from modules.shared import opts
+from modules.upscaler import Upscaler, UpscalerData
+from modules.upscaler_utils import upscale_with_model
+
+
+class UpscalerHAT(Upscaler):
+    def __init__(self, dirname):
+        self.name = "HAT"
+        self.scalers = []
+        self.user_path = dirname
+        super().__init__()
+        for file in self.find_models(ext_filter=[".pt", ".pth"]):
+            name = modelloader.friendly_name(file)
+            scale = 4  # TODO: scale might not be 4, but we can't know without loading the model
+            scaler_data = UpscalerData(name, file, upscaler=self, scale=scale)
+            self.scalers.append(scaler_data)
+
+    def do_upscale(self, img, selected_model):
+        try:
+            model = self.load_model(selected_model)
+        except Exception as e:
+            print(f"Unable to load HAT model {selected_model}: {e}", file=sys.stderr)
+            return img
+        model.to(devices.device_esrgan)  # TODO: should probably be device_hat
+        return upscale_with_model(
+            model,
+            img,
+            tile_size=opts.ESRGAN_tile,  # TODO: should probably be HAT_tile
+            tile_overlap=opts.ESRGAN_tile_overlap,  # TODO: should probably be HAT_tile_overlap
+        )
+
+    def load_model(self, path: str):
+        if not os.path.isfile(path):
+            raise FileNotFoundError(f"Model file {path} not found")
+        return modelloader.load_spandrel_model(
+            path,
+            device=devices.device_esrgan,  # TODO: should probably be device_hat
+            expected_architecture='HAT',
+        )
diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py
index 70f1cbd26b6..17454665f28 100644
--- a/modules/hypernetworks/hypernetwork.py
+++ b/modules/hypernetworks/hypernetwork.py
@@ -11,7 +11,7 @@
 from einops import rearrange, repeat
 from ldm.util import default
 from modules import devices, sd_models, shared, sd_samplers, hashes, sd_hijack_checkpoint, errors
-from modules.textual_inversion import textual_inversion, logging
+from modules.textual_inversion import textual_inversion, saving_settings
 from modules.textual_inversion.learn_schedule import LearnRateScheduler
 from torch import einsum
 from torch.nn.init import normal_, xavier_normal_, xavier_uniform_, kaiming_normal_, kaiming_uniform_, zeros_
@@ -95,6 +95,7 @@ def __init__(self, dim, state_dict=None, layer_structure=None, activation_func=N
                         zeros_(b)
                     else:
                         raise KeyError(f"Key {weight_init} is not defined as initialization!")
+        devices.torch_npu_set_device()
         self.to(devices.device)
 
     def fix_old_state_dict(self, state_dict):
@@ -468,7 +469,7 @@ def create_hypernetwork(name, enable_sizes, overwrite_old, layer_structure=None,
     shared.reload_hypernetworks()
 
 
-def train_hypernetwork(id_task, hypernetwork_name, learn_rate, batch_size, gradient_step, data_root, log_directory, training_width, training_height, varsize, steps, clip_grad_mode, clip_grad_value, shuffle_tags, tag_drop_out, latent_sampling_method, use_weight, create_image_every, save_hypernetwork_every, template_filename, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height):
+def train_hypernetwork(id_task, hypernetwork_name: str, learn_rate: float, batch_size: int, gradient_step: int, data_root: str, log_directory: str, training_width: int, training_height: int, varsize: bool, steps: int, clip_grad_mode: str, clip_grad_value: float, shuffle_tags: bool, tag_drop_out: bool, latent_sampling_method: str, use_weight: bool, create_image_every: int, save_hypernetwork_every: int, template_filename: str, preview_from_txt2img: bool, preview_prompt: str, preview_negative_prompt: str, preview_steps: int, preview_sampler_name: str, preview_cfg_scale: float, preview_seed: int, preview_width: int, preview_height: int):
     from modules import images, processing
 
     save_hypernetwork_every = save_hypernetwork_every or 0
@@ -532,7 +533,7 @@ def train_hypernetwork(id_task, hypernetwork_name, learn_rate, batch_size, gradi
             model_name=checkpoint.model_name, model_hash=checkpoint.shorthash, num_of_dataset_images=len(ds),
             **{field: getattr(hypernetwork, field) for field in ['layer_structure', 'activation_func', 'weight_init', 'add_layer_norm', 'use_dropout', ]}
         )
-        logging.save_settings_to_file(log_directory, {**saved_params, **locals()})
+        saving_settings.save_settings_to_file(log_directory, {**saved_params, **locals()})
 
     latent_sampling_method = ds.latent_sampling_method
 
@@ -698,7 +699,7 @@ def train_hypernetwork(id_task, hypernetwork_name, learn_rate, batch_size, gradi
                         p.prompt = preview_prompt
                         p.negative_prompt = preview_negative_prompt
                         p.steps = preview_steps
-                        p.sampler_name = sd_samplers.samplers[preview_sampler_index].name
+                        p.sampler_name = sd_samplers.samplers_map[preview_sampler_name.lower()]
                         p.cfg_scale = preview_cfg_scale
                         p.seed = preview_seed
                         p.width = preview_width
diff --git a/modules/images.py b/modules/images.py
index eb644733898..cfdfb338446 100644
--- a/modules/images.py
+++ b/modules/images.py
@@ -1,7 +1,7 @@
 from __future__ import annotations
 
 import datetime
-
+import functools
 import pytz
 import io
 import math
@@ -12,7 +12,9 @@
 import numpy as np
 import piexif
 import piexif.helper
-from PIL import Image, ImageFont, ImageDraw, ImageColor, PngImagePlugin
+from PIL import Image, ImageFont, ImageDraw, ImageColor, PngImagePlugin, ImageOps
+# pillow_avif needs to be imported somewhere in code for it to work
+import pillow_avif # noqa: F401
 import string
 import json
 import hashlib
@@ -52,21 +54,29 @@ def image_grid(imgs, batch_size=1, rows=None):
     params = script_callbacks.ImageGridLoopParams(imgs, cols, rows)
     script_callbacks.image_grid_callback(params)
 
-    w, h = imgs[0].size
-    grid = Image.new('RGB', size=(params.cols * w, params.rows * h), color='black')
+    w, h = map(max, zip(*(img.size for img in imgs)))
+    grid_background_color = ImageColor.getcolor(opts.grid_background_color, 'RGB')
+    grid = Image.new('RGB', size=(params.cols * w, params.rows * h), color=grid_background_color)
 
     for i, img in enumerate(params.imgs):
-        grid.paste(img, box=(i % params.cols * w, i // params.cols * h))
+        img_w, img_h = img.size
+        w_offset, h_offset = 0 if img_w == w else (w - img_w) // 2, 0 if img_h == h else (h - img_h) // 2
+        grid.paste(img, box=(i % params.cols * w + w_offset, i // params.cols * h + h_offset))
 
     return grid
 
 
-Grid = namedtuple("Grid", ["tiles", "tile_w", "tile_h", "image_w", "image_h", "overlap"])
+class Grid(namedtuple("_Grid", ["tiles", "tile_w", "tile_h", "image_w", "image_h", "overlap"])):
+    @property
+    def tile_count(self) -> int:
+        """
+        The total number of tiles in the grid.
+        """
+        return sum(len(row[2]) for row in self.tiles)
 
 
-def split_grid(image, tile_w=512, tile_h=512, overlap=64):
-    w = image.width
-    h = image.height
+def split_grid(image: Image.Image, tile_w: int = 512, tile_h: int = 512, overlap: int = 64) -> Grid:
+    w, h = image.size
 
     non_overlap_width = tile_w - overlap
     non_overlap_height = tile_h - overlap
@@ -316,13 +326,16 @@ def resize(im, w, h):
     return res
 
 
-invalid_filename_chars = '<>:"/\\|?*\n\r\t'
+if not shared.cmd_opts.unix_filenames_sanitization:
+    invalid_filename_chars = '#<>:"/\\|?*\n\r\t'
+else:
+    invalid_filename_chars = '/'
 invalid_filename_prefix = ' '
 invalid_filename_postfix = ' .'
 re_nonletters = re.compile(r'[\s' + string.punctuation + ']+')
 re_pattern = re.compile(r"(.*?)(?:\[([^\[\]]+)\]|$)")
 re_pattern_arg = re.compile(r"(.*)<([^>]*)>$")
-max_filename_part_length = 128
+max_filename_part_length = shared.cmd_opts.filenames_max_length
 NOTHING_AND_SKIP_PREVIOUS_TEXT = object()
 
 
@@ -339,8 +352,35 @@ def sanitize_filename_part(text, replace_spaces=True):
     return text
 
 
+@functools.cache
+def get_scheduler_str(sampler_name, scheduler_name):
+    """Returns {Scheduler} if the scheduler is applicable to the sampler"""
+    if scheduler_name == 'Automatic':
+        config = sd_samplers.find_sampler_config(sampler_name)
+        scheduler_name = config.options.get('scheduler', 'Automatic')
+    return scheduler_name.capitalize()
+
+
+@functools.cache
+def get_sampler_scheduler_str(sampler_name, scheduler_name):
+    """Returns the '{Sampler} {Scheduler}' if the scheduler is applicable to the sampler"""
+    return f'{sampler_name} {get_scheduler_str(sampler_name, scheduler_name)}'
+
+
+def get_sampler_scheduler(p, sampler):
+    """Returns '{Sampler} {Scheduler}' / '{Scheduler}' / 'NOTHING_AND_SKIP_PREVIOUS_TEXT'"""
+    if hasattr(p, 'scheduler') and hasattr(p, 'sampler_name'):
+        if sampler:
+            sampler_scheduler = get_sampler_scheduler_str(p.sampler_name, p.scheduler)
+        else:
+            sampler_scheduler = get_scheduler_str(p.sampler_name, p.scheduler)
+        return sanitize_filename_part(sampler_scheduler, replace_spaces=False)
+    return NOTHING_AND_SKIP_PREVIOUS_TEXT
+
+
 class FilenameGenerator:
     replacements = {
+        'basename': lambda self: self.basename or 'img',
         'seed': lambda self: self.seed if self.seed is not None else '',
         'seed_first': lambda self: self.seed if self.p.batch_size == 1 else self.p.all_seeds[0],
         'seed_last': lambda self: NOTHING_AND_SKIP_PREVIOUS_TEXT if self.p.batch_size == 1 else self.p.all_seeds[-1],
@@ -350,6 +390,8 @@ class FilenameGenerator:
         'height': lambda self: self.image.height,
         'styles': lambda self: self.p and sanitize_filename_part(", ".join([style for style in self.p.styles if not style == "None"]) or "None", replace_spaces=False),
         'sampler': lambda self: self.p and sanitize_filename_part(self.p.sampler_name, replace_spaces=False),
+        'sampler_scheduler': lambda self: self.p and get_sampler_scheduler(self.p, True),
+        'scheduler': lambda self: self.p and get_sampler_scheduler(self.p, False),
         'model_hash': lambda self: getattr(self.p, "sd_model_hash", shared.sd_model.sd_model_hash),
         'model_name': lambda self: sanitize_filename_part(shared.sd_model.sd_checkpoint_info.name_for_extra, replace_spaces=False),
         'date': lambda self: datetime.datetime.now().strftime('%Y-%m-%d'),
@@ -375,12 +417,13 @@ class FilenameGenerator:
     }
     default_time_format = '%Y%m%d%H%M%S'
 
-    def __init__(self, p, seed, prompt, image, zip=False):
+    def __init__(self, p, seed, prompt, image, zip=False, basename=""):
         self.p = p
         self.seed = seed
         self.prompt = prompt
         self.image = image
         self.zip = zip
+        self.basename = basename
 
     def get_vae_filename(self):
         """Get the name of the VAE file."""
@@ -561,6 +604,19 @@ def save_image_with_geninfo(image, geninfo, filename, extension=None, existing_p
             })
 
             piexif.insert(exif_bytes, filename)
+    elif extension.lower() == '.avif':
+        if opts.enable_pnginfo and geninfo is not None:
+            exif_bytes = piexif.dump({
+                "Exif": {
+                    piexif.ExifIFD.UserComment: piexif.helper.UserComment.dump(geninfo or "", encoding="unicode")
+                },
+            })
+        else:
+            exif_bytes = None
+
+        image.save(filename,format=image_format, quality=opts.jpeg_quality, exif=exif_bytes)
+    elif extension.lower() == ".gif":
+        image.save(filename, format=image_format, comment=geninfo)
     else:
         image.save(filename, format=image_format, quality=opts.jpeg_quality)
 
@@ -598,12 +654,12 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i
         txt_fullfn (`str` or None):
             If a text file is saved for this image, this will be its full path. Otherwise None.
     """
-    namegen = FilenameGenerator(p, seed, prompt, image)
+    namegen = FilenameGenerator(p, seed, prompt, image, basename=basename)
 
     # WebP and JPG formats have maximum dimension limits of 16383 and 65535 respectively. switch to PNG which has a much higher limit
     if (image.height > 65535 or image.width > 65535) and extension.lower() in ("jpg", "jpeg") or (image.height > 16383 or image.width > 16383) and extension.lower() == "webp":
         print('Image dimensions too large; saving as PNG')
-        extension = ".png"
+        extension = "png"
 
     if save_to_dirs is None:
         save_to_dirs = (grid and opts.grid_save_to_dirs) or (not grid and opts.save_to_dirs and not no_prompt)
@@ -661,7 +717,13 @@ def _atomically_save_image(image_to_save, filename_without_extension, extension)
 
         save_image_with_geninfo(image_to_save, info, temp_file_path, extension, existing_pnginfo=params.pnginfo, pnginfo_section_name=pnginfo_section_name)
 
-        os.replace(temp_file_path, filename_without_extension + extension)
+        filename = filename_without_extension + extension
+        if shared.opts.save_images_replace_action != "Replace":
+            n = 0
+            while os.path.exists(filename):
+                n += 1
+                filename = f"{filename_without_extension}-{n}{extension}"
+        os.replace(temp_file_path, filename)
 
     fullfn_without_extension, extension = os.path.splitext(params.filename)
     if hasattr(os, 'statvfs'):
@@ -718,7 +780,12 @@ def read_info_from_image(image: Image.Image) -> tuple[str | None, dict]:
     geninfo = items.pop('parameters', None)
 
     if "exif" in items:
-        exif = piexif.load(items["exif"])
+        exif_data = items["exif"]
+        try:
+            exif = piexif.load(exif_data)
+        except OSError:
+            # memory / exif was not valid so piexif tried to read from a file
+            exif = None
         exif_comment = (exif or {}).get("Exif", {}).get(piexif.ExifIFD.UserComment, b'')
         try:
             exif_comment = piexif.helper.UserComment.load(exif_comment)
@@ -726,8 +793,12 @@ def read_info_from_image(image: Image.Image) -> tuple[str | None, dict]:
             exif_comment = exif_comment.decode('utf8', errors="ignore")
 
         if exif_comment:
-            items['exif comment'] = exif_comment
             geninfo = exif_comment
+    elif "comment" in items: # for gif
+        if isinstance(items["comment"], bytes):
+            geninfo = items["comment"].decode('utf8', errors="ignore")
+        else:
+            geninfo = items["comment"]
 
     for field in IGNORED_INFO_KEYS:
         items.pop(field, None)
@@ -750,7 +821,7 @@ def image_data(data):
     import gradio as gr
 
     try:
-        image = Image.open(io.BytesIO(data))
+        image = read(io.BytesIO(data))
         textinfo, _ = read_info_from_image(image)
         return textinfo, None
     except Exception:
@@ -776,3 +847,31 @@ def flatten(img, bgcolor):
         img = background
 
     return img.convert('RGB')
+
+
+def read(fp, **kwargs):
+    image = Image.open(fp, **kwargs)
+    image = fix_image(image)
+
+    return image
+
+
+def fix_image(image: Image.Image):
+    if image is None:
+        return None
+
+    try:
+        image = ImageOps.exif_transpose(image)
+        image = fix_png_transparency(image)
+    except Exception:
+        pass
+
+    return image
+
+
+def fix_png_transparency(image: Image.Image):
+    if image.mode not in ("RGB", "P") or not isinstance(image.info.get("transparency"), bytes):
+        return image
+
+    image = image.convert("RGBA")
+    return image
diff --git a/modules/img2img.py b/modules/img2img.py
index 1519e132b2b..24f869f5c6a 100644
--- a/modules/img2img.py
+++ b/modules/img2img.py
@@ -6,21 +6,25 @@
 from PIL import Image, ImageOps, ImageFilter, ImageEnhance, UnidentifiedImageError
 import gradio as gr
 
-from modules import images as imgutil
-from modules.generation_parameters_copypaste import create_override_settings_dict, parse_generation_parameters
+from modules import images
+from modules.infotext_utils import create_override_settings_dict, parse_generation_parameters
 from modules.processing import Processed, StableDiffusionProcessingImg2Img, process_images
 from modules.shared import opts, state
+from modules.sd_models import get_closet_checkpoint_match
 import modules.shared as shared
 import modules.processing as processing
 from modules.ui import plaintext_to_html
 import modules.scripts
 
 
-def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=False, scale_by=1.0, use_png_info=False, png_info_props=None, png_info_dir=None):
+def process_batch(p, input, output_dir, inpaint_mask_dir, args, to_scale=False, scale_by=1.0, use_png_info=False, png_info_props=None, png_info_dir=None):
     output_dir = output_dir.strip()
     processing.fix_seed(p)
 
-    images = list(shared.walk_files(input_dir, allowed_extensions=(".png", ".jpg", ".jpeg", ".webp", ".tif", ".tiff")))
+    if isinstance(input, str):
+        batch_images = list(shared.walk_files(input, allowed_extensions=(".png", ".jpg", ".jpeg", ".webp", ".tif", ".tiff")))
+    else:
+        batch_images = [os.path.abspath(x.name) for x in input]
 
     is_inpaint_batch = False
     if inpaint_mask_dir:
@@ -30,9 +34,9 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=Fal
         if is_inpaint_batch:
             print(f"\nInpaint batch is enabled. {len(inpaint_masks)} masks found.")
 
-    print(f"Will process {len(images)} images, creating {p.n_iter * p.batch_size} new images for each.")
+    print(f"Will process {len(batch_images)} images, creating {p.n_iter * p.batch_size} new images for each.")
 
-    state.job_count = len(images) * p.n_iter
+    state.job_count = len(batch_images) * p.n_iter
 
     # extract "default" params to use in case getting png info fails
     prompt = p.prompt
@@ -41,17 +45,20 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=Fal
     cfg_scale = p.cfg_scale
     sampler_name = p.sampler_name
     steps = p.steps
-
-    for i, image in enumerate(images):
-        state.job = f"{i+1} out of {len(images)}"
+    override_settings = p.override_settings
+    sd_model_checkpoint_override = get_closet_checkpoint_match(override_settings.get("sd_model_checkpoint", None))
+    batch_results = None
+    discard_further_results = False
+    for i, image in enumerate(batch_images):
+        state.job = f"{i+1} out of {len(batch_images)}"
         if state.skipped:
             state.skipped = False
 
-        if state.interrupted:
+        if state.interrupted or state.stopping_generation:
             break
 
         try:
-            img = Image.open(image)
+            img = images.read(image)
         except UnidentifiedImageError as e:
             print(e)
             continue
@@ -82,7 +89,7 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=Fal
                 # otherwise user has many masks with the same name but different extensions
                 mask_image_path = masks_found[0]
 
-            mask_image = Image.open(mask_image_path)
+            mask_image = images.read(mask_image_path)
             p.image_mask = mask_image
 
         if use_png_info:
@@ -90,8 +97,8 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=Fal
                 info_img = img
                 if png_info_dir:
                     info_img_path = os.path.join(png_info_dir, os.path.basename(image))
-                    info_img = Image.open(info_img_path)
-                geninfo, _ = imgutil.read_info_from_image(info_img)
+                    info_img = images.read(info_img_path)
+                geninfo, _ = images.read_info_from_image(info_img)
                 parsed_parameters = parse_generation_parameters(geninfo)
                 parsed_parameters = {k: v for k, v in parsed_parameters.items() if k in (png_info_props or {})}
             except Exception:
@@ -104,19 +111,45 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=Fal
             p.sampler_name = parsed_parameters.get("Sampler", sampler_name)
             p.steps = int(parsed_parameters.get("Steps", steps))
 
+            model_info = get_closet_checkpoint_match(parsed_parameters.get("Model hash", None))
+            if model_info is not None:
+                p.override_settings['sd_model_checkpoint'] = model_info.name
+            elif sd_model_checkpoint_override:
+                p.override_settings['sd_model_checkpoint'] = sd_model_checkpoint_override
+            else:
+                p.override_settings.pop("sd_model_checkpoint", None)
+
+        if output_dir:
+            p.outpath_samples = output_dir
+            p.override_settings['save_to_dirs'] = False
+            p.override_settings['save_images_replace_action'] = "Add number suffix"
+            if p.n_iter > 1 or p.batch_size > 1:
+                p.override_settings['samples_filename_pattern'] = f'{image_path.stem}-[generation_number]'
+            else:
+                p.override_settings['samples_filename_pattern'] = f'{image_path.stem}'
+
         proc = modules.scripts.scripts_img2img.run(p, *args)
+
         if proc is None:
-            if output_dir:
-                p.outpath_samples = output_dir
-                p.override_settings['save_to_dirs'] = False
-                if p.n_iter > 1 or p.batch_size > 1:
-                    p.override_settings['samples_filename_pattern'] = f'{image_path.stem}-[generation_number]'
-                else:
-                    p.override_settings['samples_filename_pattern'] = f'{image_path.stem}'
-            process_images(p)
+            p.override_settings.pop('save_images_replace_action', None)
+            proc = process_images(p)
 
+        if not discard_further_results and proc:
+            if batch_results:
+                batch_results.images.extend(proc.images)
+                batch_results.infotexts.extend(proc.infotexts)
+            else:
+                batch_results = proc
 
-def img2img(id_task: str, mode: int, prompt: str, negative_prompt: str, prompt_styles, init_img, sketch, init_img_with_mask, inpaint_color_sketch, inpaint_color_sketch_orig, init_img_inpaint, init_mask_inpaint, steps: int, sampler_name: str, mask_blur: int, mask_alpha: float, inpainting_fill: int, n_iter: int, batch_size: int, cfg_scale: float, image_cfg_scale: float, denoising_strength: float, selected_scale_tab: int, height: int, width: int, scale_by: float, resize_mode: int, inpaint_full_res: bool, inpaint_full_res_padding: int, inpainting_mask_invert: int, img2img_batch_input_dir: str, img2img_batch_output_dir: str, img2img_batch_inpaint_mask_dir: str, override_settings_texts, img2img_batch_use_png_info: bool, img2img_batch_png_info_props: list, img2img_batch_png_info_dir: str, request: gr.Request, *args):
+            if 0 <= shared.opts.img2img_batch_show_results_limit < len(batch_results.images):
+                discard_further_results = True
+                batch_results.images = batch_results.images[:int(shared.opts.img2img_batch_show_results_limit)]
+                batch_results.infotexts = batch_results.infotexts[:int(shared.opts.img2img_batch_show_results_limit)]
+
+    return batch_results
+
+
+def img2img(id_task: str, request: gr.Request, mode: int, prompt: str, negative_prompt: str, prompt_styles, init_img, sketch, init_img_with_mask, inpaint_color_sketch, inpaint_color_sketch_orig, init_img_inpaint, init_mask_inpaint, mask_blur: int, mask_alpha: float, inpainting_fill: int, n_iter: int, batch_size: int, cfg_scale: float, image_cfg_scale: float, denoising_strength: float, selected_scale_tab: int, height: int, width: int, scale_by: float, resize_mode: int, inpaint_full_res: bool, inpaint_full_res_padding: int, inpainting_mask_invert: int, img2img_batch_input_dir: str, img2img_batch_output_dir: str, img2img_batch_inpaint_mask_dir: str, override_settings_texts, img2img_batch_use_png_info: bool, img2img_batch_png_info_props: list, img2img_batch_png_info_dir: str, img2img_batch_source_type: str, img2img_batch_upload: list, *args):
     override_settings = create_override_settings_dict(override_settings_texts)
 
     is_batch = mode == 5
@@ -145,9 +178,8 @@ def img2img(id_task: str, mode: int, prompt: str, negative_prompt: str, prompt_s
         image = None
         mask = None
 
-    # Use the EXIF orientation of photos taken by smartphones.
-    if image is not None:
-        image = ImageOps.exif_transpose(image)
+    image = images.fix_image(image)
+    mask = images.fix_image(mask)
 
     if selected_scale_tab == 1 and not is_batch:
         assert image, "Can't scale by because no image is selected"
@@ -164,10 +196,8 @@ def img2img(id_task: str, mode: int, prompt: str, negative_prompt: str, prompt_s
         prompt=prompt,
         negative_prompt=negative_prompt,
         styles=prompt_styles,
-        sampler_name=sampler_name,
         batch_size=batch_size,
         n_iter=n_iter,
-        steps=steps,
         cfg_scale=cfg_scale,
         width=width,
         height=height,
@@ -189,19 +219,23 @@ def img2img(id_task: str, mode: int, prompt: str, negative_prompt: str, prompt_s
 
     p.user = request.username
 
-    if shared.cmd_opts.enable_console_prompts:
+    if shared.opts.enable_console_prompts:
         print(f"\nimg2img: {prompt}", file=shared.progress_print_out)
 
-    if mask:
-        p.extra_generation_params["Mask blur"] = mask_blur
-
     with closing(p):
         if is_batch:
-            assert not shared.cmd_opts.hide_ui_dir_config, "Launched with --hide-ui-dir-config, batch img2img disabled"
-
-            process_batch(p, img2img_batch_input_dir, img2img_batch_output_dir, img2img_batch_inpaint_mask_dir, args, to_scale=selected_scale_tab == 1, scale_by=scale_by, use_png_info=img2img_batch_use_png_info, png_info_props=img2img_batch_png_info_props, png_info_dir=img2img_batch_png_info_dir)
+            if img2img_batch_source_type == "upload":
+                assert isinstance(img2img_batch_upload, list) and img2img_batch_upload
+                output_dir = ""
+                inpaint_mask_dir = ""
+                png_info_dir = img2img_batch_png_info_dir if not shared.cmd_opts.hide_ui_dir_config else ""
+                processed = process_batch(p, img2img_batch_upload, output_dir, inpaint_mask_dir, args, to_scale=selected_scale_tab == 1, scale_by=scale_by, use_png_info=img2img_batch_use_png_info, png_info_props=img2img_batch_png_info_props, png_info_dir=png_info_dir)
+            else: # "from dir"
+                assert not shared.cmd_opts.hide_ui_dir_config, "Launched with --hide-ui-dir-config, batch img2img disabled"
+                processed = process_batch(p, img2img_batch_input_dir, img2img_batch_output_dir, img2img_batch_inpaint_mask_dir, args, to_scale=selected_scale_tab == 1, scale_by=scale_by, use_png_info=img2img_batch_use_png_info, png_info_props=img2img_batch_png_info_props, png_info_dir=img2img_batch_png_info_dir)
 
-            processed = Processed(p, [], p.seed, "")
+            if processed is None:
+                processed = Processed(p, [], p.seed, "")
         else:
             processed = modules.scripts.scripts_img2img.run(p, *args)
             if processed is None:
diff --git a/modules/import_hook.py b/modules/import_hook.py
index 28c67dfa897..eba9a372929 100644
--- a/modules/import_hook.py
+++ b/modules/import_hook.py
@@ -3,3 +3,14 @@
 # this will break any attempt to import xformers which will prevent stability diffusion repo from trying to use it
 if "--xformers" not in "".join(sys.argv):
     sys.modules["xformers"] = None
+
+# Hack to fix a changed import in torchvision 0.17+, which otherwise breaks
+# basicsr; see https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/13985
+try:
+    import torchvision.transforms.functional_tensor  # noqa: F401
+except ImportError:
+    try:
+        import torchvision.transforms.functional as functional
+        sys.modules["torchvision.transforms.functional_tensor"] = functional
+    except ImportError:
+        pass  # shrug...
diff --git a/modules/generation_parameters_copypaste.py b/modules/infotext_utils.py
similarity index 67%
rename from modules/generation_parameters_copypaste.py
rename to modules/infotext_utils.py
index d39f2ebac36..32dbafa6518 100644
--- a/modules/generation_parameters_copypaste.py
+++ b/modules/infotext_utils.py
@@ -1,23 +1,24 @@
+from __future__ import annotations
 import base64
 import io
 import json
 import os
 import re
+import sys
 
 import gradio as gr
 from modules.paths import data_path
-from modules import shared, ui_tempdir, script_callbacks, processing
+from modules import shared, ui_tempdir, script_callbacks, processing, infotext_versions, images, prompt_parser, errors
 from PIL import Image
 
-re_param_code = r'\s*([\w ]+):\s*("(?:\\.|[^\\"])+"|[^,]*)(?:,|$)'
+sys.modules['modules.generation_parameters_copypaste'] = sys.modules[__name__]  # alias for old name
+
+re_param_code = r'\s*(\w[\w \-/]+):\s*("(?:\\.|[^\\"])+"|[^,]*)(?:,|$)'
 re_param = re.compile(re_param_code)
 re_imagesize = re.compile(r"^(\d+)x(\d+)$")
 re_hypernet_hash = re.compile("\(([0-9a-f]+)\)$")
 type_of_gr_update = type(gr.update())
 
-paste_fields = {}
-registered_param_bindings = []
-
 
 class ParamBinding:
     def __init__(self, paste_button, tabname, source_text_component=None, source_image_component=None, source_tabname=None, override_settings_component=None, paste_field_names=None):
@@ -30,6 +31,23 @@ def __init__(self, paste_button, tabname, source_text_component=None, source_ima
         self.paste_field_names = paste_field_names or []
 
 
+class PasteField(tuple):
+    def __new__(cls, component, target, *, api=None):
+        return super().__new__(cls, (component, target))
+
+    def __init__(self, component, target, *, api=None):
+        super().__init__()
+
+        self.api = api
+        self.component = component
+        self.label = target if isinstance(target, str) else None
+        self.function = target if callable(target) else None
+
+
+paste_fields: dict[str, dict] = {}
+registered_param_bindings: list[ParamBinding] = []
+
+
 def reset():
     paste_fields.clear()
     registered_param_bindings.clear()
@@ -65,7 +83,7 @@ def image_from_url_text(filedata):
         assert is_in_right_dir, 'trying to open image file outside of allowed directories'
 
         filename = filename.rsplit('?', 1)[0]
-        return Image.open(filename)
+        return images.read(filename)
 
     if type(filedata) == list:
         if len(filedata) == 0:
@@ -77,11 +95,17 @@ def image_from_url_text(filedata):
         filedata = filedata[len("data:image/png;base64,"):]
 
     filedata = base64.decodebytes(filedata.encode('utf-8'))
-    image = Image.open(io.BytesIO(filedata))
+    image = images.read(io.BytesIO(filedata))
     return image
 
 
 def add_paste_fields(tabname, init_img, fields, override_settings_component=None):
+
+    if fields:
+        for i in range(len(fields)):
+            if not isinstance(fields[i], PasteField):
+                fields[i] = PasteField(*fields[i])
+
     paste_fields[tabname] = {"init_img": init_img, "fields": fields, "override_settings_component": override_settings_component}
 
     # backwards compatibility for existing extensions
@@ -113,7 +137,6 @@ def register_paste_params_button(binding: ParamBinding):
 
 
 def connect_paste_params_buttons():
-    binding: ParamBinding
     for binding in registered_param_bindings:
         destination_image_component = paste_fields[binding.tabname]["init_img"]
         fields = paste_fields[binding.tabname]["fields"]
@@ -123,18 +146,19 @@ def connect_paste_params_buttons():
         destination_height_component = next(iter([field for field, name in fields if name == "Size-2"] if fields else []), None)
 
         if binding.source_image_component and destination_image_component:
+            need_send_dementions = destination_width_component and binding.tabname != 'inpaint'
             if isinstance(binding.source_image_component, gr.Gallery):
-                func = send_image_and_dimensions if destination_width_component else image_from_url_text
+                func = send_image_and_dimensions if need_send_dementions else image_from_url_text
                 jsfunc = "extract_image_from_gallery"
             else:
-                func = send_image_and_dimensions if destination_width_component else lambda x: x
+                func = send_image_and_dimensions if need_send_dementions else lambda x: x
                 jsfunc = None
 
             binding.paste_button.click(
                 fn=func,
                 _js=jsfunc,
                 inputs=[binding.source_image_component],
-                outputs=[destination_image_component, destination_width_component, destination_height_component] if destination_width_component else [destination_image_component],
+                outputs=[destination_image_component, destination_width_component, destination_height_component] if need_send_dementions else [destination_image_component],
                 show_progress=False,
             )
 
@@ -207,7 +231,7 @@ def restore_old_hires_fix_params(res):
     res['Hires resize-2'] = height
 
 
-def parse_generation_parameters(x: str):
+def parse_generation_parameters(x: str, skip_fields: list[str] | None = None):
     """parses generation parameters string, the one you see in text field under the picture in UI:
 ```
 girl with an artist's beret, determined, blue eyes, desert scene, computer monitors, heavy makeup, by Alphonse Mucha and Charlie Bowater, ((eyeshadow)), (coquettish), detailed, intricate
@@ -217,6 +241,8 @@ def parse_generation_parameters(x: str):
 
     returns a dict with field values
     """
+    if skip_fields is None:
+        skip_fields = shared.opts.infotext_skip_pasting
 
     res = {}
 
@@ -240,17 +266,6 @@ def parse_generation_parameters(x: str):
         else:
             prompt += ("" if prompt == "" else "\n") + line
 
-    if shared.opts.infotext_styles != "Ignore":
-        found_styles, prompt, negative_prompt = shared.prompt_styles.extract_styles_from_prompt(prompt, negative_prompt)
-
-        if shared.opts.infotext_styles == "Apply":
-            res["Styles array"] = found_styles
-        elif shared.opts.infotext_styles == "Apply if any" and found_styles:
-            res["Styles array"] = found_styles
-
-    res["Prompt"] = prompt
-    res["Negative prompt"] = negative_prompt
-
     for k, v in re_param.findall(lastline):
         try:
             if v[0] == '"' and v[-1] == '"':
@@ -265,6 +280,26 @@ def parse_generation_parameters(x: str):
         except Exception:
             print(f"Error parsing \"{k}: {v}\"")
 
+    # Extract styles from prompt
+    if shared.opts.infotext_styles != "Ignore":
+        found_styles, prompt_no_styles, negative_prompt_no_styles = shared.prompt_styles.extract_styles_from_prompt(prompt, negative_prompt)
+
+        same_hr_styles = True
+        if ("Hires prompt" in res or "Hires negative prompt" in res) and (infotext_ver > infotext_versions.v180_hr_styles if (infotext_ver := infotext_versions.parse_version(res.get("Version"))) else True):
+            hr_prompt, hr_negative_prompt = res.get("Hires prompt", prompt), res.get("Hires negative prompt", negative_prompt)
+            hr_found_styles, hr_prompt_no_styles, hr_negative_prompt_no_styles = shared.prompt_styles.extract_styles_from_prompt(hr_prompt, hr_negative_prompt)
+            if same_hr_styles := found_styles == hr_found_styles:
+                res["Hires prompt"] = '' if hr_prompt_no_styles == prompt_no_styles else hr_prompt_no_styles
+                res['Hires negative prompt'] = '' if hr_negative_prompt_no_styles == negative_prompt_no_styles else hr_negative_prompt_no_styles
+
+        if same_hr_styles:
+            prompt, negative_prompt = prompt_no_styles, negative_prompt_no_styles
+            if (shared.opts.infotext_styles == "Apply if any" and found_styles) or shared.opts.infotext_styles == "Apply":
+                res['Styles array'] = found_styles
+
+    res["Prompt"] = prompt
+    res["Negative prompt"] = negative_prompt
+
     # Missing CLIP skip means it was set to 1 (the default)
     if "Clip skip" not in res:
         res["Clip skip"] = "1"
@@ -280,6 +315,9 @@ def parse_generation_parameters(x: str):
     if "Hires sampler" not in res:
         res["Hires sampler"] = "Use same sampler"
 
+    if "Hires schedule type" not in res:
+        res["Hires schedule type"] = "Use same scheduler"
+
     if "Hires checkpoint" not in res:
         res["Hires checkpoint"] = "Use same checkpoint"
 
@@ -289,6 +327,18 @@ def parse_generation_parameters(x: str):
     if "Hires negative prompt" not in res:
         res["Hires negative prompt"] = ""
 
+    if "Mask mode" not in res:
+        res["Mask mode"] = "Inpaint masked"
+
+    if "Masked content" not in res:
+        res["Masked content"] = 'original'
+
+    if "Inpaint area" not in res:
+        res["Inpaint area"] = "Whole picture"
+
+    if "Masked area padding" not in res:
+        res["Masked area padding"] = 32
+
     restore_old_hires_fix_params(res)
 
     # Missing RNG means the default was set, which is GPU RNG
@@ -313,6 +363,26 @@ def parse_generation_parameters(x: str):
     if "VAE Decoder" not in res:
         res["VAE Decoder"] = "Full"
 
+    if "FP8 weight" not in res:
+        res["FP8 weight"] = "Disable"
+
+    if "Cache FP16 weight for LoRA" not in res and res["FP8 weight"] != "Disable":
+        res["Cache FP16 weight for LoRA"] = False
+
+    prompt_attention = prompt_parser.parse_prompt_attention(prompt)
+    prompt_attention += prompt_parser.parse_prompt_attention(negative_prompt)
+    prompt_uses_emphasis = len(prompt_attention) != len([p for p in prompt_attention if p[1] == 1.0 or p[0] == 'BREAK'])
+    if "Emphasis" not in res and prompt_uses_emphasis:
+        res["Emphasis"] = "Original"
+
+    if "Refiner switch by sampling steps" not in res:
+        res["Refiner switch by sampling steps"] = False
+
+    infotext_versions.backcompat(res)
+
+    for key in skip_fields:
+        res.pop(key, None)
+
     return res
 
 
@@ -361,13 +431,57 @@ def create_override_settings_dict(text_pairs):
     return res
 
 
+def get_override_settings(params, *, skip_fields=None):
+    """Returns a list of settings overrides from the infotext parameters dictionary.
+
+    This function checks the `params` dictionary for any keys that correspond to settings in `shared.opts` and returns
+    a list of tuples containing the parameter name, setting name, and new value cast to correct type.
+
+    It checks for conditions before adding an override:
+    - ignores settings that match the current value
+    - ignores parameter keys present in skip_fields argument.
+
+    Example input:
+        {"Clip skip": "2"}
+
+    Example output:
+        [("Clip skip", "CLIP_stop_at_last_layers", 2)]
+    """
+
+    res = []
+
+    mapping = [(info.infotext, k) for k, info in shared.opts.data_labels.items() if info.infotext]
+    for param_name, setting_name in mapping + infotext_to_setting_name_mapping:
+        if param_name in (skip_fields or {}):
+            continue
+
+        v = params.get(param_name, None)
+        if v is None:
+            continue
+
+        if setting_name == "sd_model_checkpoint" and shared.opts.disable_weights_auto_swap:
+            continue
+
+        v = shared.opts.cast_value(setting_name, v)
+        current_value = getattr(shared.opts, setting_name, None)
+
+        if v == current_value:
+            continue
+
+        res.append((param_name, setting_name, v))
+
+    return res
+
+
 def connect_paste(button, paste_fields, input_comp, override_settings_component, tabname):
     def paste_func(prompt):
-        if not prompt and not shared.cmd_opts.hide_ui_dir_config:
+        if not prompt and not shared.cmd_opts.hide_ui_dir_config and not shared.cmd_opts.no_prompt_history:
             filename = os.path.join(data_path, "params.txt")
-            if os.path.exists(filename):
+            try:
                 with open(filename, "r", encoding="utf8") as file:
                     prompt = file.read()
+            except OSError:
+                pass
 
         params = parse_generation_parameters(prompt)
         script_callbacks.infotext_pasted_callback(prompt, params)
@@ -375,7 +489,11 @@ def paste_func(prompt):
 
         for output, key in paste_fields:
             if callable(key):
-                v = key(params)
+                try:
+                    v = key(params)
+                except Exception:
+                    errors.report(f"Error executing {key}", exc_info=True)
+                    v = None
             else:
                 v = params.get(key, None)
 
@@ -389,6 +507,8 @@ def paste_func(prompt):
 
                     if valtype == bool and v == "False":
                         val = False
+                    elif valtype == int:
+                        val = float(v)
                     else:
                         val = valtype(v)
 
@@ -402,29 +522,9 @@ def paste_func(prompt):
         already_handled_fields = {key: 1 for _, key in paste_fields}
 
         def paste_settings(params):
-            vals = {}
-
-            mapping = [(info.infotext, k) for k, info in shared.opts.data_labels.items() if info.infotext]
-            for param_name, setting_name in mapping + infotext_to_setting_name_mapping:
-                if param_name in already_handled_fields:
-                    continue
-
-                v = params.get(param_name, None)
-                if v is None:
-                    continue
-
-                if setting_name == "sd_model_checkpoint" and shared.opts.disable_weights_auto_swap:
-                    continue
+            vals = get_override_settings(params, skip_fields=already_handled_fields)
 
-                v = shared.opts.cast_value(setting_name, v)
-                current_value = getattr(shared.opts, setting_name, None)
-
-                if v == current_value:
-                    continue
-
-                vals[param_name] = v
-
-            vals_pairs = [f"{k}: {v}" for k, v in vals.items()]
+            vals_pairs = [f"{infotext_text}: {value}" for infotext_text, setting_name, value in vals]
 
             return gr.Dropdown.update(value=vals_pairs, choices=vals_pairs, visible=bool(vals_pairs))
 
@@ -443,3 +543,4 @@ def paste_settings(params):
         outputs=[],
         show_progress=False,
     )
+
diff --git a/modules/infotext_versions.py b/modules/infotext_versions.py
new file mode 100644
index 00000000000..cea676cda61
--- /dev/null
+++ b/modules/infotext_versions.py
@@ -0,0 +1,46 @@
+from modules import shared
+from packaging import version
+import re
+
+
+v160 = version.parse("1.6.0")
+v170_tsnr = version.parse("v1.7.0-225")
+v180 = version.parse("1.8.0")
+v180_hr_styles = version.parse("1.8.0-139")
+
+
+def parse_version(text):
+    if text is None:
+        return None
+
+    m = re.match(r'([^-]+-[^-]+)-.*', text)
+    if m:
+        text = m.group(1)
+
+    try:
+        return version.parse(text)
+    except Exception:
+        return None
+
+
+def backcompat(d):
+    """Checks infotext Version field, and enables backwards compatibility options according to it."""
+
+    if not shared.opts.auto_backcompat:
+        return
+
+    ver = parse_version(d.get("Version"))
+    if ver is None:
+        return
+
+    if ver < v160 and '[' in d.get('Prompt', ''):
+        d["Old prompt editing timelines"] = True
+
+    if ver < v160 and d.get('Sampler', '') in ('DDIM', 'PLMS'):
+        d["Pad conds v0"] = True
+
+    if ver < v170_tsnr:
+        d["Downcast alphas_cumprod"] = True
+
+    if ver < v180 and d.get('Refiner'):
+        d["Refiner switch by sampling steps"] = True
diff --git a/modules/initialize.py b/modules/initialize.py
index f24f76375db..0365bbb3093 100644
--- a/modules/initialize.py
+++ b/modules/initialize.py
@@ -1,5 +1,6 @@
 import importlib
 import logging
+import os
 import sys
 import warnings
 from threading import Thread
@@ -18,6 +19,7 @@ def imports():
     warnings.filterwarnings(action="ignore", category=DeprecationWarning, module="pytorch_lightning")
     warnings.filterwarnings(action="ignore", category=UserWarning, module="torchvision")
 
+    os.environ.setdefault('GRADIO_ANALYTICS_ENABLED', 'False')
     import gradio  # noqa: F401
     startup_timer.record("import gradio")
 
@@ -49,14 +51,12 @@ def check_versions():
 def initialize():
     from modules import initialize_util
     initialize_util.fix_torch_version()
+    initialize_util.fix_pytorch_lightning()
     initialize_util.fix_asyncio_event_loop_policy()
     initialize_util.validate_tls_options()
     initialize_util.configure_sigint_handler()
     initialize_util.configure_opts_onchange()
 
-    from modules import modelloader
-    modelloader.cleanup_models()
-
     from modules import sd_models
     sd_models.setup_model()
     startup_timer.record("setup SD model")
@@ -110,7 +110,7 @@ def initialize_rest(*, reload_script_modules=False):
     with startup_timer.subcategory("load scripts"):
         scripts.load_scripts()
 
-    if reload_script_modules:
+    if reload_script_modules and shared.opts.enable_reloading_ui_scripts:
         for module in [module for name, module in sys.modules.items() if name.startswith("modules.ui")]:
             importlib.reload(module)
         startup_timer.record("reload script modules")
@@ -140,19 +140,20 @@ def load_model():
         """
         Accesses shared.sd_model property to load model.
         After it's available, if it has been loaded before this access by some extension,
-        its optimization may be None because the list of optimizaers has neet been filled
+        its optimization may be None because the list of optimizers has not been filled
         by that time, so we apply optimization again.
         """
+        from modules import devices
+        devices.torch_npu_set_device()
 
         shared.sd_model  # noqa: B018
 
         if sd_hijack.current_optimizer is None:
             sd_hijack.apply_optimizations()
 
-        from modules import devices
         devices.first_time_calculation()
-
-    Thread(target=load_model).start()
+    if not shared.cmd_opts.skip_load_model_at_start:
+        Thread(target=load_model).start()
 
     from modules import shared_items
     shared_items.reload_hypernetworks()
diff --git a/modules/initialize_util.py b/modules/initialize_util.py
index 2894eee4c1a..79a72cb3aba 100644
--- a/modules/initialize_util.py
+++ b/modules/initialize_util.py
@@ -24,6 +24,13 @@ def fix_torch_version():
         torch.__long_version__ = torch.__version__
         torch.__version__ = re.search(r'[\d.]+[\d]', torch.__version__).group(0)
 
+def fix_pytorch_lightning():
+    # Checks if pytorch_lightning.utilities.distributed already exists in the sys.modules cache
+    if 'pytorch_lightning.utilities.distributed' not in sys.modules:
+        import pytorch_lightning
+        # Lets the user know that the library was not found and then will set it to pytorch_lightning.utilities.rank_zero
+        print("Pytorch_lightning.distributed not found, attempting pytorch_lightning.rank_zero")
+        sys.modules["pytorch_lightning.utilities.distributed"] = pytorch_lightning.utilities.rank_zero
 
 def fix_asyncio_event_loop_policy():
     """
@@ -150,10 +157,14 @@ def dumpstacks():
 
 def configure_sigint_handler():
     # make the program just exit at ctrl+c without waiting for anything
+
+    from modules import shared
+
     def sigint_handler(sig, frame):
         print(f'Interrupted with signal {sig} in {frame}')
 
-        dumpstacks()
+        if shared.opts.dump_stacks_on_signal:
+            dumpstacks()
 
         os._exit(0)
 
@@ -173,6 +184,8 @@ def configure_opts_onchange():
     shared.opts.onchange("temp_dir", ui_tempdir.on_tmpdir_changed)
     shared.opts.onchange("gradio_theme", shared.reload_gradio_theme)
     shared.opts.onchange("cross_attention_optimization", wrap_queued_call(lambda: sd_hijack.model_hijack.redo_hijack(shared.sd_model)), call=False)
+    shared.opts.onchange("fp8_storage", wrap_queued_call(lambda: sd_models.reload_model_weights()), call=False)
+    shared.opts.onchange("cache_fp16_weight", wrap_queued_call(lambda: sd_models.reload_model_weights(forced_reload=True)), call=False)
     startup_timer.record("opts onchange")
 
 
diff --git a/modules/interrogate.py b/modules/interrogate.py
index 3045560d0ae..c93e7aa861d 100644
--- a/modules/interrogate.py
+++ b/modules/interrogate.py
@@ -10,14 +10,14 @@
 from torchvision import transforms
 from torchvision.transforms.functional import InterpolationMode
 
-from modules import devices, paths, shared, lowvram, modelloader, errors
+from modules import devices, paths, shared, lowvram, modelloader, errors, torch_utils
 
 blip_image_eval_size = 384
 clip_model_name = 'ViT-L/14'
 
 Category = namedtuple("Category", ["name", "topn", "items"])
 
-re_topn = re.compile(r"\.top(\d+)\.")
+re_topn = re.compile(r"\.top(\d+)$")
 
 def category_types():
     return [f.stem for f in Path(shared.interrogator.content_dir).glob('*.txt')]
@@ -131,7 +131,7 @@ def load(self):
 
         self.clip_model = self.clip_model.to(devices.device_interrogate)
 
-        self.dtype = next(self.clip_model.parameters()).dtype
+        self.dtype = torch_utils.get_param(self.clip_model).dtype
 
     def send_clip_to_ram(self):
         if not shared.opts.interrogate_keep_models_in_memory:
diff --git a/modules/launch_utils.py b/modules/launch_utils.py
index 6e54d06367c..20c7dc127a7 100644
--- a/modules/launch_utils.py
+++ b/modules/launch_utils.py
@@ -6,8 +6,10 @@
 import shutil
 import sys
 import importlib.util
+import importlib.metadata
 import platform
 import json
+import shlex
 from functools import lru_cache
 
 from modules import cmd_args, errors
@@ -26,8 +28,7 @@
 # Whether to default to printing command output
 default_command_live = (os.environ.get('WEBUI_LAUNCH_LIVE_OUTPUT') == "1")
 
-if 'GRADIO_ANALYTICS_ENABLED' not in os.environ:
-    os.environ['GRADIO_ANALYTICS_ENABLED'] = 'False'
+os.environ.setdefault('GRADIO_ANALYTICS_ENABLED', 'False')
 
 
 def check_python_version():
@@ -55,7 +56,7 @@ def check_python_version():
 
 You can download 3.10 Python from here: https://www.python.org/downloads/release/python-3106/
 
-{"Alternatively, use a binary release of WebUI: https://github.com/AUTOMATIC1111/stable-diffusion-webui/releases" if is_windows else ""}
+{"Alternatively, use a binary release of WebUI: https://github.com/AUTOMATIC1111/stable-diffusion-webui/releases/tag/v1.0.0-pre" if is_windows else ""}
 
 Use --skip-python-version-check to suppress this warning.
 """)
@@ -64,7 +65,7 @@ def check_python_version():
 @lru_cache()
 def commit_hash():
     try:
-        return subprocess.check_output([git, "rev-parse", "HEAD"], shell=False, encoding='utf8').strip()
+        return subprocess.check_output([git, "-C", script_path, "rev-parse", "HEAD"], shell=False, encoding='utf8').strip()
     except Exception:
         return "<none>"
 
@@ -72,11 +73,11 @@ def commit_hash():
 @lru_cache()
 def git_tag():
     try:
-        return subprocess.check_output([git, "describe", "--tags"], shell=False, encoding='utf8').strip()
+        return subprocess.check_output([git, "-C", script_path, "describe", "--tags"], shell=False, encoding='utf8').strip()
     except Exception:
         try:
 
-            changelog_md = os.path.join(os.path.dirname(os.path.dirname(__file__)), "CHANGELOG.md")
+            changelog_md = os.path.join(script_path, "CHANGELOG.md")
             with open(changelog_md, "r", encoding="utf-8") as file:
                 line = next((line.strip() for line in file if line.strip()), "<none>")
                 line = line.replace("## ", "")
@@ -119,11 +120,16 @@ def run(command, desc=None, errdesc=None, custom_env=None, live: bool = default_
 
 def is_installed(package):
     try:
-        spec = importlib.util.find_spec(package)
-    except ModuleNotFoundError:
-        return False
+        dist = importlib.metadata.distribution(package)
+    except importlib.metadata.PackageNotFoundError:
+        try:
+            spec = importlib.util.find_spec(package)
+        except ModuleNotFoundError:
+            return False
+
+        return spec is not None
 
-    return spec is not None
+    return dist is not None
 
 
 def repo_dir(name):
@@ -183,7 +189,7 @@ def git_clone(url, dir, name, commithash=None):
         return
 
     try:
-        run(f'"{git}" clone "{url}" "{dir}"', f"Cloning {name} into {dir}...", f"Couldn't clone {name}", live=True)
+        run(f'"{git}" clone --config core.filemode=false "{url}" "{dir}"', f"Cloning {name} into {dir}...", f"Couldn't clone {name}", live=True)
     except RuntimeError:
         shutil.rmtree(dir, ignore_errors=True)
         raise
@@ -226,7 +232,7 @@ def run_extension_installer(extension_dir):
 
     try:
         env = os.environ.copy()
-        env['PYTHONPATH'] = f"{os.path.abspath('.')}{os.pathsep}{env.get('PYTHONPATH', '')}"
+        env['PYTHONPATH'] = f"{script_path}{os.pathsep}{env.get('PYTHONPATH', '')}"
 
         stdout = run(f'"{python}" "{path_installer}"', errdesc=f"Error running install.py for extension {extension_dir}", custom_env=env).strip()
         if stdout:
@@ -239,11 +245,13 @@ def list_extensions(settings_file):
     settings = {}
 
     try:
-        if os.path.isfile(settings_file):
-            with open(settings_file, "r", encoding="utf8") as file:
-                settings = json.load(file)
+        with open(settings_file, "r", encoding="utf8") as file:
+            settings = json.load(file)
+    except FileNotFoundError:
+        pass
     except Exception:
-        errors.report("Could not load settings", exc_info=True)
+        errors.report(f'\nCould not load settings\nThe config file "{settings_file}" is likely corrupted\nIt has been moved to the "tmp/config.json"\nReverting config to default\n\n''', exc_info=True)
+        os.replace(settings_file, os.path.join(script_path, "tmp", "config.json"))
 
     disabled_extensions = set(settings.get('disabled_extensions', []))
     disable_all_extensions = settings.get('disable_all_extensions', 'none')
@@ -308,24 +316,45 @@ def requirements_met(requirements_file):
 
 
 def prepare_environment():
-    torch_index_url = os.environ.get('TORCH_INDEX_URL', "https://download.pytorch.org/whl/cu118")
-    torch_command = os.environ.get('TORCH_COMMAND', f"pip install torch==2.0.1 torchvision==0.15.2 --extra-index-url {torch_index_url}")
+    torch_index_url = os.environ.get('TORCH_INDEX_URL', "https://download.pytorch.org/whl/cu121")
+    torch_command = os.environ.get('TORCH_COMMAND', f"pip install torch==2.1.2 torchvision==0.16.2 --extra-index-url {torch_index_url}")
+    if args.use_ipex:
+        if platform.system() == "Windows":
+            # The "Nuullll/intel-extension-for-pytorch" wheels were built from IPEX source for Intel Arc GPU: https://github.com/intel/intel-extension-for-pytorch/tree/xpu-main
+            # This is NOT an Intel official release so please use it at your own risk!!
+            # See https://github.com/Nuullll/intel-extension-for-pytorch/releases/tag/v2.0.110%2Bxpu-master%2Bdll-bundle for details.
+            #
+            # Strengths (over official IPEX 2.0.110 windows release):
+            #   - AOT build (for Arc GPU only) to eliminate JIT compilation overhead: https://github.com/intel/intel-extension-for-pytorch/issues/399
+            #   - Bundles minimal oneAPI 2023.2 dependencies into the python wheels, so users don't need to install oneAPI for the whole system.
+            #   - Provides a compatible torchvision wheel: https://github.com/intel/intel-extension-for-pytorch/issues/465
+            # Limitation:
+            #   - Only works for python 3.10
+            url_prefix = "https://github.com/Nuullll/intel-extension-for-pytorch/releases/download/v2.0.110%2Bxpu-master%2Bdll-bundle"
+            torch_command = os.environ.get('TORCH_COMMAND', f"pip install {url_prefix}/torch-2.0.0a0+gite9ebda2-cp310-cp310-win_amd64.whl {url_prefix}/torchvision-0.15.2a0+fa99a53-cp310-cp310-win_amd64.whl {url_prefix}/intel_extension_for_pytorch-2.0.110+gitc6ea20b-cp310-cp310-win_amd64.whl")
+        else:
+            # Using official IPEX release for linux since it's already an AOT build.
+            # However, users still have to install oneAPI toolkit and activate oneAPI environment manually.
+            # See https://intel.github.io/intel-extension-for-pytorch/index.html#installation for details.
+            torch_index_url = os.environ.get('TORCH_INDEX_URL', "https://pytorch-extension.intel.com/release-whl/stable/xpu/us/")
+            torch_command = os.environ.get('TORCH_COMMAND', f"pip install torch==2.0.0a0 intel-extension-for-pytorch==2.0.110+gitba7f6c1 --extra-index-url {torch_index_url}")
     requirements_file = os.environ.get('REQS_FILE', "requirements_versions.txt")
+    requirements_file_for_npu = os.environ.get('REQS_FILE_FOR_NPU', "requirements_npu.txt")
 
-    xformers_package = os.environ.get('XFORMERS_PACKAGE', 'xformers==0.0.20')
+    xformers_package = os.environ.get('XFORMERS_PACKAGE', 'xformers==0.0.23.post1')
     clip_package = os.environ.get('CLIP_PACKAGE', "https://github.com/openai/CLIP/archive/d50d76daa670286dd6cacf3bcd80b5e4823fc8e1.zip")
     openclip_package = os.environ.get('OPENCLIP_PACKAGE', "https://github.com/mlfoundations/open_clip/archive/bb6e834e9c70d9c27d0dc3ecedeebeaeb1ffad6b.zip")
 
+    assets_repo = os.environ.get('ASSETS_REPO', "https://github.com/AUTOMATIC1111/stable-diffusion-webui-assets.git")
     stable_diffusion_repo = os.environ.get('STABLE_DIFFUSION_REPO', "https://github.com/Stability-AI/stablediffusion.git")
     stable_diffusion_xl_repo = os.environ.get('STABLE_DIFFUSION_XL_REPO', "https://github.com/Stability-AI/generative-models.git")
     k_diffusion_repo = os.environ.get('K_DIFFUSION_REPO', 'https://github.com/crowsonkb/k-diffusion.git')
-    codeformer_repo = os.environ.get('CODEFORMER_REPO', 'https://github.com/sczhou/CodeFormer.git')
     blip_repo = os.environ.get('BLIP_REPO', 'https://github.com/salesforce/BLIP.git')
 
+    assets_commit_hash = os.environ.get('ASSETS_COMMIT_HASH', "6f7db241d2f8ba7457bac5ca9753331f0c266917")
     stable_diffusion_commit_hash = os.environ.get('STABLE_DIFFUSION_COMMIT_HASH', "cf1d67a6fd5ea1aa600c4df58e5b47da45f6bdbf")
     stable_diffusion_xl_commit_hash = os.environ.get('STABLE_DIFFUSION_XL_COMMIT_HASH', "45c443b316737a4ab6e40413d7794a7f5657c19f")
     k_diffusion_commit_hash = os.environ.get('K_DIFFUSION_COMMIT_HASH', "ab527a9a6d347f364e3d185ba6d714e22d80cb3c")
-    codeformer_commit_hash = os.environ.get('CODEFORMER_COMMIT_HASH', "c5b4593074ba6214284d6acd5f1719b6c5d739af")
     blip_commit_hash = os.environ.get('BLIP_COMMIT_HASH', "48211a1594f1321b00f14c9f7a5b4813144b2fb9")
 
     try:
@@ -352,6 +381,8 @@ def prepare_environment():
         run(f'"{python}" -m {torch_command}', "Installing torch and torchvision", "Couldn't install torch", live=True)
         startup_timer.record("install torch")
 
+    if args.use_ipex:
+        args.skip_torch_cuda_test = True
     if not args.skip_torch_cuda_test and not check_run_python("import torch; assert torch.cuda.is_available()"):
         raise RuntimeError(
             'Torch is not able to use GPU; '
@@ -377,18 +408,14 @@ def prepare_environment():
 
     os.makedirs(os.path.join(script_path, dir_repos), exist_ok=True)
 
+    git_clone(assets_repo, repo_dir('stable-diffusion-webui-assets'), "assets", assets_commit_hash)
     git_clone(stable_diffusion_repo, repo_dir('stable-diffusion-stability-ai'), "Stable Diffusion", stable_diffusion_commit_hash)
     git_clone(stable_diffusion_xl_repo, repo_dir('generative-models'), "Stable Diffusion XL", stable_diffusion_xl_commit_hash)
     git_clone(k_diffusion_repo, repo_dir('k-diffusion'), "K-diffusion", k_diffusion_commit_hash)
-    git_clone(codeformer_repo, repo_dir('CodeFormer'), "CodeFormer", codeformer_commit_hash)
     git_clone(blip_repo, repo_dir('BLIP'), "BLIP", blip_commit_hash)
 
     startup_timer.record("clone repositores")
 
-    if not is_installed("lpips"):
-        run_pip(f"install -r \"{os.path.join(repo_dir('CodeFormer'), 'requirements.txt')}\"", "requirements for CodeFormer")
-        startup_timer.record("install CodeFormer requirements")
-
     if not os.path.isfile(requirements_file):
         requirements_file = os.path.join(script_path, requirements_file)
 
@@ -396,6 +423,13 @@ def prepare_environment():
         run_pip(f"install -r \"{requirements_file}\"", "requirements")
         startup_timer.record("install requirements")
 
+    if not os.path.isfile(requirements_file_for_npu):
+        requirements_file_for_npu = os.path.join(script_path, requirements_file_for_npu)
+
+    if "torch_npu" in torch_command and not requirements_met(requirements_file_for_npu):
+        run_pip(f"install -r \"{requirements_file_for_npu}\"", "requirements_for_npu")
+        startup_timer.record("install requirements_for_npu")
+
     if not args.skip_install:
         run_extensions_installers(settings_file=args.ui_settings_file)
 
@@ -412,7 +446,6 @@ def prepare_environment():
         exit(0)
 
 
-
 def configure_for_tests():
     if "--api" not in sys.argv:
         sys.argv.append("--api")
@@ -428,7 +461,7 @@ def configure_for_tests():
 
 
 def start():
-    print(f"Launching {'API server' if '--nowebui' in sys.argv else 'Web UI'} with arguments: {' '.join(sys.argv[1:])}")
+    print(f"Launching {'API server' if '--nowebui' in sys.argv else 'Web UI'} with arguments: {shlex.join(sys.argv[1:])}")
     import webui
     if '--nowebui' in sys.argv:
         webui.api_only()
@@ -441,7 +474,7 @@ def dump_sysinfo():
     import datetime
 
     text = sysinfo.get()
-    filename = f"sysinfo-{datetime.datetime.utcnow().strftime('%Y-%m-%d-%H-%M')}.txt"
+    filename = f"sysinfo-{datetime.datetime.utcnow().strftime('%Y-%m-%d-%H-%M')}.json"
 
     with open(filename, "w", encoding="utf8") as file:
         file.write(text)
diff --git a/modules/localization.py b/modules/localization.py
index c132028856f..108f792e96f 100644
--- a/modules/localization.py
+++ b/modules/localization.py
@@ -14,21 +14,24 @@ def list_localizations(dirname):
         if ext.lower() != ".json":
             continue
 
-        localizations[fn] = os.path.join(dirname, file)
+        localizations[fn] = [os.path.join(dirname, file)]
 
     for file in scripts.list_scripts("localizations", ".json"):
         fn, ext = os.path.splitext(file.filename)
-        localizations[fn] = file.path
+        if fn not in localizations:
+            localizations[fn] = []
+        localizations[fn].append(file.path)
 
 
 def localization_js(current_localization_name: str) -> str:
-    fn = localizations.get(current_localization_name, None)
+    fns = localizations.get(current_localization_name, None)
     data = {}
-    if fn is not None:
-        try:
-            with open(fn, "r", encoding="utf8") as file:
-                data = json.load(file)
-        except Exception:
-            errors.report(f"Error loading localization from {fn}", exc_info=True)
+    if fns is not None:
+        for fn in fns:
+            try:
+                with open(fn, "r", encoding="utf8") as file:
+                    data.update(json.load(file))
+            except Exception:
+                errors.report(f"Error loading localization from {fn}", exc_info=True)
 
     return f"window.localization = {json.dumps(data)}"
diff --git a/modules/logging_config.py b/modules/logging_config.py
index 7db23d4b6e5..8e31d8c9fd1 100644
--- a/modules/logging_config.py
+++ b/modules/logging_config.py
@@ -1,16 +1,58 @@
-import os
 import logging
+import os
+
+try:
+    from tqdm import tqdm
+
+
+    class TqdmLoggingHandler(logging.Handler):
+        def __init__(self, fallback_handler: logging.Handler):
+            super().__init__()
+            self.fallback_handler = fallback_handler
+
+        def emit(self, record):
+            try:
+                # If there are active tqdm progress bars,
+                # attempt to not interfere with them.
+                if tqdm._instances:
+                    tqdm.write(self.format(record))
+                else:
+                    self.fallback_handler.emit(record)
+            except Exception:
+                self.fallback_handler.emit(record)
+
+except ImportError:
+    TqdmLoggingHandler = None
 
 
 def setup_logging(loglevel):
     if loglevel is None:
         loglevel = os.environ.get("SD_WEBUI_LOG_LEVEL")
 
-    if loglevel:
-        log_level = getattr(logging, loglevel.upper(), None) or logging.INFO
-        logging.basicConfig(
-            level=log_level,
-            format='%(asctime)s %(levelname)s [%(name)s] %(message)s',
-            datefmt='%Y-%m-%d %H:%M:%S',
-        )
+    if not loglevel:
+        return
+
+    if logging.root.handlers:
+        # Already configured, do not interfere
+        return
+
+    formatter = logging.Formatter(
+        '%(asctime)s %(levelname)s [%(name)s] %(message)s',
+        '%Y-%m-%d %H:%M:%S',
+    )
+
+    if os.environ.get("SD_WEBUI_RICH_LOG"):
+        from rich.logging import RichHandler
+        handler = RichHandler()
+    else:
+        handler = logging.StreamHandler()
+        handler.setFormatter(formatter)
+
+    if TqdmLoggingHandler:
+        handler = TqdmLoggingHandler(handler)
+
+    handler.setFormatter(formatter)
 
+    log_level = getattr(logging, loglevel.upper(), None) or logging.INFO
+    logging.root.setLevel(log_level)
+    logging.root.addHandler(handler)
diff --git a/modules/lowvram.py b/modules/lowvram.py
index 45701046b54..6728c337b64 100644
--- a/modules/lowvram.py
+++ b/modules/lowvram.py
@@ -1,9 +1,12 @@
+from collections import namedtuple
+
 import torch
 from modules import devices, shared
 
 module_in_gpu = None
 cpu = torch.device("cpu")
 
+ModuleWithParent = namedtuple('ModuleWithParent', ['module', 'parent'], defaults=['None'])
 
 def send_everything_to_cpu():
     global module_in_gpu
@@ -75,13 +78,14 @@ def first_stage_model_decode_wrap(z):
         (sd_model, 'depth_model'),
         (sd_model, 'embedder'),
         (sd_model, 'model'),
-        (sd_model, 'embedder'),
     ]
 
     is_sdxl = hasattr(sd_model, 'conditioner')
     is_sd2 = not is_sdxl and hasattr(sd_model.cond_stage_model, 'model')
 
-    if is_sdxl:
+    if hasattr(sd_model, 'medvram_fields'):
+        to_remain_in_cpu = sd_model.medvram_fields()
+    elif is_sdxl:
         to_remain_in_cpu.append((sd_model, 'conditioner'))
     elif is_sd2:
         to_remain_in_cpu.append((sd_model.cond_stage_model, 'model'))
@@ -103,7 +107,21 @@ def first_stage_model_decode_wrap(z):
         setattr(obj, field, module)
 
     # register hooks for those the first three models
-    if is_sdxl:
+    if hasattr(sd_model, "cond_stage_model") and hasattr(sd_model.cond_stage_model, "medvram_modules"):
+        for module in sd_model.cond_stage_model.medvram_modules():
+            if isinstance(module, ModuleWithParent):
+                parent = module.parent
+                module = module.module
+            else:
+                parent = None
+
+            if module:
+                module.register_forward_pre_hook(send_me_to_gpu)
+
+                if parent:
+                    parents[module] = parent
+
+    elif is_sdxl:
         sd_model.conditioner.register_forward_pre_hook(send_me_to_gpu)
     elif is_sd2:
         sd_model.cond_stage_model.model.register_forward_pre_hook(send_me_to_gpu)
@@ -117,9 +135,9 @@ def first_stage_model_decode_wrap(z):
     sd_model.first_stage_model.register_forward_pre_hook(send_me_to_gpu)
     sd_model.first_stage_model.encode = first_stage_model_encode_wrap
     sd_model.first_stage_model.decode = first_stage_model_decode_wrap
-    if sd_model.depth_model:
+    if getattr(sd_model, 'depth_model', None) is not None:
         sd_model.depth_model.register_forward_pre_hook(send_me_to_gpu)
-    if sd_model.embedder:
+    if getattr(sd_model, 'embedder', None) is not None:
         sd_model.embedder.register_forward_pre_hook(send_me_to_gpu)
 
     if use_medvram:
diff --git a/modules/mac_specific.py b/modules/mac_specific.py
index 89256c5b060..039689f32e1 100644
--- a/modules/mac_specific.py
+++ b/modules/mac_specific.py
@@ -1,6 +1,7 @@
 import logging
 
 import torch
+from torch import Tensor
 import platform
 from modules.sd_hijack_utils import CondFunc
 from packaging import version
@@ -11,7 +12,7 @@
 
 # before torch version 1.13, has_mps is only available in nightly pytorch and macOS 12.3+,
 # use check `getattr` and try it for compatibility.
-# in torch version 1.13, backends.mps.is_available() and backends.mps.is_built() are introduced in to check mps availabilty,
+# in torch version 1.13, backends.mps.is_available() and backends.mps.is_built() are introduced in to check mps availability,
 # since torch 2.0.1+ nightly build, getattr(torch, 'has_mps', False) was deprecated, see https://github.com/pytorch/pytorch/pull/103279
 def check_for_mps() -> bool:
     if version.parse(torch.__version__) <= version.parse("2.0.1"):
@@ -51,6 +52,17 @@ def cumsum_fix(input, cumsum_func, *args, **kwargs):
     return cumsum_func(input, *args, **kwargs)
 
 
+# MPS workaround for https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14046
+def interpolate_with_fp32_fallback(orig_func, *args, **kwargs) -> Tensor:
+    try:
+        return orig_func(*args, **kwargs)
+    except RuntimeError as e:
+        if "not implemented for" in str(e) and "Half" in str(e):
+            input_tensor = args[0]
+            return orig_func(input_tensor.to(torch.float32), *args[1:], **kwargs).to(input_tensor.dtype)
+        else:
+            print(f"An unexpected RuntimeError occurred: {str(e)}")
+
 if has_mps:
     if platform.mac_ver()[0].startswith("13.2."):
         # MPS workaround for https://github.com/pytorch/pytorch/issues/95188, thanks to danieldk (https://github.com/explosion/curated-transformers/pull/124)
@@ -77,6 +89,9 @@ def cumsum_fix(input, cumsum_func, *args, **kwargs):
         # MPS workaround for https://github.com/pytorch/pytorch/issues/96113
         CondFunc('torch.nn.functional.layer_norm', lambda orig_func, x, normalized_shape, weight, bias, eps, **kwargs: orig_func(x.float(), normalized_shape, weight.float() if weight is not None else None, bias.float() if bias is not None else bias, eps).to(x.dtype), lambda _, input, *args, **kwargs: len(args) == 4 and input.device.type == 'mps')
 
+        # MPS workaround for https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14046
+        CondFunc('torch.nn.functional.interpolate', interpolate_with_fp32_fallback, None)
+
         # MPS workaround for https://github.com/pytorch/pytorch/issues/92311
         if platform.processor() == 'i386':
             for funcName in ['torch.argmax', 'torch.Tensor.argmax']:
diff --git a/modules/masking.py b/modules/masking.py
index be9f84c76f6..2fc83031953 100644
--- a/modules/masking.py
+++ b/modules/masking.py
@@ -1,42 +1,39 @@
 from PIL import Image, ImageFilter, ImageOps
 
 
+def get_crop_region_v2(mask, pad=0):
+    """
+    Finds a rectangular region that contains all masked ares in a mask.
+    Returns None if mask is completely black mask (all 0)
+
+    Parameters:
+    mask: PIL.Image.Image L mode or numpy 1d array
+    pad: int number of pixels that the region will be extended on all sides
+    Returns: (x1, y1, x2, y2) | None
+
+    Introduced post 1.9.0
+    """
+    mask = mask if isinstance(mask, Image.Image) else Image.fromarray(mask)
+    if box := mask.getbbox():
+        x1, y1, x2, y2 = box
+        return (max(x1 - pad, 0), max(y1 - pad, 0), min(x2 + pad, mask.size[0]), min(y2 + pad, mask.size[1])) if pad else box
+
+
 def get_crop_region(mask, pad=0):
-    """finds a rectangular region that contains all masked ares in an image. Returns (x1, y1, x2, y2) coordinates of the rectangle.
-    For example, if a user has painted the top-right part of a 512x512 image", the result may be (256, 0, 512, 256)"""
-
-    h, w = mask.shape
-
-    crop_left = 0
-    for i in range(w):
-        if not (mask[:, i] == 0).all():
-            break
-        crop_left += 1
-
-    crop_right = 0
-    for i in reversed(range(w)):
-        if not (mask[:, i] == 0).all():
-            break
-        crop_right += 1
-
-    crop_top = 0
-    for i in range(h):
-        if not (mask[i] == 0).all():
-            break
-        crop_top += 1
-
-    crop_bottom = 0
-    for i in reversed(range(h)):
-        if not (mask[i] == 0).all():
-            break
-        crop_bottom += 1
-
-    return (
-        int(max(crop_left-pad, 0)),
-        int(max(crop_top-pad, 0)),
-        int(min(w - crop_right + pad, w)),
-        int(min(h - crop_bottom + pad, h))
-    )
+    """
+    Same function as get_crop_region_v2 but handles completely black mask (all 0) differently
+    when mask all black still return coordinates but the coordinates may be invalid ie x2>x1 or y2>y1
+    Notes: it is possible for the coordinates to be "valid" again if pad size is sufficiently large
+    (mask_size.x-pad, mask_size.y-pad, pad, pad)
+
+    Extension developer should use get_crop_region_v2 instead unless for compatibility considerations.
+    """
+    mask = mask if isinstance(mask, Image.Image) else Image.fromarray(mask)
+    if box := get_crop_region_v2(mask, pad):
+        return box
+    x1, y1 = mask.size
+    x2 = y2 = 0
+    return max(x1 - pad, 0), max(y1 - pad, 0), min(x2 + pad, mask.size[0]), min(y2 + pad, mask.size[1])
 
 
 def expand_crop_region(crop_region, processing_width, processing_height, image_width, image_height):
diff --git a/modules/modelloader.py b/modules/modelloader.py
index 098bcb79336..36e7415af43 100644
--- a/modules/modelloader.py
+++ b/modules/modelloader.py
@@ -1,13 +1,20 @@
 from __future__ import annotations
 
-import os
-import shutil
 import importlib
+import logging
+import os
+from typing import TYPE_CHECKING
 from urllib.parse import urlparse
 
+import torch
+
 from modules import shared
 from modules.upscaler import Upscaler, UpscalerLanczos, UpscalerNearest, UpscalerNone
-from modules.paths import script_path, models_path
+
+if TYPE_CHECKING:
+    import spandrel
+
+logger = logging.getLogger(__name__)
 
 
 def load_file_from_url(
@@ -16,6 +23,7 @@ def load_file_from_url(
     model_dir: str,
     progress: bool = True,
     file_name: str | None = None,
+    hash_prefix: str | None = None,
 ) -> str:
     """Download a file from `url` into `model_dir`, using the file present if possible.
 
@@ -29,11 +37,11 @@ def load_file_from_url(
     if not os.path.exists(cached_file):
         print(f'Downloading: "{url}" to {cached_file}\n')
         from torch.hub import download_url_to_file
-        download_url_to_file(url, cached_file, progress=progress)
+        download_url_to_file(url, cached_file, progress=progress, hash_prefix=hash_prefix)
     return cached_file
 
 
-def load_models(model_path: str, model_url: str = None, command_path: str = None, ext_filter=None, download_name=None, ext_blacklist=None) -> list:
+def load_models(model_path: str, model_url: str = None, command_path: str = None, ext_filter=None, download_name=None, ext_blacklist=None, hash_prefix=None) -> list:
     """
     A one-and done loader to try finding the desired models in specified directories.
 
@@ -42,6 +50,7 @@ def load_models(model_path: str, model_url: str = None, command_path: str = None
     @param model_path: The location to store/find models in.
     @param command_path: A command-line argument to search for models in first.
     @param ext_filter: An optional list of filename extensions to filter by
+    @param hash_prefix: the expected sha256 of the model_url
     @return: A list of paths containing the desired model(s)
     """
     output = []
@@ -71,7 +80,7 @@ def load_models(model_path: str, model_url: str = None, command_path: str = None
 
         if model_url is not None and len(output) == 0:
             if download_name is not None:
-                output.append(load_file_from_url(model_url, model_dir=places[0], file_name=download_name))
+                output.append(load_file_from_url(model_url, model_dir=places[0], file_name=download_name, hash_prefix=hash_prefix))
             else:
                 output.append(model_url)
 
@@ -90,54 +99,6 @@ def friendly_name(file: str):
     return model_name
 
 
-def cleanup_models():
-    # This code could probably be more efficient if we used a tuple list or something to store the src/destinations
-    # and then enumerate that, but this works for now. In the future, it'd be nice to just have every "model" scaler
-    # somehow auto-register and just do these things...
-    root_path = script_path
-    src_path = models_path
-    dest_path = os.path.join(models_path, "Stable-diffusion")
-    move_files(src_path, dest_path, ".ckpt")
-    move_files(src_path, dest_path, ".safetensors")
-    src_path = os.path.join(root_path, "ESRGAN")
-    dest_path = os.path.join(models_path, "ESRGAN")
-    move_files(src_path, dest_path)
-    src_path = os.path.join(models_path, "BSRGAN")
-    dest_path = os.path.join(models_path, "ESRGAN")
-    move_files(src_path, dest_path, ".pth")
-    src_path = os.path.join(root_path, "gfpgan")
-    dest_path = os.path.join(models_path, "GFPGAN")
-    move_files(src_path, dest_path)
-    src_path = os.path.join(root_path, "SwinIR")
-    dest_path = os.path.join(models_path, "SwinIR")
-    move_files(src_path, dest_path)
-    src_path = os.path.join(root_path, "repositories/latent-diffusion/experiments/pretrained_models/")
-    dest_path = os.path.join(models_path, "LDSR")
-    move_files(src_path, dest_path)
-
-
-def move_files(src_path: str, dest_path: str, ext_filter: str = None):
-    try:
-        os.makedirs(dest_path, exist_ok=True)
-        if os.path.exists(src_path):
-            for file in os.listdir(src_path):
-                fullpath = os.path.join(src_path, file)
-                if os.path.isfile(fullpath):
-                    if ext_filter is not None:
-                        if ext_filter not in file:
-                            continue
-                    print(f"Moving {file} from {src_path} to {dest_path}.")
-                    try:
-                        shutil.move(fullpath, dest_path)
-                    except Exception:
-                        pass
-            if len(os.listdir(src_path)) == 0:
-                print(f"Removing empty folder: {src_path}")
-                shutil.rmtree(src_path, True)
-    except Exception:
-        pass
-
-
 def load_upscalers():
     # We can only do this 'magic' method to dynamically load upscalers if they are referenced,
     # so we'll try to import any _model.py files before looking in __subclasses__
@@ -151,7 +112,7 @@ def load_upscalers():
             except Exception:
                 pass
 
-    datas = []
+    data = []
     commandline_options = vars(shared.cmd_opts)
 
     # some of upscaler classes will not go away after reloading their modules, and we'll end
@@ -170,10 +131,67 @@ def load_upscalers():
         scaler = cls(commandline_model_path)
         scaler.user_path = commandline_model_path
         scaler.model_download_path = commandline_model_path or scaler.model_path
-        datas += scaler.scalers
+        data += scaler.scalers
 
     shared.sd_upscalers = sorted(
-        datas,
+        data,
         # Special case for UpscalerNone keeps it at the beginning of the list.
         key=lambda x: x.name.lower() if not isinstance(x.scaler, (UpscalerNone, UpscalerLanczos, UpscalerNearest)) else ""
     )
+
+# None: not loaded, False: failed to load, True: loaded
+_spandrel_extra_init_state = None
+
+
+def _init_spandrel_extra_archs() -> None:
+    """
+    Try to initialize `spandrel_extra_archs` (exactly once).
+    """
+    global _spandrel_extra_init_state
+    if _spandrel_extra_init_state is not None:
+        return
+
+    try:
+        import spandrel
+        import spandrel_extra_arches
+        spandrel.MAIN_REGISTRY.add(*spandrel_extra_arches.EXTRA_REGISTRY)
+        _spandrel_extra_init_state = True
+    except Exception:
+        logger.warning("Failed to load spandrel_extra_arches", exc_info=True)
+        _spandrel_extra_init_state = False
+
+
+def load_spandrel_model(
+    path: str | os.PathLike,
+    *,
+    device: str | torch.device | None,
+    prefer_half: bool = False,
+    dtype: str | torch.dtype | None = None,
+    expected_architecture: str | None = None,
+) -> spandrel.ModelDescriptor:
+    global _spandrel_extra_init_state
+
+    import spandrel
+    _init_spandrel_extra_archs()
+
+    model_descriptor = spandrel.ModelLoader(device=device).load_from_file(str(path))
+    arch = model_descriptor.architecture
+    if expected_architecture and arch.name != expected_architecture:
+        logger.warning(
+            f"Model {path!r} is not a {expected_architecture!r} model (got {arch.name!r})",
+        )
+    half = False
+    if prefer_half:
+        if model_descriptor.supports_half:
+            model_descriptor.model.half()
+            half = True
+        else:
+            logger.info("Model %s does not support half precision, ignoring --half", path)
+    if dtype:
+        model_descriptor.model.to(dtype=dtype)
+    model_descriptor.model.eval()
+    logger.debug(
+        "Loaded %s from %s (device=%s, half=%s, dtype=%s)",
+        arch, path, device, half, dtype,
+    )
+    return model_descriptor
diff --git a/modules/models/diffusion/ddpm_edit.py b/modules/models/diffusion/ddpm_edit.py
index b892d5fc7b0..7b51c83c5d9 100644
--- a/modules/models/diffusion/ddpm_edit.py
+++ b/modules/models/diffusion/ddpm_edit.py
@@ -24,10 +24,15 @@
 from ldm.util import log_txt_as_img, exists, default, ismap, isimage, mean_flat, count_params, instantiate_from_config
 from ldm.modules.ema import LitEma
 from ldm.modules.distributions.distributions import normal_kl, DiagonalGaussianDistribution
-from ldm.models.autoencoder import VQModelInterface, IdentityFirstStage, AutoencoderKL
+from ldm.models.autoencoder import IdentityFirstStage, AutoencoderKL
 from ldm.modules.diffusionmodules.util import make_beta_schedule, extract_into_tensor, noise_like
 from ldm.models.diffusion.ddim import DDIMSampler
 
+try:
+    from ldm.models.autoencoder import VQModelInterface
+except Exception:
+    class VQModelInterface:
+        pass
 
 __conditioning_keys__ = {'concat': 'c_concat',
                          'crossattn': 'c_crossattn',
@@ -336,7 +341,7 @@ def p_losses(self, x_start, t, noise=None):
         elif self.parameterization == "x0":
             target = x_start
         else:
-            raise NotImplementedError(f"Paramterization {self.parameterization} not yet supported")
+            raise NotImplementedError(f"Parameterization {self.parameterization} not yet supported")
 
         loss = self.get_loss(model_out, target, mean=False).mean(dim=[1, 2, 3])
 
@@ -896,7 +901,7 @@ def forward(self, x, c, *args, **kwargs):
     def apply_model(self, x_noisy, t, cond, return_ids=False):
 
         if isinstance(cond, dict):
-            # hybrid case, cond is exptected to be a dict
+            # hybrid case, cond is expected to be a dict
             pass
         else:
             if not isinstance(cond, list):
@@ -932,7 +937,7 @@ def apply_model(self, x_noisy, t, cond, return_ids=False):
                 cond_list = [{c_key: [c[:, :, :, :, i]]} for i in range(c.shape[-1])]
 
             elif self.cond_stage_key == 'coordinates_bbox':
-                assert 'original_image_size' in self.split_input_params, 'BoudingBoxRescaling is missing original_image_size'
+                assert 'original_image_size' in self.split_input_params, 'BoundingBoxRescaling is missing original_image_size'
 
                 # assuming padding of unfold is always 0 and its dilation is always 1
                 n_patches_per_row = int((w - ks[0]) / stride[0] + 1)
@@ -942,7 +947,7 @@ def apply_model(self, x_noisy, t, cond, return_ids=False):
                 num_downs = self.first_stage_model.encoder.num_resolutions - 1
                 rescale_latent = 2 ** (num_downs)
 
-                # get top left postions of patches as conforming for the bbbox tokenizer, therefore we
+                # get top left positions of patches as conforming for the bbbox tokenizer, therefore we
                 # need to rescale the tl patch coordinates to be in between (0,1)
                 tl_patch_coordinates = [(rescale_latent * stride[0] * (patch_nr % n_patches_per_row) / full_img_w,
                                          rescale_latent * stride[1] * (patch_nr // n_patches_per_row) / full_img_h)
diff --git a/modules/models/diffusion/uni_pc/uni_pc.py b/modules/models/diffusion/uni_pc/uni_pc.py
index d257a7286fc..3333bc808d0 100644
--- a/modules/models/diffusion/uni_pc/uni_pc.py
+++ b/modules/models/diffusion/uni_pc/uni_pc.py
@@ -323,7 +323,7 @@ def cond_grad_fn(x, t_input, condition):
 
     def model_fn(x, t_continuous, condition, unconditional_condition):
         """
-        The noise predicition model function that is used for DPM-Solver.
+        The noise prediction model function that is used for DPM-Solver.
         """
         if t_continuous.reshape((-1,)).shape[0] == 1:
             t_continuous = t_continuous.expand((x.shape[0]))
diff --git a/modules/models/sd3/mmdit.py b/modules/models/sd3/mmdit.py
new file mode 100644
index 00000000000..8ddf49a4e3e
--- /dev/null
+++ b/modules/models/sd3/mmdit.py
@@ -0,0 +1,622 @@
+### This file contains impls for MM-DiT, the core model component of SD3
+
+import math
+from typing import Dict, Optional
+import numpy as np
+import torch
+import torch.nn as nn
+from einops import rearrange, repeat
+from modules.models.sd3.other_impls import attention, Mlp
+
+
+class PatchEmbed(nn.Module):
+    """ 2D Image to Patch Embedding"""
+    def __init__(
+            self,
+            img_size: Optional[int] = 224,
+            patch_size: int = 16,
+            in_chans: int = 3,
+            embed_dim: int = 768,
+            flatten: bool = True,
+            bias: bool = True,
+            strict_img_size: bool = True,
+            dynamic_img_pad: bool = False,
+            dtype=None,
+            device=None,
+    ):
+        super().__init__()
+        self.patch_size = (patch_size, patch_size)
+        if img_size is not None:
+            self.img_size = (img_size, img_size)
+            self.grid_size = tuple([s // p for s, p in zip(self.img_size, self.patch_size)])
+            self.num_patches = self.grid_size[0] * self.grid_size[1]
+        else:
+            self.img_size = None
+            self.grid_size = None
+            self.num_patches = None
+
+        # flatten spatial dim and transpose to channels last, kept for bwd compat
+        self.flatten = flatten
+        self.strict_img_size = strict_img_size
+        self.dynamic_img_pad = dynamic_img_pad
+
+        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size, bias=bias, dtype=dtype, device=device)
+
+    def forward(self, x):
+        B, C, H, W = x.shape
+        x = self.proj(x)
+        if self.flatten:
+            x = x.flatten(2).transpose(1, 2)  # NCHW -> NLC
+        return x
+
+
+def modulate(x, shift, scale):
+    if shift is None:
+        shift = torch.zeros_like(scale)
+    return x * (1 + scale.unsqueeze(1)) + shift.unsqueeze(1)
+
+
+#################################################################################
+#                   Sine/Cosine Positional Embedding Functions                  #
+#################################################################################
+
+
+def get_2d_sincos_pos_embed(embed_dim, grid_size, cls_token=False, extra_tokens=0, scaling_factor=None, offset=None):
+    """
+    grid_size: int of the grid height and width
+    return:
+    pos_embed: [grid_size*grid_size, embed_dim] or [1+grid_size*grid_size, embed_dim] (w/ or w/o cls_token)
+    """
+    grid_h = np.arange(grid_size, dtype=np.float32)
+    grid_w = np.arange(grid_size, dtype=np.float32)
+    grid = np.meshgrid(grid_w, grid_h)  # here w goes first
+    grid = np.stack(grid, axis=0)
+    if scaling_factor is not None:
+        grid = grid / scaling_factor
+    if offset is not None:
+        grid = grid - offset
+    grid = grid.reshape([2, 1, grid_size, grid_size])
+    pos_embed = get_2d_sincos_pos_embed_from_grid(embed_dim, grid)
+    if cls_token and extra_tokens > 0:
+        pos_embed = np.concatenate([np.zeros([extra_tokens, embed_dim]), pos_embed], axis=0)
+    return pos_embed
+
+
+def get_2d_sincos_pos_embed_from_grid(embed_dim, grid):
+    assert embed_dim % 2 == 0
+    # use half of dimensions to encode grid_h
+    emb_h = get_1d_sincos_pos_embed_from_grid(embed_dim // 2, grid[0])  # (H*W, D/2)
+    emb_w = get_1d_sincos_pos_embed_from_grid(embed_dim // 2, grid[1])  # (H*W, D/2)
+    emb = np.concatenate([emb_h, emb_w], axis=1)  # (H*W, D)
+    return emb
+
+
+def get_1d_sincos_pos_embed_from_grid(embed_dim, pos):
+    """
+    embed_dim: output dimension for each position
+    pos: a list of positions to be encoded: size (M,)
+    out: (M, D)
+    """
+    assert embed_dim % 2 == 0
+    omega = np.arange(embed_dim // 2, dtype=np.float64)
+    omega /= embed_dim / 2.0
+    omega = 1.0 / 10000**omega  # (D/2,)
+    pos = pos.reshape(-1)  # (M,)
+    out = np.einsum("m,d->md", pos, omega)  # (M, D/2), outer product
+    emb_sin = np.sin(out)  # (M, D/2)
+    emb_cos = np.cos(out)  # (M, D/2)
+    return np.concatenate([emb_sin, emb_cos], axis=1)  # (M, D)
+
+
+#################################################################################
+#               Embedding Layers for Timesteps and Class Labels                 #
+#################################################################################
+
+
+class TimestepEmbedder(nn.Module):
+    """Embeds scalar timesteps into vector representations."""
+
+    def __init__(self, hidden_size, frequency_embedding_size=256, dtype=None, device=None):
+        super().__init__()
+        self.mlp = nn.Sequential(
+            nn.Linear(frequency_embedding_size, hidden_size, bias=True, dtype=dtype, device=device),
+            nn.SiLU(),
+            nn.Linear(hidden_size, hidden_size, bias=True, dtype=dtype, device=device),
+        )
+        self.frequency_embedding_size = frequency_embedding_size
+
+    @staticmethod
+    def timestep_embedding(t, dim, max_period=10000):
+        """
+        Create sinusoidal timestep embeddings.
+        :param t: a 1-D Tensor of N indices, one per batch element.
+                          These may be fractional.
+        :param dim: the dimension of the output.
+        :param max_period: controls the minimum frequency of the embeddings.
+        :return: an (N, D) Tensor of positional embeddings.
+        """
+        half = dim // 2
+        freqs = torch.exp(
+            -math.log(max_period)
+            * torch.arange(start=0, end=half, dtype=torch.float32)
+            / half
+        ).to(device=t.device)
+        args = t[:, None].float() * freqs[None]
+        embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
+        if dim % 2:
+            embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
+        if torch.is_floating_point(t):
+            embedding = embedding.to(dtype=t.dtype)
+        return embedding
+
+    def forward(self, t, dtype, **kwargs):
+        t_freq = self.timestep_embedding(t, self.frequency_embedding_size).to(dtype)
+        t_emb = self.mlp(t_freq)
+        return t_emb
+
+
+class VectorEmbedder(nn.Module):
+    """Embeds a flat vector of dimension input_dim"""
+
+    def __init__(self, input_dim: int, hidden_size: int, dtype=None, device=None):
+        super().__init__()
+        self.mlp = nn.Sequential(
+            nn.Linear(input_dim, hidden_size, bias=True, dtype=dtype, device=device),
+            nn.SiLU(),
+            nn.Linear(hidden_size, hidden_size, bias=True, dtype=dtype, device=device),
+        )
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        return self.mlp(x)
+
+
+#################################################################################
+#                                 Core DiT Model                                #
+#################################################################################
+
+
+class QkvLinear(torch.nn.Linear):
+    pass
+
+def split_qkv(qkv, head_dim):
+    qkv = qkv.reshape(qkv.shape[0], qkv.shape[1], 3, -1, head_dim).movedim(2, 0)
+    return qkv[0], qkv[1], qkv[2]
+
+def optimized_attention(qkv, num_heads):
+    return attention(qkv[0], qkv[1], qkv[2], num_heads)
+
+class SelfAttention(nn.Module):
+    ATTENTION_MODES = ("xformers", "torch", "torch-hb", "math", "debug")
+
+    def __init__(
+        self,
+        dim: int,
+        num_heads: int = 8,
+        qkv_bias: bool = False,
+        qk_scale: Optional[float] = None,
+        attn_mode: str = "xformers",
+        pre_only: bool = False,
+        qk_norm: Optional[str] = None,
+        rmsnorm: bool = False,
+        dtype=None,
+        device=None,
+    ):
+        super().__init__()
+        self.num_heads = num_heads
+        self.head_dim = dim // num_heads
+
+        self.qkv = QkvLinear(dim, dim * 3, bias=qkv_bias, dtype=dtype, device=device)
+        if not pre_only:
+            self.proj = nn.Linear(dim, dim, dtype=dtype, device=device)
+        assert attn_mode in self.ATTENTION_MODES
+        self.attn_mode = attn_mode
+        self.pre_only = pre_only
+
+        if qk_norm == "rms":
+            self.ln_q = RMSNorm(self.head_dim, elementwise_affine=True, eps=1.0e-6, dtype=dtype, device=device)
+            self.ln_k = RMSNorm(self.head_dim, elementwise_affine=True, eps=1.0e-6, dtype=dtype, device=device)
+        elif qk_norm == "ln":
+            self.ln_q = nn.LayerNorm(self.head_dim, elementwise_affine=True, eps=1.0e-6, dtype=dtype, device=device)
+            self.ln_k = nn.LayerNorm(self.head_dim, elementwise_affine=True, eps=1.0e-6, dtype=dtype, device=device)
+        elif qk_norm is None:
+            self.ln_q = nn.Identity()
+            self.ln_k = nn.Identity()
+        else:
+            raise ValueError(qk_norm)
+
+    def pre_attention(self, x: torch.Tensor):
+        B, L, C = x.shape
+        qkv = self.qkv(x)
+        q, k, v = split_qkv(qkv, self.head_dim)
+        q = self.ln_q(q).reshape(q.shape[0], q.shape[1], -1)
+        k = self.ln_k(k).reshape(q.shape[0], q.shape[1], -1)
+        return (q, k, v)
+
+    def post_attention(self, x: torch.Tensor) -> torch.Tensor:
+        assert not self.pre_only
+        x = self.proj(x)
+        return x
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        (q, k, v) = self.pre_attention(x)
+        x = attention(q, k, v, self.num_heads)
+        x = self.post_attention(x)
+        return x
+
+
+class RMSNorm(torch.nn.Module):
+    def __init__(
+        self, dim: int, elementwise_affine: bool = False, eps: float = 1e-6, device=None, dtype=None
+    ):
+        """
+        Initialize the RMSNorm normalization layer.
+        Args:
+            dim (int): The dimension of the input tensor.
+            eps (float, optional): A small value added to the denominator for numerical stability. Default is 1e-6.
+        Attributes:
+            eps (float): A small value added to the denominator for numerical stability.
+            weight (nn.Parameter): Learnable scaling parameter.
+        """
+        super().__init__()
+        self.eps = eps
+        self.learnable_scale = elementwise_affine
+        if self.learnable_scale:
+            self.weight = nn.Parameter(torch.empty(dim, device=device, dtype=dtype))
+        else:
+            self.register_parameter("weight", None)
+
+    def _norm(self, x):
+        """
+        Apply the RMSNorm normalization to the input tensor.
+        Args:
+            x (torch.Tensor): The input tensor.
+        Returns:
+            torch.Tensor: The normalized tensor.
+        """
+        return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
+
+    def forward(self, x):
+        """
+        Forward pass through the RMSNorm layer.
+        Args:
+            x (torch.Tensor): The input tensor.
+        Returns:
+            torch.Tensor: The output tensor after applying RMSNorm.
+        """
+        x = self._norm(x)
+        if self.learnable_scale:
+            return x * self.weight.to(device=x.device, dtype=x.dtype)
+        else:
+            return x
+
+
+class SwiGLUFeedForward(nn.Module):
+    def __init__(
+        self,
+        dim: int,
+        hidden_dim: int,
+        multiple_of: int,
+        ffn_dim_multiplier: Optional[float] = None,
+    ):
+        """
+        Initialize the FeedForward module.
+
+        Args:
+            dim (int): Input dimension.
+            hidden_dim (int): Hidden dimension of the feedforward layer.
+            multiple_of (int): Value to ensure hidden dimension is a multiple of this value.
+            ffn_dim_multiplier (float, optional): Custom multiplier for hidden dimension. Defaults to None.
+
+        Attributes:
+            w1 (ColumnParallelLinear): Linear transformation for the first layer.
+            w2 (RowParallelLinear): Linear transformation for the second layer.
+            w3 (ColumnParallelLinear): Linear transformation for the third layer.
+
+        """
+        super().__init__()
+        hidden_dim = int(2 * hidden_dim / 3)
+        # custom dim factor multiplier
+        if ffn_dim_multiplier is not None:
+            hidden_dim = int(ffn_dim_multiplier * hidden_dim)
+        hidden_dim = multiple_of * ((hidden_dim + multiple_of - 1) // multiple_of)
+
+        self.w1 = nn.Linear(dim, hidden_dim, bias=False)
+        self.w2 = nn.Linear(hidden_dim, dim, bias=False)
+        self.w3 = nn.Linear(dim, hidden_dim, bias=False)
+
+    def forward(self, x):
+        return self.w2(nn.functional.silu(self.w1(x)) * self.w3(x))
+
+
+class DismantledBlock(nn.Module):
+    """A DiT block with gated adaptive layer norm (adaLN) conditioning."""
+
+    ATTENTION_MODES = ("xformers", "torch", "torch-hb", "math", "debug")
+
+    def __init__(
+        self,
+        hidden_size: int,
+        num_heads: int,
+        mlp_ratio: float = 4.0,
+        attn_mode: str = "xformers",
+        qkv_bias: bool = False,
+        pre_only: bool = False,
+        rmsnorm: bool = False,
+        scale_mod_only: bool = False,
+        swiglu: bool = False,
+        qk_norm: Optional[str] = None,
+        dtype=None,
+        device=None,
+        **block_kwargs,
+    ):
+        super().__init__()
+        assert attn_mode in self.ATTENTION_MODES
+        if not rmsnorm:
+            self.norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6, dtype=dtype, device=device)
+        else:
+            self.norm1 = RMSNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias, attn_mode=attn_mode, pre_only=pre_only, qk_norm=qk_norm, rmsnorm=rmsnorm, dtype=dtype, device=device)
+        if not pre_only:
+            if not rmsnorm:
+                self.norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6, dtype=dtype, device=device)
+            else:
+                self.norm2 = RMSNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        mlp_hidden_dim = int(hidden_size * mlp_ratio)
+        if not pre_only:
+            if not swiglu:
+                self.mlp = Mlp(in_features=hidden_size, hidden_features=mlp_hidden_dim, act_layer=nn.GELU(approximate="tanh"), dtype=dtype, device=device)
+            else:
+                self.mlp = SwiGLUFeedForward(dim=hidden_size, hidden_dim=mlp_hidden_dim, multiple_of=256)
+        self.scale_mod_only = scale_mod_only
+        if not scale_mod_only:
+            n_mods = 6 if not pre_only else 2
+        else:
+            n_mods = 4 if not pre_only else 1
+        self.adaLN_modulation = nn.Sequential(nn.SiLU(), nn.Linear(hidden_size, n_mods * hidden_size, bias=True, dtype=dtype, device=device))
+        self.pre_only = pre_only
+
+    def pre_attention(self, x: torch.Tensor, c: torch.Tensor):
+        assert x is not None, "pre_attention called with None input"
+        if not self.pre_only:
+            if not self.scale_mod_only:
+                shift_msa, scale_msa, gate_msa, shift_mlp, scale_mlp, gate_mlp = self.adaLN_modulation(c).chunk(6, dim=1)
+            else:
+                shift_msa = None
+                shift_mlp = None
+                scale_msa, gate_msa, scale_mlp, gate_mlp = self.adaLN_modulation(c).chunk(4, dim=1)
+            qkv = self.attn.pre_attention(modulate(self.norm1(x), shift_msa, scale_msa))
+            return qkv, (x, gate_msa, shift_mlp, scale_mlp, gate_mlp)
+        else:
+            if not self.scale_mod_only:
+                shift_msa, scale_msa = self.adaLN_modulation(c).chunk(2, dim=1)
+            else:
+                shift_msa = None
+                scale_msa = self.adaLN_modulation(c)
+            qkv = self.attn.pre_attention(modulate(self.norm1(x), shift_msa, scale_msa))
+            return qkv, None
+
+    def post_attention(self, attn, x, gate_msa, shift_mlp, scale_mlp, gate_mlp):
+        assert not self.pre_only
+        x = x + gate_msa.unsqueeze(1) * self.attn.post_attention(attn)
+        x = x + gate_mlp.unsqueeze(1) * self.mlp(modulate(self.norm2(x), shift_mlp, scale_mlp))
+        return x
+
+    def forward(self, x: torch.Tensor, c: torch.Tensor) -> torch.Tensor:
+        assert not self.pre_only
+        (q, k, v), intermediates = self.pre_attention(x, c)
+        attn = attention(q, k, v, self.attn.num_heads)
+        return self.post_attention(attn, *intermediates)
+
+
+def block_mixing(context, x, context_block, x_block, c):
+    assert context is not None, "block_mixing called with None context"
+    context_qkv, context_intermediates = context_block.pre_attention(context, c)
+
+    x_qkv, x_intermediates = x_block.pre_attention(x, c)
+
+    o = []
+    for t in range(3):
+        o.append(torch.cat((context_qkv[t], x_qkv[t]), dim=1))
+    q, k, v = tuple(o)
+
+    attn = attention(q, k, v, x_block.attn.num_heads)
+    context_attn, x_attn = (attn[:, : context_qkv[0].shape[1]], attn[:, context_qkv[0].shape[1] :])
+
+    if not context_block.pre_only:
+        context = context_block.post_attention(context_attn, *context_intermediates)
+    else:
+        context = None
+    x = x_block.post_attention(x_attn, *x_intermediates)
+    return context, x
+
+
+class JointBlock(nn.Module):
+    """just a small wrapper to serve as a fsdp unit"""
+
+    def __init__(self, *args, **kwargs):
+        super().__init__()
+        pre_only = kwargs.pop("pre_only")
+        qk_norm = kwargs.pop("qk_norm", None)
+        self.context_block = DismantledBlock(*args, pre_only=pre_only, qk_norm=qk_norm, **kwargs)
+        self.x_block = DismantledBlock(*args, pre_only=False, qk_norm=qk_norm, **kwargs)
+
+    def forward(self, *args, **kwargs):
+        return block_mixing(*args, context_block=self.context_block, x_block=self.x_block, **kwargs)
+
+
+class FinalLayer(nn.Module):
+    """
+    The final layer of DiT.
+    """
+
+    def __init__(self, hidden_size: int, patch_size: int, out_channels: int, total_out_channels: Optional[int] = None, dtype=None, device=None):
+        super().__init__()
+        self.norm_final = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6, dtype=dtype, device=device)
+        self.linear = (
+            nn.Linear(hidden_size, patch_size * patch_size * out_channels, bias=True, dtype=dtype, device=device)
+            if (total_out_channels is None)
+            else nn.Linear(hidden_size, total_out_channels, bias=True, dtype=dtype, device=device)
+        )
+        self.adaLN_modulation = nn.Sequential(nn.SiLU(), nn.Linear(hidden_size, 2 * hidden_size, bias=True, dtype=dtype, device=device))
+
+    def forward(self, x: torch.Tensor, c: torch.Tensor) -> torch.Tensor:
+        shift, scale = self.adaLN_modulation(c).chunk(2, dim=1)
+        x = modulate(self.norm_final(x), shift, scale)
+        x = self.linear(x)
+        return x
+
+
+class MMDiT(nn.Module):
+    """Diffusion model with a Transformer backbone."""
+
+    def __init__(
+        self,
+        input_size: int = 32,
+        patch_size: int = 2,
+        in_channels: int = 4,
+        depth: int = 28,
+        mlp_ratio: float = 4.0,
+        learn_sigma: bool = False,
+        adm_in_channels: Optional[int] = None,
+        context_embedder_config: Optional[Dict] = None,
+        register_length: int = 0,
+        attn_mode: str = "torch",
+        rmsnorm: bool = False,
+        scale_mod_only: bool = False,
+        swiglu: bool = False,
+        out_channels: Optional[int] = None,
+        pos_embed_scaling_factor: Optional[float] = None,
+        pos_embed_offset: Optional[float] = None,
+        pos_embed_max_size: Optional[int] = None,
+        num_patches = None,
+        qk_norm: Optional[str] = None,
+        qkv_bias: bool = True,
+        dtype = None,
+        device = None,
+    ):
+        super().__init__()
+        self.dtype = dtype
+        self.learn_sigma = learn_sigma
+        self.in_channels = in_channels
+        default_out_channels = in_channels * 2 if learn_sigma else in_channels
+        self.out_channels = out_channels if out_channels is not None else default_out_channels
+        self.patch_size = patch_size
+        self.pos_embed_scaling_factor = pos_embed_scaling_factor
+        self.pos_embed_offset = pos_embed_offset
+        self.pos_embed_max_size = pos_embed_max_size
+
+        # apply magic --> this defines a head_size of 64
+        hidden_size = 64 * depth
+        num_heads = depth
+
+        self.num_heads = num_heads
+
+        self.x_embedder = PatchEmbed(input_size, patch_size, in_channels, hidden_size, bias=True, strict_img_size=self.pos_embed_max_size is None, dtype=dtype, device=device)
+        self.t_embedder = TimestepEmbedder(hidden_size, dtype=dtype, device=device)
+
+        if adm_in_channels is not None:
+            assert isinstance(adm_in_channels, int)
+            self.y_embedder = VectorEmbedder(adm_in_channels, hidden_size, dtype=dtype, device=device)
+
+        self.context_embedder = nn.Identity()
+        if context_embedder_config is not None:
+            if context_embedder_config["target"] == "torch.nn.Linear":
+                self.context_embedder = nn.Linear(**context_embedder_config["params"], dtype=dtype, device=device)
+
+        self.register_length = register_length
+        if self.register_length > 0:
+            self.register = nn.Parameter(torch.randn(1, register_length, hidden_size, dtype=dtype, device=device))
+
+        # num_patches = self.x_embedder.num_patches
+        # Will use fixed sin-cos embedding:
+        # just use a buffer already
+        if num_patches is not None:
+            self.register_buffer(
+                "pos_embed",
+                torch.zeros(1, num_patches, hidden_size, dtype=dtype, device=device),
+            )
+        else:
+            self.pos_embed = None
+
+        self.joint_blocks = nn.ModuleList(
+            [
+                JointBlock(hidden_size, num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, attn_mode=attn_mode, pre_only=i == depth - 1, rmsnorm=rmsnorm, scale_mod_only=scale_mod_only, swiglu=swiglu, qk_norm=qk_norm, dtype=dtype, device=device)
+                for i in range(depth)
+            ]
+        )
+
+        self.final_layer = FinalLayer(hidden_size, patch_size, self.out_channels, dtype=dtype, device=device)
+
+    def cropped_pos_embed(self, hw):
+        assert self.pos_embed_max_size is not None
+        p = self.x_embedder.patch_size[0]
+        h, w = hw
+        # patched size
+        h = h // p
+        w = w // p
+        assert h <= self.pos_embed_max_size, (h, self.pos_embed_max_size)
+        assert w <= self.pos_embed_max_size, (w, self.pos_embed_max_size)
+        top = (self.pos_embed_max_size - h) // 2
+        left = (self.pos_embed_max_size - w) // 2
+        spatial_pos_embed = rearrange(
+            self.pos_embed,
+            "1 (h w) c -> 1 h w c",
+            h=self.pos_embed_max_size,
+            w=self.pos_embed_max_size,
+        )
+        spatial_pos_embed = spatial_pos_embed[:, top : top + h, left : left + w, :]
+        spatial_pos_embed = rearrange(spatial_pos_embed, "1 h w c -> 1 (h w) c")
+        return spatial_pos_embed
+
+    def unpatchify(self, x, hw=None):
+        """
+        x: (N, T, patch_size**2 * C)
+        imgs: (N, H, W, C)
+        """
+        c = self.out_channels
+        p = self.x_embedder.patch_size[0]
+        if hw is None:
+            h = w = int(x.shape[1] ** 0.5)
+        else:
+            h, w = hw
+            h = h // p
+            w = w // p
+        assert h * w == x.shape[1]
+
+        x = x.reshape(shape=(x.shape[0], h, w, p, p, c))
+        x = torch.einsum("nhwpqc->nchpwq", x)
+        imgs = x.reshape(shape=(x.shape[0], c, h * p, w * p))
+        return imgs
+
+    def forward_core_with_concat(self, x: torch.Tensor, c_mod: torch.Tensor, context: Optional[torch.Tensor] = None) -> torch.Tensor:
+        if self.register_length > 0:
+            context = torch.cat((repeat(self.register, "1 ... -> b ...", b=x.shape[0]), context if context is not None else torch.Tensor([]).type_as(x)), 1)
+
+        # context is B, L', D
+        # x is B, L, D
+        for block in self.joint_blocks:
+            context, x = block(context, x, c=c_mod)
+
+        x = self.final_layer(x, c_mod)  # (N, T, patch_size ** 2 * out_channels)
+        return x
+
+    def forward(self, x: torch.Tensor, t: torch.Tensor, y: Optional[torch.Tensor] = None, context: Optional[torch.Tensor] = None) -> torch.Tensor:
+        """
+        Forward pass of DiT.
+        x: (N, C, H, W) tensor of spatial inputs (images or latent representations of images)
+        t: (N,) tensor of diffusion timesteps
+        y: (N,) tensor of class labels
+        """
+        hw = x.shape[-2:]
+        x = self.x_embedder(x) + self.cropped_pos_embed(hw)
+        c = self.t_embedder(t, dtype=x.dtype)  # (N, D)
+        if y is not None:
+            y = self.y_embedder(y)  # (N, D)
+            c = c + y  # (N, D)
+
+        context = self.context_embedder(context)
+
+        x = self.forward_core_with_concat(x, c, context)
+
+        x = self.unpatchify(x, hw=hw)  # (N, out_channels, H, W)
+        return x
diff --git a/modules/models/sd3/other_impls.py b/modules/models/sd3/other_impls.py
new file mode 100644
index 00000000000..78c1dc68758
--- /dev/null
+++ b/modules/models/sd3/other_impls.py
@@ -0,0 +1,510 @@
+### This file contains impls for underlying related models (CLIP, T5, etc)
+
+import torch
+import math
+from torch import nn
+from transformers import CLIPTokenizer, T5TokenizerFast
+
+from modules import sd_hijack
+
+
+#################################################################################################
+### Core/Utility
+#################################################################################################
+
+
+class AutocastLinear(nn.Linear):
+    """Same as usual linear layer, but casts its weights to whatever the parameter type is.
+
+    This is different from torch.autocast in a way that float16 layer processing float32 input
+    will return float16 with autocast on, and float32 with this. T5 seems to be fucked
+    if you do it in full float16 (returning almost all zeros in the final output).
+    """
+
+    def forward(self, x):
+        return torch.nn.functional.linear(x, self.weight.to(x.dtype), self.bias.to(x.dtype) if self.bias is not None else None)
+
+
+def attention(q, k, v, heads, mask=None):
+    """Convenience wrapper around a basic attention operation"""
+    b, _, dim_head = q.shape
+    dim_head //= heads
+    q, k, v = [t.view(b, -1, heads, dim_head).transpose(1, 2) for t in (q, k, v)]
+    out = torch.nn.functional.scaled_dot_product_attention(q, k, v, attn_mask=mask, dropout_p=0.0, is_causal=False)
+    return out.transpose(1, 2).reshape(b, -1, heads * dim_head)
+
+
+class Mlp(nn.Module):
+    """ MLP as used in Vision Transformer, MLP-Mixer and related networks"""
+    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, bias=True, dtype=None, device=None):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+
+        self.fc1 = nn.Linear(in_features, hidden_features, bias=bias, dtype=dtype, device=device)
+        self.act = act_layer
+        self.fc2 = nn.Linear(hidden_features, out_features, bias=bias, dtype=dtype, device=device)
+
+    def forward(self, x):
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.fc2(x)
+        return x
+
+
+#################################################################################################
+### CLIP
+#################################################################################################
+
+
+class CLIPAttention(torch.nn.Module):
+    def __init__(self, embed_dim, heads, dtype, device):
+        super().__init__()
+        self.heads = heads
+        self.q_proj = nn.Linear(embed_dim, embed_dim, bias=True, dtype=dtype, device=device)
+        self.k_proj = nn.Linear(embed_dim, embed_dim, bias=True, dtype=dtype, device=device)
+        self.v_proj = nn.Linear(embed_dim, embed_dim, bias=True, dtype=dtype, device=device)
+        self.out_proj = nn.Linear(embed_dim, embed_dim, bias=True, dtype=dtype, device=device)
+
+    def forward(self, x, mask=None):
+        q = self.q_proj(x)
+        k = self.k_proj(x)
+        v = self.v_proj(x)
+        out = attention(q, k, v, self.heads, mask)
+        return self.out_proj(out)
+
+
+ACTIVATIONS = {
+    "quick_gelu": lambda a: a * torch.sigmoid(1.702 * a),
+    "gelu": torch.nn.functional.gelu,
+}
+
+class CLIPLayer(torch.nn.Module):
+    def __init__(self, embed_dim, heads, intermediate_size, intermediate_activation, dtype, device):
+        super().__init__()
+        self.layer_norm1 = nn.LayerNorm(embed_dim, dtype=dtype, device=device)
+        self.self_attn = CLIPAttention(embed_dim, heads, dtype, device)
+        self.layer_norm2 = nn.LayerNorm(embed_dim, dtype=dtype, device=device)
+        #self.mlp = CLIPMLP(embed_dim, intermediate_size, intermediate_activation, dtype, device)
+        self.mlp = Mlp(embed_dim, intermediate_size, embed_dim, act_layer=ACTIVATIONS[intermediate_activation], dtype=dtype, device=device)
+
+    def forward(self, x, mask=None):
+        x += self.self_attn(self.layer_norm1(x), mask)
+        x += self.mlp(self.layer_norm2(x))
+        return x
+
+
+class CLIPEncoder(torch.nn.Module):
+    def __init__(self, num_layers, embed_dim, heads, intermediate_size, intermediate_activation, dtype, device):
+        super().__init__()
+        self.layers = torch.nn.ModuleList([CLIPLayer(embed_dim, heads, intermediate_size, intermediate_activation, dtype, device) for i in range(num_layers)])
+
+    def forward(self, x, mask=None, intermediate_output=None):
+        if intermediate_output is not None:
+            if intermediate_output < 0:
+                intermediate_output = len(self.layers) + intermediate_output
+        intermediate = None
+        for i, layer in enumerate(self.layers):
+            x = layer(x, mask)
+            if i == intermediate_output:
+                intermediate = x.clone()
+        return x, intermediate
+
+
+class CLIPEmbeddings(torch.nn.Module):
+    def __init__(self, embed_dim, vocab_size=49408, num_positions=77, dtype=None, device=None, textual_inversion_key="clip_l"):
+        super().__init__()
+        self.token_embedding = sd_hijack.TextualInversionEmbeddings(vocab_size, embed_dim, dtype=dtype, device=device, textual_inversion_key=textual_inversion_key)
+        self.position_embedding = torch.nn.Embedding(num_positions, embed_dim, dtype=dtype, device=device)
+
+    def forward(self, input_tokens):
+        return self.token_embedding(input_tokens) + self.position_embedding.weight
+
+
+class CLIPTextModel_(torch.nn.Module):
+    def __init__(self, config_dict, dtype, device):
+        num_layers = config_dict["num_hidden_layers"]
+        embed_dim = config_dict["hidden_size"]
+        heads = config_dict["num_attention_heads"]
+        intermediate_size = config_dict["intermediate_size"]
+        intermediate_activation = config_dict["hidden_act"]
+        super().__init__()
+        self.embeddings = CLIPEmbeddings(embed_dim, dtype=torch.float32, device=device, textual_inversion_key=config_dict.get('textual_inversion_key', 'clip_l'))
+        self.encoder = CLIPEncoder(num_layers, embed_dim, heads, intermediate_size, intermediate_activation, dtype, device)
+        self.final_layer_norm = nn.LayerNorm(embed_dim, dtype=dtype, device=device)
+
+    def forward(self, input_tokens, intermediate_output=None, final_layer_norm_intermediate=True):
+        x = self.embeddings(input_tokens)
+        causal_mask = torch.empty(x.shape[1], x.shape[1], dtype=x.dtype, device=x.device).fill_(float("-inf")).triu_(1)
+        x, i = self.encoder(x, mask=causal_mask, intermediate_output=intermediate_output)
+        x = self.final_layer_norm(x)
+        if i is not None and final_layer_norm_intermediate:
+            i = self.final_layer_norm(i)
+        pooled_output = x[torch.arange(x.shape[0], device=x.device), input_tokens.to(dtype=torch.int, device=x.device).argmax(dim=-1),]
+        return x, i, pooled_output
+
+
+class CLIPTextModel(torch.nn.Module):
+    def __init__(self, config_dict, dtype, device):
+        super().__init__()
+        self.num_layers = config_dict["num_hidden_layers"]
+        self.text_model = CLIPTextModel_(config_dict, dtype, device)
+        embed_dim = config_dict["hidden_size"]
+        self.text_projection = nn.Linear(embed_dim, embed_dim, bias=False, dtype=dtype, device=device)
+        self.text_projection.weight.copy_(torch.eye(embed_dim))
+        self.dtype = dtype
+
+    def get_input_embeddings(self):
+        return self.text_model.embeddings.token_embedding
+
+    def set_input_embeddings(self, embeddings):
+        self.text_model.embeddings.token_embedding = embeddings
+
+    def forward(self, *args, **kwargs):
+        x = self.text_model(*args, **kwargs)
+        out = self.text_projection(x[2])
+        return (x[0], x[1], out, x[2])
+
+
+class SDTokenizer:
+    def __init__(self, max_length=77, pad_with_end=True, tokenizer=None, has_start_token=True, pad_to_max_length=True, min_length=None):
+        self.tokenizer = tokenizer
+        self.max_length = max_length
+        self.min_length = min_length
+        empty = self.tokenizer('')["input_ids"]
+        if has_start_token:
+            self.tokens_start = 1
+            self.start_token = empty[0]
+            self.end_token = empty[1]
+        else:
+            self.tokens_start = 0
+            self.start_token = None
+            self.end_token = empty[0]
+        self.pad_with_end = pad_with_end
+        self.pad_to_max_length = pad_to_max_length
+        vocab = self.tokenizer.get_vocab()
+        self.inv_vocab = {v: k for k, v in vocab.items()}
+        self.max_word_length = 8
+
+
+    def tokenize_with_weights(self, text:str):
+        """Tokenize the text, with weight values - presume 1.0 for all and ignore other features here. The details aren't relevant for a reference impl, and weights themselves has weak effect on SD3."""
+        if self.pad_with_end:
+            pad_token = self.end_token
+        else:
+            pad_token = 0
+        batch = []
+        if self.start_token is not None:
+            batch.append((self.start_token, 1.0))
+        to_tokenize = text.replace("\n", " ").split(' ')
+        to_tokenize = [x for x in to_tokenize if x != ""]
+        for word in to_tokenize:
+            batch.extend([(t, 1) for t in self.tokenizer(word)["input_ids"][self.tokens_start:-1]])
+        batch.append((self.end_token, 1.0))
+        if self.pad_to_max_length:
+            batch.extend([(pad_token, 1.0)] * (self.max_length - len(batch)))
+        if self.min_length is not None and len(batch) < self.min_length:
+            batch.extend([(pad_token, 1.0)] * (self.min_length - len(batch)))
+        return [batch]
+
+
+class SDXLClipGTokenizer(SDTokenizer):
+    def __init__(self, tokenizer):
+        super().__init__(pad_with_end=False, tokenizer=tokenizer)
+
+
+class SD3Tokenizer:
+    def __init__(self):
+        clip_tokenizer = CLIPTokenizer.from_pretrained("openai/clip-vit-large-patch14")
+        self.clip_l = SDTokenizer(tokenizer=clip_tokenizer)
+        self.clip_g = SDXLClipGTokenizer(clip_tokenizer)
+        self.t5xxl = T5XXLTokenizer()
+
+    def tokenize_with_weights(self, text:str):
+        out = {}
+        out["g"] = self.clip_g.tokenize_with_weights(text)
+        out["l"] = self.clip_l.tokenize_with_weights(text)
+        out["t5xxl"] = self.t5xxl.tokenize_with_weights(text)
+        return out
+
+
+class ClipTokenWeightEncoder:
+    def encode_token_weights(self, token_weight_pairs):
+        tokens = [a[0] for a in token_weight_pairs[0]]
+        out, pooled = self([tokens])
+        if pooled is not None:
+            first_pooled = pooled[0:1].cpu()
+        else:
+            first_pooled = pooled
+        output = [out[0:1]]
+        return torch.cat(output, dim=-2).cpu(), first_pooled
+
+
+class SDClipModel(torch.nn.Module, ClipTokenWeightEncoder):
+    """Uses the CLIP transformer encoder for text (from huggingface)"""
+    LAYERS = ["last", "pooled", "hidden"]
+    def __init__(self, device="cpu", max_length=77, layer="last", layer_idx=None, textmodel_json_config=None, dtype=None, model_class=CLIPTextModel,
+                 special_tokens=None, layer_norm_hidden_state=True, return_projected_pooled=True):
+        super().__init__()
+        assert layer in self.LAYERS
+        self.transformer = model_class(textmodel_json_config, dtype, device)
+        self.num_layers = self.transformer.num_layers
+        self.max_length = max_length
+        self.transformer = self.transformer.eval()
+        for param in self.parameters():
+            param.requires_grad = False
+        self.layer = layer
+        self.layer_idx = None
+        self.special_tokens = special_tokens if special_tokens is not None else {"start": 49406, "end": 49407, "pad": 49407}
+        self.logit_scale = torch.nn.Parameter(torch.tensor(4.6055))
+        self.layer_norm_hidden_state = layer_norm_hidden_state
+        self.return_projected_pooled = return_projected_pooled
+        if layer == "hidden":
+            assert layer_idx is not None
+            assert abs(layer_idx) < self.num_layers
+            self.set_clip_options({"layer": layer_idx})
+        self.options_default = (self.layer, self.layer_idx, self.return_projected_pooled)
+
+    def set_clip_options(self, options):
+        layer_idx = options.get("layer", self.layer_idx)
+        self.return_projected_pooled = options.get("projected_pooled", self.return_projected_pooled)
+        if layer_idx is None or abs(layer_idx) > self.num_layers:
+            self.layer = "last"
+        else:
+            self.layer = "hidden"
+            self.layer_idx = layer_idx
+
+    def forward(self, tokens):
+        backup_embeds = self.transformer.get_input_embeddings()
+        tokens = torch.asarray(tokens, dtype=torch.int64, device=backup_embeds.weight.device)
+        outputs = self.transformer(tokens, intermediate_output=self.layer_idx, final_layer_norm_intermediate=self.layer_norm_hidden_state)
+        self.transformer.set_input_embeddings(backup_embeds)
+        if self.layer == "last":
+            z = outputs[0]
+        else:
+            z = outputs[1]
+        pooled_output = None
+        if len(outputs) >= 3:
+            if not self.return_projected_pooled and len(outputs) >= 4 and outputs[3] is not None:
+                pooled_output = outputs[3].float()
+            elif outputs[2] is not None:
+                pooled_output = outputs[2].float()
+        return z.float(), pooled_output
+
+
+class SDXLClipG(SDClipModel):
+    """Wraps the CLIP-G model into the SD-CLIP-Model interface"""
+    def __init__(self, config, device="cpu", layer="penultimate", layer_idx=None, dtype=None):
+        if layer == "penultimate":
+            layer="hidden"
+            layer_idx=-2
+        super().__init__(device=device, layer=layer, layer_idx=layer_idx, textmodel_json_config=config, dtype=dtype, special_tokens={"start": 49406, "end": 49407, "pad": 0}, layer_norm_hidden_state=False)
+
+
+class T5XXLModel(SDClipModel):
+    """Wraps the T5-XXL model into the SD-CLIP-Model interface for convenience"""
+    def __init__(self, config, device="cpu", layer="last", layer_idx=None, dtype=None):
+        super().__init__(device=device, layer=layer, layer_idx=layer_idx, textmodel_json_config=config, dtype=dtype, special_tokens={"end": 1, "pad": 0}, model_class=T5)
+
+
+#################################################################################################
+### T5 implementation, for the T5-XXL text encoder portion, largely pulled from upstream impl
+#################################################################################################
+
+class T5XXLTokenizer(SDTokenizer):
+    """Wraps the T5 Tokenizer from HF into the SDTokenizer interface"""
+    def __init__(self):
+        super().__init__(pad_with_end=False, tokenizer=T5TokenizerFast.from_pretrained("google/t5-v1_1-xxl"), has_start_token=False, pad_to_max_length=False, max_length=99999999, min_length=77)
+
+
+class T5LayerNorm(torch.nn.Module):
+    def __init__(self, hidden_size, eps=1e-6, dtype=None, device=None):
+        super().__init__()
+        self.weight = torch.nn.Parameter(torch.ones(hidden_size, dtype=dtype, device=device))
+        self.variance_epsilon = eps
+
+    def forward(self, x):
+        variance = x.pow(2).mean(-1, keepdim=True)
+        x = x * torch.rsqrt(variance + self.variance_epsilon)
+        return self.weight.to(device=x.device, dtype=x.dtype) * x
+
+
+class T5DenseGatedActDense(torch.nn.Module):
+    def __init__(self, model_dim, ff_dim, dtype, device):
+        super().__init__()
+        self.wi_0 = AutocastLinear(model_dim, ff_dim, bias=False, dtype=dtype, device=device)
+        self.wi_1 = AutocastLinear(model_dim, ff_dim, bias=False, dtype=dtype, device=device)
+        self.wo = AutocastLinear(ff_dim, model_dim, bias=False, dtype=dtype, device=device)
+
+    def forward(self, x):
+        hidden_gelu = torch.nn.functional.gelu(self.wi_0(x), approximate="tanh")
+        hidden_linear = self.wi_1(x)
+        x = hidden_gelu * hidden_linear
+        x = self.wo(x)
+        return x
+
+
+class T5LayerFF(torch.nn.Module):
+    def __init__(self, model_dim, ff_dim, dtype, device):
+        super().__init__()
+        self.DenseReluDense = T5DenseGatedActDense(model_dim, ff_dim, dtype, device)
+        self.layer_norm = T5LayerNorm(model_dim, dtype=dtype, device=device)
+
+    def forward(self, x):
+        forwarded_states = self.layer_norm(x)
+        forwarded_states = self.DenseReluDense(forwarded_states)
+        x += forwarded_states
+        return x
+
+
+class T5Attention(torch.nn.Module):
+    def __init__(self, model_dim, inner_dim, num_heads, relative_attention_bias, dtype, device):
+        super().__init__()
+        # Mesh TensorFlow initialization to avoid scaling before softmax
+        self.q = AutocastLinear(model_dim, inner_dim, bias=False, dtype=dtype, device=device)
+        self.k = AutocastLinear(model_dim, inner_dim, bias=False, dtype=dtype, device=device)
+        self.v = AutocastLinear(model_dim, inner_dim, bias=False, dtype=dtype, device=device)
+        self.o = AutocastLinear(inner_dim, model_dim, bias=False, dtype=dtype, device=device)
+        self.num_heads = num_heads
+        self.relative_attention_bias = None
+        if relative_attention_bias:
+            self.relative_attention_num_buckets = 32
+            self.relative_attention_max_distance = 128
+            self.relative_attention_bias = torch.nn.Embedding(self.relative_attention_num_buckets, self.num_heads, device=device)
+
+    @staticmethod
+    def _relative_position_bucket(relative_position, bidirectional=True, num_buckets=32, max_distance=128):
+        """
+        Adapted from Mesh Tensorflow:
+        https://github.com/tensorflow/mesh/blob/0cb87fe07da627bf0b7e60475d59f95ed6b5be3d/mesh_tensorflow/transformer/transformer_layers.py#L593
+
+        Translate relative position to a bucket number for relative attention. The relative position is defined as
+        memory_position - query_position, i.e. the distance in tokens from the attending position to the attended-to
+        position. If bidirectional=False, then positive relative positions are invalid. We use smaller buckets for
+        small absolute relative_position and larger buckets for larger absolute relative_positions. All relative
+        positions >=max_distance map to the same bucket. All relative positions <=-max_distance map to the same bucket.
+        This should allow for more graceful generalization to longer sequences than the model has been trained on
+
+        Args:
+            relative_position: an int32 Tensor
+            bidirectional: a boolean - whether the attention is bidirectional
+            num_buckets: an integer
+            max_distance: an integer
+
+        Returns:
+            a Tensor with the same shape as relative_position, containing int32 values in the range [0, num_buckets)
+        """
+        relative_buckets = 0
+        if bidirectional:
+            num_buckets //= 2
+            relative_buckets += (relative_position > 0).to(torch.long) * num_buckets
+            relative_position = torch.abs(relative_position)
+        else:
+            relative_position = -torch.min(relative_position, torch.zeros_like(relative_position))
+        # now relative_position is in the range [0, inf)
+        # half of the buckets are for exact increments in positions
+        max_exact = num_buckets // 2
+        is_small = relative_position < max_exact
+        # The other half of the buckets are for logarithmically bigger bins in positions up to max_distance
+        relative_position_if_large = max_exact + (
+            torch.log(relative_position.float() / max_exact)
+            / math.log(max_distance / max_exact)
+            * (num_buckets - max_exact)
+        ).to(torch.long)
+        relative_position_if_large = torch.min(relative_position_if_large, torch.full_like(relative_position_if_large, num_buckets - 1))
+        relative_buckets += torch.where(is_small, relative_position, relative_position_if_large)
+        return relative_buckets
+
+    def compute_bias(self, query_length, key_length, device):
+        """Compute binned relative position bias"""
+        context_position = torch.arange(query_length, dtype=torch.long, device=device)[:, None]
+        memory_position = torch.arange(key_length, dtype=torch.long, device=device)[None, :]
+        relative_position = memory_position - context_position  # shape (query_length, key_length)
+        relative_position_bucket = self._relative_position_bucket(
+            relative_position,  # shape (query_length, key_length)
+            bidirectional=True,
+            num_buckets=self.relative_attention_num_buckets,
+            max_distance=self.relative_attention_max_distance,
+        )
+        values = self.relative_attention_bias(relative_position_bucket)  # shape (query_length, key_length, num_heads)
+        values = values.permute([2, 0, 1]).unsqueeze(0)  # shape (1, num_heads, query_length, key_length)
+        return values
+
+    def forward(self, x, past_bias=None):
+        q = self.q(x)
+        k = self.k(x)
+        v = self.v(x)
+
+        if self.relative_attention_bias is not None:
+            past_bias = self.compute_bias(x.shape[1], x.shape[1], x.device)
+        if past_bias is not None:
+            mask = past_bias
+        else:
+            mask = None
+
+        out = attention(q, k * ((k.shape[-1] / self.num_heads) ** 0.5), v, self.num_heads, mask.to(x.dtype) if mask is not None else None)
+
+        return self.o(out), past_bias
+
+
+class T5LayerSelfAttention(torch.nn.Module):
+    def __init__(self, model_dim, inner_dim, ff_dim, num_heads, relative_attention_bias, dtype, device):
+        super().__init__()
+        self.SelfAttention = T5Attention(model_dim, inner_dim, num_heads, relative_attention_bias, dtype, device)
+        self.layer_norm = T5LayerNorm(model_dim, dtype=dtype, device=device)
+
+    def forward(self, x, past_bias=None):
+        output, past_bias = self.SelfAttention(self.layer_norm(x), past_bias=past_bias)
+        x += output
+        return x, past_bias
+
+
+class T5Block(torch.nn.Module):
+    def __init__(self, model_dim, inner_dim, ff_dim, num_heads, relative_attention_bias, dtype, device):
+        super().__init__()
+        self.layer = torch.nn.ModuleList()
+        self.layer.append(T5LayerSelfAttention(model_dim, inner_dim, ff_dim, num_heads, relative_attention_bias, dtype, device))
+        self.layer.append(T5LayerFF(model_dim, ff_dim, dtype, device))
+
+    def forward(self, x, past_bias=None):
+        x, past_bias = self.layer[0](x, past_bias)
+        x = self.layer[-1](x)
+        return x, past_bias
+
+
+class T5Stack(torch.nn.Module):
+    def __init__(self, num_layers, model_dim, inner_dim, ff_dim, num_heads, vocab_size, dtype, device):
+        super().__init__()
+        self.embed_tokens = torch.nn.Embedding(vocab_size, model_dim, device=device)
+        self.block = torch.nn.ModuleList([T5Block(model_dim, inner_dim, ff_dim, num_heads, relative_attention_bias=(i == 0), dtype=dtype, device=device) for i in range(num_layers)])
+        self.final_layer_norm = T5LayerNorm(model_dim, dtype=dtype, device=device)
+
+    def forward(self, input_ids, intermediate_output=None, final_layer_norm_intermediate=True):
+        intermediate = None
+        x = self.embed_tokens(input_ids).to(torch.float32)  # needs float32 or else T5 returns all zeroes
+        past_bias = None
+        for i, layer in enumerate(self.block):
+            x, past_bias = layer(x, past_bias)
+            if i == intermediate_output:
+                intermediate = x.clone()
+        x = self.final_layer_norm(x)
+        if intermediate is not None and final_layer_norm_intermediate:
+            intermediate = self.final_layer_norm(intermediate)
+        return x, intermediate
+
+
+class T5(torch.nn.Module):
+    def __init__(self, config_dict, dtype, device):
+        super().__init__()
+        self.num_layers = config_dict["num_layers"]
+        self.encoder = T5Stack(self.num_layers, config_dict["d_model"], config_dict["d_model"], config_dict["d_ff"], config_dict["num_heads"], config_dict["vocab_size"], dtype, device)
+        self.dtype = dtype
+
+    def get_input_embeddings(self):
+        return self.encoder.embed_tokens
+
+    def set_input_embeddings(self, embeddings):
+        self.encoder.embed_tokens = embeddings
+
+    def forward(self, *args, **kwargs):
+        return self.encoder(*args, **kwargs)
diff --git a/modules/models/sd3/sd3_cond.py b/modules/models/sd3/sd3_cond.py
new file mode 100644
index 00000000000..325c512d594
--- /dev/null
+++ b/modules/models/sd3/sd3_cond.py
@@ -0,0 +1,222 @@
+import os
+import safetensors
+import torch
+import typing
+
+from transformers import CLIPTokenizer, T5TokenizerFast
+
+from modules import shared, devices, modelloader, sd_hijack_clip, prompt_parser
+from modules.models.sd3.other_impls import SDClipModel, SDXLClipG, T5XXLModel, SD3Tokenizer
+
+
+class SafetensorsMapping(typing.Mapping):
+    def __init__(self, file):
+        self.file = file
+
+    def __len__(self):
+        return len(self.file.keys())
+
+    def __iter__(self):
+        for key in self.file.keys():
+            yield key
+
+    def __getitem__(self, key):
+        return self.file.get_tensor(key)
+
+
+CLIPL_URL = "https://huggingface.co/AUTOMATIC/stable-diffusion-3-medium-text-encoders/resolve/main/clip_l.safetensors"
+CLIPL_CONFIG = {
+    "hidden_act": "quick_gelu",
+    "hidden_size": 768,
+    "intermediate_size": 3072,
+    "num_attention_heads": 12,
+    "num_hidden_layers": 12,
+}
+
+CLIPG_URL = "https://huggingface.co/AUTOMATIC/stable-diffusion-3-medium-text-encoders/resolve/main/clip_g.safetensors"
+CLIPG_CONFIG = {
+    "hidden_act": "gelu",
+    "hidden_size": 1280,
+    "intermediate_size": 5120,
+    "num_attention_heads": 20,
+    "num_hidden_layers": 32,
+    "textual_inversion_key": "clip_g",
+}
+
+T5_URL = "https://huggingface.co/AUTOMATIC/stable-diffusion-3-medium-text-encoders/resolve/main/t5xxl_fp16.safetensors"
+T5_CONFIG = {
+    "d_ff": 10240,
+    "d_model": 4096,
+    "num_heads": 64,
+    "num_layers": 24,
+    "vocab_size": 32128,
+}
+
+
+class Sd3ClipLG(sd_hijack_clip.TextConditionalModel):
+    def __init__(self, clip_l, clip_g):
+        super().__init__()
+
+        self.clip_l = clip_l
+        self.clip_g = clip_g
+
+        self.tokenizer = CLIPTokenizer.from_pretrained("openai/clip-vit-large-patch14")
+
+        empty = self.tokenizer('')["input_ids"]
+        self.id_start = empty[0]
+        self.id_end = empty[1]
+        self.id_pad = empty[1]
+
+        self.return_pooled = True
+
+    def tokenize(self, texts):
+        return self.tokenizer(texts, truncation=False, add_special_tokens=False)["input_ids"]
+
+    def encode_with_transformers(self, tokens):
+        tokens_g = tokens.clone()
+
+        for batch_pos in range(tokens_g.shape[0]):
+            index = tokens_g[batch_pos].cpu().tolist().index(self.id_end)
+            tokens_g[batch_pos, index+1:tokens_g.shape[1]] = 0
+
+        l_out, l_pooled = self.clip_l(tokens)
+        g_out, g_pooled = self.clip_g(tokens_g)
+
+        lg_out = torch.cat([l_out, g_out], dim=-1)
+        lg_out = torch.nn.functional.pad(lg_out, (0, 4096 - lg_out.shape[-1]))
+
+        vector_out = torch.cat((l_pooled, g_pooled), dim=-1)
+
+        lg_out.pooled = vector_out
+        return lg_out
+
+    def encode_embedding_init_text(self, init_text, nvpt):
+        return torch.zeros((nvpt, 768+1280), device=devices.device) # XXX
+
+
+class Sd3T5(torch.nn.Module):
+    def __init__(self, t5xxl):
+        super().__init__()
+
+        self.t5xxl = t5xxl
+        self.tokenizer = T5TokenizerFast.from_pretrained("google/t5-v1_1-xxl")
+
+        empty = self.tokenizer('', padding='max_length', max_length=2)["input_ids"]
+        self.id_end = empty[0]
+        self.id_pad = empty[1]
+
+    def tokenize(self, texts):
+        return self.tokenizer(texts, truncation=False, add_special_tokens=False)["input_ids"]
+
+    def tokenize_line(self, line, *, target_token_count=None):
+        if shared.opts.emphasis != "None":
+            parsed = prompt_parser.parse_prompt_attention(line)
+        else:
+            parsed = [[line, 1.0]]
+
+        tokenized = self.tokenize([text for text, _ in parsed])
+
+        tokens = []
+        multipliers = []
+
+        for text_tokens, (text, weight) in zip(tokenized, parsed):
+            if text == 'BREAK' and weight == -1:
+                continue
+
+            tokens += text_tokens
+            multipliers += [weight] * len(text_tokens)
+
+        tokens += [self.id_end]
+        multipliers += [1.0]
+
+        if target_token_count is not None:
+            if len(tokens) < target_token_count:
+                tokens += [self.id_pad] * (target_token_count - len(tokens))
+                multipliers += [1.0] * (target_token_count - len(tokens))
+            else:
+                tokens = tokens[0:target_token_count]
+                multipliers = multipliers[0:target_token_count]
+
+        return tokens, multipliers
+
+    def forward(self, texts, *, token_count):
+        if not self.t5xxl or not shared.opts.sd3_enable_t5:
+            return torch.zeros((len(texts), token_count, 4096), device=devices.device, dtype=devices.dtype)
+
+        tokens_batch = []
+
+        for text in texts:
+            tokens, multipliers = self.tokenize_line(text, target_token_count=token_count)
+            tokens_batch.append(tokens)
+
+        t5_out, t5_pooled = self.t5xxl(tokens_batch)
+
+        return t5_out
+
+    def encode_embedding_init_text(self, init_text, nvpt):
+        return torch.zeros((nvpt, 4096), device=devices.device) # XXX
+
+
+class SD3Cond(torch.nn.Module):
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+
+        self.tokenizer = SD3Tokenizer()
+
+        with torch.no_grad():
+            self.clip_g = SDXLClipG(CLIPG_CONFIG, device="cpu", dtype=devices.dtype)
+            self.clip_l = SDClipModel(layer="hidden", layer_idx=-2, device="cpu", dtype=devices.dtype, layer_norm_hidden_state=False, return_projected_pooled=False, textmodel_json_config=CLIPL_CONFIG)
+
+            if shared.opts.sd3_enable_t5:
+                self.t5xxl = T5XXLModel(T5_CONFIG, device="cpu", dtype=devices.dtype)
+            else:
+                self.t5xxl = None
+
+            self.model_lg = Sd3ClipLG(self.clip_l, self.clip_g)
+            self.model_t5 = Sd3T5(self.t5xxl)
+
+    def forward(self, prompts: list[str]):
+        with devices.without_autocast():
+            lg_out, vector_out = self.model_lg(prompts)
+            t5_out = self.model_t5(prompts, token_count=lg_out.shape[1])
+            lgt_out = torch.cat([lg_out, t5_out], dim=-2)
+
+        return {
+            'crossattn': lgt_out,
+            'vector': vector_out,
+        }
+
+    def before_load_weights(self, state_dict):
+        clip_path = os.path.join(shared.models_path, "CLIP")
+
+        if 'text_encoders.clip_g.transformer.text_model.embeddings.position_embedding.weight' not in state_dict:
+            clip_g_file = modelloader.load_file_from_url(CLIPG_URL, model_dir=clip_path, file_name="clip_g.safetensors")
+            with safetensors.safe_open(clip_g_file, framework="pt") as file:
+                self.clip_g.transformer.load_state_dict(SafetensorsMapping(file))
+
+        if 'text_encoders.clip_l.transformer.text_model.embeddings.position_embedding.weight' not in state_dict:
+            clip_l_file = modelloader.load_file_from_url(CLIPL_URL, model_dir=clip_path, file_name="clip_l.safetensors")
+            with safetensors.safe_open(clip_l_file, framework="pt") as file:
+                self.clip_l.transformer.load_state_dict(SafetensorsMapping(file), strict=False)
+
+        if self.t5xxl and 'text_encoders.t5xxl.transformer.encoder.embed_tokens.weight' not in state_dict:
+            t5_file = modelloader.load_file_from_url(T5_URL, model_dir=clip_path, file_name="t5xxl_fp16.safetensors")
+            with safetensors.safe_open(t5_file, framework="pt") as file:
+                self.t5xxl.transformer.load_state_dict(SafetensorsMapping(file), strict=False)
+
+    def encode_embedding_init_text(self, init_text, nvpt):
+        return self.model_lg.encode_embedding_init_text(init_text, nvpt)
+
+    def tokenize(self, texts):
+        return self.model_lg.tokenize(texts)
+
+    def medvram_modules(self):
+        return [self.clip_g, self.clip_l, self.t5xxl]
+
+    def get_token_count(self, text):
+        _, token_count = self.model_lg.process_texts([text])
+
+        return token_count
+
+    def get_target_prompt_token_count(self, token_count):
+        return self.model_lg.get_target_prompt_token_count(token_count)
diff --git a/modules/models/sd3/sd3_impls.py b/modules/models/sd3/sd3_impls.py
new file mode 100644
index 00000000000..59f11b2cbe1
--- /dev/null
+++ b/modules/models/sd3/sd3_impls.py
@@ -0,0 +1,374 @@
+### Impls of the SD3 core diffusion model and VAE
+
+import torch
+import math
+import einops
+from modules.models.sd3.mmdit import MMDiT
+from PIL import Image
+
+
+#################################################################################################
+### MMDiT Model Wrapping
+#################################################################################################
+
+
+class ModelSamplingDiscreteFlow(torch.nn.Module):
+    """Helper for sampler scheduling (ie timestep/sigma calculations) for Discrete Flow models"""
+    def __init__(self, shift=1.0):
+        super().__init__()
+        self.shift = shift
+        timesteps = 1000
+        ts = self.sigma(torch.arange(1, timesteps + 1, 1))
+        self.register_buffer('sigmas', ts)
+
+    @property
+    def sigma_min(self):
+        return self.sigmas[0]
+
+    @property
+    def sigma_max(self):
+        return self.sigmas[-1]
+
+    def timestep(self, sigma):
+        return sigma * 1000
+
+    def sigma(self, timestep: torch.Tensor):
+        timestep = timestep / 1000.0
+        if self.shift == 1.0:
+            return timestep
+        return self.shift * timestep / (1 + (self.shift - 1) * timestep)
+
+    def calculate_denoised(self, sigma, model_output, model_input):
+        sigma = sigma.view(sigma.shape[:1] + (1,) * (model_output.ndim - 1))
+        return model_input - model_output * sigma
+
+    def noise_scaling(self, sigma, noise, latent_image, max_denoise=False):
+        return sigma * noise + (1.0 - sigma) * latent_image
+
+
+class BaseModel(torch.nn.Module):
+    """Wrapper around the core MM-DiT model"""
+    def __init__(self, shift=1.0, device=None, dtype=torch.float32, state_dict=None, prefix=""):
+        super().__init__()
+        # Important configuration values can be quickly determined by checking shapes in the source file
+        # Some of these will vary between models (eg 2B vs 8B primarily differ in their depth, but also other details change)
+        patch_size = state_dict[f"{prefix}x_embedder.proj.weight"].shape[2]
+        depth = state_dict[f"{prefix}x_embedder.proj.weight"].shape[0] // 64
+        num_patches = state_dict[f"{prefix}pos_embed"].shape[1]
+        pos_embed_max_size = round(math.sqrt(num_patches))
+        adm_in_channels = state_dict[f"{prefix}y_embedder.mlp.0.weight"].shape[1]
+        context_shape = state_dict[f"{prefix}context_embedder.weight"].shape
+        context_embedder_config = {
+            "target": "torch.nn.Linear",
+            "params": {
+                "in_features": context_shape[1],
+                "out_features": context_shape[0]
+            }
+        }
+        self.diffusion_model = MMDiT(input_size=None, pos_embed_scaling_factor=None, pos_embed_offset=None, pos_embed_max_size=pos_embed_max_size, patch_size=patch_size, in_channels=16, depth=depth, num_patches=num_patches, adm_in_channels=adm_in_channels, context_embedder_config=context_embedder_config, device=device, dtype=dtype)
+        self.model_sampling = ModelSamplingDiscreteFlow(shift=shift)
+        self.depth = depth
+
+    def apply_model(self, x, sigma, c_crossattn=None, y=None):
+        dtype = self.get_dtype()
+        timestep = self.model_sampling.timestep(sigma).float()
+        model_output = self.diffusion_model(x.to(dtype), timestep, context=c_crossattn.to(dtype), y=y.to(dtype)).float()
+        return self.model_sampling.calculate_denoised(sigma, model_output, x)
+
+    def forward(self, *args, **kwargs):
+        return self.apply_model(*args, **kwargs)
+
+    def get_dtype(self):
+        return self.diffusion_model.dtype
+
+
+class CFGDenoiser(torch.nn.Module):
+    """Helper for applying CFG Scaling to diffusion outputs"""
+    def __init__(self, model):
+        super().__init__()
+        self.model = model
+
+    def forward(self, x, timestep, cond, uncond, cond_scale):
+        # Run cond and uncond in a batch together
+        batched = self.model.apply_model(torch.cat([x, x]), torch.cat([timestep, timestep]), c_crossattn=torch.cat([cond["c_crossattn"], uncond["c_crossattn"]]), y=torch.cat([cond["y"], uncond["y"]]))
+        # Then split and apply CFG Scaling
+        pos_out, neg_out = batched.chunk(2)
+        scaled = neg_out + (pos_out - neg_out) * cond_scale
+        return scaled
+
+
+class SD3LatentFormat:
+    """Latents are slightly shifted from center - this class must be called after VAE Decode to correct for the shift"""
+    def __init__(self):
+        self.scale_factor = 1.5305
+        self.shift_factor = 0.0609
+
+    def process_in(self, latent):
+        return (latent - self.shift_factor) * self.scale_factor
+
+    def process_out(self, latent):
+        return (latent / self.scale_factor) + self.shift_factor
+
+    def decode_latent_to_preview(self, x0):
+        """Quick RGB approximate preview of sd3 latents"""
+        factors = torch.tensor([
+            [-0.0645,  0.0177,  0.1052], [ 0.0028,  0.0312,  0.0650],
+            [ 0.1848,  0.0762,  0.0360], [ 0.0944,  0.0360,  0.0889],
+            [ 0.0897,  0.0506, -0.0364], [-0.0020,  0.1203,  0.0284],
+            [ 0.0855,  0.0118,  0.0283], [-0.0539,  0.0658,  0.1047],
+            [-0.0057,  0.0116,  0.0700], [-0.0412,  0.0281, -0.0039],
+            [ 0.1106,  0.1171,  0.1220], [-0.0248,  0.0682, -0.0481],
+            [ 0.0815,  0.0846,  0.1207], [-0.0120, -0.0055, -0.0867],
+            [-0.0749, -0.0634, -0.0456], [-0.1418, -0.1457, -0.1259]
+        ], device="cpu")
+        latent_image = x0[0].permute(1, 2, 0).cpu() @ factors
+
+        latents_ubyte = (((latent_image + 1) / 2)
+                            .clamp(0, 1)  # change scale from -1..1 to 0..1
+                            .mul(0xFF)  # to 0..255
+                            .byte()).cpu()
+
+        return Image.fromarray(latents_ubyte.numpy())
+
+
+#################################################################################################
+### K-Diffusion Sampling
+#################################################################################################
+
+
+def append_dims(x, target_dims):
+    """Appends dimensions to the end of a tensor until it has target_dims dimensions."""
+    dims_to_append = target_dims - x.ndim
+    return x[(...,) + (None,) * dims_to_append]
+
+
+def to_d(x, sigma, denoised):
+    """Converts a denoiser output to a Karras ODE derivative."""
+    return (x - denoised) / append_dims(sigma, x.ndim)
+
+
+@torch.no_grad()
+@torch.autocast("cuda", dtype=torch.float16)
+def sample_euler(model, x, sigmas, extra_args=None):
+    """Implements Algorithm 2 (Euler steps) from Karras et al. (2022)."""
+    extra_args = {} if extra_args is None else extra_args
+    s_in = x.new_ones([x.shape[0]])
+    for i in range(len(sigmas) - 1):
+        sigma_hat = sigmas[i]
+        denoised = model(x, sigma_hat * s_in, **extra_args)
+        d = to_d(x, sigma_hat, denoised)
+        dt = sigmas[i + 1] - sigma_hat
+        # Euler method
+        x = x + d * dt
+    return x
+
+
+#################################################################################################
+### VAE
+#################################################################################################
+
+
+def Normalize(in_channels, num_groups=32, dtype=torch.float32, device=None):
+    return torch.nn.GroupNorm(num_groups=num_groups, num_channels=in_channels, eps=1e-6, affine=True, dtype=dtype, device=device)
+
+
+class ResnetBlock(torch.nn.Module):
+    def __init__(self, *, in_channels, out_channels=None, dtype=torch.float32, device=None):
+        super().__init__()
+        self.in_channels = in_channels
+        out_channels = in_channels if out_channels is None else out_channels
+        self.out_channels = out_channels
+
+        self.norm1 = Normalize(in_channels, dtype=dtype, device=device)
+        self.conv1 = torch.nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1, padding=1, dtype=dtype, device=device)
+        self.norm2 = Normalize(out_channels, dtype=dtype, device=device)
+        self.conv2 = torch.nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=1, padding=1, dtype=dtype, device=device)
+        if self.in_channels != self.out_channels:
+            self.nin_shortcut = torch.nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, padding=0, dtype=dtype, device=device)
+        else:
+            self.nin_shortcut = None
+        self.swish = torch.nn.SiLU(inplace=True)
+
+    def forward(self, x):
+        hidden = x
+        hidden = self.norm1(hidden)
+        hidden = self.swish(hidden)
+        hidden = self.conv1(hidden)
+        hidden = self.norm2(hidden)
+        hidden = self.swish(hidden)
+        hidden = self.conv2(hidden)
+        if self.in_channels != self.out_channels:
+            x = self.nin_shortcut(x)
+        return x + hidden
+
+
+class AttnBlock(torch.nn.Module):
+    def __init__(self, in_channels, dtype=torch.float32, device=None):
+        super().__init__()
+        self.norm = Normalize(in_channels, dtype=dtype, device=device)
+        self.q = torch.nn.Conv2d(in_channels, in_channels, kernel_size=1, stride=1, padding=0, dtype=dtype, device=device)
+        self.k = torch.nn.Conv2d(in_channels, in_channels, kernel_size=1, stride=1, padding=0, dtype=dtype, device=device)
+        self.v = torch.nn.Conv2d(in_channels, in_channels, kernel_size=1, stride=1, padding=0, dtype=dtype, device=device)
+        self.proj_out = torch.nn.Conv2d(in_channels, in_channels, kernel_size=1, stride=1, padding=0, dtype=dtype, device=device)
+
+    def forward(self, x):
+        hidden = self.norm(x)
+        q = self.q(hidden)
+        k = self.k(hidden)
+        v = self.v(hidden)
+        b, c, h, w = q.shape
+        q, k, v = [einops.rearrange(x, "b c h w -> b 1 (h w) c").contiguous() for x in (q, k, v)]
+        hidden = torch.nn.functional.scaled_dot_product_attention(q, k, v)  # scale is dim ** -0.5 per default
+        hidden = einops.rearrange(hidden, "b 1 (h w) c -> b c h w", h=h, w=w, c=c, b=b)
+        hidden = self.proj_out(hidden)
+        return x + hidden
+
+
+class Downsample(torch.nn.Module):
+    def __init__(self, in_channels, dtype=torch.float32, device=None):
+        super().__init__()
+        self.conv = torch.nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=2, padding=0, dtype=dtype, device=device)
+
+    def forward(self, x):
+        pad = (0,1,0,1)
+        x = torch.nn.functional.pad(x, pad, mode="constant", value=0)
+        x = self.conv(x)
+        return x
+
+
+class Upsample(torch.nn.Module):
+    def __init__(self, in_channels, dtype=torch.float32, device=None):
+        super().__init__()
+        self.conv = torch.nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=1, padding=1, dtype=dtype, device=device)
+
+    def forward(self, x):
+        x = torch.nn.functional.interpolate(x, scale_factor=2.0, mode="nearest")
+        x = self.conv(x)
+        return x
+
+
+class VAEEncoder(torch.nn.Module):
+    def __init__(self, ch=128, ch_mult=(1,2,4,4), num_res_blocks=2, in_channels=3, z_channels=16, dtype=torch.float32, device=None):
+        super().__init__()
+        self.num_resolutions = len(ch_mult)
+        self.num_res_blocks = num_res_blocks
+        # downsampling
+        self.conv_in = torch.nn.Conv2d(in_channels, ch, kernel_size=3, stride=1, padding=1, dtype=dtype, device=device)
+        in_ch_mult = (1,) + tuple(ch_mult)
+        self.in_ch_mult = in_ch_mult
+        self.down = torch.nn.ModuleList()
+        for i_level in range(self.num_resolutions):
+            block = torch.nn.ModuleList()
+            attn = torch.nn.ModuleList()
+            block_in = ch*in_ch_mult[i_level]
+            block_out = ch*ch_mult[i_level]
+            for _ in range(num_res_blocks):
+                block.append(ResnetBlock(in_channels=block_in, out_channels=block_out, dtype=dtype, device=device))
+                block_in = block_out
+            down = torch.nn.Module()
+            down.block = block
+            down.attn = attn
+            if i_level != self.num_resolutions - 1:
+                down.downsample = Downsample(block_in, dtype=dtype, device=device)
+            self.down.append(down)
+        # middle
+        self.mid = torch.nn.Module()
+        self.mid.block_1 = ResnetBlock(in_channels=block_in, out_channels=block_in, dtype=dtype, device=device)
+        self.mid.attn_1 = AttnBlock(block_in, dtype=dtype, device=device)
+        self.mid.block_2 = ResnetBlock(in_channels=block_in, out_channels=block_in, dtype=dtype, device=device)
+        # end
+        self.norm_out = Normalize(block_in, dtype=dtype, device=device)
+        self.conv_out = torch.nn.Conv2d(block_in, 2 * z_channels, kernel_size=3, stride=1, padding=1, dtype=dtype, device=device)
+        self.swish = torch.nn.SiLU(inplace=True)
+
+    def forward(self, x):
+        # downsampling
+        hs = [self.conv_in(x)]
+        for i_level in range(self.num_resolutions):
+            for i_block in range(self.num_res_blocks):
+                h = self.down[i_level].block[i_block](hs[-1])
+                hs.append(h)
+            if i_level != self.num_resolutions-1:
+                hs.append(self.down[i_level].downsample(hs[-1]))
+        # middle
+        h = hs[-1]
+        h = self.mid.block_1(h)
+        h = self.mid.attn_1(h)
+        h = self.mid.block_2(h)
+        # end
+        h = self.norm_out(h)
+        h = self.swish(h)
+        h = self.conv_out(h)
+        return h
+
+
+class VAEDecoder(torch.nn.Module):
+    def __init__(self, ch=128, out_ch=3, ch_mult=(1, 2, 4, 4), num_res_blocks=2, resolution=256, z_channels=16, dtype=torch.float32, device=None):
+        super().__init__()
+        self.num_resolutions = len(ch_mult)
+        self.num_res_blocks = num_res_blocks
+        block_in = ch * ch_mult[self.num_resolutions - 1]
+        curr_res = resolution // 2 ** (self.num_resolutions - 1)
+        # z to block_in
+        self.conv_in = torch.nn.Conv2d(z_channels, block_in, kernel_size=3, stride=1, padding=1, dtype=dtype, device=device)
+        # middle
+        self.mid = torch.nn.Module()
+        self.mid.block_1 = ResnetBlock(in_channels=block_in, out_channels=block_in, dtype=dtype, device=device)
+        self.mid.attn_1 = AttnBlock(block_in, dtype=dtype, device=device)
+        self.mid.block_2 = ResnetBlock(in_channels=block_in, out_channels=block_in, dtype=dtype, device=device)
+        # upsampling
+        self.up = torch.nn.ModuleList()
+        for i_level in reversed(range(self.num_resolutions)):
+            block = torch.nn.ModuleList()
+            block_out = ch * ch_mult[i_level]
+            for _ in range(self.num_res_blocks + 1):
+                block.append(ResnetBlock(in_channels=block_in, out_channels=block_out, dtype=dtype, device=device))
+                block_in = block_out
+            up = torch.nn.Module()
+            up.block = block
+            if i_level != 0:
+                up.upsample = Upsample(block_in, dtype=dtype, device=device)
+                curr_res = curr_res * 2
+            self.up.insert(0, up) # prepend to get consistent order
+        # end
+        self.norm_out = Normalize(block_in, dtype=dtype, device=device)
+        self.conv_out = torch.nn.Conv2d(block_in, out_ch, kernel_size=3, stride=1, padding=1, dtype=dtype, device=device)
+        self.swish = torch.nn.SiLU(inplace=True)
+
+    def forward(self, z):
+        # z to block_in
+        hidden = self.conv_in(z)
+        # middle
+        hidden = self.mid.block_1(hidden)
+        hidden = self.mid.attn_1(hidden)
+        hidden = self.mid.block_2(hidden)
+        # upsampling
+        for i_level in reversed(range(self.num_resolutions)):
+            for i_block in range(self.num_res_blocks + 1):
+                hidden = self.up[i_level].block[i_block](hidden)
+            if i_level != 0:
+                hidden = self.up[i_level].upsample(hidden)
+        # end
+        hidden = self.norm_out(hidden)
+        hidden = self.swish(hidden)
+        hidden = self.conv_out(hidden)
+        return hidden
+
+
+class SDVAE(torch.nn.Module):
+    def __init__(self, dtype=torch.float32, device=None):
+        super().__init__()
+        self.encoder = VAEEncoder(dtype=dtype, device=device)
+        self.decoder = VAEDecoder(dtype=dtype, device=device)
+
+    @torch.autocast("cuda", dtype=torch.float16)
+    def decode(self, latent):
+        return self.decoder(latent)
+
+    @torch.autocast("cuda", dtype=torch.float16)
+    def encode(self, image):
+        hidden = self.encoder(image)
+        mean, logvar = torch.chunk(hidden, 2, dim=1)
+        logvar = torch.clamp(logvar, -30.0, 20.0)
+        std = torch.exp(0.5 * logvar)
+        return mean + std * torch.randn_like(mean)
diff --git a/modules/models/sd3/sd3_model.py b/modules/models/sd3/sd3_model.py
new file mode 100644
index 00000000000..37cf85eb36f
--- /dev/null
+++ b/modules/models/sd3/sd3_model.py
@@ -0,0 +1,96 @@
+import contextlib
+
+import torch
+
+import k_diffusion
+from modules.models.sd3.sd3_impls import BaseModel, SDVAE, SD3LatentFormat
+from modules.models.sd3.sd3_cond import SD3Cond
+
+from modules import shared, devices
+
+
+class SD3Denoiser(k_diffusion.external.DiscreteSchedule):
+    def __init__(self, inner_model, sigmas):
+        super().__init__(sigmas, quantize=shared.opts.enable_quantization)
+        self.inner_model = inner_model
+
+    def forward(self, input, sigma, **kwargs):
+        return self.inner_model.apply_model(input, sigma, **kwargs)
+
+
+class SD3Inferencer(torch.nn.Module):
+    def __init__(self, state_dict, shift=3, use_ema=False):
+        super().__init__()
+
+        self.shift = shift
+
+        with torch.no_grad():
+            self.model = BaseModel(shift=shift, state_dict=state_dict, prefix="model.diffusion_model.", device="cpu", dtype=devices.dtype)
+            self.first_stage_model = SDVAE(device="cpu", dtype=devices.dtype_vae)
+            self.first_stage_model.dtype = self.model.diffusion_model.dtype
+
+        self.alphas_cumprod = 1 / (self.model.model_sampling.sigmas ** 2 + 1)
+
+        self.text_encoders = SD3Cond()
+        self.cond_stage_key = 'txt'
+
+        self.parameterization = "eps"
+        self.model.conditioning_key = "crossattn"
+
+        self.latent_format = SD3LatentFormat()
+        self.latent_channels = 16
+
+    @property
+    def cond_stage_model(self):
+        return self.text_encoders
+
+    def before_load_weights(self, state_dict):
+        self.cond_stage_model.before_load_weights(state_dict)
+
+    def ema_scope(self):
+        return contextlib.nullcontext()
+
+    def get_learned_conditioning(self, batch: list[str]):
+        return self.cond_stage_model(batch)
+
+    def apply_model(self, x, t, cond):
+        return self.model(x, t, c_crossattn=cond['crossattn'], y=cond['vector'])
+
+    def decode_first_stage(self, latent):
+        latent = self.latent_format.process_out(latent)
+        return self.first_stage_model.decode(latent)
+
+    def encode_first_stage(self, image):
+        latent = self.first_stage_model.encode(image)
+        return self.latent_format.process_in(latent)
+
+    def get_first_stage_encoding(self, x):
+        return x
+
+    def create_denoiser(self):
+        return SD3Denoiser(self, self.model.model_sampling.sigmas)
+
+    def medvram_fields(self):
+        return [
+            (self, 'first_stage_model'),
+            (self, 'text_encoders'),
+            (self, 'model'),
+        ]
+
+    def add_noise_to_latent(self, x, noise, amount):
+        return x * (1 - amount) + noise * amount
+
+    def fix_dimensions(self, width, height):
+        return width // 16 * 16, height // 16 * 16
+
+    def diffusers_weight_mapping(self):
+        for i in range(self.model.depth):
+            yield f"transformer.transformer_blocks.{i}.attn.to_q", f"diffusion_model_joint_blocks_{i}_x_block_attn_qkv_q_proj"
+            yield f"transformer.transformer_blocks.{i}.attn.to_k", f"diffusion_model_joint_blocks_{i}_x_block_attn_qkv_k_proj"
+            yield f"transformer.transformer_blocks.{i}.attn.to_v", f"diffusion_model_joint_blocks_{i}_x_block_attn_qkv_v_proj"
+            yield f"transformer.transformer_blocks.{i}.attn.to_out.0", f"diffusion_model_joint_blocks_{i}_x_block_attn_proj"
+
+            yield f"transformer.transformer_blocks.{i}.attn.add_q_proj", f"diffusion_model_joint_blocks_{i}_context_block.attn_qkv_q_proj"
+            yield f"transformer.transformer_blocks.{i}.attn.add_k_proj", f"diffusion_model_joint_blocks_{i}_context_block.attn_qkv_k_proj"
+            yield f"transformer.transformer_blocks.{i}.attn.add_v_proj", f"diffusion_model_joint_blocks_{i}_context_block.attn_qkv_v_proj"
+            yield f"transformer.transformer_blocks.{i}.attn.add_out_proj.0", f"diffusion_model_joint_blocks_{i}_context_block_attn_proj"
diff --git a/modules/npu_specific.py b/modules/npu_specific.py
new file mode 100644
index 00000000000..9410069110f
--- /dev/null
+++ b/modules/npu_specific.py
@@ -0,0 +1,31 @@
+import importlib
+import torch
+
+from modules import shared
+
+
+def check_for_npu():
+    if importlib.util.find_spec("torch_npu") is None:
+        return False
+    import torch_npu
+
+    try:
+        # Will raise a RuntimeError if no NPU is found
+        _ = torch_npu.npu.device_count()
+        return torch.npu.is_available()
+    except RuntimeError:
+        return False
+
+
+def get_npu_device_string():
+    if shared.cmd_opts.device_id is not None:
+        return f"npu:{shared.cmd_opts.device_id}"
+    return "npu:0"
+
+
+def torch_npu_gc():
+    with torch.npu.device(get_npu_device_string()):
+        torch.npu.empty_cache()
+
+
+has_npu = check_for_npu()
diff --git a/modules/options.py b/modules/options.py
index 758b1ce5f24..2a78a825ee2 100644
--- a/modules/options.py
+++ b/modules/options.py
@@ -1,20 +1,24 @@
+import os
 import json
 import sys
+from dataclasses import dataclass
 
 import gradio as gr
 
 from modules import errors
 from modules.shared_cmd_options import cmd_opts
+from modules.paths_internal import script_path
 
 
 class OptionInfo:
-    def __init__(self, default=None, label="", component=None, component_args=None, onchange=None, section=None, refresh=None, comment_before='', comment_after='', infotext=None, restrict_api=False):
+    def __init__(self, default=None, label="", component=None, component_args=None, onchange=None, section=None, refresh=None, comment_before='', comment_after='', infotext=None, restrict_api=False, category_id=None):
         self.default = default
         self.label = label
         self.component = component
         self.component_args = component_args
         self.onchange = onchange
         self.section = section
+        self.category_id = category_id
         self.refresh = refresh
         self.do_not_save = False
 
@@ -63,7 +67,11 @@ def __init__(self, text):
 
 def options_section(section_identifier, options_dict):
     for v in options_dict.values():
-        v.section = section_identifier
+        if len(section_identifier) == 2:
+            v.section = section_identifier
+        elif len(section_identifier) == 3:
+            v.section = section_identifier[0:2]
+            v.category_id = section_identifier[2]
 
     return options_dict
 
@@ -76,7 +84,7 @@ class Options:
 
     def __init__(self, data_labels: dict[str, OptionInfo], restricted_opts):
         self.data_labels = data_labels
-        self.data = {k: v.default for k, v in self.data_labels.items()}
+        self.data = {k: v.default for k, v in self.data_labels.items() if not v.do_not_save}
         self.restricted_opts = restricted_opts
 
     def __setattr__(self, key, value):
@@ -85,18 +93,35 @@ def __setattr__(self, key, value):
 
         if self.data is not None:
             if key in self.data or key in self.data_labels:
+
+                # Check that settings aren't globally frozen
                 assert not cmd_opts.freeze_settings, "changing settings is disabled"
 
+                # Get the info related to the setting being changed
                 info = self.data_labels.get(key, None)
                 if info.do_not_save:
                     return
 
+                # Restrict component arguments
                 comp_args = info.component_args if info else None
                 if isinstance(comp_args, dict) and comp_args.get('visible', True) is False:
-                    raise RuntimeError(f"not possible to set {key} because it is restricted")
+                    raise RuntimeError(f"not possible to set '{key}' because it is restricted")
+
+                # Check that this section isn't frozen
+                if cmd_opts.freeze_settings_in_sections is not None:
+                    frozen_sections = list(map(str.strip, cmd_opts.freeze_settings_in_sections.split(','))) # Trim whitespace from section names
+                    section_key = info.section[0]
+                    section_name = info.section[1]
+                    assert section_key not in frozen_sections, f"not possible to set '{key}' because settings in section '{section_name}' ({section_key}) are frozen with --freeze-settings-in-sections"
 
+                # Check that this section of the settings isn't frozen
+                if cmd_opts.freeze_specific_settings is not None:
+                    frozen_keys = list(map(str.strip, cmd_opts.freeze_specific_settings.split(','))) # Trim whitespace from setting keys
+                    assert key not in frozen_keys, f"not possible to set '{key}' because this setting is frozen with --freeze-specific-settings"
+
+                # Check shorthand option which disables editing options in "saving-paths"
                 if cmd_opts.hide_ui_dir_config and key in self.restricted_opts:
-                    raise RuntimeError(f"not possible to set {key} because it is restricted")
+                    raise RuntimeError(f"not possible to set '{key}' because it is restricted with --hide_ui_dir_config")
 
                 self.data[key] = value
                 return
@@ -158,7 +183,7 @@ def save(self, filename):
         assert not cmd_opts.freeze_settings, "saving settings is disabled"
 
         with open(filename, "w", encoding="utf8") as file:
-            json.dump(self.data, file, indent=4)
+            json.dump(self.data, file, indent=4, ensure_ascii=False)
 
     def same_type(self, x, y):
         if x is None or y is None:
@@ -170,9 +195,15 @@ def same_type(self, x, y):
         return type_x == type_y
 
     def load(self, filename):
-        with open(filename, "r", encoding="utf8") as file:
-            self.data = json.load(file)
-
+        try:
+            with open(filename, "r", encoding="utf8") as file:
+                self.data = json.load(file)
+        except FileNotFoundError:
+            self.data = {}
+        except Exception:
+            errors.report(f'\nCould not load settings\nThe config file "{filename}" is likely corrupted\nIt has been moved to the "tmp/config.json"\nReverting config to default\n\n''', exc_info=True)
+            os.replace(filename, os.path.join(script_path, "tmp", "config.json"))
+            self.data = {}
         # 1.6.0 VAE defaults
         if self.data.get('sd_vae_as_default') is not None and self.data.get('sd_vae_overrides_per_model_preferences') is None:
             self.data['sd_vae_overrides_per_model_preferences'] = not self.data.get('sd_vae_as_default')
@@ -206,21 +237,62 @@ def dumpjson(self):
         d = {k: self.data.get(k, v.default) for k, v in self.data_labels.items()}
         d["_comments_before"] = {k: v.comment_before for k, v in self.data_labels.items() if v.comment_before is not None}
         d["_comments_after"] = {k: v.comment_after for k, v in self.data_labels.items() if v.comment_after is not None}
+
+        item_categories = {}
+        for item in self.data_labels.values():
+            if item.section[0] is None:
+                continue
+
+            category = categories.mapping.get(item.category_id)
+            category = "Uncategorized" if category is None else category.label
+            if category not in item_categories:
+                item_categories[category] = item.section[1]
+
+        # _categories is a list of pairs: [section, category]. Each section (a setting page) will get a special heading above it with the category as text.
+        d["_categories"] = [[v, k] for k, v in item_categories.items()] + [["Defaults", "Other"]]
+
         return json.dumps(d)
 
     def add_option(self, key, info):
         self.data_labels[key] = info
+        if key not in self.data and not info.do_not_save:
+            self.data[key] = info.default
 
     def reorder(self):
-        """reorder settings so that all items related to section always go together"""
+        """Reorder settings so that:
+            - all items related to section always go together
+            - all sections belonging to a category go together
+            - sections inside a category are ordered alphabetically
+            - categories are ordered by creation order
+
+        Category is a superset of sections: for category "postprocessing" there could be multiple sections: "face restoration", "upscaling".
+
+        This function also changes items' category_id so that all items belonging to a section have the same category_id.
+        """
+
+        category_ids = {}
+        section_categories = {}
 
-        section_ids = {}
         settings_items = self.data_labels.items()
         for _, item in settings_items:
-            if item.section not in section_ids:
-                section_ids[item.section] = len(section_ids)
+            if item.section not in section_categories:
+                section_categories[item.section] = item.category_id
 
-        self.data_labels = dict(sorted(settings_items, key=lambda x: section_ids[x[1].section]))
+        for _, item in settings_items:
+            item.category_id = section_categories.get(item.section)
+
+        for category_id in categories.mapping:
+            if category_id not in category_ids:
+                category_ids[category_id] = len(category_ids)
+
+        def sort_key(x):
+            item: OptionInfo = x[1]
+            category_order = category_ids.get(item.category_id, len(category_ids))
+            section_order = item.section[1]
+
+            return category_order, section_order
+
+        self.data_labels = dict(sorted(settings_items, key=sort_key))
 
     def cast_value(self, key, value):
         """casts an arbitrary to the same type as this setting's value with key
@@ -243,3 +315,22 @@ def cast_value(self, key, value):
             value = expected_type(value)
 
         return value
+
+
+@dataclass
+class OptionsCategory:
+    id: str
+    label: str
+
+class OptionsCategories:
+    def __init__(self):
+        self.mapping = {}
+
+    def register_category(self, category_id, label):
+        if category_id in self.mapping:
+            return category_id
+
+        self.mapping[category_id] = OptionsCategory(category_id, label)
+
+
+categories = OptionsCategories()
diff --git a/modules/paths.py b/modules/paths.py
index 2505233999b..030646519c3 100644
--- a/modules/paths.py
+++ b/modules/paths.py
@@ -1,6 +1,6 @@
 import os
 import sys
-from modules.paths_internal import models_path, script_path, data_path, extensions_dir, extensions_builtin_dir  # noqa: F401
+from modules.paths_internal import models_path, script_path, data_path, extensions_dir, extensions_builtin_dir, cwd  # noqa: F401
 
 import modules.safe  # noqa: F401
 
@@ -38,7 +38,6 @@ class Dummy:
 path_dirs = [
     (sd_path, 'ldm', 'Stable Diffusion', []),
     (os.path.join(sd_path, '../generative-models'), 'sgm', 'Stable Diffusion XL', ["sgm"]),
-    (os.path.join(sd_path, '../CodeFormer'), 'inference_codeformer.py', 'CodeFormer', []),
     (os.path.join(sd_path, '../BLIP'), 'models/blip.py', 'BLIP', []),
     (os.path.join(sd_path, '../k-diffusion'), 'k_diffusion/sampling.py', 'k_diffusion', ["atstart"]),
 ]
diff --git a/modules/paths_internal.py b/modules/paths_internal.py
index 005a9b0aa75..67521f5cd5d 100644
--- a/modules/paths_internal.py
+++ b/modules/paths_internal.py
@@ -4,10 +4,15 @@
 import os
 import sys
 import shlex
+from pathlib import Path
+
+
+normalized_filepath = lambda filepath: str(Path(filepath).absolute())
 
 commandline_args = os.environ.get('COMMANDLINE_ARGS', "")
 sys.argv += shlex.split(commandline_args)
 
+cwd = os.getcwd()
 modules_path = os.path.dirname(os.path.realpath(__file__))
 script_path = os.path.dirname(modules_path)
 
@@ -19,13 +24,15 @@
 # Parse the --data-dir flag first so we can use it as a base for our other argument default values
 parser_pre = argparse.ArgumentParser(add_help=False)
 parser_pre.add_argument("--data-dir", type=str, default=os.path.dirname(modules_path), help="base path where all user data is stored", )
+parser_pre.add_argument("--models-dir", type=str, default=None, help="base path where models are stored; overrides --data-dir", )
 cmd_opts_pre = parser_pre.parse_known_args()[0]
 
 data_path = cmd_opts_pre.data_dir
 
-models_path = os.path.join(data_path, "models")
+models_path = cmd_opts_pre.models_dir if cmd_opts_pre.models_dir else os.path.join(data_path, "models")
 extensions_dir = os.path.join(data_path, "extensions")
 extensions_builtin_dir = os.path.join(script_path, "extensions-builtin")
 config_states_dir = os.path.join(script_path, "config_states")
+default_output_dir = os.path.join(data_path, "outputs")
 
 roboto_ttf_file = os.path.join(modules_path, 'Roboto-Regular.ttf')
diff --git a/modules/postprocessing.py b/modules/postprocessing.py
index cf04d38b059..a413d1027c7 100644
--- a/modules/postprocessing.py
+++ b/modules/postprocessing.py
@@ -2,7 +2,7 @@
 
 from PIL import Image
 
-from modules import shared, images, devices, scripts, scripts_postprocessing, ui_common, generation_parameters_copypaste
+from modules import shared, images, devices, scripts, scripts_postprocessing, ui_common, infotext_utils
 from modules.shared import opts
 
 
@@ -17,10 +17,10 @@ def get_images(extras_mode, image, image_folder, input_dir):
         if extras_mode == 1:
             for img in image_folder:
                 if isinstance(img, Image.Image):
-                    image = img
+                    image = images.fix_image(img)
                     fn = ''
                 else:
-                    image = Image.open(os.path.abspath(img.name))
+                    image = images.read(os.path.abspath(img.name))
                     fn = os.path.splitext(img.orig_name)[0]
                 yield image, fn
         elif extras_mode == 2:
@@ -29,11 +29,7 @@ def get_images(extras_mode, image, image_folder, input_dir):
 
             image_list = shared.listfiles(input_dir)
             for filename in image_list:
-                try:
-                    image = Image.open(filename)
-                except Exception:
-                    continue
-                yield image, filename
+                yield filename, filename
         else:
             assert image, 'image not selected'
             yield image, None
@@ -45,50 +41,107 @@ def get_images(extras_mode, image, image_folder, input_dir):
 
     infotext = ''
 
-    for image_data, name in get_images(extras_mode, image, image_folder, input_dir):
+    data_to_process = list(get_images(extras_mode, image, image_folder, input_dir))
+    shared.state.job_count = len(data_to_process)
+
+    for image_placeholder, name in data_to_process:
         image_data: Image.Image
 
+        shared.state.nextjob()
         shared.state.textinfo = name
+        shared.state.skipped = False
+
+        if shared.state.interrupted or shared.state.stopping_generation:
+            break
+
+        if isinstance(image_placeholder, str):
+            try:
+                image_data = images.read(image_placeholder)
+            except Exception:
+                continue
+        else:
+            image_data = image_placeholder
+
+        image_data = image_data if image_data.mode in ("RGBA", "RGB") else image_data.convert("RGB")
 
         parameters, existing_pnginfo = images.read_info_from_image(image_data)
         if parameters:
             existing_pnginfo["parameters"] = parameters
 
-        pp = scripts_postprocessing.PostprocessedImage(image_data.convert("RGB"))
+        initial_pp = scripts_postprocessing.PostprocessedImage(image_data)
 
-        scripts.scripts_postproc.run(pp, args)
+        scripts.scripts_postproc.run(initial_pp, args)
 
-        if opts.use_original_name_batch and name is not None:
-            basename = os.path.splitext(os.path.basename(name))[0]
-        else:
-            basename = ''
+        if shared.state.skipped:
+            continue
 
-        infotext = ", ".join([k if k == v else f'{k}: {generation_parameters_copypaste.quote(v)}' for k, v in pp.info.items() if v is not None])
+        used_suffixes = {}
+        for pp in [initial_pp, *initial_pp.extra_images]:
+            suffix = pp.get_suffix(used_suffixes)
 
-        if opts.enable_pnginfo:
-            pp.image.info = existing_pnginfo
-            pp.image.info["postprocessing"] = infotext
+            if opts.use_original_name_batch and name is not None:
+                basename = os.path.splitext(os.path.basename(name))[0]
+                forced_filename = basename + suffix
+            else:
+                basename = ''
+                forced_filename = None
 
-        if save_output:
-            images.save_image(pp.image, path=outpath, basename=basename, seed=None, prompt=None, extension=opts.samples_format, info=infotext, short_filename=True, no_prompt=True, grid=False, pnginfo_section_name="extras", existing_info=existing_pnginfo, forced_filename=None)
+            infotext = ", ".join([k if k == v else f'{k}: {infotext_utils.quote(v)}' for k, v in pp.info.items() if v is not None])
 
-        if extras_mode != 2 or show_extras_results:
-            outputs.append(pp.image)
+            if opts.enable_pnginfo:
+                pp.image.info = existing_pnginfo
+                pp.image.info["postprocessing"] = infotext
 
-        image_data.close()
+            shared.state.assign_current_image(pp.image)
 
-    devices.torch_gc()
+            if save_output:
+                fullfn, _ = images.save_image(pp.image, path=outpath, basename=basename, extension=opts.samples_format, info=infotext, short_filename=True, no_prompt=True, grid=False, pnginfo_section_name="extras", existing_info=existing_pnginfo, forced_filename=forced_filename, suffix=suffix)
 
+                if pp.caption:
+                    caption_filename = os.path.splitext(fullfn)[0] + ".txt"
+                    existing_caption = ""
+                    try:
+                        with open(caption_filename, encoding="utf8") as file:
+                            existing_caption = file.read().strip()
+                    except FileNotFoundError:
+                        pass
+
+                    action = shared.opts.postprocessing_existing_caption_action
+                    if action == 'Prepend' and existing_caption:
+                        caption = f"{existing_caption} {pp.caption}"
+                    elif action == 'Append' and existing_caption:
+                        caption = f"{pp.caption} {existing_caption}"
+                    elif action == 'Keep' and existing_caption:
+                        caption = existing_caption
+                    else:
+                        caption = pp.caption
+
+                    caption = caption.strip()
+                    if caption:
+                        with open(caption_filename, "w", encoding="utf8") as file:
+                            file.write(caption)
+
+            if extras_mode != 2 or show_extras_results:
+                outputs.append(pp.image)
+
+    devices.torch_gc()
+    shared.state.end()
     return outputs, ui_common.plaintext_to_html(infotext), ''
 
 
-def run_extras(extras_mode, resize_mode, image, image_folder, input_dir, output_dir, show_extras_results, gfpgan_visibility, codeformer_visibility, codeformer_weight, upscaling_resize, upscaling_resize_w, upscaling_resize_h, upscaling_crop, extras_upscaler_1, extras_upscaler_2, extras_upscaler_2_visibility, upscale_first: bool, save_output: bool = True):
+def run_postprocessing_webui(id_task, *args, **kwargs):
+    return run_postprocessing(*args, **kwargs)
+
+
+def run_extras(extras_mode, resize_mode, image, image_folder, input_dir, output_dir, show_extras_results, gfpgan_visibility, codeformer_visibility, codeformer_weight, upscaling_resize, upscaling_resize_w, upscaling_resize_h, upscaling_crop, extras_upscaler_1, extras_upscaler_2, extras_upscaler_2_visibility, upscale_first: bool, save_output: bool = True, max_side_length: int = 0):
     """old handler for API"""
 
     args = scripts.scripts_postproc.create_args_for_run({
         "Upscale": {
+            "upscale_enabled": True,
             "upscale_mode": resize_mode,
             "upscale_by": upscaling_resize,
+            "max_side_length": max_side_length,
             "upscale_to_width": upscaling_resize_w,
             "upscale_to_height": upscaling_resize_h,
             "upscale_crop": upscaling_crop,
@@ -97,9 +150,11 @@ def run_extras(extras_mode, resize_mode, image, image_folder, input_dir, output_
             "upscaler_2_visibility": extras_upscaler_2_visibility,
         },
         "GFPGAN": {
+            "enable": True,
             "gfpgan_visibility": gfpgan_visibility,
         },
         "CodeFormer": {
+            "enable": True,
             "codeformer_visibility": codeformer_visibility,
             "codeformer_weight": codeformer_weight,
         },
diff --git a/modules/processing.py b/modules/processing.py
index e124e7f0dd2..7535b56e18c 100644
--- a/modules/processing.py
+++ b/modules/processing.py
@@ -16,7 +16,7 @@
 from typing import Any
 
 import modules.sd_hijack
-from modules import devices, prompt_parser, masking, sd_samplers, lowvram, generation_parameters_copypaste, extra_networks, sd_vae_approx, scripts, sd_samplers_common, sd_unet, errors, rng
+from modules import devices, prompt_parser, masking, sd_samplers, lowvram, infotext_utils, extra_networks, sd_vae_approx, scripts, sd_samplers_common, sd_unet, errors, rng, profiling
 from modules.rng import slerp # noqa: F401
 from modules.sd_hijack import model_hijack
 from modules.sd_samplers_common import images_tensor_to_samples, decode_first_stage, approximation_indexes
@@ -62,28 +62,37 @@ def apply_color_correction(correction, original_image):
     return image.convert('RGB')
 
 
-def apply_overlay(image, paste_loc, index, overlays):
-    if overlays is None or index >= len(overlays):
-        return image
+def uncrop(image, dest_size, paste_loc):
+    x, y, w, h = paste_loc
+    base_image = Image.new('RGBA', dest_size)
+    image = images.resize_image(1, image, w, h)
+    base_image.paste(image, (x, y))
+    image = base_image
+
+    return image
 
-    overlay = overlays[index]
+
+def apply_overlay(image, paste_loc, overlay):
+    if overlay is None:
+        return image, image.copy()
 
     if paste_loc is not None:
-        x, y, w, h = paste_loc
-        base_image = Image.new('RGBA', (overlay.width, overlay.height))
-        image = images.resize_image(1, image, w, h)
-        base_image.paste(image, (x, y))
-        image = base_image
+        image = uncrop(image, (overlay.width, overlay.height), paste_loc)
+
+    original_denoised_image = image.copy()
 
     image = image.convert('RGBA')
     image.alpha_composite(overlay)
     image = image.convert('RGB')
 
-    return image
+    return image, original_denoised_image
 
-def create_binary_mask(image):
+def create_binary_mask(image, round=True):
     if image.mode == 'RGBA' and image.getextrema()[-1] != (255, 255):
-        image = image.split()[-1].convert("L").point(lambda x: 255 if x > 128 else 0)
+        if round:
+            image = image.split()[-1].convert("L").point(lambda x: 255 if x > 128 else 0)
+        else:
+            image = image.split()[-1].convert("L")
     else:
         image = image.convert('L')
     return image
@@ -106,6 +115,18 @@ def txt2img_image_conditioning(sd_model, x, width, height):
         return x.new_zeros(x.shape[0], 2*sd_model.noise_augmentor.time_embed.dim, dtype=x.dtype, device=x.device)
 
     else:
+        if sd_model.is_sdxl_inpaint:
+            # The "masked-image" in this case will just be all 0.5 since the entire image is masked.
+            image_conditioning = torch.ones(x.shape[0], 3, height, width, device=x.device) * 0.5
+            image_conditioning = images_tensor_to_samples(image_conditioning,
+                                                            approximation_indexes.get(opts.sd_vae_encode_method))
+
+            # Add the fake full 1s mask to the first dimension.
+            image_conditioning = torch.nn.functional.pad(image_conditioning, (0, 0, 0, 0, 1, 0), value=1.0)
+            image_conditioning = image_conditioning.to(x.dtype)
+
+            return image_conditioning
+
         # Dummy zero conditioning if we're not using inpainting or unclip models.
         # Still takes up a bit of memory, but no encoder call.
         # Pretty sure we can just make this a 1x1 image since its not going to be used besides its batch size.
@@ -128,6 +149,7 @@ class StableDiffusionProcessing:
     seed_resize_from_w: int = -1
     seed_enable_extras: bool = True
     sampler_name: str = None
+    scheduler: str = None
     batch_size: int = 1
     n_iter: int = 1
     steps: int = 50
@@ -142,7 +164,7 @@ class StableDiffusionProcessing:
     overlay_images: list = None
     eta: float = None
     do_not_reload_embeddings: bool = False
-    denoising_strength: float = 0
+    denoising_strength: float = None
     ddim_discretize: str = None
     s_min_uncond: float = None
     s_churn: float = None
@@ -157,6 +179,7 @@ class StableDiffusionProcessing:
     token_merging_ratio = 0
     token_merging_ratio_hr = 0
     disable_extra_networks: bool = False
+    firstpass_image: Image = None
 
     scripts_value: scripts.ScriptRunner = field(default=None, init=False)
     script_args_value: list = field(default=None, init=False)
@@ -212,11 +235,6 @@ def __post_init__(self):
             self.styles = []
 
         self.sampler_noise_scheduler_override = None
-        self.s_min_uncond = self.s_min_uncond if self.s_min_uncond is not None else opts.s_min_uncond
-        self.s_churn = self.s_churn if self.s_churn is not None else opts.s_churn
-        self.s_tmin = self.s_tmin if self.s_tmin is not None else opts.s_tmin
-        self.s_tmax = (self.s_tmax if self.s_tmax is not None else opts.s_tmax) or float('inf')
-        self.s_noise = self.s_noise if self.s_noise is not None else opts.s_noise
 
         self.extra_generation_params = self.extra_generation_params or {}
         self.override_settings = self.override_settings or {}
@@ -233,6 +251,13 @@ def __post_init__(self):
         self.cached_uc = StableDiffusionProcessing.cached_uc
         self.cached_c = StableDiffusionProcessing.cached_c
 
+    def fill_fields_from_opts(self):
+        self.s_min_uncond = self.s_min_uncond if self.s_min_uncond is not None else opts.s_min_uncond
+        self.s_churn = self.s_churn if self.s_churn is not None else opts.s_churn
+        self.s_tmin = self.s_tmin if self.s_tmin is not None else opts.s_tmin
+        self.s_tmax = (self.s_tmax if self.s_tmax is not None else opts.s_tmax) or float('inf')
+        self.s_noise = self.s_noise if self.s_noise is not None else opts.s_noise
+
     @property
     def sd_model(self):
         return shared.sd_model
@@ -296,7 +321,7 @@ def depth2img_image_conditioning(self, source_image):
         return conditioning
 
     def edit_image_conditioning(self, source_image):
-        conditioning_image = images_tensor_to_samples(source_image*0.5+0.5, approximation_indexes.get(opts.sd_vae_encode_method))
+        conditioning_image = shared.sd_model.encode_first_stage(source_image).mode()
 
         return conditioning_image
 
@@ -308,7 +333,7 @@ def unclip_image_conditioning(self, source_image):
             c_adm = torch.cat((c_adm, noise_level_emb), 1)
         return c_adm
 
-    def inpainting_image_conditioning(self, source_image, latent_image, image_mask=None):
+    def inpainting_image_conditioning(self, source_image, latent_image, image_mask=None, round_image_mask=True):
         self.is_using_inpainting_conditioning = True
 
         # Handle the different mask inputs
@@ -320,8 +345,10 @@ def inpainting_image_conditioning(self, source_image, latent_image, image_mask=N
                 conditioning_mask = conditioning_mask.astype(np.float32) / 255.0
                 conditioning_mask = torch.from_numpy(conditioning_mask[None, None])
 
-                # Inpainting model uses a discretized mask as input, so we round to either 1.0 or 0.0
-                conditioning_mask = torch.round(conditioning_mask)
+                if round_image_mask:
+                    # Caller is requesting a discretized mask as input, so we round to either 1.0 or 0.0
+                    conditioning_mask = torch.round(conditioning_mask)
+
         else:
             conditioning_mask = source_image.new_ones(1, 1, *source_image.shape[-2:])
 
@@ -345,7 +372,7 @@ def inpainting_image_conditioning(self, source_image, latent_image, image_mask=N
 
         return image_conditioning
 
-    def img2img_image_conditioning(self, source_image, latent_image, image_mask=None):
+    def img2img_image_conditioning(self, source_image, latent_image, image_mask=None, round_image_mask=True):
         source_image = devices.cond_cast_float(source_image)
 
         # HACK: Using introspection as the Depth2Image model doesn't appear to uniquely
@@ -357,11 +384,14 @@ def img2img_image_conditioning(self, source_image, latent_image, image_mask=None
             return self.edit_image_conditioning(source_image)
 
         if self.sampler.conditioning_key in {'hybrid', 'concat'}:
-            return self.inpainting_image_conditioning(source_image, latent_image, image_mask=image_mask)
+            return self.inpainting_image_conditioning(source_image, latent_image, image_mask=image_mask, round_image_mask=round_image_mask)
 
         if self.sampler.conditioning_key == "crossattn-adm":
             return self.unclip_image_conditioning(source_image)
 
+        if self.sampler.model_wrap.inner_model.is_sdxl_inpaint:
+            return self.inpainting_image_conditioning(source_image, latent_image, image_mask=image_mask)
+
         # Dummy zero conditioning if we're not using inpainting or depth model.
         return latent_image.new_zeros(latent_image.shape[0], 5, 1, 1)
 
@@ -422,6 +452,9 @@ def cached_params(self, required_prompts, steps, extra_network_data, hires_steps
             opts.sdxl_crop_top,
             self.width,
             self.height,
+            opts.fp8_storage,
+            opts.cache_fp16_weight,
+            opts.emphasis,
         )
 
     def get_conds_with_caching(self, function, required_prompts, steps, caches, extra_network_data, hires_steps=None):
@@ -532,7 +565,8 @@ def __init__(self, p: StableDiffusionProcessing, images_list, seed=-1, info="",
         self.all_negative_prompts = all_negative_prompts or p.all_negative_prompts or [self.negative_prompt]
         self.all_seeds = all_seeds or p.all_seeds or [self.seed]
         self.all_subseeds = all_subseeds or p.all_subseeds or [self.subseed]
-        self.infotexts = infotexts or [info]
+        self.infotexts = infotexts or [info] * len(images_list)
+        self.version = program_version()
 
     def js(self):
         obj = {
@@ -567,9 +601,10 @@ def js(self):
             "job_timestamp": self.job_timestamp,
             "clip_skip": self.clip_skip,
             "is_using_inpainting_conditioning": self.is_using_inpainting_conditioning,
+            "version": self.version,
         }
 
-        return json.dumps(obj)
+        return json.dumps(obj, default=lambda o: None)
 
     def infotext(self, p: StableDiffusionProcessing, index):
         return create_infotext(p, self.all_prompts, self.all_seeds, self.all_subseeds, comments=[], position_in_batch=index % self.batch_size, iteration=index // self.batch_size)
@@ -590,24 +625,40 @@ class DecodedSamples(list):
 def decode_latent_batch(model, batch, target_device=None, check_for_nans=False):
     samples = DecodedSamples()
 
+    if check_for_nans:
+        devices.test_for_nans(batch, "unet")
+
     for i in range(batch.shape[0]):
         sample = decode_first_stage(model, batch[i:i + 1])[0]
 
         if check_for_nans:
+
             try:
                 devices.test_for_nans(sample, "vae")
             except devices.NansException as e:
-                if devices.dtype_vae == torch.float32 or not shared.opts.auto_vae_precision:
+                if shared.opts.auto_vae_precision_bfloat16:
+                    autofix_dtype = torch.bfloat16
+                    autofix_dtype_text = "bfloat16"
+                    autofix_dtype_setting = "Automatically convert VAE to bfloat16"
+                    autofix_dtype_comment = ""
+                elif shared.opts.auto_vae_precision:
+                    autofix_dtype = torch.float32
+                    autofix_dtype_text = "32-bit float"
+                    autofix_dtype_setting = "Automatically revert VAE to 32-bit floats"
+                    autofix_dtype_comment = "\nTo always start with 32-bit VAE, use --no-half-vae commandline flag."
+                else:
+                    raise e
+
+                if devices.dtype_vae == autofix_dtype:
                     raise e
 
                 errors.print_error_explanation(
                     "A tensor with all NaNs was produced in VAE.\n"
-                    "Web UI will now convert VAE into 32-bit float and retry.\n"
-                    "To disable this behavior, disable the 'Automatically revert VAE to 32-bit floats' setting.\n"
-                    "To always start with 32-bit VAE, use --no-half-vae commandline flag."
+                    f"Web UI will now convert VAE into {autofix_dtype_text} and retry.\n"
+                    f"To disable this behavior, disable the '{autofix_dtype_setting}' setting.{autofix_dtype_comment}"
                 )
 
-                devices.dtype_vae = torch.float32
+                devices.dtype_vae = autofix_dtype
                 model.first_stage_model.to(devices.dtype_vae)
                 batch = batch.to(devices.dtype_vae)
 
@@ -652,7 +703,53 @@ def program_version():
 
 
 def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments=None, iteration=0, position_in_batch=0, use_main_prompt=False, index=None, all_negative_prompts=None):
-    if index is None:
+    """
+    this function is used to generate the infotext that is stored in the generated images, it's contains the parameters that are required to generate the imagee
+    Args:
+        p: StableDiffusionProcessing
+        all_prompts: list[str]
+        all_seeds: list[int]
+        all_subseeds: list[int]
+        comments: list[str]
+        iteration: int
+        position_in_batch: int
+        use_main_prompt: bool
+        index: int
+        all_negative_prompts: list[str]
+
+    Returns: str
+
+    Extra generation params
+    p.extra_generation_params dictionary allows for additional parameters to be added to the infotext
+    this can be use by the base webui or extensions.
+    To add a new entry, add a new key value pair, the dictionary key will be used as the key of the parameter in the infotext
+    the value generation_params can be defined as:
+        - str | None
+        - List[str|None]
+        - callable func(**kwargs) -> str | None
+
+    When defined as a string, it will be used as without extra processing; this is this most common use case.
+
+    Defining as a list allows for parameter that changes across images in the job, for example, the 'Seed' parameter.
+    The list should have the same length as the total number of images in the entire job.
+
+    Defining as a callable function allows parameter cannot be generated earlier or when extra logic is required.
+    For example 'Hires prompt', due to reasons the hr_prompt might be changed by process in the pipeline or extensions
+    and may vary across different images, defining as a static string or list would not work.
+
+    The function takes locals() as **kwargs, as such will have access to variables like 'p' and 'index'.
+    the base signature of the function should be:
+        func(**kwargs) -> str | None
+    optionally it can have additional arguments that will be used in the function:
+        func(p, index, **kwargs) -> str | None
+    note: for better future compatibility even though this function will have access to all variables in the locals(),
+        it is recommended to only use the arguments present in the function signature of create_infotext.
+    For actual implementation examples, see StableDiffusionProcessingTxt2Img.init > get_hr_prompt.
+    """
+
+    if use_main_prompt:
+        index = 0
+    elif index is None:
         index = position_in_batch + iteration * p.batch_size
 
     if all_negative_prompts is None:
@@ -663,6 +760,9 @@ def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments=None, iter
     token_merging_ratio = p.get_token_merging_ratio()
     token_merging_ratio_hr = p.get_token_merging_ratio(for_hr=True)
 
+    prompt_text = p.main_prompt if use_main_prompt else all_prompts[index]
+    negative_prompt = p.main_negative_prompt if use_main_prompt else all_negative_prompts[index]
+
     uses_ensd = opts.eta_noise_seed_delta != 0
     if uses_ensd:
         uses_ensd = sd_samplers_common.is_sampler_using_eta_noise_seed_delta(p)
@@ -670,6 +770,7 @@ def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments=None, iter
     generation_params = {
         "Steps": p.steps,
         "Sampler": p.sampler_name,
+        "Schedule type": p.scheduler,
         "CFG scale": p.cfg_scale,
         "Image CFG scale": getattr(p, 'image_cfg_scale', None),
         "Seed": p.all_seeds[0] if use_main_prompt else all_seeds[index],
@@ -677,12 +778,14 @@ def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments=None, iter
         "Size": f"{p.width}x{p.height}",
         "Model hash": p.sd_model_hash if opts.add_model_hash_to_info else None,
         "Model": p.sd_model_name if opts.add_model_name_to_info else None,
-        "VAE hash": p.sd_vae_hash if opts.add_model_hash_to_info else None,
-        "VAE": p.sd_vae_name if opts.add_model_name_to_info else None,
+        "FP8 weight": opts.fp8_storage if devices.fp8 else None,
+        "Cache FP16 weight for LoRA": opts.cache_fp16_weight if devices.fp8 else None,
+        "VAE hash": p.sd_vae_hash if opts.add_vae_hash_to_info else None,
+        "VAE": p.sd_vae_name if opts.add_vae_name_to_info else None,
         "Variation seed": (None if p.subseed_strength == 0 else (p.all_subseeds[0] if use_main_prompt else all_subseeds[index])),
         "Variation seed strength": (None if p.subseed_strength == 0 else p.subseed_strength),
         "Seed resize from": (None if p.seed_resize_from_w <= 0 or p.seed_resize_from_h <= 0 else f"{p.seed_resize_from_w}x{p.seed_resize_from_h}"),
-        "Denoising strength": getattr(p, 'denoising_strength', None),
+        "Denoising strength": p.extra_generation_params.get("Denoising strength"),
         "Conditional mask weight": getattr(p, "inpainting_mask_weight", shared.opts.inpainting_mask_weight) if p.is_using_inpainting_conditioning else None,
         "Clip skip": None if clip_skip <= 1 else clip_skip,
         "ENSD": opts.eta_noise_seed_delta if uses_ensd else None,
@@ -690,17 +793,25 @@ def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments=None, iter
         "Token merging ratio hr": None if not enable_hr or token_merging_ratio_hr == 0 else token_merging_ratio_hr,
         "Init image hash": getattr(p, 'init_img_hash', None),
         "RNG": opts.randn_source if opts.randn_source != "GPU" else None,
-        "NGMS": None if p.s_min_uncond == 0 else p.s_min_uncond,
         "Tiling": "True" if p.tiling else None,
         **p.extra_generation_params,
         "Version": program_version() if opts.add_version_to_infotext else None,
         "User": p.user if opts.add_user_name_to_info else None,
     }
 
-    generation_params_text = ", ".join([k if k == v else f'{k}: {generation_parameters_copypaste.quote(v)}' for k, v in generation_params.items() if v is not None])
+    for key, value in generation_params.items():
+        try:
+            if isinstance(value, list):
+                generation_params[key] = value[index]
+            elif callable(value):
+                generation_params[key] = value(**locals())
+        except Exception:
+            errors.report(f'Error creating infotext for key "{key}"', exc_info=True)
+            generation_params[key] = None
 
-    prompt_text = p.main_prompt if use_main_prompt else all_prompts[index]
-    negative_prompt_text = f"\nNegative prompt: {p.main_negative_prompt if use_main_prompt else all_negative_prompts[index]}" if all_negative_prompts[index] else ""
+    generation_params_text = ", ".join([k if k == v else f'{k}: {infotext_utils.quote(v)}' for k, v in generation_params.items() if v is not None])
+
+    negative_prompt_text = f"\nNegative prompt: {negative_prompt}" if negative_prompt else ""
 
     return f"{prompt_text}{negative_prompt_text}\n{generation_params_text}".strip()
 
@@ -709,7 +820,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed:
     if p.scripts is not None:
         p.scripts.before_process(p)
 
-    stored_opts = {k: opts.data[k] for k in p.override_settings.keys()}
+    stored_opts = {k: opts.data[k] if k in opts.data else opts.get_default(k) for k in p.override_settings.keys() if k in opts.data}
 
     try:
         # if no checkpoint override or the override checkpoint can't be found, remove override entry and load opts checkpoint
@@ -729,7 +840,11 @@ def process_images(p: StableDiffusionProcessing) -> Processed:
 
         sd_models.apply_token_merging(p.sd_model, p.get_token_merging_ratio())
 
-        res = process_images_inner(p)
+        # backwards compatibility, fix sampler and scheduler if invalid
+        sd_samplers.fix_p_invalid_sampler_and_scheduler(p)
+
+        with profiling.Profiler():
+            res = process_images_inner(p)
 
     finally:
         sd_models.apply_token_merging(p.sd_model, 0)
@@ -769,6 +884,9 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
         if p.refiner_checkpoint_info is None:
             raise Exception(f'Could not find checkpoint with name {p.refiner_checkpoint}')
 
+    if hasattr(shared.sd_model, 'fix_dimensions'):
+        p.width, p.height = shared.sd_model.fix_dimensions(p.width, p.height)
+
     p.sd_model_name = shared.sd_model.sd_checkpoint_info.name_for_extra
     p.sd_model_hash = shared.sd_model.sd_model_hash
     p.sd_vae_name = sd_vae.get_loaded_vae_name()
@@ -777,6 +895,7 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
     modules.sd_hijack.model_hijack.apply_circular(p.tiling)
     modules.sd_hijack.model_hijack.clear_comments()
 
+    p.fill_fields_from_opts()
     p.setup_prompts()
 
     if isinstance(seed, list):
@@ -797,7 +916,6 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
 
     infotexts = []
     output_images = []
-
     with torch.no_grad(), p.sd_model.ema_scope():
         with devices.autocast():
             p.init(p.all_prompts, p.all_seeds, p.all_subseeds)
@@ -817,7 +935,7 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
             if state.skipped:
                 state.skipped = False
 
-            if state.interrupted:
+            if state.interrupted or state.stopping_generation:
                 break
 
             sd_models.reload_model_weights()  # model can be changed for example by refiner
@@ -827,7 +945,8 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
             p.seeds = p.all_seeds[n * p.batch_size:(n + 1) * p.batch_size]
             p.subseeds = p.all_subseeds[n * p.batch_size:(n + 1) * p.batch_size]
 
-            p.rng = rng.ImageRNG((opt_C, p.height // opt_f, p.width // opt_f), p.seeds, subseeds=p.subseeds, subseed_strength=p.subseed_strength, seed_resize_from_h=p.seed_resize_from_h, seed_resize_from_w=p.seed_resize_from_w)
+            latent_channels = getattr(shared.sd_model, 'latent_channels', opt_C)
+            p.rng = rng.ImageRNG((latent_channels, p.height // opt_f, p.width // opt_f), p.seeds, subseeds=p.subseeds, subseed_strength=p.subseed_strength, seed_resize_from_h=p.seed_resize_from_h, seed_resize_from_w=p.seed_resize_from_w)
 
             if p.scripts is not None:
                 p.scripts.before_process_batch(p, batch_number=n, prompts=p.prompts, seeds=p.seeds, subseeds=p.subseeds)
@@ -844,34 +963,42 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
             if p.scripts is not None:
                 p.scripts.process_batch(p, batch_number=n, prompts=p.prompts, seeds=p.seeds, subseeds=p.subseeds)
 
+            p.setup_conds()
+
+            p.extra_generation_params.update(model_hijack.extra_generation_params)
+
             # params.txt should be saved after scripts.process_batch, since the
             # infotext could be modified by that callback
             # Example: a wildcard processed by process_batch sets an extra model
             # strength, which is saved as "Model Strength: 1.0" in the infotext
-            if n == 0:
+            if n == 0 and not cmd_opts.no_prompt_history:
                 with open(os.path.join(paths.data_path, "params.txt"), "w", encoding="utf8") as file:
                     processed = Processed(p, [])
                     file.write(processed.infotext(p, 0))
 
-            p.setup_conds()
-
             for comment in model_hijack.comments:
                 p.comment(comment)
 
-            p.extra_generation_params.update(model_hijack.extra_generation_params)
-
             if p.n_iter > 1:
                 shared.state.job = f"Batch {n+1} out of {p.n_iter}"
 
+            sd_models.apply_alpha_schedule_override(p.sd_model, p)
+
             with devices.without_autocast() if devices.unet_needs_upcast else devices.autocast():
                 samples_ddim = p.sample(conditioning=p.c, unconditional_conditioning=p.uc, seeds=p.seeds, subseeds=p.subseeds, subseed_strength=p.subseed_strength, prompts=p.prompts)
 
+            if p.scripts is not None:
+                ps = scripts.PostSampleArgs(samples_ddim)
+                p.scripts.post_sample(p, ps)
+                samples_ddim = ps.samples
+
             if getattr(samples_ddim, 'already_decoded', False):
                 x_samples_ddim = samples_ddim
             else:
+                devices.test_for_nans(samples_ddim, "unet")
+
                 if opts.sd_vae_decode_method != 'Full':
                     p.extra_generation_params['VAE Decoder'] = opts.sd_vae_decode_method
-
                 x_samples_ddim = decode_latent_batch(p.sd_model, samples_ddim, target_device=devices.cpu, check_for_nans=True)
 
             x_samples_ddim = torch.stack(x_samples_ddim).float()
@@ -884,6 +1011,8 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
 
             devices.torch_gc()
 
+            state.nextjob()
+
             if p.scripts is not None:
                 p.scripts.postprocess_batch(p, x_samples_ddim, batch_number=n)
 
@@ -920,13 +1049,37 @@ def infotext(index=0, use_main_prompt=False):
                     pp = scripts.PostprocessImageArgs(image)
                     p.scripts.postprocess_image(p, pp)
                     image = pp.image
+
+                mask_for_overlay = getattr(p, "mask_for_overlay", None)
+
+                if not shared.opts.overlay_inpaint:
+                    overlay_image = None
+                elif getattr(p, "overlay_images", None) is not None and i < len(p.overlay_images):
+                    overlay_image = p.overlay_images[i]
+                else:
+                    overlay_image = None
+
+                if p.scripts is not None:
+                    ppmo = scripts.PostProcessMaskOverlayArgs(i, mask_for_overlay, overlay_image)
+                    p.scripts.postprocess_maskoverlay(p, ppmo)
+                    mask_for_overlay, overlay_image = ppmo.mask_for_overlay, ppmo.overlay_image
+
                 if p.color_corrections is not None and i < len(p.color_corrections):
                     if save_samples and opts.save_images_before_color_correction:
-                        image_without_cc = apply_overlay(image, p.paste_to, i, p.overlay_images)
+                        image_without_cc, _ = apply_overlay(image, p.paste_to, overlay_image)
                         images.save_image(image_without_cc, p.outpath_samples, "", p.seeds[i], p.prompts[i], opts.samples_format, info=infotext(i), p=p, suffix="-before-color-correction")
                     image = apply_color_correction(p.color_corrections[i], image)
 
-                image = apply_overlay(image, p.paste_to, i, p.overlay_images)
+                # If the intention is to show the output from the model
+                # that is being composited over the original image,
+                # we need to keep the original image around
+                # and use it in the composite step.
+                image, original_denoised_image = apply_overlay(image, p.paste_to, overlay_image)
+
+                if p.scripts is not None:
+                    pp = scripts.PostprocessImageArgs(image)
+                    p.scripts.postprocess_image_after_composite(p, pp)
+                    image = pp.image
 
                 if save_samples:
                     images.save_image(image, p.outpath_samples, "", p.seeds[i], p.prompts[i], opts.samples_format, info=infotext(i), p=p)
@@ -936,27 +1089,28 @@ def infotext(index=0, use_main_prompt=False):
                 if opts.enable_pnginfo:
                     image.info["parameters"] = text
                 output_images.append(image)
-                if save_samples and hasattr(p, 'mask_for_overlay') and p.mask_for_overlay and any([opts.save_mask, opts.save_mask_composite, opts.return_mask, opts.return_mask_composite]):
-                    image_mask = p.mask_for_overlay.convert('RGB')
-                    image_mask_composite = Image.composite(image.convert('RGBA').convert('RGBa'), Image.new('RGBa', image.size), images.resize_image(2, p.mask_for_overlay, image.width, image.height).convert('L')).convert('RGBA')
-
-                    if opts.save_mask:
-                        images.save_image(image_mask, p.outpath_samples, "", p.seeds[i], p.prompts[i], opts.samples_format, info=infotext(i), p=p, suffix="-mask")
-
-                    if opts.save_mask_composite:
-                        images.save_image(image_mask_composite, p.outpath_samples, "", p.seeds[i], p.prompts[i], opts.samples_format, info=infotext(i), p=p, suffix="-mask-composite")
 
-                    if opts.return_mask:
-                        output_images.append(image_mask)
-
-                    if opts.return_mask_composite:
-                        output_images.append(image_mask_composite)
+                if mask_for_overlay is not None:
+                    if opts.return_mask or opts.save_mask:
+                        image_mask = mask_for_overlay.convert('RGB')
+                        if save_samples and opts.save_mask:
+                            images.save_image(image_mask, p.outpath_samples, "", p.seeds[i], p.prompts[i], opts.samples_format, info=infotext(i), p=p, suffix="-mask")
+                        if opts.return_mask:
+                            output_images.append(image_mask)
+
+                    if opts.return_mask_composite or opts.save_mask_composite:
+                        image_mask_composite = Image.composite(original_denoised_image.convert('RGBA').convert('RGBa'), Image.new('RGBa', image.size), images.resize_image(2, mask_for_overlay, image.width, image.height).convert('L')).convert('RGBA')
+                        if save_samples and opts.save_mask_composite:
+                            images.save_image(image_mask_composite, p.outpath_samples, "", p.seeds[i], p.prompts[i], opts.samples_format, info=infotext(i), p=p, suffix="-mask-composite")
+                        if opts.return_mask_composite:
+                            output_images.append(image_mask_composite)
 
             del x_samples_ddim
 
             devices.torch_gc()
 
-            state.nextjob()
+        if not infotexts:
+            infotexts.append(Processed(p, []).infotext(p, 0))
 
         p.color_corrections = None
 
@@ -1021,8 +1175,10 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing):
     hr_resize_y: int = 0
     hr_checkpoint_name: str = None
     hr_sampler_name: str = None
+    hr_scheduler: str = None
     hr_prompt: str = ''
     hr_negative_prompt: str = ''
+    force_task_id: str = None
 
     cached_hr_uc = [None, None]
     cached_hr_c = [None, None]
@@ -1095,7 +1251,9 @@ def calculate_target_resolution(self):
 
     def init(self, all_prompts, all_seeds, all_subseeds):
         if self.enable_hr:
-            if self.hr_checkpoint_name:
+            self.extra_generation_params["Denoising strength"] = self.denoising_strength
+
+            if self.hr_checkpoint_name and self.hr_checkpoint_name != 'Use same checkpoint':
                 self.hr_checkpoint_info = sd_models.get_closet_checkpoint_match(self.hr_checkpoint_name)
 
                 if self.hr_checkpoint_info is None:
@@ -1106,11 +1264,21 @@ def init(self, all_prompts, all_seeds, all_subseeds):
             if self.hr_sampler_name is not None and self.hr_sampler_name != self.sampler_name:
                 self.extra_generation_params["Hires sampler"] = self.hr_sampler_name
 
-            if tuple(self.hr_prompt) != tuple(self.prompt):
-                self.extra_generation_params["Hires prompt"] = self.hr_prompt
+            def get_hr_prompt(p, index, prompt_text, **kwargs):
+                hr_prompt = p.all_hr_prompts[index]
+                return hr_prompt if hr_prompt != prompt_text else None
+
+            def get_hr_negative_prompt(p, index, negative_prompt, **kwargs):
+                hr_negative_prompt = p.all_hr_negative_prompts[index]
+                return hr_negative_prompt if hr_negative_prompt != negative_prompt else None
 
-            if tuple(self.hr_negative_prompt) != tuple(self.negative_prompt):
-                self.extra_generation_params["Hires negative prompt"] = self.hr_negative_prompt
+            self.extra_generation_params["Hires prompt"] = get_hr_prompt
+            self.extra_generation_params["Hires negative prompt"] = get_hr_negative_prompt
+
+            self.extra_generation_params["Hires schedule type"] = None  # to be set in sd_samplers_kdiffusion.py
+
+            if self.hr_scheduler is None:
+                self.hr_scheduler = self.scheduler
 
             self.latent_scale_mode = shared.latent_upscale_modes.get(self.hr_upscaler, None) if self.hr_upscaler is not None else shared.latent_upscale_modes.get(shared.latent_upscale_default_mode, "nearest")
             if self.enable_hr and self.latent_scale_mode is None:
@@ -1122,8 +1290,11 @@ def init(self, all_prompts, all_seeds, all_subseeds):
             if not state.processing_has_refined_job_count:
                 if state.job_count == -1:
                     state.job_count = self.n_iter
-
-                shared.total_tqdm.updateTotal((self.steps + (self.hr_second_pass_steps or self.steps)) * state.job_count)
+                if getattr(self, 'txt2img_upscale', False):
+                    total_steps = (self.hr_second_pass_steps or self.steps) * state.job_count
+                else:
+                    total_steps = (self.steps + (self.hr_second_pass_steps or self.steps)) * state.job_count
+                shared.total_tqdm.updateTotal(total_steps)
                 state.job_count = state.job_count * 2
                 state.processing_has_refined_job_count = True
 
@@ -1136,23 +1307,58 @@ def init(self, all_prompts, all_seeds, all_subseeds):
     def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts):
         self.sampler = sd_samplers.create_sampler(self.sampler_name, self.sd_model)
 
-        x = self.rng.next()
-        samples = self.sampler.sample(self, x, conditioning, unconditional_conditioning, image_conditioning=self.txt2img_image_conditioning(x))
-        del x
+        if self.firstpass_image is not None and self.enable_hr:
+            # here we don't need to generate image, we just take self.firstpass_image and prepare it for hires fix
 
-        if not self.enable_hr:
-            return samples
+            if self.latent_scale_mode is None:
+                image = np.array(self.firstpass_image).astype(np.float32) / 255.0 * 2.0 - 1.0
+                image = np.moveaxis(image, 2, 0)
+
+                samples = None
+                decoded_samples = torch.asarray(np.expand_dims(image, 0))
+
+            else:
+                image = np.array(self.firstpass_image).astype(np.float32) / 255.0
+                image = np.moveaxis(image, 2, 0)
+                image = torch.from_numpy(np.expand_dims(image, axis=0))
+                image = image.to(shared.device, dtype=devices.dtype_vae)
+
+                if opts.sd_vae_encode_method != 'Full':
+                    self.extra_generation_params['VAE Encoder'] = opts.sd_vae_encode_method
+
+                samples = images_tensor_to_samples(image, approximation_indexes.get(opts.sd_vae_encode_method), self.sd_model)
+                decoded_samples = None
+                devices.torch_gc()
 
-        if self.latent_scale_mode is None:
-            decoded_samples = torch.stack(decode_latent_batch(self.sd_model, samples, target_device=devices.cpu, check_for_nans=True)).to(dtype=torch.float32)
         else:
-            decoded_samples = None
+            # here we generate an image normally
+
+            x = self.rng.next()
+            if self.scripts is not None:
+                self.scripts.process_before_every_sampling(
+                    p=self,
+                    x=x,
+                    noise=x,
+                    c=conditioning,
+                    uc=unconditional_conditioning
+                )
+
+            samples = self.sampler.sample(self, x, conditioning, unconditional_conditioning, image_conditioning=self.txt2img_image_conditioning(x))
+            del x
+
+            if not self.enable_hr:
+                return samples
+
+            devices.torch_gc()
+
+            if self.latent_scale_mode is None:
+                decoded_samples = torch.stack(decode_latent_batch(self.sd_model, samples, target_device=devices.cpu, check_for_nans=True)).to(dtype=torch.float32)
+            else:
+                decoded_samples = None
 
         with sd_models.SkipWritingToConfig():
             sd_models.reload_model_weights(info=self.hr_checkpoint_info)
 
-        devices.torch_gc()
-
         return self.sample_hr_pass(samples, decoded_samples, seeds, subseeds, subseed_strength, prompts)
 
     def sample_hr_pass(self, samples, decoded_samples, seeds, subseeds, subseed_strength, prompts):
@@ -1160,7 +1366,6 @@ def sample_hr_pass(self, samples, decoded_samples, seeds, subseeds, subseed_stre
             return samples
 
         self.is_hr_pass = True
-
         target_width = self.hr_upscale_to_x
         target_height = self.hr_upscale_to_y
 
@@ -1238,6 +1443,13 @@ def save_intermediate(image, index):
 
         if self.scripts is not None:
             self.scripts.before_hr(self)
+            self.scripts.process_before_every_sampling(
+                p=self,
+                x=samples,
+                noise=noise,
+                c=self.hr_c,
+                uc=self.hr_uc,
+            )
 
         samples = self.sampler.sample_img2img(self, samples, noise, self.hr_c, self.hr_uc, steps=self.hr_second_pass_steps or self.steps, image_conditioning=image_conditioning)
 
@@ -1249,7 +1461,6 @@ def save_intermediate(image, index):
         decoded_samples = decode_latent_batch(self.sd_model, samples, target_device=devices.cpu, check_for_nans=True)
 
         self.is_hr_pass = False
-
         return decoded_samples
 
     def close(self):
@@ -1352,12 +1563,14 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing):
     mask_blur_x: int = 4
     mask_blur_y: int = 4
     mask_blur: int = None
+    mask_round: bool = True
     inpainting_fill: int = 0
     inpaint_full_res: bool = True
     inpaint_full_res_padding: int = 0
     inpainting_mask_invert: int = 0
     initial_noise_multiplier: float = None
     latent_mask: Image = None
+    force_task_id: str = None
 
     image_mask: Any = field(default=None, init=False)
 
@@ -1387,6 +1600,8 @@ def mask_blur(self, value):
             self.mask_blur_y = value
 
     def init(self, all_prompts, all_seeds, all_subseeds):
+        self.extra_generation_params["Denoising strength"] = self.denoising_strength
+
         self.image_cfg_scale: float = self.image_cfg_scale if shared.sd_model.cond_stage_key == "edit" else None
 
         self.sampler = sd_samplers.create_sampler(self.sampler_name, self.sd_model)
@@ -1397,10 +1612,11 @@ def init(self, all_prompts, all_seeds, all_subseeds):
         if image_mask is not None:
             # image_mask is passed in as RGBA by Gradio to support alpha masks,
             # but we still want to support binary masks.
-            image_mask = create_binary_mask(image_mask)
+            image_mask = create_binary_mask(image_mask, round=self.mask_round)
 
             if self.inpainting_mask_invert:
                 image_mask = ImageOps.invert(image_mask)
+                self.extra_generation_params["Mask mode"] = "Inpaint not masked"
 
             if self.mask_blur_x > 0:
                 np_mask = np.array(image_mask)
@@ -1414,16 +1630,29 @@ def init(self, all_prompts, all_seeds, all_subseeds):
                 np_mask = cv2.GaussianBlur(np_mask, (1, kernel_size), self.mask_blur_y)
                 image_mask = Image.fromarray(np_mask)
 
+            if self.mask_blur_x > 0 or self.mask_blur_y > 0:
+                self.extra_generation_params["Mask blur"] = self.mask_blur
+
             if self.inpaint_full_res:
                 self.mask_for_overlay = image_mask
                 mask = image_mask.convert('L')
-                crop_region = masking.get_crop_region(np.array(mask), self.inpaint_full_res_padding)
-                crop_region = masking.expand_crop_region(crop_region, self.width, self.height, mask.width, mask.height)
-                x1, y1, x2, y2 = crop_region
-
-                mask = mask.crop(crop_region)
-                image_mask = images.resize_image(2, mask, self.width, self.height)
-                self.paste_to = (x1, y1, x2-x1, y2-y1)
+                crop_region = masking.get_crop_region_v2(mask, self.inpaint_full_res_padding)
+                if crop_region:
+                    crop_region = masking.expand_crop_region(crop_region, self.width, self.height, mask.width, mask.height)
+                    x1, y1, x2, y2 = crop_region
+                    mask = mask.crop(crop_region)
+                    image_mask = images.resize_image(2, mask, self.width, self.height)
+                    self.paste_to = (x1, y1, x2-x1, y2-y1)
+                    self.extra_generation_params["Inpaint area"] = "Only masked"
+                    self.extra_generation_params["Masked area padding"] = self.inpaint_full_res_padding
+                else:
+                    crop_region = None
+                    image_mask = None
+                    self.mask_for_overlay = None
+                    self.inpaint_full_res = False
+                    massage = 'Unable to perform "Inpaint Only mask" because mask is blank, switch to img2img mode.'
+                    model_hijack.comments.append(massage)
+                    logging.info(massage)
             else:
                 image_mask = images.resize_image(self.resize_mode, image_mask, self.width, self.height)
                 np_mask = np.array(image_mask)
@@ -1443,7 +1672,7 @@ def init(self, all_prompts, all_seeds, all_subseeds):
             # Save init image
             if opts.save_init_img:
                 self.init_img_hash = hashlib.md5(img.tobytes()).hexdigest()
-                images.save_image(img, path=opts.outdir_init_images, basename=None, forced_filename=self.init_img_hash, save_to_dirs=False)
+                images.save_image(img, path=opts.outdir_init_images, basename=None, forced_filename=self.init_img_hash, save_to_dirs=False, existing_info=img.info)
 
             image = images.flatten(img, opts.img2img_background_color)
 
@@ -1451,6 +1680,8 @@ def init(self, all_prompts, all_seeds, all_subseeds):
                 image = images.resize_image(self.resize_mode, image, self.width, self.height)
 
             if image_mask is not None:
+                if self.mask_for_overlay.size != (image.width, image.height):
+                    self.mask_for_overlay = images.resize_image(self.resize_mode, self.mask_for_overlay, image.width, image.height)
                 image_masked = Image.new('RGBa', (image.width, image.height))
                 image_masked.paste(image.convert("RGBA").convert("RGBa"), mask=ImageOps.invert(self.mask_for_overlay.convert('L')))
 
@@ -1465,6 +1696,9 @@ def init(self, all_prompts, all_seeds, all_subseeds):
                 if self.inpainting_fill != 1:
                     image = masking.fill(image, latent_mask)
 
+                    if self.inpainting_fill == 0:
+                        self.extra_generation_params["Masked content"] = 'fill'
+
             if add_color_corrections:
                 self.color_corrections.append(setup_color_correction(image))
 
@@ -1504,19 +1738,23 @@ def init(self, all_prompts, all_seeds, all_subseeds):
             latmask = init_mask.convert('RGB').resize((self.init_latent.shape[3], self.init_latent.shape[2]))
             latmask = np.moveaxis(np.array(latmask, dtype=np.float32), 2, 0) / 255
             latmask = latmask[0]
-            latmask = np.around(latmask)
-            latmask = np.tile(latmask[None], (4, 1, 1))
+            if self.mask_round:
+                latmask = np.around(latmask)
+            latmask = np.tile(latmask[None], (self.init_latent.shape[1], 1, 1))
 
-            self.mask = torch.asarray(1.0 - latmask).to(shared.device).type(self.sd_model.dtype)
-            self.nmask = torch.asarray(latmask).to(shared.device).type(self.sd_model.dtype)
+            self.mask = torch.asarray(1.0 - latmask).to(shared.device).type(devices.dtype)
+            self.nmask = torch.asarray(latmask).to(shared.device).type(devices.dtype)
 
             # this needs to be fixed to be done in sample() using actual seeds for batches
             if self.inpainting_fill == 2:
                 self.init_latent = self.init_latent * self.mask + create_random_tensors(self.init_latent.shape[1:], all_seeds[0:self.init_latent.shape[0]]) * self.nmask
+                self.extra_generation_params["Masked content"] = 'latent noise'
+
             elif self.inpainting_fill == 3:
                 self.init_latent = self.init_latent * self.mask
+                self.extra_generation_params["Masked content"] = 'latent nothing'
 
-        self.image_conditioning = self.img2img_image_conditioning(image * 2 - 1, self.init_latent, image_mask)
+        self.image_conditioning = self.img2img_image_conditioning(image * 2 - 1, self.init_latent, image_mask, self.mask_round)
 
     def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength, prompts):
         x = self.rng.next()
@@ -1525,10 +1763,25 @@ def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subs
             self.extra_generation_params["Noise multiplier"] = self.initial_noise_multiplier
             x *= self.initial_noise_multiplier
 
+        if self.scripts is not None:
+            self.scripts.process_before_every_sampling(
+                p=self,
+                x=self.init_latent,
+                noise=x,
+                c=conditioning,
+                uc=unconditional_conditioning
+            )
         samples = self.sampler.sample_img2img(self, self.init_latent, x, conditioning, unconditional_conditioning, image_conditioning=self.image_conditioning)
 
         if self.mask is not None:
-            samples = samples * self.nmask + self.init_latent * self.mask
+            blended_samples = samples * self.nmask + self.init_latent * self.mask
+
+            if self.scripts is not None:
+                mba = scripts.MaskBlendArgs(samples, self.nmask, self.init_latent, self.mask, blended_samples)
+                self.scripts.on_mask_blend(self, mba)
+                blended_samples = mba.blended_latent
+
+            samples = blended_samples
 
         del x
         devices.torch_gc()
diff --git a/modules/processing_scripts/comments.py b/modules/processing_scripts/comments.py
new file mode 100644
index 00000000000..cf81dfd8b49
--- /dev/null
+++ b/modules/processing_scripts/comments.py
@@ -0,0 +1,49 @@
+from modules import scripts, shared, script_callbacks
+import re
+
+
+def strip_comments(text):
+    text = re.sub('(^|\n)#[^\n]*(\n|$)', '\n', text)  # while line comment
+    text = re.sub('#[^\n]*(\n|$)', '\n', text)  # in the middle of the line comment
+
+    return text
+
+
+class ScriptStripComments(scripts.Script):
+    def title(self):
+        return "Comments"
+
+    def show(self, is_img2img):
+        return scripts.AlwaysVisible
+
+    def process(self, p, *args):
+        if not shared.opts.enable_prompt_comments:
+            return
+
+        p.all_prompts = [strip_comments(x) for x in p.all_prompts]
+        p.all_negative_prompts = [strip_comments(x) for x in p.all_negative_prompts]
+
+        p.main_prompt = strip_comments(p.main_prompt)
+        p.main_negative_prompt = strip_comments(p.main_negative_prompt)
+
+        if getattr(p, 'enable_hr', False):
+            p.all_hr_prompts = [strip_comments(x) for x in p.all_hr_prompts]
+            p.all_hr_negative_prompts = [strip_comments(x) for x in p.all_hr_negative_prompts]
+
+            p.hr_prompt = strip_comments(p.hr_prompt)
+            p.hr_negative_prompt = strip_comments(p.hr_negative_prompt)
+
+
+def before_token_counter(params: script_callbacks.BeforeTokenCounterParams):
+    if not shared.opts.enable_prompt_comments:
+        return
+
+    params.prompt = strip_comments(params.prompt)
+
+
+script_callbacks.on_before_token_counter(before_token_counter)
+
+
+shared.options_templates.update(shared.options_section(('sd', "Stable Diffusion", "sd"), {
+    "enable_prompt_comments": shared.OptionInfo(True, "Enable comments").info("Use # anywhere in the prompt to hide the text between # and the end of the line from the generation."),
+}))
diff --git a/modules/processing_scripts/refiner.py b/modules/processing_scripts/refiner.py
index 29ccb78f903..ba33d8a4b80 100644
--- a/modules/processing_scripts/refiner.py
+++ b/modules/processing_scripts/refiner.py
@@ -1,6 +1,7 @@
 import gradio as gr
 
 from modules import scripts, sd_models
+from modules.infotext_utils import PasteField
 from modules.ui_common import create_refresh_button
 from modules.ui_components import InputAccordion
 
@@ -31,9 +32,9 @@ def lookup_checkpoint(title):
             return None if info is None else info.title
 
         self.infotext_fields = [
-            (enable_refiner, lambda d: 'Refiner' in d),
-            (refiner_checkpoint, lambda d: lookup_checkpoint(d.get('Refiner'))),
-            (refiner_switch_at, 'Refiner switch at'),
+            PasteField(enable_refiner, lambda d: 'Refiner' in d),
+            PasteField(refiner_checkpoint, lambda d: lookup_checkpoint(d.get('Refiner')), api="refiner_checkpoint"),
+            PasteField(refiner_switch_at, 'Refiner switch at', api="refiner_switch_at"),
         ]
 
         return enable_refiner, refiner_checkpoint, refiner_switch_at
diff --git a/modules/processing_scripts/sampler.py b/modules/processing_scripts/sampler.py
new file mode 100644
index 00000000000..5d50a162cee
--- /dev/null
+++ b/modules/processing_scripts/sampler.py
@@ -0,0 +1,45 @@
+import gradio as gr
+
+from modules import scripts, sd_samplers, sd_schedulers, shared
+from modules.infotext_utils import PasteField
+from modules.ui_components import FormRow, FormGroup
+
+
+class ScriptSampler(scripts.ScriptBuiltinUI):
+    section = "sampler"
+
+    def __init__(self):
+        self.steps = None
+        self.sampler_name = None
+        self.scheduler = None
+
+    def title(self):
+        return "Sampler"
+
+    def ui(self, is_img2img):
+        sampler_names = [x.name for x in sd_samplers.visible_samplers()]
+        scheduler_names = [x.label for x in sd_schedulers.schedulers]
+
+        if shared.opts.samplers_in_dropdown:
+            with FormRow(elem_id=f"sampler_selection_{self.tabname}"):
+                self.sampler_name = gr.Dropdown(label='Sampling method', elem_id=f"{self.tabname}_sampling", choices=sampler_names, value=sampler_names[0])
+                self.scheduler = gr.Dropdown(label='Schedule type', elem_id=f"{self.tabname}_scheduler", choices=scheduler_names, value=scheduler_names[0])
+                self.steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{self.tabname}_steps", label="Sampling steps", value=20)
+        else:
+            with FormGroup(elem_id=f"sampler_selection_{self.tabname}"):
+                self.steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{self.tabname}_steps", label="Sampling steps", value=20)
+                self.sampler_name = gr.Radio(label='Sampling method', elem_id=f"{self.tabname}_sampling", choices=sampler_names, value=sampler_names[0])
+                self.scheduler = gr.Dropdown(label='Schedule type', elem_id=f"{self.tabname}_scheduler", choices=scheduler_names, value=scheduler_names[0])
+
+        self.infotext_fields = [
+            PasteField(self.steps, "Steps", api="steps"),
+            PasteField(self.sampler_name, sd_samplers.get_sampler_from_infotext, api="sampler_name"),
+            PasteField(self.scheduler, sd_samplers.get_scheduler_from_infotext, api="scheduler"),
+        ]
+
+        return self.steps, self.sampler_name, self.scheduler
+
+    def setup(self, p, steps, sampler_name, scheduler):
+        p.steps = steps
+        p.sampler_name = sampler_name
+        p.scheduler = scheduler
diff --git a/modules/processing_scripts/seed.py b/modules/processing_scripts/seed.py
index 6b6ff987d2d..7a4c0159831 100644
--- a/modules/processing_scripts/seed.py
+++ b/modules/processing_scripts/seed.py
@@ -3,8 +3,10 @@
 import gradio as gr
 
 from modules import scripts, ui, errors
+from modules.infotext_utils import PasteField
 from modules.shared import cmd_opts
 from modules.ui_components import ToolButton
+from modules import infotext_utils
 
 
 class ScriptSeed(scripts.ScriptBuiltinUI):
@@ -29,8 +31,8 @@ def ui(self, is_img2img):
             else:
                 self.seed = gr.Number(label='Seed', value=-1, elem_id=self.elem_id("seed"), min_width=100, precision=0)
 
-            random_seed = ToolButton(ui.random_symbol, elem_id=self.elem_id("random_seed"), label='Random seed')
-            reuse_seed = ToolButton(ui.reuse_symbol, elem_id=self.elem_id("reuse_seed"), label='Reuse seed')
+            random_seed = ToolButton(ui.random_symbol, elem_id=self.elem_id("random_seed"), tooltip="Set seed to -1, which will cause a new random number to be used every time")
+            reuse_seed = ToolButton(ui.reuse_symbol, elem_id=self.elem_id("reuse_seed"), tooltip="Reuse seed from last generation, mostly useful if it was randomized")
 
             seed_checkbox = gr.Checkbox(label='Extra', elem_id=self.elem_id("subseed_show"), value=False)
 
@@ -51,12 +53,12 @@ def ui(self, is_img2img):
         seed_checkbox.change(lambda x: gr.update(visible=x), show_progress=False, inputs=[seed_checkbox], outputs=[seed_extras])
 
         self.infotext_fields = [
-            (self.seed, "Seed"),
-            (seed_checkbox, lambda d: "Variation seed" in d or "Seed resize from-1" in d),
-            (subseed, "Variation seed"),
-            (subseed_strength, "Variation seed strength"),
-            (seed_resize_from_w, "Seed resize from-1"),
-            (seed_resize_from_h, "Seed resize from-2"),
+            PasteField(self.seed, "Seed", api="seed"),
+            PasteField(seed_checkbox, lambda d: "Variation seed" in d or "Seed resize from-1" in d),
+            PasteField(subseed, "Variation seed", api="subseed"),
+            PasteField(subseed_strength, "Variation seed strength", api="subseed_strength"),
+            PasteField(seed_resize_from_w, "Seed resize from-1", api="seed_resize_from_h"),
+            PasteField(seed_resize_from_h, "Seed resize from-2", api="seed_resize_from_w"),
         ]
 
         self.on_after_component(lambda x: connect_reuse_seed(self.seed, reuse_seed, x.component, False), elem_id=f'generation_info_{self.tabname}')
@@ -76,7 +78,6 @@ def setup(self, p, seed, seed_checkbox, subseed, subseed_strength, seed_resize_f
             p.seed_resize_from_h = seed_resize_from_h
 
 
-
 def connect_reuse_seed(seed: gr.Number, reuse_seed: gr.Button, generation_info: gr.Textbox, is_subseed):
     """ Connects a 'reuse (sub)seed' button's click event so that it copies last used
         (sub)seed value from generation info the to the seed field. If copying subseed and subseed strength
@@ -84,21 +85,14 @@ def connect_reuse_seed(seed: gr.Number, reuse_seed: gr.Button, generation_info:
 
     def copy_seed(gen_info_string: str, index):
         res = -1
-
         try:
             gen_info = json.loads(gen_info_string)
-            index -= gen_info.get('index_of_first_image', 0)
-
-            if is_subseed and gen_info.get('subseed_strength', 0) > 0:
-                all_subseeds = gen_info.get('all_subseeds', [-1])
-                res = all_subseeds[index if 0 <= index < len(all_subseeds) else 0]
-            else:
-                all_seeds = gen_info.get('all_seeds', [-1])
-                res = all_seeds[index if 0 <= index < len(all_seeds) else 0]
-
-        except json.decoder.JSONDecodeError:
+            infotext = gen_info.get('infotexts')[index]
+            gen_parameters = infotext_utils.parse_generation_parameters(infotext, [])
+            res = int(gen_parameters.get('Variation seed' if is_subseed else 'Seed', -1))
+        except Exception:
             if gen_info_string:
-                errors.report(f"Error parsing JSON generation info: {gen_info_string}")
+                errors.report(f"Error retrieving seed from generation info: {gen_info_string}", exc_info=True)
 
         return [res, gr.update()]
 
diff --git a/modules/profiling.py b/modules/profiling.py
new file mode 100644
index 00000000000..95b59f71a20
--- /dev/null
+++ b/modules/profiling.py
@@ -0,0 +1,46 @@
+import torch
+
+from modules import shared, ui_gradio_extensions
+
+
+class Profiler:
+    def __init__(self):
+        if not shared.opts.profiling_enable:
+            self.profiler = None
+            return
+
+        activities = []
+        if "CPU" in shared.opts.profiling_activities:
+            activities.append(torch.profiler.ProfilerActivity.CPU)
+        if "CUDA" in shared.opts.profiling_activities:
+            activities.append(torch.profiler.ProfilerActivity.CUDA)
+
+        if not activities:
+            self.profiler = None
+            return
+
+        self.profiler = torch.profiler.profile(
+            activities=activities,
+            record_shapes=shared.opts.profiling_record_shapes,
+            profile_memory=shared.opts.profiling_profile_memory,
+            with_stack=shared.opts.profiling_with_stack
+        )
+
+    def __enter__(self):
+        if self.profiler:
+            self.profiler.__enter__()
+
+        return self
+
+    def __exit__(self, exc_type, exc, exc_tb):
+        if self.profiler:
+            shared.state.textinfo = "Finishing profile..."
+
+            self.profiler.__exit__(exc_type, exc, exc_tb)
+
+            self.profiler.export_chrome_trace(shared.opts.profiling_filename)
+
+
+def webpath():
+    return ui_gradio_extensions.webpath(shared.opts.profiling_filename)
+
diff --git a/modules/progress.py b/modules/progress.py
index 69921de7281..85255e821f5 100644
--- a/modules/progress.py
+++ b/modules/progress.py
@@ -8,10 +8,13 @@
 from modules.shared import opts
 
 import modules.shared as shared
-
+from collections import OrderedDict
+import string
+import random
+from typing import List
 
 current_task = None
-pending_tasks = {}
+pending_tasks = OrderedDict()
 finished_tasks = []
 recorded_results = []
 recorded_results_limit = 2
@@ -34,6 +37,11 @@ def finish_task(id_task):
     if len(finished_tasks) > 16:
         finished_tasks.pop(0)
 
+def create_task_id(task_type):
+    N = 7
+    res = ''.join(random.choices(string.ascii_uppercase +
+    string.digits, k=N))
+    return f"task({task_type}-{res})"
 
 def record_results(id_task, res):
     recorded_results.append((id_task, res))
@@ -44,6 +52,9 @@ def record_results(id_task, res):
 def add_task_to_queue(id_job):
     pending_tasks[id_job] = time.time()
 
+class PendingTasksResponse(BaseModel):
+    size: int = Field(title="Pending task size")
+    tasks: List[str] = Field(title="Pending task ids")
 
 class ProgressRequest(BaseModel):
     id_task: str = Field(default=None, title="Task ID", description="id of the task to get progress for")
@@ -63,9 +74,16 @@ class ProgressResponse(BaseModel):
 
 
 def setup_progress_api(app):
+    app.add_api_route("/internal/pending-tasks", get_pending_tasks, methods=["GET"])
     return app.add_api_route("/internal/progress", progressapi, methods=["POST"], response_model=ProgressResponse)
 
 
+def get_pending_tasks():
+    pending_tasks_ids = list(pending_tasks)
+    pending_len = len(pending_tasks_ids)
+    return PendingTasksResponse(size=pending_len, tasks=pending_tasks_ids)
+
+
 def progressapi(req: ProgressRequest):
     active = req.id_task == current_task
     queued = req.id_task in pending_tasks
diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py
index 334efeef317..4e393d2866f 100644
--- a/modules/prompt_parser.py
+++ b/modules/prompt_parser.py
@@ -2,10 +2,9 @@
 
 import re
 from collections import namedtuple
-from typing import List
 import lark
 
-# a prompt like this: "fantasy landscape with a [mountain:lake:0.25] and [an oak:a christmas tree:0.75][ in foreground::0.6][ in background:0.25] [shoddy:masterful:0.5]"
+# a prompt like this: "fantasy landscape with a [mountain:lake:0.25] and [an oak:a christmas tree:0.75][ in foreground::0.6][: in background:0.25] [shoddy:masterful:0.5]"
 # will be represented with prompt_schedule like this (assuming steps=100):
 # [25, 'fantasy landscape with a mountain and an oak in foreground shoddy']
 # [50, 'fantasy landscape with a lake and an oak in foreground in background shoddy']
@@ -240,14 +239,14 @@ def get_multicond_prompt_list(prompts: SdConditioning | list[str]):
 
 class ComposableScheduledPromptConditioning:
     def __init__(self, schedules, weight=1.0):
-        self.schedules: List[ScheduledPromptConditioning] = schedules
+        self.schedules: list[ScheduledPromptConditioning] = schedules
         self.weight: float = weight
 
 
 class MulticondLearnedConditioning:
     def __init__(self, shape, batch):
         self.shape: tuple = shape  # the shape field is needed to send this object to DDIM/PLMS
-        self.batch: List[List[ComposableScheduledPromptConditioning]] = batch
+        self.batch: list[list[ComposableScheduledPromptConditioning]] = batch
 
 
 def get_multicond_learned_conditioning(model, prompts, steps, hires_steps=None, use_old_scheduling=False) -> MulticondLearnedConditioning:
@@ -269,7 +268,7 @@ def get_multicond_learned_conditioning(model, prompts, steps, hires_steps=None,
 
 
 class DictWithShape(dict):
-    def __init__(self, x, shape):
+    def __init__(self, x, shape=None):
         super().__init__()
         self.update(x)
 
@@ -278,7 +277,7 @@ def shape(self):
         return self["crossattn"].shape
 
 
-def reconstruct_cond_batch(c: List[List[ScheduledPromptConditioning]], current_step):
+def reconstruct_cond_batch(c: list[list[ScheduledPromptConditioning]], current_step):
     param = c[0][0].cond
     is_dict = isinstance(param, dict)
 
diff --git a/modules/realesrgan_model.py b/modules/realesrgan_model.py
index 02841c30289..ff9d8ac0d69 100644
--- a/modules/realesrgan_model.py
+++ b/modules/realesrgan_model.py
@@ -1,12 +1,9 @@
 import os
 
-import numpy as np
-from PIL import Image
-from realesrgan import RealESRGANer
-
-from modules.upscaler import Upscaler, UpscalerData
-from modules.shared import cmd_opts, opts
 from modules import modelloader, errors
+from modules.shared import cmd_opts, opts
+from modules.upscaler import Upscaler, UpscalerData
+from modules.upscaler_utils import upscale_with_model
 
 
 class UpscalerRealESRGAN(Upscaler):
@@ -14,29 +11,20 @@ def __init__(self, path):
         self.name = "RealESRGAN"
         self.user_path = path
         super().__init__()
-        try:
-            from basicsr.archs.rrdbnet_arch import RRDBNet  # noqa: F401
-            from realesrgan import RealESRGANer  # noqa: F401
-            from realesrgan.archs.srvgg_arch import SRVGGNetCompact  # noqa: F401
-            self.enable = True
-            self.scalers = []
-            scalers = self.load_models(path)
+        self.enable = True
+        self.scalers = []
+        scalers = get_realesrgan_models(self)
 
-            local_model_paths = self.find_models(ext_filter=[".pth"])
-            for scaler in scalers:
-                if scaler.local_data_path.startswith("http"):
-                    filename = modelloader.friendly_name(scaler.local_data_path)
-                    local_model_candidates = [local_model for local_model in local_model_paths if local_model.endswith(f"{filename}.pth")]
-                    if local_model_candidates:
-                        scaler.local_data_path = local_model_candidates[0]
+        local_model_paths = self.find_models(ext_filter=[".pth"])
+        for scaler in scalers:
+            if scaler.local_data_path.startswith("http"):
+                filename = modelloader.friendly_name(scaler.local_data_path)
+                local_model_candidates = [local_model for local_model in local_model_paths if local_model.endswith(f"{filename}.pth")]
+                if local_model_candidates:
+                    scaler.local_data_path = local_model_candidates[0]
 
-                if scaler.name in opts.realesrgan_enabled_models:
-                    self.scalers.append(scaler)
-
-        except Exception:
-            errors.report("Error importing Real-ESRGAN", exc_info=True)
-            self.enable = False
-            self.scalers = []
+            if scaler.name in opts.realesrgan_enabled_models:
+                self.scalers.append(scaler)
 
     def do_upscale(self, img, path):
         if not self.enable:
@@ -48,20 +36,19 @@ def do_upscale(self, img, path):
             errors.report(f"Unable to load RealESRGAN model {path}", exc_info=True)
             return img
 
-        upsampler = RealESRGANer(
-            scale=info.scale,
-            model_path=info.local_data_path,
-            model=info.model(),
-            half=not cmd_opts.no_half and not cmd_opts.upcast_sampling,
-            tile=opts.ESRGAN_tile,
-            tile_pad=opts.ESRGAN_tile_overlap,
+        model_descriptor = modelloader.load_spandrel_model(
+            info.local_data_path,
             device=self.device,
+            prefer_half=(not cmd_opts.no_half and not cmd_opts.upcast_sampling),
+            expected_architecture="ESRGAN",  # "RealESRGAN" isn't a specific thing for Spandrel
+        )
+        return upscale_with_model(
+            model_descriptor,
+            img,
+            tile_size=opts.ESRGAN_tile,
+            tile_overlap=opts.ESRGAN_tile_overlap,
+            # TODO: `outscale`?
         )
-
-        upsampled = upsampler.enhance(np.array(img), outscale=info.scale)[0]
-
-        image = Image.fromarray(upsampled)
-        return image
 
     def load_model(self, path):
         for scaler in self.scalers:
@@ -76,58 +63,43 @@ def load_model(self, path):
                 return scaler
         raise ValueError(f"Unable to find model info: {path}")
 
-    def load_models(self, _):
-        return get_realesrgan_models(self)
-
 
-def get_realesrgan_models(scaler):
-    try:
-        from basicsr.archs.rrdbnet_arch import RRDBNet
-        from realesrgan.archs.srvgg_arch import SRVGGNetCompact
-        models = [
-            UpscalerData(
-                name="R-ESRGAN General 4xV3",
-                path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-general-x4v3.pth",
-                scale=4,
-                upscaler=scaler,
-                model=lambda: SRVGGNetCompact(num_in_ch=3, num_out_ch=3, num_feat=64, num_conv=32, upscale=4, act_type='prelu')
-            ),
-            UpscalerData(
-                name="R-ESRGAN General WDN 4xV3",
-                path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-general-wdn-x4v3.pth",
-                scale=4,
-                upscaler=scaler,
-                model=lambda: SRVGGNetCompact(num_in_ch=3, num_out_ch=3, num_feat=64, num_conv=32, upscale=4, act_type='prelu')
-            ),
-            UpscalerData(
-                name="R-ESRGAN AnimeVideo",
-                path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-animevideov3.pth",
-                scale=4,
-                upscaler=scaler,
-                model=lambda: SRVGGNetCompact(num_in_ch=3, num_out_ch=3, num_feat=64, num_conv=16, upscale=4, act_type='prelu')
-            ),
-            UpscalerData(
-                name="R-ESRGAN 4x+",
-                path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth",
-                scale=4,
-                upscaler=scaler,
-                model=lambda: RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=4)
-            ),
-            UpscalerData(
-                name="R-ESRGAN 4x+ Anime6B",
-                path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth",
-                scale=4,
-                upscaler=scaler,
-                model=lambda: RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=6, num_grow_ch=32, scale=4)
-            ),
-            UpscalerData(
-                name="R-ESRGAN 2x+",
-                path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.1/RealESRGAN_x2plus.pth",
-                scale=2,
-                upscaler=scaler,
-                model=lambda: RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=2)
-            ),
-        ]
-        return models
-    except Exception:
-        errors.report("Error making Real-ESRGAN models list", exc_info=True)
+def get_realesrgan_models(scaler: UpscalerRealESRGAN):
+    return [
+        UpscalerData(
+            name="R-ESRGAN General 4xV3",
+            path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-general-x4v3.pth",
+            scale=4,
+            upscaler=scaler,
+        ),
+        UpscalerData(
+            name="R-ESRGAN General WDN 4xV3",
+            path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-general-wdn-x4v3.pth",
+            scale=4,
+            upscaler=scaler,
+        ),
+        UpscalerData(
+            name="R-ESRGAN AnimeVideo",
+            path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-animevideov3.pth",
+            scale=4,
+            upscaler=scaler,
+        ),
+        UpscalerData(
+            name="R-ESRGAN 4x+",
+            path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth",
+            scale=4,
+            upscaler=scaler,
+        ),
+        UpscalerData(
+            name="R-ESRGAN 4x+ Anime6B",
+            path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth",
+            scale=4,
+            upscaler=scaler,
+        ),
+        UpscalerData(
+            name="R-ESRGAN 2x+",
+            path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.1/RealESRGAN_x2plus.pth",
+            scale=2,
+            upscaler=scaler,
+        ),
+    ]
diff --git a/modules/restart.py b/modules/restart.py
index 18eacaf377e..2dd6493b450 100644
--- a/modules/restart.py
+++ b/modules/restart.py
@@ -14,7 +14,9 @@ def is_restartable() -> bool:
 def restart_program() -> None:
     """creates file tmp/restart and immediately stops the process, which webui.bat/webui.sh interpret as a command to start webui again"""
 
-    (Path(script_path) / "tmp" / "restart").touch()
+    tmpdir = Path(script_path) / "tmp"
+    tmpdir.mkdir(parents=True, exist_ok=True)
+    (tmpdir / "restart").touch()
 
     stop_program()
 
diff --git a/modules/rng.py b/modules/rng.py
index 9e8ba2ee9d7..5390d1bb73e 100644
--- a/modules/rng.py
+++ b/modules/rng.py
@@ -34,7 +34,7 @@ def randn_local(seed, shape):
 
 
 def randn_like(x):
-    """Generate a tensor with random numbers from a normal distribution using the previously initialized genrator.
+    """Generate a tensor with random numbers from a normal distribution using the previously initialized generator.
 
     Use either randn() or manual_seed() to initialize the generator."""
 
@@ -48,7 +48,7 @@ def randn_like(x):
 
 
 def randn_without_seed(shape, generator=None):
-    """Generate a tensor with random numbers from a normal distribution using the previously initialized genrator.
+    """Generate a tensor with random numbers from a normal distribution using the previously initialized generator.
 
     Use either randn() or manual_seed() to initialize the generator."""
 
@@ -110,7 +110,7 @@ def __init__(self, shape, seeds, subseeds=None, subseed_strength=0.0, seed_resiz
         self.is_first = True
 
     def first(self):
-        noise_shape = self.shape if self.seed_resize_from_h <= 0 or self.seed_resize_from_w <= 0 else (self.shape[0], self.seed_resize_from_h // 8, self.seed_resize_from_w // 8)
+        noise_shape = self.shape if self.seed_resize_from_h <= 0 or self.seed_resize_from_w <= 0 else (self.shape[0], int(self.seed_resize_from_h) // 8, int(self.seed_resize_from_w // 8))
 
         xs = []
 
diff --git a/modules/safe.py b/modules/safe.py
index b1d08a7928e..af019ffd980 100644
--- a/modules/safe.py
+++ b/modules/safe.py
@@ -64,8 +64,8 @@ def find_class(self, module, name):
         raise Exception(f"global '{module}/{name}' is forbidden")
 
 
-# Regular expression that accepts 'dirname/version', 'dirname/data.pkl', and 'dirname/data/<number>'
-allowed_zip_names_re = re.compile(r"^([^/]+)/((data/\d+)|version|(data\.pkl))$")
+# Regular expression that accepts 'dirname/version', 'dirname/byteorder', 'dirname/data.pkl', '.data/serialization_id', and 'dirname/data/<number>'
+allowed_zip_names_re = re.compile(r"^([^/]+)/((data/\d+)|version|byteorder|.data/serialization_id|(data\.pkl))$")
 data_pkl_re = re.compile(r"^([^/]+)/data\.pkl$")
 
 def check_zip_filenames(filename, names):
diff --git a/modules/script_callbacks.py b/modules/script_callbacks.py
index c99695eb3d9..9059d4d9385 100644
--- a/modules/script_callbacks.py
+++ b/modules/script_callbacks.py
@@ -1,12 +1,14 @@
+from __future__ import annotations
+
+import dataclasses
 import inspect
 import os
-from collections import namedtuple
-from typing import Optional, Dict, Any
+from typing import Optional, Any
 
 from fastapi import FastAPI
 from gradio import Blocks
 
-from modules import errors, timer
+from modules import errors, timer, extensions, shared, util
 
 
 def report_exception(c, job):
@@ -41,7 +43,7 @@ def __init__(self, noise, x, xi):
 
 
 class CFGDenoiserParams:
-    def __init__(self, x, image_cond, sigma, sampling_step, total_sampling_steps, text_cond, text_uncond):
+    def __init__(self, x, image_cond, sigma, sampling_step, total_sampling_steps, text_cond, text_uncond, denoiser=None):
         self.x = x
         """Latent image representation in the process of being denoised"""
 
@@ -63,6 +65,9 @@ def __init__(self, x, image_cond, sigma, sampling_step, total_sampling_steps, te
         self.text_uncond = text_uncond
         """ Encoder hidden states of text conditioning from negative prompt"""
 
+        self.denoiser = denoiser
+        """Current CFGDenoiser object with processing parameters"""
+
 
 class CFGDenoisedParams:
     def __init__(self, x, sampling_step, total_sampling_steps, inner_model):
@@ -103,7 +108,114 @@ def __init__(self, imgs, cols, rows):
         self.rows = rows
 
 
-ScriptCallback = namedtuple("ScriptCallback", ["script", "callback"])
+@dataclasses.dataclass
+class BeforeTokenCounterParams:
+    prompt: str
+    steps: int
+    styles: list
+
+    is_positive: bool = True
+
+
+@dataclasses.dataclass
+class ScriptCallback:
+    script: str
+    callback: any
+    name: str = "unnamed"
+
+
+def add_callback(callbacks, fun, *, name=None, category='unknown', filename=None):
+    if filename is None:
+        stack = [x for x in inspect.stack() if x.filename != __file__]
+        filename = stack[0].filename if stack else 'unknown file'
+
+    extension = extensions.find_extension(filename)
+    extension_name = extension.canonical_name if extension else 'base'
+
+    callback_name = f"{extension_name}/{os.path.basename(filename)}/{category}"
+    if name is not None:
+        callback_name += f'/{name}'
+
+    unique_callback_name = callback_name
+    for index in range(1000):
+        existing = any(x.name == unique_callback_name for x in callbacks)
+        if not existing:
+            break
+
+        unique_callback_name = f'{callback_name}-{index+1}'
+
+    callbacks.append(ScriptCallback(filename, fun, unique_callback_name))
+
+
+def sort_callbacks(category, unordered_callbacks, *, enable_user_sort=True):
+    callbacks = unordered_callbacks.copy()
+    callback_lookup = {x.name: x for x in callbacks}
+    dependencies = {}
+
+    order_instructions = {}
+    for extension in extensions.extensions:
+        for order_instruction in extension.metadata.list_callback_order_instructions():
+            if order_instruction.name in callback_lookup:
+                if order_instruction.name not in order_instructions:
+                    order_instructions[order_instruction.name] = []
+
+                order_instructions[order_instruction.name].append(order_instruction)
+
+    if order_instructions:
+        for callback in callbacks:
+            dependencies[callback.name] = []
+
+        for callback in callbacks:
+            for order_instruction in order_instructions.get(callback.name, []):
+                for after in order_instruction.after:
+                    if after not in callback_lookup:
+                        continue
+
+                    dependencies[callback.name].append(after)
+
+                for before in order_instruction.before:
+                    if before not in callback_lookup:
+                        continue
+
+                    dependencies[before].append(callback.name)
+
+        sorted_names = util.topological_sort(dependencies)
+        callbacks = [callback_lookup[x] for x in sorted_names]
+
+    if enable_user_sort:
+        for name in reversed(getattr(shared.opts, 'prioritized_callbacks_' + category, [])):
+            index = next((i for i, callback in enumerate(callbacks) if callback.name == name), None)
+            if index is not None:
+                callbacks.insert(0, callbacks.pop(index))
+
+    return callbacks
+
+
+def ordered_callbacks(category, unordered_callbacks=None, *, enable_user_sort=True):
+    if unordered_callbacks is None:
+        unordered_callbacks = callback_map.get('callbacks_' + category, [])
+
+    if not enable_user_sort:
+        return sort_callbacks(category, unordered_callbacks, enable_user_sort=False)
+
+    callbacks = ordered_callbacks_map.get(category)
+    if callbacks is not None and len(callbacks) == len(unordered_callbacks):
+        return callbacks
+
+    callbacks = sort_callbacks(category, unordered_callbacks)
+
+    ordered_callbacks_map[category] = callbacks
+    return callbacks
+
+
+def enumerate_callbacks():
+    for category, callbacks in callback_map.items():
+        if category.startswith('callbacks_'):
+            category = category[10:]
+
+        yield category, callbacks
+
+
 callback_map = dict(
     callbacks_app_started=[],
     callbacks_model_loaded=[],
@@ -125,16 +237,21 @@ def __init__(self, imgs, cols, rows):
     callbacks_on_reload=[],
     callbacks_list_optimizers=[],
     callbacks_list_unets=[],
+    callbacks_before_token_counter=[],
 )
 
+ordered_callbacks_map = {}
+
 
 def clear_callbacks():
     for callback_list in callback_map.values():
         callback_list.clear()
 
+    ordered_callbacks_map.clear()
+
 
 def app_started_callback(demo: Optional[Blocks], app: FastAPI):
-    for c in callback_map['callbacks_app_started']:
+    for c in ordered_callbacks('app_started'):
         try:
             c.callback(demo, app)
             timer.startup_timer.record(os.path.basename(c.script))
@@ -143,7 +260,7 @@ def app_started_callback(demo: Optional[Blocks], app: FastAPI):
 
 
 def app_reload_callback():
-    for c in callback_map['callbacks_on_reload']:
+    for c in ordered_callbacks('on_reload'):
         try:
             c.callback()
         except Exception:
@@ -151,7 +268,7 @@ def app_reload_callback():
 
 
 def model_loaded_callback(sd_model):
-    for c in callback_map['callbacks_model_loaded']:
+    for c in ordered_callbacks('model_loaded'):
         try:
             c.callback(sd_model)
         except Exception:
@@ -161,7 +278,7 @@ def model_loaded_callback(sd_model):
 def ui_tabs_callback():
     res = []
 
-    for c in callback_map['callbacks_ui_tabs']:
+    for c in ordered_callbacks('ui_tabs'):
         try:
             res += c.callback() or []
         except Exception:
@@ -171,7 +288,7 @@ def ui_tabs_callback():
 
 
 def ui_train_tabs_callback(params: UiTrainTabParams):
-    for c in callback_map['callbacks_ui_train_tabs']:
+    for c in ordered_callbacks('ui_train_tabs'):
         try:
             c.callback(params)
         except Exception:
@@ -179,7 +296,7 @@ def ui_train_tabs_callback(params: UiTrainTabParams):
 
 
 def ui_settings_callback():
-    for c in callback_map['callbacks_ui_settings']:
+    for c in ordered_callbacks('ui_settings'):
         try:
             c.callback()
         except Exception:
@@ -187,7 +304,7 @@ def ui_settings_callback():
 
 
 def before_image_saved_callback(params: ImageSaveParams):
-    for c in callback_map['callbacks_before_image_saved']:
+    for c in ordered_callbacks('before_image_saved'):
         try:
             c.callback(params)
         except Exception:
@@ -195,7 +312,7 @@ def before_image_saved_callback(params: ImageSaveParams):
 
 
 def image_saved_callback(params: ImageSaveParams):
-    for c in callback_map['callbacks_image_saved']:
+    for c in ordered_callbacks('image_saved'):
         try:
             c.callback(params)
         except Exception:
@@ -203,7 +320,7 @@ def image_saved_callback(params: ImageSaveParams):
 
 
 def extra_noise_callback(params: ExtraNoiseParams):
-    for c in callback_map['callbacks_extra_noise']:
+    for c in ordered_callbacks('extra_noise'):
         try:
             c.callback(params)
         except Exception:
@@ -211,7 +328,7 @@ def extra_noise_callback(params: ExtraNoiseParams):
 
 
 def cfg_denoiser_callback(params: CFGDenoiserParams):
-    for c in callback_map['callbacks_cfg_denoiser']:
+    for c in ordered_callbacks('cfg_denoiser'):
         try:
             c.callback(params)
         except Exception:
@@ -219,7 +336,7 @@ def cfg_denoiser_callback(params: CFGDenoiserParams):
 
 
 def cfg_denoised_callback(params: CFGDenoisedParams):
-    for c in callback_map['callbacks_cfg_denoised']:
+    for c in ordered_callbacks('cfg_denoised'):
         try:
             c.callback(params)
         except Exception:
@@ -227,7 +344,7 @@ def cfg_denoised_callback(params: CFGDenoisedParams):
 
 
 def cfg_after_cfg_callback(params: AfterCFGCallbackParams):
-    for c in callback_map['callbacks_cfg_after_cfg']:
+    for c in ordered_callbacks('cfg_after_cfg'):
         try:
             c.callback(params)
         except Exception:
@@ -235,7 +352,7 @@ def cfg_after_cfg_callback(params: AfterCFGCallbackParams):
 
 
 def before_component_callback(component, **kwargs):
-    for c in callback_map['callbacks_before_component']:
+    for c in ordered_callbacks('before_component'):
         try:
             c.callback(component, **kwargs)
         except Exception:
@@ -243,7 +360,7 @@ def before_component_callback(component, **kwargs):
 
 
 def after_component_callback(component, **kwargs):
-    for c in callback_map['callbacks_after_component']:
+    for c in ordered_callbacks('after_component'):
         try:
             c.callback(component, **kwargs)
         except Exception:
@@ -251,15 +368,15 @@ def after_component_callback(component, **kwargs):
 
 
 def image_grid_callback(params: ImageGridLoopParams):
-    for c in callback_map['callbacks_image_grid']:
+    for c in ordered_callbacks('image_grid'):
         try:
             c.callback(params)
         except Exception:
             report_exception(c, 'image_grid')
 
 
-def infotext_pasted_callback(infotext: str, params: Dict[str, Any]):
-    for c in callback_map['callbacks_infotext_pasted']:
+def infotext_pasted_callback(infotext: str, params: dict[str, Any]):
+    for c in ordered_callbacks('infotext_pasted'):
         try:
             c.callback(infotext, params)
         except Exception:
@@ -267,7 +384,7 @@ def infotext_pasted_callback(infotext: str, params: Dict[str, Any]):
 
 
 def script_unloaded_callback():
-    for c in reversed(callback_map['callbacks_script_unloaded']):
+    for c in reversed(ordered_callbacks('script_unloaded')):
         try:
             c.callback()
         except Exception:
@@ -275,7 +392,7 @@ def script_unloaded_callback():
 
 
 def before_ui_callback():
-    for c in reversed(callback_map['callbacks_before_ui']):
+    for c in reversed(ordered_callbacks('before_ui')):
         try:
             c.callback()
         except Exception:
@@ -285,7 +402,7 @@ def before_ui_callback():
 def list_optimizers_callback():
     res = []
 
-    for c in callback_map['callbacks_list_optimizers']:
+    for c in ordered_callbacks('list_optimizers'):
         try:
             c.callback(res)
         except Exception:
@@ -297,7 +414,7 @@ def list_optimizers_callback():
 def list_unets_callback():
     res = []
 
-    for c in callback_map['callbacks_list_unets']:
+    for c in ordered_callbacks('list_unets'):
         try:
             c.callback(res)
         except Exception:
@@ -306,11 +423,12 @@ def list_unets_callback():
     return res
 
 
-def add_callback(callbacks, fun):
-    stack = [x for x in inspect.stack() if x.filename != __file__]
-    filename = stack[0].filename if stack else 'unknown file'
-
-    callbacks.append(ScriptCallback(filename, fun))
+def before_token_counter_callback(params: BeforeTokenCounterParams):
+    for c in ordered_callbacks('before_token_counter'):
+        try:
+            c.callback(params)
+        except Exception:
+            report_exception(c, 'before_token_counter')
 
 
 def remove_current_script_callbacks():
@@ -321,32 +439,38 @@ def remove_current_script_callbacks():
     for callback_list in callback_map.values():
         for callback_to_remove in [cb for cb in callback_list if cb.script == filename]:
             callback_list.remove(callback_to_remove)
+    for ordered_callbacks_list in ordered_callbacks_map.values():
+        for callback_to_remove in [cb for cb in ordered_callbacks_list if cb.script == filename]:
+            ordered_callbacks_list.remove(callback_to_remove)
 
 
 def remove_callbacks_for_function(callback_func):
     for callback_list in callback_map.values():
         for callback_to_remove in [cb for cb in callback_list if cb.callback == callback_func]:
             callback_list.remove(callback_to_remove)
+    for ordered_callback_list in ordered_callbacks_map.values():
+        for callback_to_remove in [cb for cb in ordered_callback_list if cb.callback == callback_func]:
+            ordered_callback_list.remove(callback_to_remove)
 
 
-def on_app_started(callback):
+def on_app_started(callback, *, name=None):
     """register a function to be called when the webui started, the gradio `Block` component and
     fastapi `FastAPI` object are passed as the arguments"""
-    add_callback(callback_map['callbacks_app_started'], callback)
+    add_callback(callback_map['callbacks_app_started'], callback, name=name, category='app_started')
 
 
-def on_before_reload(callback):
+def on_before_reload(callback, *, name=None):
     """register a function to be called just before the server reloads."""
-    add_callback(callback_map['callbacks_on_reload'], callback)
+    add_callback(callback_map['callbacks_on_reload'], callback, name=name, category='on_reload')
 
 
-def on_model_loaded(callback):
+def on_model_loaded(callback, *, name=None):
     """register a function to be called when the stable diffusion model is created; the model is
     passed as an argument; this function is also called when the script is reloaded. """
-    add_callback(callback_map['callbacks_model_loaded'], callback)
+    add_callback(callback_map['callbacks_model_loaded'], callback, name=name, category='model_loaded')
 
 
-def on_ui_tabs(callback):
+def on_ui_tabs(callback, *, name=None):
     """register a function to be called when the UI is creating new tabs.
     The function must either return a None, which means no new tabs to be added, or a list, where
     each element is a tuple:
@@ -356,71 +480,71 @@ def on_ui_tabs(callback):
     title is tab text displayed to user in the UI
     elem_id is HTML id for the tab
     """
-    add_callback(callback_map['callbacks_ui_tabs'], callback)
+    add_callback(callback_map['callbacks_ui_tabs'], callback, name=name, category='ui_tabs')
 
 
-def on_ui_train_tabs(callback):
+def on_ui_train_tabs(callback, *, name=None):
     """register a function to be called when the UI is creating new tabs for the train tab.
     Create your new tabs with gr.Tab.
     """
-    add_callback(callback_map['callbacks_ui_train_tabs'], callback)
+    add_callback(callback_map['callbacks_ui_train_tabs'], callback, name=name, category='ui_train_tabs')
 
 
-def on_ui_settings(callback):
+def on_ui_settings(callback, *, name=None):
     """register a function to be called before UI settings are populated; add your settings
     by using shared.opts.add_option(shared.OptionInfo(...)) """
-    add_callback(callback_map['callbacks_ui_settings'], callback)
+    add_callback(callback_map['callbacks_ui_settings'], callback, name=name, category='ui_settings')
 
 
-def on_before_image_saved(callback):
+def on_before_image_saved(callback, *, name=None):
     """register a function to be called before an image is saved to a file.
     The callback is called with one argument:
         - params: ImageSaveParams - parameters the image is to be saved with. You can change fields in this object.
     """
-    add_callback(callback_map['callbacks_before_image_saved'], callback)
+    add_callback(callback_map['callbacks_before_image_saved'], callback, name=name, category='before_image_saved')
 
 
-def on_image_saved(callback):
+def on_image_saved(callback, *, name=None):
     """register a function to be called after an image is saved to a file.
     The callback is called with one argument:
         - params: ImageSaveParams - parameters the image was saved with. Changing fields in this object does nothing.
     """
-    add_callback(callback_map['callbacks_image_saved'], callback)
+    add_callback(callback_map['callbacks_image_saved'], callback, name=name, category='image_saved')
 
 
-def on_extra_noise(callback):
+def on_extra_noise(callback, *, name=None):
     """register a function to be called before adding extra noise in img2img or hires fix;
     The callback is called with one argument:
         - params: ExtraNoiseParams - contains noise determined by seed and latent representation of image
     """
-    add_callback(callback_map['callbacks_extra_noise'], callback)
+    add_callback(callback_map['callbacks_extra_noise'], callback, name=name, category='extra_noise')
 
 
-def on_cfg_denoiser(callback):
+def on_cfg_denoiser(callback, *, name=None):
     """register a function to be called in the kdiffussion cfg_denoiser method after building the inner model inputs.
     The callback is called with one argument:
         - params: CFGDenoiserParams - parameters to be passed to the inner model and sampling state details.
     """
-    add_callback(callback_map['callbacks_cfg_denoiser'], callback)
+    add_callback(callback_map['callbacks_cfg_denoiser'], callback, name=name, category='cfg_denoiser')
 
 
-def on_cfg_denoised(callback):
+def on_cfg_denoised(callback, *, name=None):
     """register a function to be called in the kdiffussion cfg_denoiser method after building the inner model inputs.
     The callback is called with one argument:
         - params: CFGDenoisedParams - parameters to be passed to the inner model and sampling state details.
     """
-    add_callback(callback_map['callbacks_cfg_denoised'], callback)
+    add_callback(callback_map['callbacks_cfg_denoised'], callback, name=name, category='cfg_denoised')
 
 
-def on_cfg_after_cfg(callback):
+def on_cfg_after_cfg(callback, *, name=None):
     """register a function to be called in the kdiffussion cfg_denoiser method after cfg calculations are completed.
     The callback is called with one argument:
         - params: AfterCFGCallbackParams - parameters to be passed to the script for post-processing after cfg calculation.
     """
-    add_callback(callback_map['callbacks_cfg_after_cfg'], callback)
+    add_callback(callback_map['callbacks_cfg_after_cfg'], callback, name=name, category='cfg_after_cfg')
 
 
-def on_before_component(callback):
+def on_before_component(callback, *, name=None):
     """register a function to be called before a component is created.
     The callback is called with arguments:
         - component - gradio component that is about to be created.
@@ -429,54 +553,61 @@ def on_before_component(callback):
     Use elem_id/label fields of kwargs to figure out which component it is.
     This can be useful to inject your own components somewhere in the middle of vanilla UI.
     """
-    add_callback(callback_map['callbacks_before_component'], callback)
+    add_callback(callback_map['callbacks_before_component'], callback, name=name, category='before_component')
 
 
-def on_after_component(callback):
+def on_after_component(callback, *, name=None):
     """register a function to be called after a component is created. See on_before_component for more."""
-    add_callback(callback_map['callbacks_after_component'], callback)
+    add_callback(callback_map['callbacks_after_component'], callback, name=name, category='after_component')
 
 
-def on_image_grid(callback):
+def on_image_grid(callback, *, name=None):
     """register a function to be called before making an image grid.
     The callback is called with one argument:
        - params: ImageGridLoopParams - parameters to be used for grid creation. Can be modified.
     """
-    add_callback(callback_map['callbacks_image_grid'], callback)
+    add_callback(callback_map['callbacks_image_grid'], callback, name=name, category='image_grid')
 
 
-def on_infotext_pasted(callback):
+def on_infotext_pasted(callback, *, name=None):
     """register a function to be called before applying an infotext.
     The callback is called with two arguments:
        - infotext: str - raw infotext.
-       - result: Dict[str, any] - parsed infotext parameters.
+       - result: dict[str, any] - parsed infotext parameters.
     """
-    add_callback(callback_map['callbacks_infotext_pasted'], callback)
+    add_callback(callback_map['callbacks_infotext_pasted'], callback, name=name, category='infotext_pasted')
 
 
-def on_script_unloaded(callback):
+def on_script_unloaded(callback, *, name=None):
     """register a function to be called before the script is unloaded. Any hooks/hijacks/monkeying about that
     the script did should be reverted here"""
 
-    add_callback(callback_map['callbacks_script_unloaded'], callback)
+    add_callback(callback_map['callbacks_script_unloaded'], callback, name=name, category='script_unloaded')
 
 
-def on_before_ui(callback):
+def on_before_ui(callback, *, name=None):
     """register a function to be called before the UI is created."""
 
-    add_callback(callback_map['callbacks_before_ui'], callback)
+    add_callback(callback_map['callbacks_before_ui'], callback, name=name, category='before_ui')
 
 
-def on_list_optimizers(callback):
+def on_list_optimizers(callback, *, name=None):
     """register a function to be called when UI is making a list of cross attention optimization options.
     The function will be called with one argument, a list, and shall add objects of type modules.sd_hijack_optimizations.SdOptimization
     to it."""
 
-    add_callback(callback_map['callbacks_list_optimizers'], callback)
+    add_callback(callback_map['callbacks_list_optimizers'], callback, name=name, category='list_optimizers')
 
 
-def on_list_unets(callback):
+def on_list_unets(callback, *, name=None):
     """register a function to be called when UI is making a list of alternative options for unet.
     The function will be called with one argument, a list, and shall add objects of type modules.sd_unet.SdUnetOption to it."""
 
-    add_callback(callback_map['callbacks_list_unets'], callback)
+    add_callback(callback_map['callbacks_list_unets'], callback, name=name, category='list_unets')
+
+
+def on_before_token_counter(callback, *, name=None):
+    """register a function to be called when UI is counting tokens for a prompt.
+    The function will be called with one argument of type BeforeTokenCounterParams, and should modify its fields if necessary."""
+
+    add_callback(callback_map['callbacks_before_token_counter'], callback, name=name, category='before_token_counter')
diff --git a/modules/script_loading.py b/modules/script_loading.py
index 0d55f1932ee..cccb309665f 100644
--- a/modules/script_loading.py
+++ b/modules/script_loading.py
@@ -4,11 +4,15 @@
 from modules import errors
 
 
+loaded_scripts = {}
+
+
 def load_module(path):
     module_spec = importlib.util.spec_from_file_location(os.path.basename(path), path)
     module = importlib.util.module_from_spec(module_spec)
     module_spec.loader.exec_module(module)
 
+    loaded_scripts[path] = module
     return module
 
 
diff --git a/modules/scripts.py b/modules/scripts.py
index e8518ad0fba..8eca396b140 100644
--- a/modules/scripts.py
+++ b/modules/scripts.py
@@ -7,15 +7,37 @@
 
 import gradio as gr
 
-from modules import shared, paths, script_callbacks, extensions, script_loading, scripts_postprocessing, errors, timer
+from modules import shared, paths, script_callbacks, extensions, script_loading, scripts_postprocessing, errors, timer, util
+
+topological_sort = util.topological_sort
 
 AlwaysVisible = object()
 
+class MaskBlendArgs:
+    def __init__(self, current_latent, nmask, init_latent, mask, blended_latent, denoiser=None, sigma=None):
+        self.current_latent = current_latent
+        self.nmask = nmask
+        self.init_latent = init_latent
+        self.mask = mask
+        self.blended_latent = blended_latent
+
+        self.denoiser = denoiser
+        self.is_final_blend = denoiser is None
+        self.sigma = sigma
+
+class PostSampleArgs:
+    def __init__(self, samples):
+        self.samples = samples
 
 class PostprocessImageArgs:
     def __init__(self, image):
         self.image = image
 
+class PostProcessMaskOverlayArgs:
+    def __init__(self, index, mask_for_overlay, overlay_image):
+        self.index = index
+        self.mask_for_overlay = mask_for_overlay
+        self.overlay_image = overlay_image
 
 class PostprocessBatchListArgs:
     def __init__(self, images):
@@ -71,6 +93,9 @@ class Script:
     setup_for_ui_only = False
     """If true, the script setup will only be run in Gradio UI, not in API"""
 
+    controls = None
+    """A list of controls returned by the ui()."""
+
     def title(self):
         """this function should return the title of the script. This is what will be displayed in the dropdown menu."""
 
@@ -86,7 +111,7 @@ def ui(self, is_img2img):
 
     def show(self, is_img2img):
         """
-        is_img2img is True if this function is called for the img2img interface, and Fasle otherwise
+        is_img2img is True if this function is called for the img2img interface, and False otherwise
 
         This function should return:
          - False if the script should not be shown in UI at all
@@ -115,7 +140,6 @@ def setup(self, p, *args):
         """
         pass
 
-
     def before_process(self, p, *args):
         """
         This function is called very early during processing begins for AlwaysVisible scripts.
@@ -163,6 +187,13 @@ def after_extra_networks_activate(self, p, *args, **kwargs):
         """
         pass
 
+    def process_before_every_sampling(self, p, *args, **kwargs):
+        """
+        Similar to process(), called before every sampling.
+        If you use high-res fix, this will be called two times.
+        """
+        pass
+
     def process_batch(self, p, *args, **kwargs):
         """
         Same as process(), but called for every batch.
@@ -206,6 +237,25 @@ def postprocess_batch_list(self, p, pp: PostprocessBatchListArgs, *args, **kwarg
 
         pass
 
+    def on_mask_blend(self, p, mba: MaskBlendArgs, *args):
+        """
+        Called in inpainting mode when the original content is blended with the inpainted content.
+        This is called at every step in the denoising process and once at the end.
+        If is_final_blend is true, this is called for the final blending stage.
+        Otherwise, denoiser and sigma are defined and may be used to inform the procedure.
+        """
+
+        pass
+
+    def post_sample(self, p, ps: PostSampleArgs, *args):
+        """
+        Called after the samples have been generated,
+        but before they have been decoded by the VAE, if applicable.
+        Check getattr(samples, 'already_decoded', False) to test if the images are decoded.
+        """
+
+        pass
+
     def postprocess_image(self, p, pp: PostprocessImageArgs, *args):
         """
         Called for every image after it has been generated.
@@ -213,6 +263,22 @@ def postprocess_image(self, p, pp: PostprocessImageArgs, *args):
 
         pass
 
+    def postprocess_maskoverlay(self, p, ppmo: PostProcessMaskOverlayArgs, *args):
+        """
+        Called for every image after it has been generated.
+        """
+
+        pass
+
+    def postprocess_image_after_composite(self, p, pp: PostprocessImageArgs, *args):
+        """
+        Called for every image after it has been generated.
+        Same as postprocess_image but after inpaint_full_res composite
+        So that it operates on the full image instead of the inpaint_full_res crop region.
+        """
+
+        pass
+
     def postprocess(self, p, processed, *args):
         """
         This function is called after processing ends for AlwaysVisible scripts.
@@ -293,6 +359,9 @@ def elem_id(self, item_id):
 
         return f'{tabname}{item_id}'
 
+    def show(self, is_img2img):
+        return AlwaysVisible
+
 
 current_basedir = paths.script_path
 
@@ -312,19 +381,89 @@ def basedir():
 ScriptClassData = namedtuple("ScriptClassData", ["script_class", "path", "basedir", "module"])
 
 
+@dataclass
+class ScriptWithDependencies:
+    script_canonical_name: str
+    file: ScriptFile
+    requires: list
+    load_before: list
+    load_after: list
+
+
 def list_scripts(scriptdirname, extension, *, include_extensions=True):
-    scripts_list = []
+    scripts = {}
 
-    basedir = os.path.join(paths.script_path, scriptdirname)
-    if os.path.exists(basedir):
-        for filename in sorted(os.listdir(basedir)):
-            scripts_list.append(ScriptFile(paths.script_path, filename, os.path.join(basedir, filename)))
+    loaded_extensions = {ext.canonical_name: ext for ext in extensions.active()}
+    loaded_extensions_scripts = {ext.canonical_name: [] for ext in extensions.active()}
+
+    # build script dependency map
+    root_script_basedir = os.path.join(paths.script_path, scriptdirname)
+    if os.path.exists(root_script_basedir):
+        for filename in sorted(os.listdir(root_script_basedir)):
+            if not os.path.isfile(os.path.join(root_script_basedir, filename)):
+                continue
+
+            if os.path.splitext(filename)[1].lower() != extension:
+                continue
+
+            script_file = ScriptFile(paths.script_path, filename, os.path.join(root_script_basedir, filename))
+            scripts[filename] = ScriptWithDependencies(filename, script_file, [], [], [])
 
     if include_extensions:
         for ext in extensions.active():
-            scripts_list += ext.list_files(scriptdirname, extension)
-
-    scripts_list = [x for x in scripts_list if os.path.splitext(x.path)[1].lower() == extension and os.path.isfile(x.path)]
+            extension_scripts_list = ext.list_files(scriptdirname, extension)
+            for extension_script in extension_scripts_list:
+                if not os.path.isfile(extension_script.path):
+                    continue
+
+                script_canonical_name = ("builtin/" if ext.is_builtin else "") + ext.canonical_name + "/" + extension_script.filename
+                relative_path = scriptdirname + "/" + extension_script.filename
+
+                script = ScriptWithDependencies(
+                    script_canonical_name=script_canonical_name,
+                    file=extension_script,
+                    requires=ext.metadata.get_script_requirements("Requires", relative_path, scriptdirname),
+                    load_before=ext.metadata.get_script_requirements("Before", relative_path, scriptdirname),
+                    load_after=ext.metadata.get_script_requirements("After", relative_path, scriptdirname),
+                )
+
+                scripts[script_canonical_name] = script
+                loaded_extensions_scripts[ext.canonical_name].append(script)
+
+    for script_canonical_name, script in scripts.items():
+        # load before requires inverse dependency
+        # in this case, append the script name into the load_after list of the specified script
+        for load_before in script.load_before:
+            # if this requires an individual script to be loaded before
+            other_script = scripts.get(load_before)
+            if other_script:
+                other_script.load_after.append(script_canonical_name)
+
+            # if this requires an extension
+            other_extension_scripts = loaded_extensions_scripts.get(load_before)
+            if other_extension_scripts:
+                for other_script in other_extension_scripts:
+                    other_script.load_after.append(script_canonical_name)
+
+        # if After mentions an extension, remove it and instead add all of its scripts
+        for load_after in list(script.load_after):
+            if load_after not in scripts and load_after in loaded_extensions_scripts:
+                script.load_after.remove(load_after)
+
+                for other_script in loaded_extensions_scripts.get(load_after, []):
+                    script.load_after.append(other_script.script_canonical_name)
+
+    dependencies = {}
+
+    for script_canonical_name, script in scripts.items():
+        for required_script in script.requires:
+            if required_script not in scripts and required_script not in loaded_extensions:
+                errors.report(f'Script "{script_canonical_name}" requires "{required_script}" to be loaded, but it is not.', exc_info=False)
+
+        dependencies[script_canonical_name] = script.load_after
+
+    ordered_scripts = topological_sort(dependencies)
+    scripts_list = [scripts[script_canonical_name].file for script_canonical_name in ordered_scripts]
 
     return scripts_list
 
@@ -365,15 +504,9 @@ def register_scripts_from_module(module):
             elif issubclass(script_class, scripts_postprocessing.ScriptPostprocessing):
                 postprocessing_scripts_data.append(ScriptClassData(script_class, scriptfile.path, scriptfile.basedir, module))
 
-    def orderby(basedir):
-        # 1st webui, 2nd extensions-builtin, 3rd extensions
-        priority = {os.path.join(paths.script_path, "extensions-builtin"):1, paths.script_path:0}
-        for key in priority:
-            if basedir.startswith(key):
-                return priority[key]
-        return 9999
-
-    for scriptfile in sorted(scripts_list, key=lambda x: [orderby(x.basedir), x]):
+    # here the scripts_list is already ordered
+    # processing_script is not considered though
+    for scriptfile in scripts_list:
         try:
             if scriptfile.basedir != paths.script_path:
                 sys.path = [scriptfile.basedir] + sys.path
@@ -417,6 +550,25 @@ def __init__(self):
         self.paste_field_names = []
         self.inputs = [None]
 
+        self.callback_map = {}
+        self.callback_names = [
+            'before_process',
+            'process',
+            'before_process_batch',
+            'after_extra_networks_activate',
+            'process_batch',
+            'postprocess',
+            'postprocess_batch',
+            'postprocess_batch_list',
+            'post_sample',
+            'on_mask_blend',
+            'postprocess_image',
+            'postprocess_maskoverlay',
+            'postprocess_image_after_composite',
+            'before_component',
+            'after_component',
+        ]
+
         self.on_before_component_elem_id = {}
         """dict of callbacks to be called before an element is created; key=elem_id, value=list of callbacks"""
 
@@ -433,7 +585,12 @@ def initialize_scripts(self, is_img2img):
         auto_processing_scripts = scripts_auto_postprocessing.create_auto_preprocessing_script_data()
 
         for script_data in auto_processing_scripts + scripts_data:
-            script = script_data.script_class()
+            try:
+                script = script_data.script_class()
+            except Exception:
+                errors.report(f"Error # failed to initialize Script {script_data.module}: ", exc_info=True)
+                continue
+
             script.filename = script_data.path
             script.is_txt2img = not is_img2img
             script.is_img2img = is_img2img
@@ -450,6 +607,8 @@ def initialize_scripts(self, is_img2img):
                 self.scripts.append(script)
                 self.selectable_scripts.append(script)
 
+        self.callback_map.clear()
+
         self.apply_on_before_component_callbacks()
 
     def apply_on_before_component_callbacks(self):
@@ -473,17 +632,26 @@ def apply_on_before_component_callbacks(self):
             on_after.clear()
 
     def create_script_ui(self, script):
-        import modules.api.models as api_models
 
         script.args_from = len(self.inputs)
         script.args_to = len(self.inputs)
 
+        try:
+            self.create_script_ui_inner(script)
+        except Exception:
+            errors.report(f"Error creating UI for {script.name}: ", exc_info=True)
+
+    def create_script_ui_inner(self, script):
+        import modules.api.models as api_models
+
         controls = wrap_call(script.ui, script.filename, "ui", script.is_img2img)
+        script.controls = controls
 
         if controls is None:
             return
 
         script.name = wrap_call(script.title, script.filename, "title", default=script.filename).lower()
+
         api_args = []
 
         for control in controls:
@@ -491,11 +659,15 @@ def create_script_ui(self, script):
 
             arg_info = api_models.ScriptArg(label=control.label or "")
 
-            for field in ("value", "minimum", "maximum", "step", "choices"):
+            for field in ("value", "minimum", "maximum", "step"):
                 v = getattr(control, field, None)
                 if v is not None:
                     setattr(arg_info, field, v)
 
+            choices = getattr(control, 'choices', None)  # as of gradio 3.41, some items in choices are strings, and some are tuples where the first elem is the string
+            if choices is not None:
+                arg_info.choices = [x[0] if isinstance(x, tuple) else x for x in choices]
+
             api_args.append(arg_info)
 
         script.api_info = api_models.ScriptInfo(
@@ -546,6 +718,8 @@ def setup_ui(self):
         self.setup_ui_for_section(None, self.selectable_scripts)
 
         def select_script(script_index):
+            if script_index is None:
+                script_index = 0
             selected_script = self.selectable_scripts[script_index - 1] if script_index>0 else None
 
             return [gr.update(visible=selected_script == s) for s in self.selectable_scripts]
@@ -572,12 +746,17 @@ def init_field(title):
         def onload_script_visibility(params):
             title = params.get('Script', None)
             if title:
-                title_index = self.titles.index(title)
-                visibility = title_index == self.script_load_ctr
-                self.script_load_ctr = (self.script_load_ctr + 1) % len(self.titles)
-                return gr.update(visible=visibility)
-            else:
-                return gr.update(visible=False)
+                try:
+                    title_index = self.titles.index(title)
+                    visibility = title_index == self.script_load_ctr
+                    self.script_load_ctr = (self.script_load_ctr + 1) % len(self.titles)
+                    return gr.update(visible=visibility)
+                except ValueError:
+                    params['Script'] = None
+                    massage = f'Cannot find Script: "{title}"'
+                    print(massage)
+                    gr.Warning(massage)
+            return gr.update(visible=False)
 
         self.infotext_fields.append((dropdown, lambda x: gr.update(value=x.get('Script', 'None'))))
         self.infotext_fields.extend([(script.group, onload_script_visibility) for script in self.selectable_scripts])
@@ -589,7 +768,7 @@ def onload_script_visibility(params):
     def run(self, p, *args):
         script_index = args[0]
 
-        if script_index == 0:
+        if script_index == 0 or script_index is None:
             return None
 
         script = self.selectable_scripts[script_index-1]
@@ -604,8 +783,42 @@ def run(self, p, *args):
 
         return processed
 
+    def list_scripts_for_method(self, method_name):
+        if method_name in ('before_component', 'after_component'):
+            return self.scripts
+        else:
+            return self.alwayson_scripts
+
+    def create_ordered_callbacks_list(self,  method_name, *, enable_user_sort=True):
+        script_list = self.list_scripts_for_method(method_name)
+        category = f'script_{method_name}'
+        callbacks = []
+
+        for script in script_list:
+            if getattr(script.__class__, method_name, None) == getattr(Script, method_name, None):
+                continue
+
+            script_callbacks.add_callback(callbacks, script, category=category, name=script.__class__.__name__, filename=script.filename)
+
+        return script_callbacks.sort_callbacks(category, callbacks, enable_user_sort=enable_user_sort)
+
+    def ordered_callbacks(self, method_name, *, enable_user_sort=True):
+        script_list = self.list_scripts_for_method(method_name)
+        category = f'script_{method_name}'
+
+        scrpts_len, callbacks = self.callback_map.get(category, (-1, None))
+
+        if callbacks is None or scrpts_len != len(script_list):
+            callbacks = self.create_ordered_callbacks_list(method_name, enable_user_sort=enable_user_sort)
+            self.callback_map[category] = len(script_list), callbacks
+
+        return callbacks
+
+    def ordered_scripts(self, method_name):
+        return [x.callback for x in self.ordered_callbacks(method_name)]
+
     def before_process(self, p):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('before_process'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.before_process(p, *script_args)
@@ -613,15 +826,23 @@ def before_process(self, p):
                 errors.report(f"Error running before_process: {script.filename}", exc_info=True)
 
     def process(self, p):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('process'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.process(p, *script_args)
             except Exception:
                 errors.report(f"Error running process: {script.filename}", exc_info=True)
 
+    def process_before_every_sampling(self, p, **kwargs):
+        for script in self.ordered_scripts('process_before_every_sampling'):
+            try:
+                script_args = p.script_args[script.args_from:script.args_to]
+                script.process_before_every_sampling(p, *script_args, **kwargs)
+            except Exception:
+                errors.report(f"Error running process_before_every_sampling: {script.filename}", exc_info=True)
+
     def before_process_batch(self, p, **kwargs):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('before_process_batch'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.before_process_batch(p, *script_args, **kwargs)
@@ -629,7 +850,7 @@ def before_process_batch(self, p, **kwargs):
                 errors.report(f"Error running before_process_batch: {script.filename}", exc_info=True)
 
     def after_extra_networks_activate(self, p, **kwargs):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('after_extra_networks_activate'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.after_extra_networks_activate(p, *script_args, **kwargs)
@@ -637,7 +858,7 @@ def after_extra_networks_activate(self, p, **kwargs):
                 errors.report(f"Error running after_extra_networks_activate: {script.filename}", exc_info=True)
 
     def process_batch(self, p, **kwargs):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('process_batch'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.process_batch(p, *script_args, **kwargs)
@@ -645,7 +866,7 @@ def process_batch(self, p, **kwargs):
                 errors.report(f"Error running process_batch: {script.filename}", exc_info=True)
 
     def postprocess(self, p, processed):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('postprocess'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.postprocess(p, processed, *script_args)
@@ -653,7 +874,7 @@ def postprocess(self, p, processed):
                 errors.report(f"Error running postprocess: {script.filename}", exc_info=True)
 
     def postprocess_batch(self, p, images, **kwargs):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('postprocess_batch'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.postprocess_batch(p, *script_args, images=images, **kwargs)
@@ -661,21 +882,53 @@ def postprocess_batch(self, p, images, **kwargs):
                 errors.report(f"Error running postprocess_batch: {script.filename}", exc_info=True)
 
     def postprocess_batch_list(self, p, pp: PostprocessBatchListArgs, **kwargs):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('postprocess_batch_list'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.postprocess_batch_list(p, pp, *script_args, **kwargs)
             except Exception:
                 errors.report(f"Error running postprocess_batch_list: {script.filename}", exc_info=True)
 
+    def post_sample(self, p, ps: PostSampleArgs):
+        for script in self.ordered_scripts('post_sample'):
+            try:
+                script_args = p.script_args[script.args_from:script.args_to]
+                script.post_sample(p, ps, *script_args)
+            except Exception:
+                errors.report(f"Error running post_sample: {script.filename}", exc_info=True)
+
+    def on_mask_blend(self, p, mba: MaskBlendArgs):
+        for script in self.ordered_scripts('on_mask_blend'):
+            try:
+                script_args = p.script_args[script.args_from:script.args_to]
+                script.on_mask_blend(p, mba, *script_args)
+            except Exception:
+                errors.report(f"Error running post_sample: {script.filename}", exc_info=True)
+
     def postprocess_image(self, p, pp: PostprocessImageArgs):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('postprocess_image'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.postprocess_image(p, pp, *script_args)
             except Exception:
                 errors.report(f"Error running postprocess_image: {script.filename}", exc_info=True)
 
+    def postprocess_maskoverlay(self, p, ppmo: PostProcessMaskOverlayArgs):
+        for script in self.ordered_scripts('postprocess_maskoverlay'):
+            try:
+                script_args = p.script_args[script.args_from:script.args_to]
+                script.postprocess_maskoverlay(p, ppmo, *script_args)
+            except Exception:
+                errors.report(f"Error running postprocess_image: {script.filename}", exc_info=True)
+
+    def postprocess_image_after_composite(self, p, pp: PostprocessImageArgs):
+        for script in self.ordered_scripts('postprocess_image_after_composite'):
+            try:
+                script_args = p.script_args[script.args_from:script.args_to]
+                script.postprocess_image_after_composite(p, pp, *script_args)
+            except Exception:
+                errors.report(f"Error running postprocess_image_after_composite: {script.filename}", exc_info=True)
+
     def before_component(self, component, **kwargs):
         for callback, script in self.on_before_component_elem_id.get(kwargs.get("elem_id"), []):
             try:
@@ -683,7 +936,7 @@ def before_component(self, component, **kwargs):
             except Exception:
                 errors.report(f"Error running on_before_component: {script.filename}", exc_info=True)
 
-        for script in self.scripts:
+        for script in self.ordered_scripts('before_component'):
             try:
                 script.before_component(component, **kwargs)
             except Exception:
@@ -696,7 +949,7 @@ def after_component(self, component, **kwargs):
             except Exception:
                 errors.report(f"Error running on_after_component: {script.filename}", exc_info=True)
 
-        for script in self.scripts:
+        for script in self.ordered_scripts('after_component'):
             try:
                 script.after_component(component, **kwargs)
             except Exception:
@@ -724,7 +977,7 @@ def reload_sources(self, cache):
                     self.scripts[si].args_to = args_to
 
     def before_hr(self, p):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('before_hr'):
             try:
                 script_args = p.script_args[script.args_from:script.args_to]
                 script.before_hr(p, *script_args)
@@ -732,7 +985,7 @@ def before_hr(self, p):
                 errors.report(f"Error running before_hr: {script.filename}", exc_info=True)
 
     def setup_scrips(self, p, *, is_ui=True):
-        for script in self.alwayson_scripts:
+        for script in self.ordered_scripts('setup'):
             if not is_ui and script.setup_for_ui_only:
                 continue
 
@@ -742,6 +995,35 @@ def setup_scrips(self, p, *, is_ui=True):
             except Exception:
                 errors.report(f"Error running setup: {script.filename}", exc_info=True)
 
+    def set_named_arg(self, args, script_name, arg_elem_id, value, fuzzy=False):
+        """Locate an arg of a specific script in script_args and set its value
+        Args:
+            args: all script args of process p, p.script_args
+            script_name: the name target script name to
+            arg_elem_id: the elem_id of the target arg
+            value: the value to set
+            fuzzy: if True, arg_elem_id can be a substring of the control.elem_id else exact match
+        Returns:
+            Updated script args
+        when script_name in not found or arg_elem_id is not found in script controls, raise RuntimeError
+        """
+        script = next((x for x in self.scripts if x.name == script_name), None)
+        if script is None:
+            raise RuntimeError(f"script {script_name} not found")
+
+        for i, control in enumerate(script.controls):
+            if arg_elem_id in control.elem_id if fuzzy else arg_elem_id == control.elem_id:
+                index = script.args_from + i
+
+                if isinstance(args, tuple):
+                    return args[:index] + (value,) + args[index + 1:]
+                elif isinstance(args, list):
+                    args[index] = value
+                    return args
+                else:
+                    raise RuntimeError(f"args is not a list or tuple, but {type(args)}")
+        raise RuntimeError(f"arg_elem_id {arg_elem_id} not found in script {script_name}")
+
 
 scripts_txt2img: ScriptRunner = None
 scripts_img2img: ScriptRunner = None
diff --git a/modules/scripts_postprocessing.py b/modules/scripts_postprocessing.py
index bac1335dc35..4b3b7afda1c 100644
--- a/modules/scripts_postprocessing.py
+++ b/modules/scripts_postprocessing.py
@@ -1,13 +1,56 @@
+import dataclasses
 import os
 import gradio as gr
 
 from modules import errors, shared
 
 
+@dataclasses.dataclass
+class PostprocessedImageSharedInfo:
+    target_width: int = None
+    target_height: int = None
+
+
 class PostprocessedImage:
     def __init__(self, image):
         self.image = image
         self.info = {}
+        self.shared = PostprocessedImageSharedInfo()
+        self.extra_images = []
+        self.nametags = []
+        self.disable_processing = False
+        self.caption = None
+
+    def get_suffix(self, used_suffixes=None):
+        used_suffixes = {} if used_suffixes is None else used_suffixes
+        suffix = "-".join(self.nametags)
+        if suffix:
+            suffix = "-" + suffix
+
+        if suffix not in used_suffixes:
+            used_suffixes[suffix] = 1
+            return suffix
+
+        for i in range(1, 100):
+            proposed_suffix = suffix + "-" + str(i)
+
+            if proposed_suffix not in used_suffixes:
+                used_suffixes[proposed_suffix] = 1
+                return proposed_suffix
+
+        return suffix
+
+    def create_copy(self, new_image, *, nametags=None, disable_processing=False):
+        pp = PostprocessedImage(new_image)
+        pp.shared = self.shared
+        pp.nametags = self.nametags.copy()
+        pp.info = self.info.copy()
+        pp.disable_processing = disable_processing
+
+        if nametags is not None:
+            pp.nametags += nametags
+
+        return pp
 
 
 class ScriptPostprocessing:
@@ -42,10 +85,17 @@ def process(self, pp: PostprocessedImage, **args):
 
         pass
 
-    def image_changed(self):
-        pass
+    def process_firstpass(self, pp: PostprocessedImage, **args):
+        """
+        Called for all scripts before calling process(). Scripts can examine the image here and set fields
+        of the pp object to communicate things to other scripts.
+        args contains a dictionary with all values returned by components from ui()
+        """
 
+        pass
 
+    def image_changed(self):
+        pass
 
 
 def wrap_call(func, filename, funcname, *args, default=None, **kwargs):
@@ -93,6 +143,7 @@ def scripts_in_preferred_order(self):
             self.initialize_scripts(modules.scripts.postprocessing_scripts_data)
 
         scripts_order = shared.opts.postprocessing_operation_order
+        scripts_filter_out = set(shared.opts.postprocessing_disable_in_extras)
 
         def script_score(name):
             for i, possible_match in enumerate(scripts_order):
@@ -101,9 +152,10 @@ def script_score(name):
 
             return len(self.scripts)
 
-        script_scores = {script.name: (script_score(script.name), script.order, script.name, original_index) for original_index, script in enumerate(self.scripts)}
+        filtered_scripts = [script for script in self.scripts if script.name not in scripts_filter_out]
+        script_scores = {script.name: (script_score(script.name), script.order, script.name, original_index) for original_index, script in enumerate(filtered_scripts)}
 
-        return sorted(self.scripts, key=lambda x: script_scores[x.name])
+        return sorted(filtered_scripts, key=lambda x: script_scores[x.name])
 
     def setup_ui(self):
         inputs = []
@@ -118,16 +170,42 @@ def setup_ui(self):
         return inputs
 
     def run(self, pp: PostprocessedImage, args):
-        for script in self.scripts_in_preferred_order():
-            shared.state.job = script.name
+        scripts = []
 
+        for script in self.scripts_in_preferred_order():
             script_args = args[script.args_from:script.args_to]
 
             process_args = {}
             for (name, _component), value in zip(script.controls.items(), script_args):
                 process_args[name] = value
 
-            script.process(pp, **process_args)
+            scripts.append((script, process_args))
+
+        for script, process_args in scripts:
+            script.process_firstpass(pp, **process_args)
+
+        all_images = [pp]
+
+        for script, process_args in scripts:
+            if shared.state.skipped:
+                break
+
+            shared.state.job = script.name
+
+            for single_image in all_images.copy():
+
+                if not single_image.disable_processing:
+                    script.process(single_image, **process_args)
+
+                for extra_image in single_image.extra_images:
+                    if not isinstance(extra_image, PostprocessedImage):
+                        extra_image = single_image.create_copy(extra_image)
+
+                    all_images.append(extra_image)
+
+                single_image.extra_images.clear()
+
+        pp.extra_images = all_images[1:]
 
     def create_args_for_run(self, scripts_args):
         if not self.ui_created:
diff --git a/modules/sd_disable_initialization.py b/modules/sd_disable_initialization.py
index 8863107ae6f..273a7edd8b4 100644
--- a/modules/sd_disable_initialization.py
+++ b/modules/sd_disable_initialization.py
@@ -215,7 +215,7 @@ def load_state_dict(original, module, state_dict, strict=True):
             would be on the meta device.
             """
 
-            if state_dict == sd:
+            if state_dict is sd:
                 state_dict = {k: v.to(device="meta", dtype=v.dtype) for k, v in state_dict.items()}
 
             original(module, state_dict, strict=strict)
diff --git a/modules/sd_emphasis.py b/modules/sd_emphasis.py
new file mode 100644
index 00000000000..49ef1a6acb3
--- /dev/null
+++ b/modules/sd_emphasis.py
@@ -0,0 +1,70 @@
+from __future__ import annotations
+import torch
+
+
+class Emphasis:
+    """Emphasis class decides how to death with (emphasized:1.1) text in prompts"""
+
+    name: str = "Base"
+    description: str = ""
+
+    tokens: list[list[int]]
+    """tokens from the chunk of the prompt"""
+
+    multipliers: torch.Tensor
+    """tensor with multipliers, once for each token"""
+
+    z: torch.Tensor
+    """output of cond transformers network (CLIP)"""
+
+    def after_transformers(self):
+        """Called after cond transformers network has processed the chunk of the prompt; this function should modify self.z to apply the emphasis"""
+
+        pass
+
+
+class EmphasisNone(Emphasis):
+    name = "None"
+    description = "disable the mechanism entirely and treat (:.1.1) as literal characters"
+
+
+class EmphasisIgnore(Emphasis):
+    name = "Ignore"
+    description = "treat all empasised words as if they have no emphasis"
+
+
+class EmphasisOriginal(Emphasis):
+    name = "Original"
+    description = "the original emphasis implementation"
+
+    def after_transformers(self):
+        original_mean = self.z.mean()
+        self.z = self.z * self.multipliers.reshape(self.multipliers.shape + (1,)).expand(self.z.shape)
+
+        # restoring original mean is likely not correct, but it seems to work well to prevent artifacts that happen otherwise
+        new_mean = self.z.mean()
+        self.z = self.z * (original_mean / new_mean)
+
+
+class EmphasisOriginalNoNorm(EmphasisOriginal):
+    name = "No norm"
+    description = "same as original, but without normalization (seems to work better for SDXL)"
+
+    def after_transformers(self):
+        self.z = self.z * self.multipliers.reshape(self.multipliers.shape + (1,)).expand(self.z.shape)
+
+
+def get_current_option(emphasis_option_name):
+    return next(iter([x for x in options if x.name == emphasis_option_name]), EmphasisOriginal)
+
+
+def get_options_descriptions():
+    return ", ".join(f"{x.name}: {x.description}" for x in options)
+
+
+options = [
+    EmphasisNone,
+    EmphasisIgnore,
+    EmphasisOriginal,
+    EmphasisOriginalNoNorm,
+]
diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py
index 592f00551f1..0de83054186 100644
--- a/modules/sd_hijack.py
+++ b/modules/sd_hijack.py
@@ -2,14 +2,15 @@
 from torch.nn.functional import silu
 from types import MethodType
 
-from modules import devices, sd_hijack_optimizations, shared, script_callbacks, errors, sd_unet
+from modules import devices, sd_hijack_optimizations, shared, script_callbacks, errors, sd_unet, patches
 from modules.hypernetworks import hypernetwork
 from modules.shared import cmd_opts
-from modules import sd_hijack_clip, sd_hijack_open_clip, sd_hijack_unet, sd_hijack_xlmr, xlmr
+from modules import sd_hijack_clip, sd_hijack_open_clip, sd_hijack_unet, sd_hijack_xlmr, xlmr, xlmr_m18
 
 import ldm.modules.attention
 import ldm.modules.diffusionmodules.model
 import ldm.modules.diffusionmodules.openaimodel
+import ldm.models.diffusion.ddpm
 import ldm.models.diffusion.ddim
 import ldm.models.diffusion.plms
 import ldm.modules.encoders.modules
@@ -37,6 +38,12 @@
 optimizers = []
 current_optimizer: sd_hijack_optimizations.SdOptimization = None
 
+ldm_patched_forward = sd_unet.create_unet_forward(ldm.modules.diffusionmodules.openaimodel.UNetModel.forward)
+ldm_original_forward = patches.patch(__file__, ldm.modules.diffusionmodules.openaimodel.UNetModel, "forward", ldm_patched_forward)
+
+sgm_patched_forward = sd_unet.create_unet_forward(sgm.modules.diffusionmodules.openaimodel.UNetModel.forward)
+sgm_original_forward = patches.patch(__file__, sgm.modules.diffusionmodules.openaimodel.UNetModel, "forward", sgm_patched_forward)
+
 
 def list_optimizers():
     new_optimizers = script_callbacks.list_optimizers_callback()
@@ -181,6 +188,20 @@ def apply_optimizations(self, option=None):
             errors.display(e, "applying cross attention optimization")
             undo_optimizations()
 
+    def convert_sdxl_to_ssd(self, m):
+        """Converts an SDXL model to a Segmind Stable Diffusion model (see https://huggingface.co/segmind/SSD-1B)"""
+
+        delattr(m.model.diffusion_model.middle_block, '1')
+        delattr(m.model.diffusion_model.middle_block, '2')
+        for i in ['9', '8', '7', '6', '5', '4']:
+            delattr(m.model.diffusion_model.input_blocks[7][1].transformer_blocks, i)
+            delattr(m.model.diffusion_model.input_blocks[8][1].transformer_blocks, i)
+            delattr(m.model.diffusion_model.output_blocks[0][1].transformer_blocks, i)
+            delattr(m.model.diffusion_model.output_blocks[1][1].transformer_blocks, i)
+        delattr(m.model.diffusion_model.output_blocks[4][1].transformer_blocks, '1')
+        delattr(m.model.diffusion_model.output_blocks[5][1].transformer_blocks, '1')
+        devices.torch_gc()
+
     def hijack(self, m):
         conditioner = getattr(m, 'conditioner', None)
         if conditioner:
@@ -208,7 +229,7 @@ def hijack(self, m):
             else:
                 m.cond_stage_model = conditioner
 
-        if type(m.cond_stage_model) == xlmr.BertSeriesModelWithTransformation:
+        if type(m.cond_stage_model) == xlmr.BertSeriesModelWithTransformation or type(m.cond_stage_model) == xlmr_m18.BertSeriesModelWithTransformation:
             model_embeddings = m.cond_stage_model.roberta.embeddings
             model_embeddings.token_embedding = EmbeddingsWithFixes(model_embeddings.word_embeddings, self)
             m.cond_stage_model = sd_hijack_xlmr.FrozenXLMREmbedderWithCustomWords(m.cond_stage_model, self)
@@ -239,10 +260,17 @@ def flatten(el):
 
         self.layers = flatten(m)
 
-        if not hasattr(ldm.modules.diffusionmodules.openaimodel, 'copy_of_UNetModel_forward_for_webui'):
-            ldm.modules.diffusionmodules.openaimodel.copy_of_UNetModel_forward_for_webui = ldm.modules.diffusionmodules.openaimodel.UNetModel.forward
+        import modules.models.diffusion.ddpm_edit
+
+        if isinstance(m, ldm.models.diffusion.ddpm.LatentDiffusion):
+            sd_unet.original_forward = ldm_original_forward
+        elif isinstance(m, modules.models.diffusion.ddpm_edit.LatentDiffusion):
+            sd_unet.original_forward = ldm_original_forward
+        elif isinstance(m, sgm.models.diffusion.DiffusionEngine):
+            sd_unet.original_forward = sgm_original_forward
+        else:
+            sd_unet.original_forward = None
 
-        ldm.modules.diffusionmodules.openaimodel.UNetModel.forward = sd_unet.UNetModel_forward
 
     def undo_hijack(self, m):
         conditioner = getattr(m, 'conditioner', None)
@@ -279,7 +307,6 @@ def undo_hijack(self, m):
         self.layers = None
         self.clip = None
 
-        ldm.modules.diffusionmodules.openaimodel.UNetModel.forward = ldm.modules.diffusionmodules.openaimodel.copy_of_UNetModel_forward_for_webui
 
     def apply_circular(self, enable):
         if self.circular_enabled == enable:
@@ -298,7 +325,10 @@ def get_prompt_lengths(self, text):
         if self.clip is None:
             return "-", "-"
 
-        _, token_count = self.clip.process_texts([text])
+        if hasattr(self.clip, 'get_token_count'):
+            token_count = self.clip.get_token_count(text)
+        else:
+            _, token_count = self.clip.process_texts([text])
 
         return token_count, self.clip.get_target_prompt_token_count(token_count)
 
@@ -329,13 +359,28 @@ def forward(self, input_ids):
                 vec = embedding.vec[self.textual_inversion_key] if isinstance(embedding.vec, dict) else embedding.vec
                 emb = devices.cond_cast_unet(vec)
                 emb_len = min(tensor.shape[0] - offset - 1, emb.shape[0])
-                tensor = torch.cat([tensor[0:offset + 1], emb[0:emb_len], tensor[offset + 1 + emb_len:]])
+                tensor = torch.cat([tensor[0:offset + 1], emb[0:emb_len], tensor[offset + 1 + emb_len:]]).to(dtype=inputs_embeds.dtype)
 
             vecs.append(tensor)
 
         return torch.stack(vecs)
 
 
+class TextualInversionEmbeddings(torch.nn.Embedding):
+    def __init__(self, num_embeddings: int, embedding_dim: int, textual_inversion_key='clip_l', **kwargs):
+        super().__init__(num_embeddings, embedding_dim, **kwargs)
+
+        self.embeddings = model_hijack
+        self.textual_inversion_key = textual_inversion_key
+
+    @property
+    def wrapped(self):
+        return super().forward
+
+    def forward(self, input_ids):
+        return EmbeddingsWithFixes.forward(self, input_ids)
+
+
 def add_circular_option_to_conv_2d():
     conv2d_constructor = torch.nn.Conv2d.__init__
 
diff --git a/modules/sd_hijack_clip.py b/modules/sd_hijack_clip.py
index 8f29057a9cf..a479148fc21 100644
--- a/modules/sd_hijack_clip.py
+++ b/modules/sd_hijack_clip.py
@@ -3,7 +3,7 @@
 
 import torch
 
-from modules import prompt_parser, devices, sd_hijack
+from modules import prompt_parser, devices, sd_hijack, sd_emphasis
 from modules.shared import opts
 
 
@@ -23,28 +23,25 @@ def __init__(self):
 
 PromptChunkFix = namedtuple('PromptChunkFix', ['offset', 'embedding'])
 """An object of this type is a marker showing that textual inversion embedding's vectors have to placed at offset in the prompt
-chunk. Thos objects are found in PromptChunk.fixes and, are placed into FrozenCLIPEmbedderWithCustomWordsBase.hijack.fixes, and finally
+chunk. Those objects are found in PromptChunk.fixes and, are placed into FrozenCLIPEmbedderWithCustomWordsBase.hijack.fixes, and finally
 are applied by sd_hijack.EmbeddingsWithFixes's forward function."""
 
 
-class FrozenCLIPEmbedderWithCustomWordsBase(torch.nn.Module):
-    """A pytorch module that is a wrapper for FrozenCLIPEmbedder module. it enhances FrozenCLIPEmbedder, making it possible to
-    have unlimited prompt length and assign weights to tokens in prompt.
-    """
-
-    def __init__(self, wrapped, hijack):
+class TextConditionalModel(torch.nn.Module):
+    def __init__(self):
         super().__init__()
 
-        self.wrapped = wrapped
-        """Original FrozenCLIPEmbedder module; can also be FrozenOpenCLIPEmbedder or xlmr.BertSeriesModelWithTransformation,
-        depending on model."""
-
-        self.hijack: sd_hijack.StableDiffusionModelHijack = hijack
+        self.hijack = sd_hijack.model_hijack
         self.chunk_length = 75
 
-        self.is_trainable = getattr(wrapped, 'is_trainable', False)
-        self.input_key = getattr(wrapped, 'input_key', 'txt')
-        self.legacy_ucg_val = None
+        self.is_trainable = False
+        self.input_key = 'txt'
+        self.return_pooled = False
+
+        self.comma_token = None
+        self.id_start = None
+        self.id_end = None
+        self.id_pad = None
 
     def empty_chunk(self):
         """creates an empty PromptChunk and returns it"""
@@ -66,7 +63,7 @@ def tokenize(self, texts):
 
     def encode_with_transformers(self, tokens):
         """
-        converts a batch of token ids (in python lists) into a single tensor with numeric respresentation of those tokens;
+        converts a batch of token ids (in python lists) into a single tensor with numeric representation of those tokens;
         All python lists with tokens are assumed to have same length, usually 77.
         if input is a list with B elements and each element has T tokens, expected output shape is (B, T, C), where C depends on
         model - can be 768 and 1024.
@@ -88,7 +85,7 @@ def tokenize_line(self, line):
         Returns the list and the total number of tokens in the prompt.
         """
 
-        if opts.enable_emphasis:
+        if opts.emphasis != "None":
             parsed = prompt_parser.parse_prompt_attention(line)
         else:
             parsed = [[line, 1.0]]
@@ -136,7 +133,7 @@ def next_chunk(is_last=False):
                 if token == self.comma_token:
                     last_comma = len(chunk.tokens)
 
-                # this is when we are at the end of alloted 75 tokens for the current chunk, and the current token is not a comma. opts.comma_padding_backtrack
+                # this is when we are at the end of allotted 75 tokens for the current chunk, and the current token is not a comma. opts.comma_padding_backtrack
                 # is a setting that specifies that if there is a comma nearby, the text after the comma should be moved out of this chunk and into the next.
                 elif opts.comma_padding_backtrack != 0 and len(chunk.tokens) == self.chunk_length and last_comma != -1 and len(chunk.tokens) - last_comma <= opts.comma_padding_backtrack:
                     break_location = last_comma + 1
@@ -206,14 +203,10 @@ def forward(self, texts):
         be a multiple of 77; and C is dimensionality of each token - for SD1 it's 768, for SD2 it's 1024, and for SDXL it's 1280.
         An example shape returned by this function can be: (2, 77, 768).
         For SDXL, instead of returning one tensor avobe, it returns a tuple with two: the other one with shape (B, 1280) with pooled values.
-        Webui usually sends just one text at a time through this function - the only time when texts is an array with more than one elemenet
+        Webui usually sends just one text at a time through this function - the only time when texts is an array with more than one element
         is when you do prompt editing: "a picture of a [cat:dog:0.4] eating ice cream"
         """
 
-        if opts.use_old_emphasis_implementation:
-            import modules.sd_hijack_clip_old
-            return modules.sd_hijack_clip_old.forward_old(self, texts)
-
         batch_chunks, token_count = self.process_texts(texts)
 
         used_embeddings = {}
@@ -230,7 +223,7 @@ def forward(self, texts):
             for fixes in self.hijack.fixes:
                 for _position, embedding in fixes:
                     used_embeddings[embedding.name] = embedding
-
+            devices.torch_npu_set_device()
             z = self.process_tokens(tokens, multipliers)
             zs.append(z)
 
@@ -249,7 +242,10 @@ def forward(self, texts):
                     hashes.append(self.hijack.extra_generation_params.get("TI hashes"))
                 self.hijack.extra_generation_params["TI hashes"] = ", ".join(hashes)
 
-        if getattr(self.wrapped, 'return_pooled', False):
+        if any(x for x in texts if "(" in x or "[" in x) and opts.emphasis != "Original":
+            self.hijack.extra_generation_params["Emphasis"] = opts.emphasis
+
+        if self.return_pooled:
             return torch.hstack(zs), zs[0].pooled
         else:
             return torch.hstack(zs)
@@ -274,12 +270,14 @@ def process_tokens(self, remade_batch_tokens, batch_multipliers):
 
         pooled = getattr(z, 'pooled', None)
 
-        # restoring original mean is likely not correct, but it seems to work well to prevent artifacts that happen otherwise
-        batch_multipliers = torch.asarray(batch_multipliers).to(devices.device)
-        original_mean = z.mean()
-        z = z * batch_multipliers.reshape(batch_multipliers.shape + (1,)).expand(z.shape)
-        new_mean = z.mean()
-        z = z * (original_mean / new_mean)
+        emphasis = sd_emphasis.get_current_option(opts.emphasis)()
+        emphasis.tokens = remade_batch_tokens
+        emphasis.multipliers = torch.asarray(batch_multipliers).to(devices.device)
+        emphasis.z = z
+
+        emphasis.after_transformers()
+
+        z = emphasis.z
 
         if pooled is not None:
             z.pooled = pooled
@@ -287,6 +285,34 @@ def process_tokens(self, remade_batch_tokens, batch_multipliers):
         return z
 
 
+class FrozenCLIPEmbedderWithCustomWordsBase(TextConditionalModel):
+    """A pytorch module that is a wrapper for FrozenCLIPEmbedder module. it enhances FrozenCLIPEmbedder, making it possible to
+    have unlimited prompt length and assign weights to tokens in prompt.
+    """
+
+    def __init__(self, wrapped, hijack):
+        super().__init__()
+
+        self.hijack = hijack
+
+        self.wrapped = wrapped
+        """Original FrozenCLIPEmbedder module; can also be FrozenOpenCLIPEmbedder or xlmr.BertSeriesModelWithTransformation,
+        depending on model."""
+
+        self.is_trainable = getattr(wrapped, 'is_trainable', False)
+        self.input_key = getattr(wrapped, 'input_key', 'txt')
+        self.return_pooled = getattr(self.wrapped, 'return_pooled', False)
+
+        self.legacy_ucg_val = None  # for sgm codebase
+
+    def forward(self, texts):
+        if opts.use_old_emphasis_implementation:
+            import modules.sd_hijack_clip_old
+            return modules.sd_hijack_clip_old.forward_old(self, texts)
+
+        return super().forward(texts)
+
+
 class FrozenCLIPEmbedderWithCustomWords(FrozenCLIPEmbedderWithCustomWordsBase):
     def __init__(self, wrapped, hijack):
         super().__init__(wrapped, hijack)
@@ -348,7 +374,9 @@ def __init__(self, wrapped, hijack):
     def encode_with_transformers(self, tokens):
         outputs = self.wrapped.transformer(input_ids=tokens, output_hidden_states=self.wrapped.layer == "hidden")
 
-        if self.wrapped.layer == "last":
+        if opts.sdxl_clip_l_skip is True:
+            z = outputs.hidden_states[-opts.CLIP_stop_at_last_layers]
+        elif self.wrapped.layer == "last":
             z = outputs.last_hidden_state
         else:
             z = outputs.hidden_states[self.wrapped.layer_idx]
diff --git a/modules/sd_hijack_clip_old.py b/modules/sd_hijack_clip_old.py
index c5c6270b9c4..43e9b9529e8 100644
--- a/modules/sd_hijack_clip_old.py
+++ b/modules/sd_hijack_clip_old.py
@@ -32,7 +32,7 @@ def process_text_old(self: sd_hijack_clip.FrozenCLIPEmbedderWithCustomWordsBase,
 
                 embedding, embedding_length_in_tokens = self.hijack.embedding_db.find_embedding_at_position(tokens, i)
 
-                mult_change = self.token_mults.get(token) if shared.opts.enable_emphasis else None
+                mult_change = self.token_mults.get(token) if shared.opts.emphasis != "None" else None
                 if mult_change is not None:
                     mult *= mult_change
                     i += 1
diff --git a/modules/sd_hijack_optimizations.py b/modules/sd_hijack_optimizations.py
index 7f9e328d05a..0269f1f5b4b 100644
--- a/modules/sd_hijack_optimizations.py
+++ b/modules/sd_hijack_optimizations.py
@@ -486,7 +486,8 @@ def xformers_attention_forward(self, x, context=None, mask=None, **kwargs):
     k_in = self.to_k(context_k)
     v_in = self.to_v(context_v)
 
-    q, k, v = (rearrange(t, 'b n (h d) -> b n h d', h=h) for t in (q_in, k_in, v_in))
+    q, k, v = (t.reshape(t.shape[0], t.shape[1], h, -1) for t in (q_in, k_in, v_in))
+
     del q_in, k_in, v_in
 
     dtype = q.dtype
@@ -497,7 +498,8 @@ def xformers_attention_forward(self, x, context=None, mask=None, **kwargs):
 
     out = out.to(dtype)
 
-    out = rearrange(out, 'b n h d -> b n (h d)', h=h)
+    b, n, h, d = out.shape
+    out = out.reshape(b, n, h * d)
     return self.to_out(out)
 
 
diff --git a/modules/sd_hijack_unet.py b/modules/sd_hijack_unet.py
index 2101f1a0415..b4f03b138a4 100644
--- a/modules/sd_hijack_unet.py
+++ b/modules/sd_hijack_unet.py
@@ -1,5 +1,7 @@
 import torch
 from packaging import version
+from einops import repeat
+import math
 
 from modules import devices
 from modules.sd_hijack_utils import CondFunc
@@ -36,7 +38,7 @@ def cat(self, tensors, *args, **kwargs):
 
 # Below are monkey patches to enable upcasting a float16 UNet for float32 sampling
 def apply_model(orig_func, self, x_noisy, t, cond, **kwargs):
-
+    """Always make sure inputs to unet are in correct dtype."""
     if isinstance(cond, dict):
         for y in cond.keys():
             if isinstance(cond[y], list):
@@ -45,7 +47,59 @@ def apply_model(orig_func, self, x_noisy, t, cond, **kwargs):
                 cond[y] = cond[y].to(devices.dtype_unet) if isinstance(cond[y], torch.Tensor) else cond[y]
 
     with devices.autocast():
-        return orig_func(self, x_noisy.to(devices.dtype_unet), t.to(devices.dtype_unet), cond, **kwargs).float()
+        result = orig_func(self, x_noisy.to(devices.dtype_unet), t.to(devices.dtype_unet), cond, **kwargs)
+        if devices.unet_needs_upcast:
+            return result.float()
+        else:
+            return result
+
+
+# Monkey patch to create timestep embed tensor on device, avoiding a block.
+def timestep_embedding(_, timesteps, dim, max_period=10000, repeat_only=False):
+    """
+    Create sinusoidal timestep embeddings.
+    :param timesteps: a 1-D Tensor of N indices, one per batch element.
+                      These may be fractional.
+    :param dim: the dimension of the output.
+    :param max_period: controls the minimum frequency of the embeddings.
+    :return: an [N x dim] Tensor of positional embeddings.
+    """
+    if not repeat_only:
+        half = dim // 2
+        freqs = torch.exp(
+            -math.log(max_period) * torch.arange(start=0, end=half, dtype=torch.float32, device=timesteps.device) / half
+        )
+        args = timesteps[:, None].float() * freqs[None]
+        embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
+        if dim % 2:
+            embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
+    else:
+        embedding = repeat(timesteps, 'b -> b d', d=dim)
+    return embedding
+
+
+# Monkey patch to SpatialTransformer removing unnecessary contiguous calls.
+# Prevents a lot of unnecessary aten::copy_ calls
+def spatial_transformer_forward(_, self, x: torch.Tensor, context=None):
+    # note: if no context is given, cross-attention defaults to self-attention
+    if not isinstance(context, list):
+        context = [context]
+    b, c, h, w = x.shape
+    x_in = x
+    x = self.norm(x)
+    if not self.use_linear:
+        x = self.proj_in(x)
+    x = x.permute(0, 2, 3, 1).reshape(b, h * w, c)
+    if self.use_linear:
+        x = self.proj_in(x)
+    for i, block in enumerate(self.transformer_blocks):
+        x = block(x, context=context[i])
+    if self.use_linear:
+        x = self.proj_out(x)
+    x = x.view(b, h, w, c).permute(0, 3, 1, 2)
+    if not self.use_linear:
+        x = self.proj_out(x)
+    return x + x_in
 
 
 class GELUHijack(torch.nn.GELU, torch.nn.Module):
@@ -64,12 +118,15 @@ def hijack_ddpm_edit():
     if not ddpm_edit_hijack:
         CondFunc('modules.models.diffusion.ddpm_edit.LatentDiffusion.decode_first_stage', first_stage_sub, first_stage_cond)
         CondFunc('modules.models.diffusion.ddpm_edit.LatentDiffusion.encode_first_stage', first_stage_sub, first_stage_cond)
-        ddpm_edit_hijack = CondFunc('modules.models.diffusion.ddpm_edit.LatentDiffusion.apply_model', apply_model, unet_needs_upcast)
+        ddpm_edit_hijack = CondFunc('modules.models.diffusion.ddpm_edit.LatentDiffusion.apply_model', apply_model)
 
 
 unet_needs_upcast = lambda *args, **kwargs: devices.unet_needs_upcast
 CondFunc('ldm.models.diffusion.ddpm.LatentDiffusion.apply_model', apply_model, unet_needs_upcast)
+CondFunc('ldm.modules.diffusionmodules.openaimodel.timestep_embedding', timestep_embedding)
+CondFunc('ldm.modules.attention.SpatialTransformer.forward', spatial_transformer_forward)
 CondFunc('ldm.modules.diffusionmodules.openaimodel.timestep_embedding', lambda orig_func, timesteps, *args, **kwargs: orig_func(timesteps, *args, **kwargs).to(torch.float32 if timesteps.dtype == torch.int64 else devices.dtype_unet), unet_needs_upcast)
+
 if version.parse(torch.__version__) <= version.parse("1.13.2") or torch.cuda.is_available():
     CondFunc('ldm.modules.diffusionmodules.util.GroupNorm32.forward', lambda orig_func, self, *args, **kwargs: orig_func(self.float(), *args, **kwargs), unet_needs_upcast)
     CondFunc('ldm.modules.attention.GEGLU.forward', lambda orig_func, self, x: orig_func(self.float(), x.float()).to(devices.dtype_unet), unet_needs_upcast)
@@ -81,5 +138,17 @@ def hijack_ddpm_edit():
 CondFunc('ldm.models.diffusion.ddpm.LatentDiffusion.encode_first_stage', first_stage_sub, first_stage_cond)
 CondFunc('ldm.models.diffusion.ddpm.LatentDiffusion.get_first_stage_encoding', lambda orig_func, *args, **kwargs: orig_func(*args, **kwargs).float(), first_stage_cond)
 
-CondFunc('sgm.modules.diffusionmodules.wrappers.OpenAIWrapper.forward', apply_model, unet_needs_upcast)
-CondFunc('sgm.modules.diffusionmodules.openaimodel.timestep_embedding', lambda orig_func, timesteps, *args, **kwargs: orig_func(timesteps, *args, **kwargs).to(torch.float32 if timesteps.dtype == torch.int64 else devices.dtype_unet), unet_needs_upcast)
+CondFunc('ldm.models.diffusion.ddpm.LatentDiffusion.apply_model', apply_model)
+CondFunc('sgm.modules.diffusionmodules.wrappers.OpenAIWrapper.forward', apply_model)
+
+
+def timestep_embedding_cast_result(orig_func, timesteps, *args, **kwargs):
+    if devices.unet_needs_upcast and timesteps.dtype == torch.int64:
+        dtype = torch.float32
+    else:
+        dtype = devices.dtype_unet
+    return orig_func(timesteps, *args, **kwargs).to(dtype=dtype)
+
+
+CondFunc('ldm.modules.diffusionmodules.openaimodel.timestep_embedding', timestep_embedding_cast_result)
+CondFunc('sgm.modules.diffusionmodules.openaimodel.timestep_embedding', timestep_embedding_cast_result)
diff --git a/modules/sd_hijack_utils.py b/modules/sd_hijack_utils.py
index f8684475ec5..546f2eda4ec 100644
--- a/modules/sd_hijack_utils.py
+++ b/modules/sd_hijack_utils.py
@@ -1,7 +1,11 @@
 import importlib
 
+
+always_true_func = lambda *args, **kwargs: True
+
+
 class CondFunc:
-    def __new__(cls, orig_func, sub_func, cond_func):
+    def __new__(cls, orig_func, sub_func, cond_func=always_true_func):
         self = super(CondFunc, cls).__new__(cls)
         if isinstance(orig_func, str):
             func_path = orig_func.split('.')
@@ -11,18 +15,22 @@ def __new__(cls, orig_func, sub_func, cond_func):
                     break
                 except ImportError:
                     pass
-            for attr_name in func_path[i:-1]:
-                resolved_obj = getattr(resolved_obj, attr_name)
-            orig_func = getattr(resolved_obj, func_path[-1])
-            setattr(resolved_obj, func_path[-1], lambda *args, **kwargs: self(*args, **kwargs))
+            try:
+                for attr_name in func_path[i:-1]:
+                    resolved_obj = getattr(resolved_obj, attr_name)
+                orig_func = getattr(resolved_obj, func_path[-1])
+                setattr(resolved_obj, func_path[-1], lambda *args, **kwargs: self(*args, **kwargs))
+            except AttributeError:
+                print(f"Warning: Failed to resolve {orig_func} for CondFunc hijack")
+                pass
         self.__init__(orig_func, sub_func, cond_func)
-        return lambda *args, **kwargs: self(*args, **kwargs)
-    def __init__(self, orig_func, sub_func, cond_func):
-        self.__orig_func = orig_func
-        self.__sub_func = sub_func
-        self.__cond_func = cond_func
-    def __call__(self, *args, **kwargs):
-        if not self.__cond_func or self.__cond_func(self.__orig_func, *args, **kwargs):
-            return self.__sub_func(self.__orig_func, *args, **kwargs)
-        else:
-            return self.__orig_func(*args, **kwargs)
+        return lambda *args, **kwargs: self(*args, **kwargs)
+    def __init__(self, orig_func, sub_func, cond_func):
+        self.__orig_func = orig_func
+        self.__sub_func = sub_func
+        self.__cond_func = cond_func
+    def __call__(self, *args, **kwargs):
+        if not self.__cond_func or self.__cond_func(self.__orig_func, *args, **kwargs):
+            return self.__sub_func(self.__orig_func, *args, **kwargs)
+        else:
+            return self.__orig_func(*args, **kwargs)
diff --git a/modules/sd_models.py b/modules/sd_models.py
index 930d0bee5c8..55bd9ca5e43 100644
--- a/modules/sd_models.py
+++ b/modules/sd_models.py
@@ -1,22 +1,22 @@
 import collections
-import os.path
+import importlib
+import os
 import sys
-import gc
 import threading
+import enum
 
 import torch
 import re
 import safetensors.torch
-from omegaconf import OmegaConf
-from os import mkdir
+from omegaconf import OmegaConf, ListConfig
 from urllib import request
 import ldm.modules.midas as midas
 
-from ldm.util import instantiate_from_config
-
-from modules import paths, shared, modelloader, devices, script_callbacks, sd_vae, sd_disable_initialization, errors, hashes, sd_models_config, sd_unet, sd_models_xl, cache, extra_networks, processing, lowvram, sd_hijack
+from modules import paths, shared, modelloader, devices, script_callbacks, sd_vae, sd_disable_initialization, errors, hashes, sd_models_config, sd_unet, sd_models_xl, cache, extra_networks, processing, lowvram, sd_hijack, patches
 from modules.timer import Timer
+from modules.shared import opts
 import tomesd
+import numpy as np
 
 model_dir = "Stable-diffusion"
 model_path = os.path.abspath(os.path.join(paths.models_path, model_dir))
@@ -27,6 +27,14 @@
 checkpoints_loaded = collections.OrderedDict()
 
 
+class ModelType(enum.Enum):
+    SD1 = 1
+    SD2 = 2
+    SDXL = 3
+    SSD = 4
+    SD3 = 5
+
+
 def replace_key(d, key, new_key, value):
     keys = list(d.keys())
 
@@ -49,11 +57,12 @@ class CheckpointInfo:
     def __init__(self, filename):
         self.filename = filename
         abspath = os.path.abspath(filename)
+        abs_ckpt_dir = os.path.abspath(shared.cmd_opts.ckpt_dir) if shared.cmd_opts.ckpt_dir is not None else None
 
         self.is_safetensors = os.path.splitext(filename)[1].lower() == ".safetensors"
 
-        if shared.cmd_opts.ckpt_dir is not None and abspath.startswith(shared.cmd_opts.ckpt_dir):
-            name = abspath.replace(shared.cmd_opts.ckpt_dir, '')
+        if abs_ckpt_dir and abspath.startswith(abs_ckpt_dir):
+            name = abspath.replace(abs_ckpt_dir, '')
         elif abspath.startswith(model_path):
             name = abspath.replace(model_path, '')
         else:
@@ -129,9 +138,12 @@ def calculate_shorthash(self):
 
 
 def setup_model():
+    """called once at startup to do various one-time tasks related to SD models"""
+
     os.makedirs(model_path, exist_ok=True)
 
     enable_midas_autodownload()
+    patch_given_betas()
 
 
 def checkpoint_tiles(use_short=False):
@@ -145,10 +157,12 @@ def list_models():
     cmd_ckpt = shared.cmd_opts.ckpt
     if shared.cmd_opts.no_download_sd_model or cmd_ckpt != shared.sd_model_file or os.path.exists(cmd_ckpt):
         model_url = None
+        expected_sha256 = None
     else:
-        model_url = "https://huggingface.co/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors"
+        model_url = f"{shared.hf_endpoint}/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors"
+        expected_sha256 = '6ce0161689b3853acaa03779ec93eafe75a02f4ced659bee03f50797806fa2fa'
 
-    model_list = modelloader.load_models(model_path=model_path, model_url=model_url, command_path=shared.cmd_opts.ckpt_dir, ext_filter=[".ckpt", ".safetensors"], download_name="v1-5-pruned-emaonly.safetensors", ext_blacklist=[".vae.ckpt", ".vae.safetensors"])
+    model_list = modelloader.load_models(model_path=model_path, model_url=model_url, command_path=shared.cmd_opts.ckpt_dir, ext_filter=[".ckpt", ".safetensors"], download_name="v1-5-pruned-emaonly.safetensors", ext_blacklist=[".vae.ckpt", ".vae.safetensors"], hash_prefix=expected_sha256)
 
     if os.path.exists(cmd_ckpt):
         checkpoint_info = CheckpointInfo(cmd_ckpt)
@@ -226,15 +240,19 @@ def select_checkpoint():
     return checkpoint_info
 
 
-checkpoint_dict_replacements = {
+checkpoint_dict_replacements_sd1 = {
     'cond_stage_model.transformer.embeddings.': 'cond_stage_model.transformer.text_model.embeddings.',
     'cond_stage_model.transformer.encoder.': 'cond_stage_model.transformer.text_model.encoder.',
     'cond_stage_model.transformer.final_layer_norm.': 'cond_stage_model.transformer.text_model.final_layer_norm.',
 }
 
+checkpoint_dict_replacements_sd2_turbo = { # Converts SD 2.1 Turbo from SGM to LDM format.
+    'conditioner.embedders.0.': 'cond_stage_model.',
+}
+
 
-def transform_checkpoint_dict_key(k):
-    for text, replacement in checkpoint_dict_replacements.items():
+def transform_checkpoint_dict_key(k, replacements):
+    for text, replacement in replacements.items():
         if k.startswith(text):
             k = replacement + k[len(text):]
 
@@ -245,9 +263,14 @@ def get_state_dict_from_checkpoint(pl_sd):
     pl_sd = pl_sd.pop("state_dict", pl_sd)
     pl_sd.pop("state_dict", None)
 
+    is_sd2_turbo = 'conditioner.embedders.0.model.ln_final.weight' in pl_sd and pl_sd['conditioner.embedders.0.model.ln_final.weight'].size()[0] == 1024
+
     sd = {}
     for k, v in pl_sd.items():
-        new_key = transform_checkpoint_dict_key(k)
+        if is_sd2_turbo:
+            new_key = transform_checkpoint_dict_key(k, checkpoint_dict_replacements_sd2_turbo)
+        else:
+            new_key = transform_checkpoint_dict_key(k, checkpoint_dict_replacements_sd1)
 
         if new_key is not None:
             sd[new_key] = v
@@ -267,17 +290,21 @@ def read_metadata_from_safetensors(filename):
         json_start = file.read(2)
 
         assert metadata_len > 2 and json_start in (b'{"', b"{'"), f"{filename} is not a safetensors file"
-        json_data = json_start + file.read(metadata_len-2)
-        json_obj = json.loads(json_data)
 
         res = {}
-        for k, v in json_obj.get("__metadata__", {}).items():
-            res[k] = v
-            if isinstance(v, str) and v[0:1] == '{':
-                try:
-                    res[k] = json.loads(v)
-                except Exception:
-                    pass
+
+        try:
+            json_data = json_start + file.read(metadata_len-2)
+            json_obj = json.loads(json_data)
+            for k, v in json_obj.get("__metadata__", {}).items():
+                res[k] = v
+                if isinstance(v, str) and v[0:1] == '{':
+                    try:
+                        res[k] = json.loads(v)
+                    except Exception:
+                        pass
+        except Exception:
+             errors.report(f"Error reading metadata from file: {filename}", exc_info=True)
 
         return res
 
@@ -309,6 +336,8 @@ def get_checkpoint_state_dict(checkpoint_info: CheckpointInfo, timer):
     if checkpoint_info in checkpoints_loaded:
         # use checkpoint cache
         print(f"Loading weights [{sd_model_hash}] from cache")
+        # move to end as latest
+        checkpoints_loaded.move_to_end(checkpoint_info)
         return checkpoints_loaded[checkpoint_info]
 
     print(f"Loading weights [{sd_model_hash}] from {checkpoint_info.filename}")
@@ -333,39 +362,111 @@ def __exit__(self, exc_type, exc_value, exc_traceback):
         SkipWritingToConfig.skip = self.previous
 
 
+def check_fp8(model):
+    if model is None:
+        return None
+    if devices.get_optimal_device_name() == "mps":
+        enable_fp8 = False
+    elif shared.opts.fp8_storage == "Enable":
+        enable_fp8 = True
+    elif getattr(model, "is_sdxl", False) and shared.opts.fp8_storage == "Enable for SDXL":
+        enable_fp8 = True
+    else:
+        enable_fp8 = False
+    return enable_fp8
+
+
+def set_model_type(model, state_dict):
+    model.is_sd1 = False
+    model.is_sd2 = False
+    model.is_sdxl = False
+    model.is_ssd = False
+    model.is_sd3 = False
+
+    if "model.diffusion_model.x_embedder.proj.weight" in state_dict:
+        model.is_sd3 = True
+        model.model_type = ModelType.SD3
+    elif hasattr(model, 'conditioner'):
+        model.is_sdxl = True
+
+        if 'model.diffusion_model.middle_block.1.transformer_blocks.0.attn1.to_q.weight' not in state_dict.keys():
+            model.is_ssd = True
+            model.model_type = ModelType.SSD
+        else:
+            model.model_type = ModelType.SDXL
+    elif hasattr(model.cond_stage_model, 'model'):
+        model.is_sd2 = True
+        model.model_type = ModelType.SD2
+    else:
+        model.is_sd1 = True
+        model.model_type = ModelType.SD1
+
+
+def set_model_fields(model):
+    if not hasattr(model, 'latent_channels'):
+        model.latent_channels = 4
+
+
 def load_model_weights(model, checkpoint_info: CheckpointInfo, state_dict, timer):
     sd_model_hash = checkpoint_info.calculate_shorthash()
     timer.record("calculate hash")
 
+    if devices.fp8:
+        # prevent model to load state dict in fp8
+        model.half()
+
     if not SkipWritingToConfig.skip:
         shared.opts.data["sd_model_checkpoint"] = checkpoint_info.title
 
     if state_dict is None:
         state_dict = get_checkpoint_state_dict(checkpoint_info, timer)
 
-    model.is_sdxl = hasattr(model, 'conditioner')
-    model.is_sd2 = not model.is_sdxl and hasattr(model.cond_stage_model, 'model')
-    model.is_sd1 = not model.is_sdxl and not model.is_sd2
+    set_model_type(model, state_dict)
+    set_model_fields(model)
 
     if model.is_sdxl:
         sd_models_xl.extend_sdxl(model)
 
-    model.load_state_dict(state_dict, strict=False)
-    timer.record("apply weights to model")
+    if model.is_ssd:
+        sd_hijack.model_hijack.convert_sdxl_to_ssd(model)
 
     if shared.opts.sd_checkpoint_cache > 0:
         # cache newly loaded model
-        checkpoints_loaded[checkpoint_info] = state_dict
+        checkpoints_loaded[checkpoint_info] = state_dict.copy()
+
+    if hasattr(model, "before_load_weights"):
+        model.before_load_weights(state_dict)
+
+    model.load_state_dict(state_dict, strict=False)
+    timer.record("apply weights to model")
+
+    if hasattr(model, "after_load_weights"):
+        model.after_load_weights(state_dict)
 
     del state_dict
 
+    # Set is_sdxl_inpaint flag.
+    # Checks Unet structure to detect inpaint model. The inpaint model's
+    # checkpoint state_dict does not contain the key
+    # 'diffusion_model.input_blocks.0.0.weight'.
+    diffusion_model_input = model.model.state_dict().get(
+        'diffusion_model.input_blocks.0.0.weight'
+    )
+    model.is_sdxl_inpaint = (
+        model.is_sdxl and
+        diffusion_model_input is not None and
+        diffusion_model_input.shape[1] == 9
+    )
+
     if shared.cmd_opts.opt_channelslast:
         model.to(memory_format=torch.channels_last)
         timer.record("apply channels_last")
 
     if shared.cmd_opts.no_half:
         model.float()
+        model.alphas_cumprod_original = model.alphas_cumprod
         devices.dtype_unet = torch.float32
+        assert shared.cmd_opts.precision != "half", "Cannot use --precision half with --no-half"
         timer.record("apply float()")
     else:
         vae = model.first_stage_model
@@ -378,7 +479,11 @@ def load_model_weights(model, checkpoint_info: CheckpointInfo, state_dict, timer
         if shared.cmd_opts.upcast_sampling and depth_model:
             model.depth_model = None
 
+        alphas_cumprod = model.alphas_cumprod
+        model.alphas_cumprod = None
         model.half()
+        model.alphas_cumprod = alphas_cumprod
+        model.alphas_cumprod_original = alphas_cumprod
         model.first_stage_model = vae
         if depth_model:
             model.depth_model = depth_model
@@ -386,6 +491,30 @@ def load_model_weights(model, checkpoint_info: CheckpointInfo, state_dict, timer
         devices.dtype_unet = torch.float16
         timer.record("apply half()")
 
+    apply_alpha_schedule_override(model)
+
+    for module in model.modules():
+        if hasattr(module, 'fp16_weight'):
+            del module.fp16_weight
+        if hasattr(module, 'fp16_bias'):
+            del module.fp16_bias
+
+    if check_fp8(model):
+        devices.fp8 = True
+        first_stage = model.first_stage_model
+        model.first_stage_model = None
+        for module in model.modules():
+            if isinstance(module, (torch.nn.Conv2d, torch.nn.Linear)):
+                if shared.opts.cache_fp16_weight:
+                    module.fp16_weight = module.weight.data.clone().cpu().half()
+                    if module.bias is not None:
+                        module.fp16_bias = module.bias.data.clone().cpu().half()
+                module.to(torch.float8_e4m3fn)
+        model.first_stage_model = first_stage
+        timer.record("apply fp8")
+    else:
+        devices.fp8 = False
+
     devices.unet_needs_upcast = shared.cmd_opts.upcast_sampling and devices.dtype == torch.float16 and devices.dtype_unet == torch.float16
 
     model.first_stage_model.to(devices.dtype_vae)
@@ -442,7 +571,7 @@ def load_model_wrapper(model_type):
         path = midas.api.ISL_PATHS[model_type]
         if not os.path.exists(path):
             if not os.path.exists(midas_path):
-                mkdir(midas_path)
+                os.mkdir(midas_path)
 
             print(f"Downloading midas model weights for {model_type} to {path}")
             request.urlretrieve(midas_urls[model_type], path)
@@ -453,25 +582,90 @@ def load_model_wrapper(model_type):
     midas.api.load_model = load_model_wrapper
 
 
-def repair_config(sd_config):
+def patch_given_betas():
+    import ldm.models.diffusion.ddpm
+
+    def patched_register_schedule(*args, **kwargs):
+        """a modified version of register_schedule function that converts plain list from Omegaconf into numpy"""
+
+        if isinstance(args[1], ListConfig):
+            args = (args[0], np.array(args[1]), *args[2:])
 
+        original_register_schedule(*args, **kwargs)
+
+    original_register_schedule = patches.patch(__name__, ldm.models.diffusion.ddpm.DDPM, 'register_schedule', patched_register_schedule)
+
+
+def repair_config(sd_config, state_dict=None):
     if not hasattr(sd_config.model.params, "use_ema"):
         sd_config.model.params.use_ema = False
 
     if hasattr(sd_config.model.params, 'unet_config'):
         if shared.cmd_opts.no_half:
             sd_config.model.params.unet_config.params.use_fp16 = False
-        elif shared.cmd_opts.upcast_sampling:
+        elif shared.cmd_opts.upcast_sampling or shared.cmd_opts.precision == "half":
             sd_config.model.params.unet_config.params.use_fp16 = True
 
-    if getattr(sd_config.model.params.first_stage_config.params.ddconfig, "attn_type", None) == "vanilla-xformers" and not shared.xformers_available:
-        sd_config.model.params.first_stage_config.params.ddconfig.attn_type = "vanilla"
+    if hasattr(sd_config.model.params, 'first_stage_config'):
+        if getattr(sd_config.model.params.first_stage_config.params.ddconfig, "attn_type", None) == "vanilla-xformers" and not shared.xformers_available:
+            sd_config.model.params.first_stage_config.params.ddconfig.attn_type = "vanilla"
 
     # For UnCLIP-L, override the hardcoded karlo directory
     if hasattr(sd_config.model.params, "noise_aug_config") and hasattr(sd_config.model.params.noise_aug_config.params, "clip_stats_path"):
         karlo_path = os.path.join(paths.models_path, 'karlo')
         sd_config.model.params.noise_aug_config.params.clip_stats_path = sd_config.model.params.noise_aug_config.params.clip_stats_path.replace("checkpoints/karlo_models", karlo_path)
 
+    # Do not use checkpoint for inference.
+    # This helps prevent extra performance overhead on checking parameters.
+    # The perf overhead is about 100ms/it on 4090 for SDXL.
+    if hasattr(sd_config.model.params, "network_config"):
+        sd_config.model.params.network_config.params.use_checkpoint = False
+    if hasattr(sd_config.model.params, "unet_config"):
+        sd_config.model.params.unet_config.params.use_checkpoint = False
+
+
+
+def rescale_zero_terminal_snr_abar(alphas_cumprod):
+    alphas_bar_sqrt = alphas_cumprod.sqrt()
+
+    # Store old values.
+    alphas_bar_sqrt_0 = alphas_bar_sqrt[0].clone()
+    alphas_bar_sqrt_T = alphas_bar_sqrt[-1].clone()
+
+    # Shift so the last timestep is zero.
+    alphas_bar_sqrt -= (alphas_bar_sqrt_T)
+
+    # Scale so the first timestep is back to the old value.
+    alphas_bar_sqrt *= alphas_bar_sqrt_0 / (alphas_bar_sqrt_0 - alphas_bar_sqrt_T)
+
+    # Convert alphas_bar_sqrt to betas
+    alphas_bar = alphas_bar_sqrt ** 2  # Revert sqrt
+    alphas_bar[-1] = 4.8973451890853435e-08
+    return alphas_bar
+
+
+def apply_alpha_schedule_override(sd_model, p=None):
+    """
+    Applies an override to the alpha schedule of the model according to settings.
+    - downcasts the alpha schedule to half precision
+    - rescales the alpha schedule to have zero terminal SNR
+    """
+
+    if not hasattr(sd_model, 'alphas_cumprod') or not hasattr(sd_model, 'alphas_cumprod_original'):
+        return
+
+    sd_model.alphas_cumprod = sd_model.alphas_cumprod_original.to(shared.device)
+
+    if opts.use_downcasted_alpha_bar:
+        if p is not None:
+            p.extra_generation_params['Downcast alphas_cumprod'] = opts.use_downcasted_alpha_bar
+        sd_model.alphas_cumprod = sd_model.alphas_cumprod.half().to(shared.device)
+
+    if opts.sd_noise_schedule == "Zero Terminal SNR":
+        if p is not None:
+            p.extra_generation_params['Noise Schedule'] = opts.sd_noise_schedule
+        sd_model.alphas_cumprod = rescale_zero_terminal_snr_abar(sd_model.alphas_cumprod).to(shared.device)
+
 
 sd1_clip_weight = 'cond_stage_model.transformer.text_model.embeddings.token_embedding.weight'
 sd2_clip_weight = 'cond_stage_model.model.transformer.resblocks.0.attn.in_proj_weight'
@@ -530,18 +724,23 @@ def get_empty_cond(sd_model):
     p = processing.StableDiffusionProcessingTxt2Img()
     extra_networks.activate(p, {})
 
-    if hasattr(sd_model, 'conditioner'):
+    if hasattr(sd_model, 'get_learned_conditioning'):
         d = sd_model.get_learned_conditioning([""])
-        return d['crossattn']
     else:
-        return sd_model.cond_stage_model([""])
+        d = sd_model.cond_stage_model([""])
+
+    if isinstance(d, dict):
+        d = d['crossattn']
+
+    return d
 
 
 def send_model_to_cpu(m):
-    if m.lowvram:
-        lowvram.send_everything_to_cpu()
-    else:
-        m.to(devices.cpu)
+    if m is not None:
+        if m.lowvram:
+            lowvram.send_everything_to_cpu()
+        else:
+            m.to(devices.cpu)
 
     devices.torch_gc()
 
@@ -565,6 +764,25 @@ def send_model_to_trash(m):
     devices.torch_gc()
 
 
+def instantiate_from_config(config, state_dict=None):
+    constructor = get_obj_from_str(config["target"])
+
+    params = {**config.get("params", {})}
+
+    if state_dict and "state_dict" in params and params["state_dict"] is None:
+        params["state_dict"] = state_dict
+
+    return constructor(**params)
+
+
+def get_obj_from_str(string, reload=False):
+    module, cls = string.rsplit(".", 1)
+    if reload:
+        module_imp = importlib.import_module(module)
+        importlib.reload(module_imp)
+    return getattr(importlib.import_module(module, package=None), cls)
+
+
 def load_model(checkpoint_info=None, already_loaded_state_dict=None):
     from modules import sd_hijack
     checkpoint_info = checkpoint_info or select_checkpoint()
@@ -589,7 +807,7 @@ def load_model(checkpoint_info=None, already_loaded_state_dict=None):
     timer.record("find config")
 
     sd_config = OmegaConf.load(checkpoint_config)
-    repair_config(sd_config)
+    repair_config(sd_config, state_dict)
 
     timer.record("load config")
 
@@ -599,7 +817,7 @@ def load_model(checkpoint_info=None, already_loaded_state_dict=None):
     try:
         with sd_disable_initialization.DisableInitialization(disable_clip=clip_is_included_into_sd or shared.cmd_opts.do_not_download_clip):
             with sd_disable_initialization.InitializeOnMeta():
-                sd_model = instantiate_from_config(sd_config.model)
+                sd_model = instantiate_from_config(sd_config.model, state_dict)
 
     except Exception as e:
         errors.display(e, "creating model quickly", full_traceback=True)
@@ -608,7 +826,7 @@ def load_model(checkpoint_info=None, already_loaded_state_dict=None):
         print('Failed to create model quickly; will retry using slow method.', file=sys.stderr)
 
         with sd_disable_initialization.InitializeOnMeta():
-            sd_model = instantiate_from_config(sd_config.model)
+            sd_model = instantiate_from_config(sd_config.model, state_dict)
 
     sd_model.used_config = checkpoint_config
 
@@ -619,11 +837,13 @@ def load_model(checkpoint_info=None, already_loaded_state_dict=None):
     else:
         weight_dtype_conversion = {
             'first_stage_model': None,
+            'alphas_cumprod': None,
             '': torch.float16,
         }
 
     with sd_disable_initialization.LoadStateDictOnMeta(state_dict, device=model_target_device(sd_model), weight_dtype_conversion=weight_dtype_conversion):
         load_model_weights(sd_model, checkpoint_info, state_dict, timer)
+
     timer.record("load weights from state dict")
 
     send_model_to_device(sd_model)
@@ -661,9 +881,16 @@ def reuse_model_from_already_loaded(sd_model, checkpoint_info, timer):
     If it is loaded, returns that (moving it to GPU if necessary, and moving the currently loadded model to CPU if necessary).
     If not, returns the model that can be used to load weights from checkpoint_info's file.
     If no such model exists, returns None.
-    Additionaly deletes loaded models that are over the limit set in settings (sd_checkpoints_limit).
+    Additionally deletes loaded models that are over the limit set in settings (sd_checkpoints_limit).
     """
 
+    if sd_model is not None and sd_model.sd_checkpoint_info.filename == checkpoint_info.filename:
+        return sd_model
+
+    if shared.opts.sd_checkpoints_keep_in_cpu:
+        send_model_to_cpu(sd_model)
+        timer.record("send model to cpu")
+
     already_loaded = None
     for i in reversed(range(len(model_data.loaded_sd_models))):
         loaded_model = model_data.loaded_sd_models[i]
@@ -673,14 +900,10 @@ def reuse_model_from_already_loaded(sd_model, checkpoint_info, timer):
 
         if len(model_data.loaded_sd_models) > shared.opts.sd_checkpoints_limit > 0:
             print(f"Unloading model {len(model_data.loaded_sd_models)} over the limit of {shared.opts.sd_checkpoints_limit}: {loaded_model.sd_checkpoint_info.title}")
-            model_data.loaded_sd_models.pop()
+            del model_data.loaded_sd_models[i]
             send_model_to_trash(loaded_model)
             timer.record("send model to trash")
 
-        if shared.opts.sd_checkpoints_keep_in_cpu:
-            send_model_to_cpu(sd_model)
-            timer.record("send model to cpu")
-
     if already_loaded is not None:
         send_model_to_device(already_loaded)
         timer.record("send model to device")
@@ -714,7 +937,7 @@ def reuse_model_from_already_loaded(sd_model, checkpoint_info, timer):
         return None
 
 
-def reload_model_weights(sd_model=None, info=None):
+def reload_model_weights(sd_model=None, info=None, forced_reload=False):
     checkpoint_info = info or select_checkpoint()
 
     timer = Timer()
@@ -726,11 +949,14 @@ def reload_model_weights(sd_model=None, info=None):
         current_checkpoint_info = None
     else:
         current_checkpoint_info = sd_model.sd_checkpoint_info
-        if sd_model.sd_model_checkpoint == checkpoint_info.filename:
+        if check_fp8(sd_model) != devices.fp8:
+            # load from state dict again to prevent extra numerical errors
+            forced_reload = True
+        elif sd_model.sd_model_checkpoint == checkpoint_info.filename and not forced_reload:
             return sd_model
 
     sd_model = reuse_model_from_already_loaded(sd_model, checkpoint_info, timer)
-    if sd_model is not None and sd_model.sd_checkpoint_info.filename == checkpoint_info.filename:
+    if not forced_reload and sd_model is not None and sd_model.sd_checkpoint_info.filename == checkpoint_info.filename:
         return sd_model
 
     if sd_model is not None:
@@ -761,13 +987,13 @@ def reload_model_weights(sd_model=None, info=None):
         sd_hijack.model_hijack.hijack(sd_model)
         timer.record("hijack")
 
-        script_callbacks.model_loaded_callback(sd_model)
-        timer.record("script callbacks")
-
         if not sd_model.lowvram:
             sd_model.to(devices.device)
             timer.record("move model to device")
 
+        script_callbacks.model_loaded_callback(sd_model)
+        timer.record("script callbacks")
+
     print(f"Weights loaded in {timer.summary()}.")
 
     model_data.set_sd_model(sd_model)
@@ -777,17 +1003,7 @@ def reload_model_weights(sd_model=None, info=None):
 
 
 def unload_model_weights(sd_model=None, info=None):
-    timer = Timer()
-
-    if model_data.sd_model:
-        model_data.sd_model.to(devices.cpu)
-        sd_hijack.model_hijack.undo_hijack(model_data.sd_model)
-        model_data.sd_model = None
-        sd_model = None
-        gc.collect()
-        devices.torch_gc()
-
-    print(f"Unloaded weights {timer.summary()}.")
+    send_model_to_cpu(sd_model or shared.sd_model)
 
     return sd_model
 
diff --git a/modules/sd_models_config.py b/modules/sd_models_config.py
index 08dd03f19c7..fb44c5a8d98 100644
--- a/modules/sd_models_config.py
+++ b/modules/sd_models_config.py
@@ -15,12 +15,15 @@
 config_sd2_inpainting = os.path.join(sd_repo_configs_path, "v2-inpainting-inference.yaml")
 config_sdxl = os.path.join(sd_xl_repo_configs_path, "sd_xl_base.yaml")
 config_sdxl_refiner = os.path.join(sd_xl_repo_configs_path, "sd_xl_refiner.yaml")
+config_sdxl_inpainting = os.path.join(sd_configs_path, "sd_xl_inpaint.yaml")
 config_depth_model = os.path.join(sd_repo_configs_path, "v2-midas-inference.yaml")
 config_unclip = os.path.join(sd_repo_configs_path, "v2-1-stable-unclip-l-inference.yaml")
 config_unopenclip = os.path.join(sd_repo_configs_path, "v2-1-stable-unclip-h-inference.yaml")
 config_inpainting = os.path.join(sd_configs_path, "v1-inpainting-inference.yaml")
 config_instruct_pix2pix = os.path.join(sd_configs_path, "instruct-pix2pix.yaml")
 config_alt_diffusion = os.path.join(sd_configs_path, "alt-diffusion-inference.yaml")
+config_alt_diffusion_m18 = os.path.join(sd_configs_path, "alt-diffusion-m18-inference.yaml")
+config_sd3 = os.path.join(sd_configs_path, "sd3-inference.yaml")
 
 
 def is_using_v_parameterization_for_sd2(state_dict):
@@ -30,11 +33,11 @@ def is_using_v_parameterization_for_sd2(state_dict):
 
     import ldm.modules.diffusionmodules.openaimodel
 
-    device = devices.cpu
+    device = devices.device
 
     with sd_disable_initialization.DisableInitialization():
         unet = ldm.modules.diffusionmodules.openaimodel.UNetModel(
-            use_checkpoint=True,
+            use_checkpoint=False,
             use_fp16=False,
             image_size=32,
             in_channels=4,
@@ -55,12 +58,13 @@ def is_using_v_parameterization_for_sd2(state_dict):
     with torch.no_grad():
         unet_sd = {k.replace("model.diffusion_model.", ""): v for k, v in state_dict.items() if "model.diffusion_model." in k}
         unet.load_state_dict(unet_sd, strict=True)
-        unet.to(device=device, dtype=torch.float)
+        unet.to(device=device, dtype=devices.dtype_unet)
 
         test_cond = torch.ones((1, 2, 1024), device=device) * 0.5
         x_test = torch.ones((1, 4, 8, 8), device=device) * 0.5
 
-        out = (unet(x_test, torch.asarray([999], device=device), context=test_cond) - x_test).mean().item()
+        with devices.autocast():
+            out = (unet(x_test, torch.asarray([999], device=device), context=test_cond) - x_test).mean().cpu().item()
 
     return out < -1
 
@@ -70,8 +74,15 @@ def guess_model_config_from_state_dict(sd, filename):
     diffusion_model_input = sd.get('model.diffusion_model.input_blocks.0.0.weight', None)
     sd2_variations_weight = sd.get('embedder.model.ln_final.weight', None)
 
+    if "model.diffusion_model.x_embedder.proj.weight" in sd:
+        return config_sd3
+
     if sd.get('conditioner.embedders.1.model.ln_final.weight', None) is not None:
-        return config_sdxl
+        if diffusion_model_input.shape[1] == 9:
+            return config_sdxl_inpainting
+        else:
+            return config_sdxl
+
     if sd.get('conditioner.embedders.0.model.ln_final.weight', None) is not None:
         return config_sdxl_refiner
     elif sd.get('depth_model.model.pretrained.act_postprocess3.0.project.0.bias', None) is not None:
@@ -96,6 +107,8 @@ def guess_model_config_from_state_dict(sd, filename):
             return config_instruct_pix2pix
 
     if sd.get('cond_stage_model.roberta.embeddings.word_embeddings.weight', None) is not None:
+        if sd.get('cond_stage_model.transformation.weight').size()[0] == 1024:
+            return config_alt_diffusion_m18
         return config_alt_diffusion
 
     return config_default
diff --git a/modules/sd_models_types.py b/modules/sd_models_types.py
index 5ffd2f4f9fd..2fce2777b2f 100644
--- a/modules/sd_models_types.py
+++ b/modules/sd_models_types.py
@@ -22,10 +22,19 @@ class WebuiSdModel(LatentDiffusion):
     """structure with additional information about the file with model's weights"""
 
     is_sdxl: bool
-    """True if the model's architecture is SDXL"""
+    """True if the model's architecture is SDXL or SSD"""
+
+    is_ssd: bool
+    """True if the model is SSD"""
 
     is_sd2: bool
     """True if the model's architecture is SD 2.x"""
 
     is_sd1: bool
     """True if the model's architecture is SD 1.x"""
+
+    is_sd3: bool
+    """True if the model's architecture is SD 3"""
+
+    latent_channels: int
+    """number of layer in latent image representation; will be 16 in SD3 and 4 in other version"""
diff --git a/modules/sd_models_xl.py b/modules/sd_models_xl.py
index 0112332161f..1242a59369f 100644
--- a/modules/sd_models_xl.py
+++ b/modules/sd_models_xl.py
@@ -6,14 +6,15 @@
 import sgm.modules.diffusionmodules.denoiser_scaling
 import sgm.modules.diffusionmodules.discretizer
 from modules import devices, shared, prompt_parser
+from modules import torch_utils
 
 
 def get_learned_conditioning(self: sgm.models.diffusion.DiffusionEngine, batch: prompt_parser.SdConditioning | list[str]):
     for embedder in self.conditioner.embedders:
         embedder.ucg_rate = 0.0
 
-    width = getattr(batch, 'width', 1024)
-    height = getattr(batch, 'height', 1024)
+    width = getattr(batch, 'width', 1024) or 1024
+    height = getattr(batch, 'height', 1024) or 1024
     is_negative_prompt = getattr(batch, 'is_negative_prompt', False)
     aesthetic_score = shared.opts.sdxl_refiner_low_aesthetic_score if is_negative_prompt else shared.opts.sdxl_refiner_high_aesthetic_score
 
@@ -34,6 +35,11 @@ def get_learned_conditioning(self: sgm.models.diffusion.DiffusionEngine, batch:
 
 
 def apply_model(self: sgm.models.diffusion.DiffusionEngine, x, t, cond):
+    """WARNING: This function is called once per denoising iteration. DO NOT add
+    expensive functionc calls such as `model.state_dict`. """
+    if self.is_sdxl_inpaint:
+        x = torch.cat([x] + cond['c_concat'], dim=1)
+
     return self.model(x, t, cond)
 
 
@@ -84,7 +90,7 @@ def get_target_prompt_token_count(self, token_count):
 def extend_sdxl(model):
     """this adds a bunch of parameters to make SDXL model look a bit more like SD1.5 to the rest of the codebase."""
 
-    dtype = next(model.model.diffusion_model.parameters()).dtype
+    dtype = torch_utils.get_param(model.model.diffusion_model).dtype
     model.model.diffusion_model.dtype = dtype
     model.model.conditioning_key = 'crossattn'
     model.cond_stage_key = 'txt'
@@ -93,7 +99,7 @@ def extend_sdxl(model):
     model.parameterization = "v" if isinstance(model.denoiser.scaling, sgm.modules.diffusionmodules.denoiser_scaling.VScaling) else "eps"
 
     discretization = sgm.modules.diffusionmodules.discretizer.LegacyDDPMDiscretization()
-    model.alphas_cumprod = torch.asarray(discretization.alphas_cumprod, device=devices.device, dtype=dtype)
+    model.alphas_cumprod = torch.asarray(discretization.alphas_cumprod, device=devices.device, dtype=torch.float32)
 
     model.conditioner.wrapped = torch.nn.Module()
 
diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py
index 45faae62821..963da5be0bf 100644
--- a/modules/sd_samplers.py
+++ b/modules/sd_samplers.py
@@ -1,16 +1,22 @@
-from modules import sd_samplers_kdiffusion, sd_samplers_timesteps, shared
+from __future__ import annotations
+
+import functools
+import logging
+from modules import sd_samplers_kdiffusion, sd_samplers_timesteps, sd_samplers_lcm, shared, sd_samplers_common, sd_schedulers
 
 # imports for functions that previously were here and are used by other modules
-from modules.sd_samplers_common import samples_to_image_grid, sample_to_image  # noqa: F401
+samples_to_image_grid = sd_samplers_common.samples_to_image_grid
+sample_to_image = sd_samplers_common.sample_to_image
 
 all_samplers = [
     *sd_samplers_kdiffusion.samplers_data_k_diffusion,
     *sd_samplers_timesteps.samplers_data_timesteps,
+    *sd_samplers_lcm.samplers_data_lcm,
 ]
 all_samplers_map = {x.name: x for x in all_samplers}
 
-samplers = []
-samplers_for_img2img = []
+samplers: list[sd_samplers_common.SamplerData] = []
+samplers_for_img2img: list[sd_samplers_common.SamplerData] = []
 samplers_map = {}
 samplers_hidden = {}
 
@@ -56,4 +62,71 @@ def visible_sampler_names():
     return [x.name for x in samplers if x.name not in samplers_hidden]
 
 
+def visible_samplers():
+    return [x for x in samplers if x.name not in samplers_hidden]
+
+
+def get_sampler_from_infotext(d: dict):
+    return get_sampler_and_scheduler(d.get("Sampler"), d.get("Schedule type"))[0]
+
+
+def get_scheduler_from_infotext(d: dict):
+    return get_sampler_and_scheduler(d.get("Sampler"), d.get("Schedule type"))[1]
+
+
+def get_hr_sampler_and_scheduler(d: dict):
+    hr_sampler = d.get("Hires sampler", "Use same sampler")
+    sampler = d.get("Sampler") if hr_sampler == "Use same sampler" else hr_sampler
+
+    hr_scheduler = d.get("Hires schedule type", "Use same scheduler")
+    scheduler = d.get("Schedule type") if hr_scheduler == "Use same scheduler" else hr_scheduler
+
+    sampler, scheduler = get_sampler_and_scheduler(sampler, scheduler)
+
+    sampler = sampler if sampler != d.get("Sampler") else "Use same sampler"
+    scheduler = scheduler if scheduler != d.get("Schedule type") else "Use same scheduler"
+
+    return sampler, scheduler
+
+
+def get_hr_sampler_from_infotext(d: dict):
+    return get_hr_sampler_and_scheduler(d)[0]
+
+
+def get_hr_scheduler_from_infotext(d: dict):
+    return get_hr_sampler_and_scheduler(d)[1]
+
+
+@functools.cache
+def get_sampler_and_scheduler(sampler_name, scheduler_name, *, convert_automatic=True):
+    default_sampler = samplers[0]
+    found_scheduler = sd_schedulers.schedulers_map.get(scheduler_name, sd_schedulers.schedulers[0])
+
+    name = sampler_name or default_sampler.name
+
+    for scheduler in sd_schedulers.schedulers:
+        name_options = [scheduler.label, scheduler.name, *(scheduler.aliases or [])]
+
+        for name_option in name_options:
+            if name.endswith(" " + name_option):
+                found_scheduler = scheduler
+                name = name[0:-(len(name_option) + 1)]
+                break
+
+    sampler = all_samplers_map.get(name, default_sampler)
+
+    # revert back to Automatic if it's the default scheduler for the selected sampler
+    if convert_automatic and sampler.options.get('scheduler', None) == found_scheduler.name:
+        found_scheduler = sd_schedulers.schedulers[0]
+
+    return sampler.name, found_scheduler.label
+
+
+def fix_p_invalid_sampler_and_scheduler(p):
+    i_sampler_name, i_scheduler = p.sampler_name, p.scheduler
+    p.sampler_name, p.scheduler = get_sampler_and_scheduler(p.sampler_name, p.scheduler, convert_automatic=False)
+    if p.sampler_name != i_sampler_name or i_scheduler != p.scheduler:
+        logging.warning(f'Sampler Scheduler autocorrection: "{i_sampler_name}" -> "{p.sampler_name}", "{i_scheduler}" -> "{p.scheduler}"')
+
+
 set_samplers()
diff --git a/modules/sd_samplers_cfg_denoiser.py b/modules/sd_samplers_cfg_denoiser.py
index b8101d38dc3..b6fbf337243 100644
--- a/modules/sd_samplers_cfg_denoiser.py
+++ b/modules/sd_samplers_cfg_denoiser.py
@@ -1,5 +1,5 @@
 import torch
-from modules import prompt_parser, devices, sd_samplers_common
+from modules import prompt_parser, sd_samplers_common
 
 from modules.shared import opts, state
 import modules.shared as shared
@@ -53,9 +53,18 @@ def __init__(self, sampler):
         self.step = 0
         self.image_cfg_scale = None
         self.padded_cond_uncond = False
+        self.padded_cond_uncond_v0 = False
         self.sampler = sampler
         self.model_wrap = None
         self.p = None
+
+        self.cond_scale_miltiplier = 1.0
+
+        self.need_last_noise_uncond = False
+        self.last_noise_uncond = None
+
+        # NOTE: masking before denoising can cause the original latents to be oversmoothed
+        # as the original latents do not have noise
         self.mask_before_denoising = False
 
     @property
@@ -88,11 +97,67 @@ def update_inner_model(self):
         self.sampler.sampler_extra_args['cond'] = c
         self.sampler.sampler_extra_args['uncond'] = uc
 
+    def pad_cond_uncond(self, cond, uncond):
+        empty = shared.sd_model.cond_stage_model_empty_prompt
+        num_repeats = (cond.shape[1] - uncond.shape[1]) // empty.shape[1]
+
+        if num_repeats < 0:
+            cond = pad_cond(cond, -num_repeats, empty)
+            self.padded_cond_uncond = True
+        elif num_repeats > 0:
+            uncond = pad_cond(uncond, num_repeats, empty)
+            self.padded_cond_uncond = True
+
+        return cond, uncond
+
+    def pad_cond_uncond_v0(self, cond, uncond):
+        """
+        Pads the 'uncond' tensor to match the shape of the 'cond' tensor.
+
+        If 'uncond' is a dictionary, it is assumed that the 'crossattn' key holds the tensor to be padded.
+        If 'uncond' is a tensor, it is padded directly.
+
+        If the number of columns in 'uncond' is less than the number of columns in 'cond', the last column of 'uncond'
+        is repeated to match the number of columns in 'cond'.
+
+        If the number of columns in 'uncond' is greater than the number of columns in 'cond', 'uncond' is truncated
+        to match the number of columns in 'cond'.
+
+        Args:
+            cond (torch.Tensor or DictWithShape): The condition tensor to match the shape of 'uncond'.
+            uncond (torch.Tensor or DictWithShape): The tensor to be padded, or a dictionary containing the tensor to be padded.
+
+        Returns:
+            tuple: A tuple containing the 'cond' tensor and the padded 'uncond' tensor.
+
+        Note:
+            This is the padding that was always used in DDIM before version 1.6.0
+        """
+
+        is_dict_cond = isinstance(uncond, dict)
+        uncond_vec = uncond['crossattn'] if is_dict_cond else uncond
+
+        if uncond_vec.shape[1] < cond.shape[1]:
+            last_vector = uncond_vec[:, -1:]
+            last_vector_repeated = last_vector.repeat([1, cond.shape[1] - uncond_vec.shape[1], 1])
+            uncond_vec = torch.hstack([uncond_vec, last_vector_repeated])
+            self.padded_cond_uncond_v0 = True
+        elif uncond_vec.shape[1] > cond.shape[1]:
+            uncond_vec = uncond_vec[:, :cond.shape[1]]
+            self.padded_cond_uncond_v0 = True
+
+        if is_dict_cond:
+            uncond['crossattn'] = uncond_vec
+        else:
+            uncond = uncond_vec
+
+        return cond, uncond
+
     def forward(self, x, sigma, uncond, cond, cond_scale, s_min_uncond, image_cond):
         if state.interrupted or state.skipped:
             raise sd_samplers_common.InterruptedException
 
-        if sd_samplers_common.apply_refiner(self):
+        if sd_samplers_common.apply_refiner(self, sigma):
             cond = self.sampler.sampler_extra_args['cond']
             uncond = self.sampler.sampler_extra_args['uncond']
 
@@ -105,8 +170,21 @@ def forward(self, x, sigma, uncond, cond, cond_scale, s_min_uncond, image_cond):
 
         assert not is_edit_model or all(len(conds) == 1 for conds in conds_list), "AND is not supported for InstructPix2Pix checkpoint (unless using Image CFG scale = 1.0)"
 
+        # If we use masks, blending between the denoised and original latent images occurs here.
+        def apply_blend(current_latent):
+            blended_latent = current_latent * self.nmask + self.init_latent * self.mask
+
+            if self.p.scripts is not None:
+                from modules import scripts
+                mba = scripts.MaskBlendArgs(current_latent, self.nmask, self.init_latent, self.mask, blended_latent, denoiser=self, sigma=sigma)
+                self.p.scripts.on_mask_blend(self.p, mba)
+                blended_latent = mba.blended_latent
+
+            return blended_latent
+
+        # Blend in the original latents (before)
         if self.mask_before_denoising and self.mask is not None:
-            x = self.init_latent * self.mask + self.nmask * x
+            x = apply_blend(x)
 
         batch_size = len(conds_list)
         repeats = [len(conds_list[i]) for i in range(batch_size)]
@@ -130,7 +208,7 @@ def forward(self, x, sigma, uncond, cond, cond_scale, s_min_uncond, image_cond):
             sigma_in = torch.cat([torch.stack([sigma[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [sigma] + [sigma])
             image_cond_in = torch.cat([torch.stack([image_cond[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [image_uncond] + [torch.zeros_like(self.init_latent)])
 
-        denoiser_params = CFGDenoiserParams(x_in, image_cond_in, sigma_in, state.sampling_step, state.sampling_steps, tensor, uncond)
+        denoiser_params = CFGDenoiserParams(x_in, image_cond_in, sigma_in, state.sampling_step, state.sampling_steps, tensor, uncond, self)
         cfg_denoiser_callback(denoiser_params)
         x_in = denoiser_params.x
         image_cond_in = denoiser_params.image_cond
@@ -139,23 +217,25 @@ def forward(self, x, sigma, uncond, cond, cond_scale, s_min_uncond, image_cond):
         uncond = denoiser_params.text_uncond
         skip_uncond = False
 
-        # alternating uncond allows for higher thresholds without the quality loss normally expected from raising it
-        if self.step % 2 and s_min_uncond > 0 and sigma[0] < s_min_uncond and not is_edit_model:
+        if shared.opts.skip_early_cond != 0. and self.step / self.total_steps <= shared.opts.skip_early_cond:
+            skip_uncond = True
+            self.p.extra_generation_params["Skip Early CFG"] = shared.opts.skip_early_cond
+        elif (self.step % 2 or shared.opts.s_min_uncond_all) and s_min_uncond > 0 and sigma[0] < s_min_uncond and not is_edit_model:
             skip_uncond = True
+            self.p.extra_generation_params["NGMS"] = s_min_uncond
+            if shared.opts.s_min_uncond_all:
+                self.p.extra_generation_params["NGMS all steps"] = shared.opts.s_min_uncond_all
+
+        if skip_uncond:
             x_in = x_in[:-batch_size]
             sigma_in = sigma_in[:-batch_size]
 
         self.padded_cond_uncond = False
-        if shared.opts.pad_cond_uncond and tensor.shape[1] != uncond.shape[1]:
-            empty = shared.sd_model.cond_stage_model_empty_prompt
-            num_repeats = (tensor.shape[1] - uncond.shape[1]) // empty.shape[1]
-
-            if num_repeats < 0:
-                tensor = pad_cond(tensor, -num_repeats, empty)
-                self.padded_cond_uncond = True
-            elif num_repeats > 0:
-                uncond = pad_cond(uncond, num_repeats, empty)
-                self.padded_cond_uncond = True
+        self.padded_cond_uncond_v0 = False
+        if shared.opts.pad_cond_uncond_v0 and tensor.shape[1] != uncond.shape[1]:
+            tensor, uncond = self.pad_cond_uncond_v0(tensor, uncond)
+        elif shared.opts.pad_cond_uncond and tensor.shape[1] != uncond.shape[1]:
+            tensor, uncond = self.pad_cond_uncond(tensor, uncond)
 
         if tensor.shape[1] == uncond.shape[1] or skip_uncond:
             if is_edit_model:
@@ -198,17 +278,19 @@ def forward(self, x, sigma, uncond, cond, cond_scale, s_min_uncond, image_cond):
         denoised_params = CFGDenoisedParams(x_out, state.sampling_step, state.sampling_steps, self.inner_model)
         cfg_denoised_callback(denoised_params)
 
-        devices.test_for_nans(x_out, "unet")
+        if self.need_last_noise_uncond:
+            self.last_noise_uncond = torch.clone(x_out[-uncond.shape[0]:])
 
         if is_edit_model:
-            denoised = self.combine_denoised_for_edit_model(x_out, cond_scale)
+            denoised = self.combine_denoised_for_edit_model(x_out, cond_scale * self.cond_scale_miltiplier)
         elif skip_uncond:
             denoised = self.combine_denoised(x_out, conds_list, uncond, 1.0)
         else:
-            denoised = self.combine_denoised(x_out, conds_list, uncond, cond_scale)
+            denoised = self.combine_denoised(x_out, conds_list, uncond, cond_scale * self.cond_scale_miltiplier)
 
+        # Blend in the original latents (after)
         if not self.mask_before_denoising and self.mask is not None:
-            denoised = self.init_latent * self.mask + self.nmask * denoised
+            denoised = apply_blend(denoised)
 
         self.sampler.last_latent = self.get_pred_x0(torch.cat([x_in[i:i + 1] for i in denoised_image_indexes]), torch.cat([x_out[i:i + 1] for i in denoised_image_indexes]), sigma)
 
diff --git a/modules/sd_samplers_common.py b/modules/sd_samplers_common.py
index 58efcad2374..c060cccb24b 100644
--- a/modules/sd_samplers_common.py
+++ b/modules/sd_samplers_common.py
@@ -54,7 +54,7 @@ def samples_to_images_tensor(sample, approximation=None, model=None):
     else:
         if model is None:
             model = shared.sd_model
-        with devices.without_autocast(): # fixes an issue with unstable VAEs that are flaky even in fp32
+        with torch.no_grad(), devices.without_autocast(): # fixes an issue with unstable VAEs that are flaky even in fp32
             x_sample = model.decode_first_stage(sample.to(model.first_stage_model.dtype))
 
     return x_sample
@@ -155,8 +155,19 @@ def torchsde_randn(size, dtype, device, seed):
 replace_torchsde_browinan()
 
 
-def apply_refiner(cfg_denoiser):
-    completed_ratio = cfg_denoiser.step / cfg_denoiser.total_steps
+def apply_refiner(cfg_denoiser, sigma=None):
+    if opts.refiner_switch_by_sample_steps or sigma is None:
+        completed_ratio = cfg_denoiser.step / cfg_denoiser.total_steps
+        cfg_denoiser.p.extra_generation_params["Refiner switch by sampling steps"] = True
+
+    else:
+        # torch.max(sigma) only to handle rare case where we might have different sigmas in the same batch
+        try:
+            timestep = torch.argmin(torch.abs(cfg_denoiser.inner_model.sigmas.to(sigma.device) - torch.max(sigma)))
+        except AttributeError:  # for samplers that don't use sigmas (DDIM) sigma is actually the timestep
+            timestep = torch.max(sigma).to(dtype=int)
+        completed_ratio = (999 - timestep) / 1000
+
     refiner_switch_at = cfg_denoiser.p.refiner_switch_at
     refiner_checkpoint_info = cfg_denoiser.p.refiner_checkpoint_info
 
@@ -235,7 +246,7 @@ def __init__(self, funcname):
         self.eta_infotext_field = 'Eta'
         self.eta_default = 1.0
 
-        self.conditioning_key = shared.sd_model.model.conditioning_key
+        self.conditioning_key = getattr(shared.sd_model.model, 'conditioning_key', 'crossattn')
 
         self.p = None
         self.model_wrap_cfg = None
@@ -335,3 +346,10 @@ def sample(self, p, x, conditioning, unconditional_conditioning, steps=None, ima
 
     def sample_img2img(self, p, x, noise, conditioning, unconditional_conditioning, steps=None, image_conditioning=None):
         raise NotImplementedError()
+
+    def add_infotext(self, p):
+        if self.model_wrap_cfg.padded_cond_uncond:
+            p.extra_generation_params["Pad conds"] = True
+
+        if self.model_wrap_cfg.padded_cond_uncond_v0:
+            p.extra_generation_params["Pad conds v0"] = True
diff --git a/modules/sd_samplers_extra.py b/modules/sd_samplers_extra.py
index 1b981ca80c3..72fd0aa5e60 100644
--- a/modules/sd_samplers_extra.py
+++ b/modules/sd_samplers_extra.py
@@ -60,7 +60,7 @@ def heun_step(x, old_sigma, new_sigma, second_order=True):
                 sigma_restart = get_sigmas_karras(restart_steps, sigmas[min_idx].item(), sigmas[max_idx].item(), device=sigmas.device)[:-1]
                 while restart_times > 0:
                     restart_times -= 1
-                    step_list.extend([(old_sigma, new_sigma) for (old_sigma, new_sigma) in zip(sigma_restart[:-1], sigma_restart[1:])])
+                    step_list.extend(zip(sigma_restart[:-1], sigma_restart[1:]))
 
     last_sigma = None
     for old_sigma, new_sigma in tqdm.tqdm(step_list, disable=disable):
diff --git a/modules/sd_samplers_kdiffusion.py b/modules/sd_samplers_kdiffusion.py
index 8a8c87e0d01..0c94d100d25 100644
--- a/modules/sd_samplers_kdiffusion.py
+++ b/modules/sd_samplers_kdiffusion.py
@@ -1,7 +1,7 @@
 import torch
 import inspect
 import k_diffusion.sampling
-from modules import sd_samplers_common, sd_samplers_extra, sd_samplers_cfg_denoiser
+from modules import sd_samplers_common, sd_samplers_extra, sd_samplers_cfg_denoiser, sd_schedulers, devices
 from modules.sd_samplers_cfg_denoiser import CFGDenoiser  # noqa: F401
 from modules.script_callbacks import ExtraNoiseParams, extra_noise_callback
 
@@ -9,32 +9,20 @@
 import modules.shared as shared
 
 samplers_k_diffusion = [
-    ('DPM++ 2M Karras', 'sample_dpmpp_2m', ['k_dpmpp_2m_ka'], {'scheduler': 'karras'}),
-    ('DPM++ SDE Karras', 'sample_dpmpp_sde', ['k_dpmpp_sde_ka'], {'scheduler': 'karras', "second_order": True, "brownian_noise": True}),
-    ('DPM++ 2M SDE Exponential', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_exp'], {'scheduler': 'exponential', "brownian_noise": True}),
-    ('DPM++ 2M SDE Karras', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_ka'], {'scheduler': 'karras', "brownian_noise": True}),
+    ('DPM++ 2M', 'sample_dpmpp_2m', ['k_dpmpp_2m'], {'scheduler': 'karras'}),
+    ('DPM++ SDE', 'sample_dpmpp_sde', ['k_dpmpp_sde'], {'scheduler': 'karras', "second_order": True, "brownian_noise": True}),
+    ('DPM++ 2M SDE', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde'], {'scheduler': 'exponential', "brownian_noise": True}),
+    ('DPM++ 2M SDE Heun', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_heun'], {'scheduler': 'exponential', "brownian_noise": True, "solver_type": "heun"}),
+    ('DPM++ 2S a', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a'], {'scheduler': 'karras', "uses_ensd": True, "second_order": True}),
+    ('DPM++ 3M SDE', 'sample_dpmpp_3m_sde', ['k_dpmpp_3m_sde'], {'scheduler': 'exponential', 'discard_next_to_last_sigma': True, "brownian_noise": True}),
     ('Euler a', 'sample_euler_ancestral', ['k_euler_a', 'k_euler_ancestral'], {"uses_ensd": True}),
     ('Euler', 'sample_euler', ['k_euler'], {}),
     ('LMS', 'sample_lms', ['k_lms'], {}),
     ('Heun', 'sample_heun', ['k_heun'], {"second_order": True}),
-    ('DPM2', 'sample_dpm_2', ['k_dpm_2'], {'discard_next_to_last_sigma': True, "second_order": True}),
-    ('DPM2 a', 'sample_dpm_2_ancestral', ['k_dpm_2_a'], {'discard_next_to_last_sigma': True, "uses_ensd": True, "second_order": True}),
-    ('DPM++ 2S a', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a'], {"uses_ensd": True, "second_order": True}),
-    ('DPM++ 2M', 'sample_dpmpp_2m', ['k_dpmpp_2m'], {}),
-    ('DPM++ SDE', 'sample_dpmpp_sde', ['k_dpmpp_sde'], {"second_order": True, "brownian_noise": True}),
-    ('DPM++ 2M SDE', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_ka'], {"brownian_noise": True}),
-    ('DPM++ 2M SDE Heun', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_heun'], {"brownian_noise": True, "solver_type": "heun"}),
-    ('DPM++ 2M SDE Heun Karras', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_heun_ka'], {'scheduler': 'karras', "brownian_noise": True, "solver_type": "heun"}),
-    ('DPM++ 2M SDE Heun Exponential', 'sample_dpmpp_2m_sde', ['k_dpmpp_2m_sde_heun_exp'], {'scheduler': 'exponential', "brownian_noise": True, "solver_type": "heun"}),
-    ('DPM++ 3M SDE', 'sample_dpmpp_3m_sde', ['k_dpmpp_3m_sde'], {'discard_next_to_last_sigma': True, "brownian_noise": True}),
-    ('DPM++ 3M SDE Karras', 'sample_dpmpp_3m_sde', ['k_dpmpp_3m_sde_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True, "brownian_noise": True}),
-    ('DPM++ 3M SDE Exponential', 'sample_dpmpp_3m_sde', ['k_dpmpp_3m_sde_exp'], {'scheduler': 'exponential', 'discard_next_to_last_sigma': True, "brownian_noise": True}),
+    ('DPM2', 'sample_dpm_2', ['k_dpm_2'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True, "second_order": True}),
+    ('DPM2 a', 'sample_dpm_2_ancestral', ['k_dpm_2_a'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True, "uses_ensd": True, "second_order": True}),
     ('DPM fast', 'sample_dpm_fast', ['k_dpm_fast'], {"uses_ensd": True}),
     ('DPM adaptive', 'sample_dpm_adaptive', ['k_dpm_ad'], {"uses_ensd": True}),
-    ('LMS Karras', 'sample_lms', ['k_lms_ka'], {'scheduler': 'karras'}),
-    ('DPM2 Karras', 'sample_dpm_2', ['k_dpm_2_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True, "uses_ensd": True, "second_order": True}),
-    ('DPM2 a Karras', 'sample_dpm_2_ancestral', ['k_dpm_2_a_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True, "uses_ensd": True, "second_order": True}),
-    ('DPM++ 2S a Karras', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a_ka'], {'scheduler': 'karras', "uses_ensd": True, "second_order": True}),
     ('Restart', sd_samplers_extra.restart_sampler, ['restart'], {'scheduler': 'karras', "second_order": True}),
 ]
 
@@ -58,20 +46,20 @@
 }
 
 k_diffusion_samplers_map = {x.name: x for x in samplers_data_k_diffusion}
-k_diffusion_scheduler = {
-    'Automatic': None,
-    'karras': k_diffusion.sampling.get_sigmas_karras,
-    'exponential': k_diffusion.sampling.get_sigmas_exponential,
-    'polyexponential': k_diffusion.sampling.get_sigmas_polyexponential
-}
+k_diffusion_scheduler = {x.name: x.function for x in sd_schedulers.schedulers}
 
 
 class CFGDenoiserKDiffusion(sd_samplers_cfg_denoiser.CFGDenoiser):
     @property
     def inner_model(self):
         if self.model_wrap is None:
-            denoiser = k_diffusion.external.CompVisVDenoiser if shared.sd_model.parameterization == "v" else k_diffusion.external.CompVisDenoiser
-            self.model_wrap = denoiser(shared.sd_model, quantize=shared.opts.enable_quantization)
+            denoiser_constructor = getattr(shared.sd_model, 'create_denoiser', None)
+
+            if denoiser_constructor is not None:
+                self.model_wrap = denoiser_constructor()
+            else:
+                denoiser = k_diffusion.external.CompVisVDenoiser if shared.sd_model.parameterization == "v" else k_diffusion.external.CompVisDenoiser
+                self.model_wrap = denoiser(shared.sd_model, quantize=shared.opts.enable_quantization)
 
         return self.model_wrap
 
@@ -96,47 +84,52 @@ def get_sigmas(self, p, steps):
 
         steps += 1 if discard_next_to_last_sigma else 0
 
+        scheduler_name = (p.hr_scheduler if p.is_hr_pass else p.scheduler) or 'Automatic'
+        if scheduler_name == 'Automatic':
+            scheduler_name = self.config.options.get('scheduler', None)
+
+        scheduler = sd_schedulers.schedulers_map.get(scheduler_name)
+
+        m_sigma_min, m_sigma_max = self.model_wrap.sigmas[0].item(), self.model_wrap.sigmas[-1].item()
+        sigma_min, sigma_max = (0.1, 10) if opts.use_old_karras_scheduler_sigmas else (m_sigma_min, m_sigma_max)
+
         if p.sampler_noise_scheduler_override:
             sigmas = p.sampler_noise_scheduler_override(steps)
-        elif opts.k_sched_type != "Automatic":
-            m_sigma_min, m_sigma_max = (self.model_wrap.sigmas[0].item(), self.model_wrap.sigmas[-1].item())
-            sigma_min, sigma_max = (0.1, 10) if opts.use_old_karras_scheduler_sigmas else (m_sigma_min, m_sigma_max)
-            sigmas_kwargs = {
-                'sigma_min': sigma_min,
-                'sigma_max': sigma_max,
-            }
-
-            sigmas_func = k_diffusion_scheduler[opts.k_sched_type]
-            p.extra_generation_params["Schedule type"] = opts.k_sched_type
-
-            if opts.sigma_min != m_sigma_min and opts.sigma_min != 0:
+        elif scheduler is None or scheduler.function is None:
+            sigmas = self.model_wrap.get_sigmas(steps)
+        else:
+            sigmas_kwargs = {'sigma_min': sigma_min, 'sigma_max': sigma_max}
+
+            if scheduler.label != 'Automatic' and not p.is_hr_pass:
+                p.extra_generation_params["Schedule type"] = scheduler.label
+            elif scheduler.label != p.extra_generation_params.get("Schedule type"):
+                p.extra_generation_params["Hires schedule type"] = scheduler.label
+
+            if opts.sigma_min != 0 and opts.sigma_min != m_sigma_min:
                 sigmas_kwargs['sigma_min'] = opts.sigma_min
                 p.extra_generation_params["Schedule min sigma"] = opts.sigma_min
-            if opts.sigma_max != m_sigma_max and opts.sigma_max != 0:
+
+            if opts.sigma_max != 0 and opts.sigma_max != m_sigma_max:
                 sigmas_kwargs['sigma_max'] = opts.sigma_max
                 p.extra_generation_params["Schedule max sigma"] = opts.sigma_max
 
-            default_rho = 1. if opts.k_sched_type == "polyexponential" else 7.
-
-            if opts.k_sched_type != 'exponential' and opts.rho != 0 and opts.rho != default_rho:
+            if scheduler.default_rho != -1 and opts.rho != 0 and opts.rho != scheduler.default_rho:
                 sigmas_kwargs['rho'] = opts.rho
                 p.extra_generation_params["Schedule rho"] = opts.rho
 
-            sigmas = sigmas_func(n=steps, **sigmas_kwargs, device=shared.device)
-        elif self.config is not None and self.config.options.get('scheduler', None) == 'karras':
-            sigma_min, sigma_max = (0.1, 10) if opts.use_old_karras_scheduler_sigmas else (self.model_wrap.sigmas[0].item(), self.model_wrap.sigmas[-1].item())
+            if scheduler.need_inner_model:
+                sigmas_kwargs['inner_model'] = self.model_wrap
 
-            sigmas = k_diffusion.sampling.get_sigmas_karras(n=steps, sigma_min=sigma_min, sigma_max=sigma_max, device=shared.device)
-        elif self.config is not None and self.config.options.get('scheduler', None) == 'exponential':
-            m_sigma_min, m_sigma_max = (self.model_wrap.sigmas[0].item(), self.model_wrap.sigmas[-1].item())
-            sigmas = k_diffusion.sampling.get_sigmas_exponential(n=steps, sigma_min=m_sigma_min, sigma_max=m_sigma_max, device=shared.device)
-        else:
-            sigmas = self.model_wrap.get_sigmas(steps)
+            if scheduler.label == 'Beta':
+                p.extra_generation_params["Beta schedule alpha"] = opts.beta_dist_alpha
+                p.extra_generation_params["Beta schedule beta"] = opts.beta_dist_beta
+
+            sigmas = scheduler.function(n=steps, **sigmas_kwargs, device=devices.cpu)
 
         if discard_next_to_last_sigma:
             sigmas = torch.cat([sigmas[:-2], sigmas[-1:]])
 
-        return sigmas
+        return sigmas.cpu()
 
     def sample_img2img(self, p, x, noise, conditioning, unconditional_conditioning, steps=None, image_conditioning=None):
         steps, t_enc = sd_samplers_common.setup_img2img_steps(p, steps)
@@ -144,7 +137,10 @@ def sample_img2img(self, p, x, noise, conditioning, unconditional_conditioning,
         sigmas = self.get_sigmas(p, steps)
         sigma_sched = sigmas[steps - t_enc - 1:]
 
-        xi = x + noise * sigma_sched[0]
+        if hasattr(shared.sd_model, 'add_noise_to_latent'):
+            xi = shared.sd_model.add_noise_to_latent(x, noise, sigma_sched[0])
+        else:
+            xi = x + noise * sigma_sched[0]
 
         if opts.img2img_extra_noise > 0:
             p.extra_generation_params["Extra noise"] = opts.img2img_extra_noise
@@ -187,8 +183,7 @@ def sample_img2img(self, p, x, noise, conditioning, unconditional_conditioning,
 
         samples = self.launch_sampling(t_enc + 1, lambda: self.func(self.model_wrap_cfg, xi, extra_args=self.sampler_extra_args, disable=False, callback=self.callback_state, **extra_params_kwargs))
 
-        if self.model_wrap_cfg.padded_cond_uncond:
-            p.extra_generation_params["Pad conds"] = True
+        self.add_infotext(p)
 
         return samples
 
@@ -234,8 +229,7 @@ def sample(self, p, x, conditioning, unconditional_conditioning, steps=None, ima
 
         samples = self.launch_sampling(steps, lambda: self.func(self.model_wrap_cfg, x, extra_args=self.sampler_extra_args, disable=False, callback=self.callback_state, **extra_params_kwargs))
 
-        if self.model_wrap_cfg.padded_cond_uncond:
-            p.extra_generation_params["Pad conds"] = True
+        self.add_infotext(p)
 
         return samples
 
diff --git a/modules/sd_samplers_lcm.py b/modules/sd_samplers_lcm.py
new file mode 100644
index 00000000000..59839b720dd
--- /dev/null
+++ b/modules/sd_samplers_lcm.py
@@ -0,0 +1,104 @@
+import torch
+
+from k_diffusion import utils, sampling
+from k_diffusion.external import DiscreteEpsDDPMDenoiser
+from k_diffusion.sampling import default_noise_sampler, trange
+
+from modules import shared, sd_samplers_cfg_denoiser, sd_samplers_kdiffusion, sd_samplers_common
+
+
+class LCMCompVisDenoiser(DiscreteEpsDDPMDenoiser):
+    def __init__(self, model):
+        timesteps = 1000
+        original_timesteps = 50     # LCM Original Timesteps (default=50, for current version of LCM)
+        self.skip_steps = timesteps // original_timesteps
+
+        alphas_cumprod_valid = torch.zeros((original_timesteps), dtype=torch.float32, device=model.device)
+        for x in range(original_timesteps):
+            alphas_cumprod_valid[original_timesteps - 1 - x] = model.alphas_cumprod[timesteps - 1 - x * self.skip_steps]
+
+        super().__init__(model, alphas_cumprod_valid, quantize=None)
+
+
+    def get_sigmas(self, n=None,):
+        if n is None:
+            return sampling.append_zero(self.sigmas.flip(0))
+
+        start = self.sigma_to_t(self.sigma_max)
+        end = self.sigma_to_t(self.sigma_min)
+
+        t = torch.linspace(start, end, n, device=shared.sd_model.device)
+
+        return sampling.append_zero(self.t_to_sigma(t))
+
+
+    def sigma_to_t(self, sigma, quantize=None):
+        log_sigma = sigma.log()
+        dists = log_sigma - self.log_sigmas[:, None]
+        return dists.abs().argmin(dim=0).view(sigma.shape) * self.skip_steps + (self.skip_steps - 1)
+
+
+    def t_to_sigma(self, timestep):
+        t = torch.clamp(((timestep - (self.skip_steps - 1)) / self.skip_steps).float(), min=0, max=(len(self.sigmas) - 1))
+        return super().t_to_sigma(t)
+
+
+    def get_eps(self, *args, **kwargs):
+        return self.inner_model.apply_model(*args, **kwargs)
+
+
+    def get_scaled_out(self, sigma, output, input):
+        sigma_data = 0.5
+        scaled_timestep = utils.append_dims(self.sigma_to_t(sigma), output.ndim) * 10.0
+
+        c_skip = sigma_data**2 / (scaled_timestep**2 + sigma_data**2)
+        c_out = scaled_timestep / (scaled_timestep**2 + sigma_data**2) ** 0.5
+
+        return c_out * output + c_skip * input
+
+
+    def forward(self, input, sigma, **kwargs):
+        c_out, c_in = [utils.append_dims(x, input.ndim) for x in self.get_scalings(sigma)]
+        eps = self.get_eps(input * c_in, self.sigma_to_t(sigma), **kwargs)
+        return self.get_scaled_out(sigma, input + eps * c_out, input)
+
+
+def sample_lcm(model, x, sigmas, extra_args=None, callback=None, disable=None, noise_sampler=None):
+    extra_args = {} if extra_args is None else extra_args
+    noise_sampler = default_noise_sampler(x) if noise_sampler is None else noise_sampler
+    s_in = x.new_ones([x.shape[0]])
+
+    for i in trange(len(sigmas) - 1, disable=disable):
+        denoised = model(x, sigmas[i] * s_in, **extra_args)
+
+        if callback is not None:
+            callback({'x': x, 'i': i, 'sigma': sigmas[i], 'sigma_hat': sigmas[i], 'denoised': denoised})
+
+        x = denoised
+        if sigmas[i + 1] > 0:
+            x += sigmas[i + 1] * noise_sampler(sigmas[i], sigmas[i + 1])
+    return x
+
+
+class CFGDenoiserLCM(sd_samplers_cfg_denoiser.CFGDenoiser):
+    @property
+    def inner_model(self):
+        if self.model_wrap is None:
+            denoiser = LCMCompVisDenoiser
+            self.model_wrap = denoiser(shared.sd_model)
+
+        return self.model_wrap
+
+
+class LCMSampler(sd_samplers_kdiffusion.KDiffusionSampler):
+    def __init__(self, funcname, sd_model, options=None):
+        super().__init__(funcname, sd_model, options)
+        self.model_wrap_cfg = CFGDenoiserLCM(self)
+        self.model_wrap = self.model_wrap_cfg.inner_model
+
+
+samplers_lcm = [('LCM', sample_lcm, ['k_lcm'], {})]
+samplers_data_lcm = [
+    sd_samplers_common.SamplerData(label, lambda model, funcname=funcname: LCMSampler(funcname, model), aliases, options)
+    for label, funcname, aliases, options in samplers_lcm
+]
diff --git a/modules/sd_samplers_timesteps.py b/modules/sd_samplers_timesteps.py
index b17a8f93c2b..81edd67d6e4 100644
--- a/modules/sd_samplers_timesteps.py
+++ b/modules/sd_samplers_timesteps.py
@@ -10,6 +10,7 @@
 
 samplers_timesteps = [
     ('DDIM', sd_samplers_timesteps_impl.ddim, ['ddim'], {}),
+    ('DDIM CFG++', sd_samplers_timesteps_impl.ddim_cfgpp, ['ddim_cfgpp'], {}),
     ('PLMS', sd_samplers_timesteps_impl.plms, ['plms'], {}),
     ('UniPC', sd_samplers_timesteps_impl.unipc, ['unipc'], {}),
 ]
@@ -36,7 +37,7 @@ def __init__(self, model, *args, **kwargs):
         self.inner_model = model
 
     def predict_eps_from_z_and_v(self, x_t, t, v):
-        return self.inner_model.sqrt_alphas_cumprod[t.to(torch.int), None, None, None] * v + self.inner_model.sqrt_one_minus_alphas_cumprod[t.to(torch.int), None, None, None] * x_t
+        return torch.sqrt(self.inner_model.alphas_cumprod)[t.to(torch.int), None, None, None] * v + torch.sqrt(1 - self.inner_model.alphas_cumprod)[t.to(torch.int), None, None, None] * x_t
 
     def forward(self, input, timesteps, **kwargs):
         model_output = self.inner_model.apply_model(input, timesteps, **kwargs)
@@ -80,6 +81,7 @@ def __init__(self, funcname, sd_model):
         self.eta_default = 0.0
 
         self.model_wrap_cfg = CFGDenoiserTimesteps(self)
+        self.model_wrap = self.model_wrap_cfg.inner_model
 
     def get_timesteps(self, p, steps):
         discard_next_to_last_sigma = self.config is not None and self.config.options.get('discard_next_to_last_sigma', False)
@@ -132,8 +134,7 @@ def sample_img2img(self, p, x, noise, conditioning, unconditional_conditioning,
 
         samples = self.launch_sampling(t_enc + 1, lambda: self.func(self.model_wrap_cfg, xi, extra_args=self.sampler_extra_args, disable=False, callback=self.callback_state, **extra_params_kwargs))
 
-        if self.model_wrap_cfg.padded_cond_uncond:
-            p.extra_generation_params["Pad conds"] = True
+        self.add_infotext(p)
 
         return samples
 
@@ -157,8 +158,7 @@ def sample(self, p, x, conditioning, unconditional_conditioning, steps=None, ima
         }
         samples = self.launch_sampling(steps, lambda: self.func(self.model_wrap_cfg, x, extra_args=self.sampler_extra_args, disable=False, callback=self.callback_state, **extra_params_kwargs))
 
-        if self.model_wrap_cfg.padded_cond_uncond:
-            p.extra_generation_params["Pad conds"] = True
+        self.add_infotext(p)
 
         return samples
 
diff --git a/modules/sd_samplers_timesteps_impl.py b/modules/sd_samplers_timesteps_impl.py
index a72daafd47d..180e4389988 100644
--- a/modules/sd_samplers_timesteps_impl.py
+++ b/modules/sd_samplers_timesteps_impl.py
@@ -5,13 +5,14 @@
 
 from modules import shared
 from modules.models.diffusion.uni_pc import uni_pc
+from modules.torch_utils import float64
 
 
 @torch.no_grad()
 def ddim(model, x, timesteps, extra_args=None, callback=None, disable=None, eta=0.0):
     alphas_cumprod = model.inner_model.inner_model.alphas_cumprod
     alphas = alphas_cumprod[timesteps]
-    alphas_prev = alphas_cumprod[torch.nn.functional.pad(timesteps[:-1], pad=(1, 0))].to(torch.float64 if x.device.type != 'mps' else torch.float32)
+    alphas_prev = alphas_cumprod[torch.nn.functional.pad(timesteps[:-1], pad=(1, 0))].to(float64(x))
     sqrt_one_minus_alphas = torch.sqrt(1 - alphas)
     sigmas = eta * np.sqrt((1 - alphas_prev.cpu().numpy()) / (1 - alphas.cpu()) * (1 - alphas.cpu() / alphas_prev.cpu().numpy()))
 
@@ -39,11 +40,51 @@ def ddim(model, x, timesteps, extra_args=None, callback=None, disable=None, eta=
     return x
 
 
+@torch.no_grad()
+def ddim_cfgpp(model, x, timesteps, extra_args=None, callback=None, disable=None, eta=0.0):
+    """ Implements CFG++: Manifold-constrained Classifier Free Guidance For Diffusion Models (2024).
+    Uses the unconditional noise prediction instead of the conditional noise to guide the denoising direction.
+    The CFG scale is divided by 12.5 to map CFG from [0.0, 12.5] to [0, 1.0].
+    """
+    alphas_cumprod = model.inner_model.inner_model.alphas_cumprod
+    alphas = alphas_cumprod[timesteps]
+    alphas_prev = alphas_cumprod[torch.nn.functional.pad(timesteps[:-1], pad=(1, 0))].to(float64(x))
+    sqrt_one_minus_alphas = torch.sqrt(1 - alphas)
+    sigmas = eta * np.sqrt((1 - alphas_prev.cpu().numpy()) / (1 - alphas.cpu()) * (1 - alphas.cpu() / alphas_prev.cpu().numpy()))
+
+    model.cond_scale_miltiplier = 1 / 12.5
+    model.need_last_noise_uncond = True
+
+    extra_args = {} if extra_args is None else extra_args
+    s_in = x.new_ones((x.shape[0]))
+    s_x = x.new_ones((x.shape[0], 1, 1, 1))
+    for i in tqdm.trange(len(timesteps) - 1, disable=disable):
+        index = len(timesteps) - 1 - i
+
+        e_t = model(x, timesteps[index].item() * s_in, **extra_args)
+        last_noise_uncond = model.last_noise_uncond
+
+        a_t = alphas[index].item() * s_x
+        a_prev = alphas_prev[index].item() * s_x
+        sigma_t = sigmas[index].item() * s_x
+        sqrt_one_minus_at = sqrt_one_minus_alphas[index].item() * s_x
+
+        pred_x0 = (x - sqrt_one_minus_at * e_t) / a_t.sqrt()
+        dir_xt = (1. - a_prev - sigma_t ** 2).sqrt() * last_noise_uncond
+        noise = sigma_t * k_diffusion.sampling.torch.randn_like(x)
+        x = a_prev.sqrt() * pred_x0 + dir_xt + noise
+
+        if callback is not None:
+            callback({'x': x, 'i': i, 'sigma': 0, 'sigma_hat': 0, 'denoised': pred_x0})
+
+    return x
+
+
 @torch.no_grad()
 def plms(model, x, timesteps, extra_args=None, callback=None, disable=None):
     alphas_cumprod = model.inner_model.inner_model.alphas_cumprod
     alphas = alphas_cumprod[timesteps]
-    alphas_prev = alphas_cumprod[torch.nn.functional.pad(timesteps[:-1], pad=(1, 0))].to(torch.float64 if x.device.type != 'mps' else torch.float32)
+    alphas_prev = alphas_cumprod[torch.nn.functional.pad(timesteps[:-1], pad=(1, 0))].to(float64(x))
     sqrt_one_minus_alphas = torch.sqrt(1 - alphas)
 
     extra_args = {} if extra_args is None else extra_args
diff --git a/modules/sd_schedulers.py b/modules/sd_schedulers.py
new file mode 100644
index 00000000000..f4d16e309ff
--- /dev/null
+++ b/modules/sd_schedulers.py
@@ -0,0 +1,145 @@
+import dataclasses
+import torch
+import k_diffusion
+import numpy as np
+from scipy import stats
+
+from modules import shared
+
+
+def to_d(x, sigma, denoised):
+    """Converts a denoiser output to a Karras ODE derivative."""
+    return (x - denoised) / sigma
+
+
+k_diffusion.sampling.to_d = to_d
+
+
+@dataclasses.dataclass
+class Scheduler:
+    name: str
+    label: str
+    function: any
+
+    default_rho: float = -1
+    need_inner_model: bool = False
+    aliases: list = None
+
+
+def uniform(n, sigma_min, sigma_max, inner_model, device):
+    return inner_model.get_sigmas(n).to(device)
+
+
+def sgm_uniform(n, sigma_min, sigma_max, inner_model, device):
+    start = inner_model.sigma_to_t(torch.tensor(sigma_max))
+    end = inner_model.sigma_to_t(torch.tensor(sigma_min))
+    sigs = [
+        inner_model.t_to_sigma(ts)
+        for ts in torch.linspace(start, end, n + 1)[:-1]
+    ]
+    sigs += [0.0]
+    return torch.FloatTensor(sigs).to(device)
+
+
+def get_align_your_steps_sigmas(n, sigma_min, sigma_max, device):
+    # https://research.nvidia.com/labs/toronto-ai/AlignYourSteps/howto.html
+    def loglinear_interp(t_steps, num_steps):
+        """
+        Performs log-linear interpolation of a given array of decreasing numbers.
+        """
+        xs = np.linspace(0, 1, len(t_steps))
+        ys = np.log(t_steps[::-1])
+
+        new_xs = np.linspace(0, 1, num_steps)
+        new_ys = np.interp(new_xs, xs, ys)
+
+        interped_ys = np.exp(new_ys)[::-1].copy()
+        return interped_ys
+
+    if shared.sd_model.is_sdxl:
+        sigmas = [14.615, 6.315, 3.771, 2.181, 1.342, 0.862, 0.555, 0.380, 0.234, 0.113, 0.029]
+    else:
+        # Default to SD 1.5 sigmas.
+        sigmas = [14.615, 6.475, 3.861, 2.697, 1.886, 1.396, 0.963, 0.652, 0.399, 0.152, 0.029]
+
+    if n != len(sigmas):
+        sigmas = np.append(loglinear_interp(sigmas, n), [0.0])
+    else:
+        sigmas.append(0.0)
+
+    return torch.FloatTensor(sigmas).to(device)
+
+
+def kl_optimal(n, sigma_min, sigma_max, device):
+    alpha_min = torch.arctan(torch.tensor(sigma_min, device=device))
+    alpha_max = torch.arctan(torch.tensor(sigma_max, device=device))
+    step_indices = torch.arange(n + 1, device=device)
+    sigmas = torch.tan(step_indices / n * alpha_min + (1.0 - step_indices / n) * alpha_max)
+    return sigmas
+
+
+def simple_scheduler(n, sigma_min, sigma_max, inner_model, device):
+    sigs = []
+    ss = len(inner_model.sigmas) / n
+    for x in range(n):
+        sigs += [float(inner_model.sigmas[-(1 + int(x * ss))])]
+    sigs += [0.0]
+    return torch.FloatTensor(sigs).to(device)
+
+
+def normal_scheduler(n, sigma_min, sigma_max, inner_model, device, sgm=False, floor=False):
+    start = inner_model.sigma_to_t(torch.tensor(sigma_max))
+    end = inner_model.sigma_to_t(torch.tensor(sigma_min))
+
+    if sgm:
+        timesteps = torch.linspace(start, end, n + 1)[:-1]
+    else:
+        timesteps = torch.linspace(start, end, n)
+
+    sigs = []
+    for x in range(len(timesteps)):
+        ts = timesteps[x]
+        sigs.append(inner_model.t_to_sigma(ts))
+    sigs += [0.0]
+    return torch.FloatTensor(sigs).to(device)
+
+
+def ddim_scheduler(n, sigma_min, sigma_max, inner_model, device):
+    sigs = []
+    ss = max(len(inner_model.sigmas) // n, 1)
+    x = 1
+    while x < len(inner_model.sigmas):
+        sigs += [float(inner_model.sigmas[x])]
+        x += ss
+    sigs = sigs[::-1]
+    sigs += [0.0]
+    return torch.FloatTensor(sigs).to(device)
+
+
+def beta_scheduler(n, sigma_min, sigma_max, inner_model, device):
+    # From "Beta Sampling is All You Need" [arXiv:2407.12173] (Lee et. al, 2024) """
+    alpha = shared.opts.beta_dist_alpha
+    beta = shared.opts.beta_dist_beta
+    timesteps = 1 - np.linspace(0, 1, n)
+    timesteps = [stats.beta.ppf(x, alpha, beta) for x in timesteps]
+    sigmas = [sigma_min + (x * (sigma_max-sigma_min)) for x in timesteps]
+    sigmas += [0.0]
+    return torch.FloatTensor(sigmas).to(device)
+
+
+schedulers = [
+    Scheduler('automatic', 'Automatic', None),
+    Scheduler('uniform', 'Uniform', uniform, need_inner_model=True),
+    Scheduler('karras', 'Karras', k_diffusion.sampling.get_sigmas_karras, default_rho=7.0),
+    Scheduler('exponential', 'Exponential', k_diffusion.sampling.get_sigmas_exponential),
+    Scheduler('polyexponential', 'Polyexponential', k_diffusion.sampling.get_sigmas_polyexponential, default_rho=1.0),
+    Scheduler('sgm_uniform', 'SGM Uniform', sgm_uniform, need_inner_model=True, aliases=["SGMUniform"]),
+    Scheduler('kl_optimal', 'KL Optimal', kl_optimal),
+    Scheduler('align_your_steps', 'Align Your Steps', get_align_your_steps_sigmas),
+    Scheduler('simple', 'Simple', simple_scheduler, need_inner_model=True),
+    Scheduler('normal', 'Normal', normal_scheduler, need_inner_model=True),
+    Scheduler('ddim', 'DDIM', ddim_scheduler, need_inner_model=True),
+    Scheduler('beta', 'Beta', beta_scheduler, need_inner_model=True),
+]
+
+schedulers_map = {**{x.name: x for x in schedulers}, **{x.label: x for x in schedulers}}
diff --git a/modules/sd_unet.py b/modules/sd_unet.py
index 5525cfbc3a0..a771849c8c2 100644
--- a/modules/sd_unet.py
+++ b/modules/sd_unet.py
@@ -1,12 +1,11 @@
 import torch.nn
-import ldm.modules.diffusionmodules.openaimodel
 
 from modules import script_callbacks, shared, devices
 
 unet_options = []
 current_unet_option = None
 current_unet = None
-
+original_forward = None  # not used, only left temporarily for compatibility
 
 def list_unets():
     new_unets = script_callbacks.list_unets_callback()
@@ -84,9 +83,12 @@ def deactivate(self):
         pass
 
 
-def UNetModel_forward(self, x, timesteps=None, context=None, *args, **kwargs):
-    if current_unet is not None:
-        return current_unet.forward(x, timesteps, context, *args, **kwargs)
+def create_unet_forward(original_forward):
+    def UNetModel_forward(self, x, timesteps=None, context=None, *args, **kwargs):
+        if current_unet is not None:
+            return current_unet.forward(x, timesteps, context, *args, **kwargs)
+
+        return original_forward(self, x, timesteps, context, *args, **kwargs)
 
-    return ldm.modules.diffusionmodules.openaimodel.copy_of_UNetModel_forward_for_webui(self, x, timesteps, context, *args, **kwargs)
+    return UNetModel_forward
 
diff --git a/modules/sd_vae.py b/modules/sd_vae.py
index 31306d8ba4b..43687e48dcf 100644
--- a/modules/sd_vae.py
+++ b/modules/sd_vae.py
@@ -273,10 +273,11 @@ def reload_vae_weights(sd_model=None, vae_file=unspecified):
     load_vae(sd_model, vae_file, vae_source)
 
     sd_hijack.model_hijack.hijack(sd_model)
-    script_callbacks.model_loaded_callback(sd_model)
 
     if not sd_model.lowvram:
         sd_model.to(devices.device)
 
+    script_callbacks.model_loaded_callback(sd_model)
+
     print("VAE weights loaded.")
     return sd_model
diff --git a/modules/sd_vae_approx.py b/modules/sd_vae_approx.py
index 3965e223e6f..c5dda7431f1 100644
--- a/modules/sd_vae_approx.py
+++ b/modules/sd_vae_approx.py
@@ -8,9 +8,9 @@
 
 
 class VAEApprox(nn.Module):
-    def __init__(self):
+    def __init__(self, latent_channels=4):
         super(VAEApprox, self).__init__()
-        self.conv1 = nn.Conv2d(4, 8, (7, 7))
+        self.conv1 = nn.Conv2d(latent_channels, 8, (7, 7))
         self.conv2 = nn.Conv2d(8, 16, (5, 5))
         self.conv3 = nn.Conv2d(16, 32, (3, 3))
         self.conv4 = nn.Conv2d(32, 64, (3, 3))
@@ -40,7 +40,13 @@ def download_model(model_path, model_url):
 
 
 def model():
-    model_name = "vaeapprox-sdxl.pt" if getattr(shared.sd_model, 'is_sdxl', False) else "model.pt"
+    if shared.sd_model.is_sd3:
+        model_name = "vaeapprox-sd3.pt"
+    elif shared.sd_model.is_sdxl:
+        model_name = "vaeapprox-sdxl.pt"
+    else:
+        model_name = "model.pt"
+
     loaded_model = sd_vae_approx_models.get(model_name)
 
     if loaded_model is None:
@@ -52,7 +58,7 @@ def model():
             model_path = os.path.join(paths.models_path, "VAE-approx", model_name)
             download_model(model_path, 'https://github.com/AUTOMATIC1111/stable-diffusion-webui/releases/download/v1.0.0-pre/' + model_name)
 
-        loaded_model = VAEApprox()
+        loaded_model = VAEApprox(latent_channels=shared.sd_model.latent_channels)
         loaded_model.load_state_dict(torch.load(model_path, map_location='cpu' if devices.device.type != 'cuda' else None))
         loaded_model.eval()
         loaded_model.to(devices.device, devices.dtype)
@@ -64,7 +70,18 @@ def model():
 def cheap_approximation(sample):
     # https://discuss.huggingface.co/t/decoding-latents-to-rgb-without-upscaling/23204/2
 
-    if shared.sd_model.is_sdxl:
+    if shared.sd_model.is_sd3:
+        coeffs = [
+            [-0.0645,  0.0177,  0.1052], [ 0.0028,  0.0312,  0.0650],
+            [ 0.1848,  0.0762,  0.0360], [ 0.0944,  0.0360,  0.0889],
+            [ 0.0897,  0.0506, -0.0364], [-0.0020,  0.1203,  0.0284],
+            [ 0.0855,  0.0118,  0.0283], [-0.0539,  0.0658,  0.1047],
+            [-0.0057,  0.0116,  0.0700], [-0.0412,  0.0281, -0.0039],
+            [ 0.1106,  0.1171,  0.1220], [-0.0248,  0.0682, -0.0481],
+            [ 0.0815,  0.0846,  0.1207], [-0.0120, -0.0055, -0.0867],
+            [-0.0749, -0.0634, -0.0456], [-0.1418, -0.1457, -0.1259],
+        ]
+    elif shared.sd_model.is_sdxl:
         coeffs = [
             [ 0.3448,  0.4168,  0.4395],
             [-0.1953, -0.0290,  0.0250],
diff --git a/modules/sd_vae_taesd.py b/modules/sd_vae_taesd.py
index 808eb3624fd..d06253d2a88 100644
--- a/modules/sd_vae_taesd.py
+++ b/modules/sd_vae_taesd.py
@@ -34,9 +34,9 @@ def forward(self, x):
         return self.fuse(self.conv(x) + self.skip(x))
 
 
-def decoder():
+def decoder(latent_channels=4):
     return nn.Sequential(
-        Clamp(), conv(4, 64), nn.ReLU(),
+        Clamp(), conv(latent_channels, 64), nn.ReLU(),
         Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False),
         Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False),
         Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False),
@@ -44,13 +44,13 @@ def decoder():
     )
 
 
-def encoder():
+def encoder(latent_channels=4):
     return nn.Sequential(
         conv(3, 64), Block(64, 64),
         conv(64, 64, stride=2, bias=False), Block(64, 64), Block(64, 64), Block(64, 64),
         conv(64, 64, stride=2, bias=False), Block(64, 64), Block(64, 64), Block(64, 64),
         conv(64, 64, stride=2, bias=False), Block(64, 64), Block(64, 64), Block(64, 64),
-        conv(64, 4),
+        conv(64, latent_channels),
     )
 
 
@@ -58,10 +58,14 @@ class TAESDDecoder(nn.Module):
     latent_magnitude = 3
     latent_shift = 0.5
 
-    def __init__(self, decoder_path="taesd_decoder.pth"):
+    def __init__(self, decoder_path="taesd_decoder.pth", latent_channels=None):
         """Initialize pretrained TAESD on the given device from the given checkpoints."""
         super().__init__()
-        self.decoder = decoder()
+
+        if latent_channels is None:
+            latent_channels = 16 if "taesd3" in str(decoder_path) else 4
+
+        self.decoder = decoder(latent_channels)
         self.decoder.load_state_dict(
             torch.load(decoder_path, map_location='cpu' if devices.device.type != 'cuda' else None))
 
@@ -70,10 +74,14 @@ class TAESDEncoder(nn.Module):
     latent_magnitude = 3
     latent_shift = 0.5
 
-    def __init__(self, encoder_path="taesd_encoder.pth"):
+    def __init__(self, encoder_path="taesd_encoder.pth", latent_channels=None):
         """Initialize pretrained TAESD on the given device from the given checkpoints."""
         super().__init__()
-        self.encoder = encoder()
+
+        if latent_channels is None:
+            latent_channels = 16 if "taesd3" in str(encoder_path) else 4
+
+        self.encoder = encoder(latent_channels)
         self.encoder.load_state_dict(
             torch.load(encoder_path, map_location='cpu' if devices.device.type != 'cuda' else None))
 
@@ -87,7 +95,13 @@ def download_model(model_path, model_url):
 
 
 def decoder_model():
-    model_name = "taesdxl_decoder.pth" if getattr(shared.sd_model, 'is_sdxl', False) else "taesd_decoder.pth"
+    if shared.sd_model.is_sd3:
+        model_name = "taesd3_decoder.pth"
+    elif shared.sd_model.is_sdxl:
+        model_name = "taesdxl_decoder.pth"
+    else:
+        model_name = "taesd_decoder.pth"
+
     loaded_model = sd_vae_taesd_models.get(model_name)
 
     if loaded_model is None:
@@ -106,7 +120,13 @@ def decoder_model():
 
 
 def encoder_model():
-    model_name = "taesdxl_encoder.pth" if getattr(shared.sd_model, 'is_sdxl', False) else "taesd_encoder.pth"
+    if shared.sd_model.is_sd3:
+        model_name = "taesd3_encoder.pth"
+    elif shared.sd_model.is_sdxl:
+        model_name = "taesdxl_encoder.pth"
+    else:
+        model_name = "taesd_encoder.pth"
+
     loaded_model = sd_vae_taesd_models.get(model_name)
 
     if loaded_model is None:
diff --git a/modules/shared.py b/modules/shared.py
index 636619391fc..2a3787f990d 100644
--- a/modules/shared.py
+++ b/modules/shared.py
@@ -1,3 +1,4 @@
+import os
 import sys
 
 import gradio as gr
@@ -5,21 +6,25 @@
 from modules import shared_cmd_options, shared_gradio_themes, options, shared_items, sd_models_types
 from modules.paths_internal import models_path, script_path, data_path, sd_configs_path, sd_default_config, sd_model_file, default_sd_model_file, extensions_dir, extensions_builtin_dir  # noqa: F401
 from modules import util
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from modules import shared_state, styles, interrogate, shared_total_tqdm, memmon
 
 cmd_opts = shared_cmd_options.cmd_opts
 parser = shared_cmd_options.parser
 
 batch_cond_uncond = True  # old field, unused now in favor of shared.opts.batch_cond_uncond
 parallel_processing_allowed = True
-styles_filename = cmd_opts.styles_file
+styles_filename = cmd_opts.styles_file = cmd_opts.styles_file if len(cmd_opts.styles_file) > 0 else [os.path.join(data_path, 'styles.csv')]
 config_filename = cmd_opts.ui_settings_file
 hide_dirs = {"visible": not cmd_opts.hide_ui_dir_config}
 
-demo = None
+demo: gr.Blocks = None
 
-device = None
+device: str = None
 
-weight_load_location = None
+weight_load_location: str = None
 
 xformers_available = False
 
@@ -27,22 +32,22 @@
 
 loaded_hypernetworks = []
 
-state = None
+state: 'shared_state.State' = None
 
-prompt_styles = None
+prompt_styles: 'styles.StyleDatabase' = None
 
-interrogator = None
+interrogator: 'interrogate.InterrogateModels' = None
 
 face_restorers = []
 
-options_templates = None
-opts = None
-restricted_opts = None
+options_templates: dict = None
+opts: options.Options = None
+restricted_opts: set[str] = None
 
 sd_model: sd_models_types.WebuiSdModel = None
 
-settings_components = None
-"""assinged from ui.py, a mapping on setting names to gradio components repsponsible for those settings"""
+settings_components: dict = None
+"""assigned from ui.py, a mapping on setting names to gradio components responsible for those settings"""
 
 tab_names = []
 
@@ -64,9 +69,9 @@
 
 gradio_theme = gr.themes.Base()
 
-total_tqdm = None
+total_tqdm: 'shared_total_tqdm.TotalTQDM' = None
 
-mem_mon = None
+mem_mon: 'memmon.MemUsageMonitor' = None
 
 options_section = options.options_section
 OptionInfo = options.OptionInfo
@@ -85,3 +90,5 @@
 refresh_checkpoints = shared_items.refresh_checkpoints
 list_samplers = shared_items.list_samplers
 reload_hypernetworks = shared_items.reload_hypernetworks
+
+hf_endpoint = os.getenv('HF_ENDPOINT', 'https://huggingface.co')
diff --git a/modules/shared_cmd_options.py b/modules/shared_cmd_options.py
index dd93f5206ce..c9626667fb8 100644
--- a/modules/shared_cmd_options.py
+++ b/modules/shared_cmd_options.py
@@ -14,5 +14,5 @@
 else:
     cmd_opts, _ = parser.parse_known_args()
 
-
-cmd_opts.disable_extension_access = any([cmd_opts.share, cmd_opts.listen, cmd_opts.ngrok, cmd_opts.server_name]) and not cmd_opts.enable_insecure_extension_access
+cmd_opts.webui_is_non_local = any([cmd_opts.share, cmd_opts.listen, cmd_opts.ngrok, cmd_opts.server_name])
+cmd_opts.disable_extension_access = cmd_opts.webui_is_non_local and not cmd_opts.enable_insecure_extension_access
diff --git a/modules/shared_gradio_themes.py b/modules/shared_gradio_themes.py
index 822db0a951d..b4e3f32bc9f 100644
--- a/modules/shared_gradio_themes.py
+++ b/modules/shared_gradio_themes.py
@@ -65,3 +65,48 @@ def reload_gradio_theme(theme_name=None):
         except Exception as e:
             errors.display(e, "changing gradio theme")
             shared.gradio_theme = gr.themes.Default(**default_theme_args)
+
+    # append additional values gradio_theme
+    shared.gradio_theme.sd_webui_modal_lightbox_toolbar_opacity = shared.opts.sd_webui_modal_lightbox_toolbar_opacity
+    shared.gradio_theme.sd_webui_modal_lightbox_icon_opacity = shared.opts.sd_webui_modal_lightbox_icon_opacity
+
+
+def resolve_var(name: str, gradio_theme=None, history=None):
+    """
+    Attempt to resolve a theme variable name to its value
+
+    Parameters:
+        name (str): The name of the theme variable
+            ie "background_fill_primary", "background_fill_primary_dark"
+            spaces and asterisk (*) prefix is removed from name before lookup
+        gradio_theme (gradio.themes.ThemeClass): The theme object to resolve the variable from
+            blank to use the webui default shared.gradio_theme
+        history (list): A list of previously resolved variables to prevent circular references
+            for regular use leave blank
+    Returns:
+        str: The resolved value
+
+    Error handling:
+        return either #000000 or #ffffff depending on initial name ending with "_dark"
+    """
+    try:
+        if history is None:
+            history = []
+        if gradio_theme is None:
+            gradio_theme = shared.gradio_theme
+
+        name = name.strip()
+        name = name[1:] if name.startswith("*") else name
+
+        if name in history:
+            raise ValueError(f'Circular references: name "{name}" in {history}')
+
+        if value := getattr(gradio_theme, name, None):
+            return resolve_var(value, gradio_theme, history + [name])
+        else:
+            return name
+
+    except Exception:
+        name = history[0] if history else name
+        errors.report(f'resolve_color({name})', exc_info=True)
+        return '#000000' if name.endswith("_dark") else '#ffffff'
diff --git a/modules/shared_init.py b/modules/shared_init.py
index d3fb687e0cd..a6ad0433d6f 100644
--- a/modules/shared_init.py
+++ b/modules/shared_init.py
@@ -18,8 +18,10 @@ def initialize():
     shared.options_templates = shared_options.options_templates
     shared.opts = options.Options(shared_options.options_templates, shared_options.restricted_opts)
     shared.restricted_opts = shared_options.restricted_opts
-    if os.path.exists(shared.config_filename):
+    try:
         shared.opts.load(shared.config_filename)
+    except FileNotFoundError:
+        pass
 
     from modules import devices
     devices.device, devices.device_interrogate, devices.device_gfpgan, devices.device_esrgan, devices.device_codeformer = \
@@ -27,6 +29,15 @@ def initialize():
 
     devices.dtype = torch.float32 if cmd_opts.no_half else torch.float16
     devices.dtype_vae = torch.float32 if cmd_opts.no_half or cmd_opts.no_half_vae else torch.float16
+    devices.dtype_inference = torch.float32 if cmd_opts.precision == 'full' else devices.dtype
+
+    if cmd_opts.precision == "half":
+        msg = "--no-half and --no-half-vae conflict with --precision half"
+        assert devices.dtype == torch.float16, msg
+        assert devices.dtype_vae == torch.float16, msg
+        assert devices.dtype_inference == torch.float16, msg
+        devices.force_fp16 = True
+        devices.force_model_fp16()
 
     shared.device = devices.device
     shared.weight_load_location = None if cmd_opts.lowram else "cpu"
diff --git a/modules/shared_items.py b/modules/shared_items.py
index 84d69c8df43..11f10b3f7b1 100644
--- a/modules/shared_items.py
+++ b/modules/shared_items.py
@@ -1,5 +1,8 @@
+import html
 import sys
 
+from modules import script_callbacks, scripts, ui_components
+from modules.options import OptionHTML, OptionInfo
 from modules.shared_cmd_options import cmd_opts
 
 
@@ -8,6 +11,11 @@ def realesrgan_models_names():
     return [x.name for x in modules.realesrgan_model.get_realesrgan_models(None)]
 
 
+def dat_models_names():
+    import modules.dat_model
+    return [x.name for x in modules.dat_model.get_dat_models(None)]
+
+
 def postprocessing_scripts():
     import modules.scripts
 
@@ -44,9 +52,9 @@ def refresh_unet_list():
     modules.sd_unet.list_unets()
 
 
-def list_checkpoint_tiles():
+def list_checkpoint_tiles(use_short=False):
     import modules.sd_models
-    return modules.sd_models.checkpoint_tiles()
+    return modules.sd_models.checkpoint_tiles(use_short)
 
 
 def refresh_checkpoints():
@@ -66,7 +74,25 @@ def reload_hypernetworks():
     shared.hypernetworks = hypernetwork.list_hypernetworks(cmd_opts.hypernetwork_dir)
 
 
+def get_infotext_names():
+    from modules import infotext_utils, shared
+    res = {}
+
+    for info in shared.opts.data_labels.values():
+        if info.infotext:
+            res[info.infotext] = 1
+
+    for tab_data in infotext_utils.paste_fields.values():
+        for _, name in tab_data.get("fields") or []:
+            if isinstance(name, str):
+                res[name] = 1
+
+    return list(res)
+
+
 ui_reorder_categories_builtin_items = [
+    "prompt",
+    "image",
     "inpaint",
     "sampler",
     "accordions",
@@ -95,6 +121,45 @@ def ui_reorder_categories():
     yield "scripts"
 
 
+def callbacks_order_settings():
+    options = {
+        "sd_vae_explanation": OptionHTML("""
+    For categories below, callbacks added to dropdowns happen before others, in order listed.
+    """),
+
+    }
+
+    callback_options = {}
+
+    for category, _ in script_callbacks.enumerate_callbacks():
+        callback_options[category] = script_callbacks.ordered_callbacks(category, enable_user_sort=False)
+
+    for method_name in scripts.scripts_txt2img.callback_names:
+        callback_options["script_" + method_name] = scripts.scripts_txt2img.create_ordered_callbacks_list(method_name, enable_user_sort=False)
+
+    for method_name in scripts.scripts_img2img.callback_names:
+        callbacks = callback_options.get("script_" + method_name, [])
+
+        for addition in scripts.scripts_img2img.create_ordered_callbacks_list(method_name, enable_user_sort=False):
+            if any(x.name == addition.name for x in callbacks):
+                continue
+
+            callbacks.append(addition)
+
+        callback_options["script_" + method_name] = callbacks
+
+    for category, callbacks in callback_options.items():
+        if not callbacks:
+            continue
+
+        option_info = OptionInfo([], f"{category} callback priority", ui_components.DropdownMulti, {"choices": [x.name for x in callbacks]})
+        option_info.needs_restart()
+        option_info.html("<div class='info'>Default order: <ol>" + "".join(f"<li>{html.escape(x.name)}</li>\n" for x in callbacks) + "</ol></div>")
+        options['prioritized_callbacks_' + category] = option_info
+
+    return options
+
+
 class Shared(sys.modules[__name__].__class__):
     """
     this class is here to provide sd_model field as a property, so that it can be created and loaded on demand rather than
diff --git a/modules/shared_options.py b/modules/shared_options.py
index 00b273faa54..9f4520274b1 100644
--- a/modules/shared_options.py
+++ b/modules/shared_options.py
@@ -1,9 +1,10 @@
+import os
 import gradio as gr
 
-from modules import localization, ui_components, shared_items, shared, interrogate, shared_gradio_themes
-from modules.paths_internal import models_path, script_path, data_path, sd_configs_path, sd_default_config, sd_model_file, default_sd_model_file, extensions_dir, extensions_builtin_dir  # noqa: F401
+from modules import localization, ui_components, shared_items, shared, interrogate, shared_gradio_themes, util, sd_emphasis
+from modules.paths_internal import models_path, script_path, data_path, sd_configs_path, sd_default_config, sd_model_file, default_sd_model_file, extensions_dir, extensions_builtin_dir, default_output_dir  # noqa: F401
 from modules.shared_cmd_options import cmd_opts
-from modules.options import options_section, OptionInfo, OptionHTML
+from modules.options import options_section, OptionInfo, OptionHTML, categories
 
 options_templates = {}
 hide_dirs = shared.hide_dirs
@@ -18,15 +19,24 @@
     "outdir_grids",
     "outdir_txt2img_grids",
     "outdir_save",
-    "outdir_init_images"
+    "outdir_init_images",
+    "temp_dir",
+    "clean_temp_dir_at_start",
 }
 
-options_templates.update(options_section(('saving-images', "Saving images/grids"), {
+categories.register_category("saving", "Saving images")
+categories.register_category("sd", "Stable Diffusion")
+categories.register_category("ui", "User Interface")
+categories.register_category("system", "System")
+categories.register_category("postprocessing", "Postprocessing")
+categories.register_category("training", "Training")
+
+options_templates.update(options_section(('saving-images', "Saving images/grids", "saving"), {
     "samples_save": OptionInfo(True, "Always save all generated images"),
     "samples_format": OptionInfo('png', 'File format for images'),
     "samples_filename_pattern": OptionInfo("", "Images filename pattern", component_args=hide_dirs).link("wiki", "https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Custom-Images-Filename-Name-and-Subdirectory"),
     "save_images_add_number": OptionInfo(True, "Add number to filename when saving", component_args=hide_dirs),
-
+    "save_images_replace_action": OptionInfo("Replace", "Saving the image to an existing file", gr.Radio, {"choices": ["Replace", "Add number suffix"], **hide_dirs}),
     "grid_save": OptionInfo(True, "Always save all generated image grids"),
     "grid_format": OptionInfo('png', 'File format for grids'),
     "grid_extended_filename": OptionInfo(False, "Add extended info (seed, prompt) to filename when saving grid"),
@@ -39,14 +49,12 @@
     "grid_text_inactive_color": OptionInfo("#999999", "Inactive text color for image grids", ui_components.FormColorPicker, {}),
     "grid_background_color": OptionInfo("#ffffff", "Background color for image grids", ui_components.FormColorPicker, {}),
 
-    "enable_pnginfo": OptionInfo(True, "Save text information about generation parameters as chunks to png files"),
-    "save_txt": OptionInfo(False, "Create a text file next to every image with generation parameters."),
     "save_images_before_face_restoration": OptionInfo(False, "Save a copy of image before doing face restoration."),
     "save_images_before_highres_fix": OptionInfo(False, "Save a copy of image before applying highres fix."),
     "save_images_before_color_correction": OptionInfo(False, "Save a copy of image before applying color correction to img2img results"),
     "save_mask": OptionInfo(False, "For inpainting, save a copy of the greyscale mask"),
     "save_mask_composite": OptionInfo(False, "For inpainting, save a masked composite"),
-    "jpeg_quality": OptionInfo(80, "Quality for saved jpeg images", gr.Slider, {"minimum": 1, "maximum": 100, "step": 1}),
+    "jpeg_quality": OptionInfo(80, "Quality for saved jpeg and avif images", gr.Slider, {"minimum": 1, "maximum": 100, "step": 1}),
     "webp_lossless": OptionInfo(False, "Use lossless compression for webp images"),
     "export_for_4chan": OptionInfo(True, "Save copy of large images as JPG").info("if the file size is above the limit, or either width or height are above the limit"),
     "img_downscale_threshold": OptionInfo(4.0, "File size limit for the above option, MB", gr.Number),
@@ -56,27 +64,31 @@
     "use_original_name_batch": OptionInfo(True, "Use original name for output filename during batch process in extras tab"),
     "use_upscaler_name_as_suffix": OptionInfo(False, "Use upscaler name as filename suffix in the extras tab"),
     "save_selected_only": OptionInfo(True, "When using 'Save' button, only save a single selected image"),
+    "save_write_log_csv": OptionInfo(True, "Write log.csv when saving images using 'Save' button"),
     "save_init_img": OptionInfo(False, "Save init images when using img2img"),
 
     "temp_dir":  OptionInfo("", "Directory for temporary images; leave empty for default"),
     "clean_temp_dir_at_start": OptionInfo(False, "Cleanup non-default temporary directory when starting webui"),
 
     "save_incomplete_images": OptionInfo(False, "Save incomplete images").info("save images that has been interrupted in mid-generation; even if not saved, they will still show up in webui output."),
+
+    "notification_audio": OptionInfo(True, "Play notification sound after image generation").info("notification.mp3 should be present in the root directory").needs_reload_ui(),
+    "notification_volume": OptionInfo(100, "Notification sound volume", gr.Slider, {"minimum": 0, "maximum": 100, "step": 1}).info("in %"),
 }))
 
-options_templates.update(options_section(('saving-paths', "Paths for saving"), {
+options_templates.update(options_section(('saving-paths', "Paths for saving", "saving"), {
     "outdir_samples": OptionInfo("", "Output directory for images; if empty, defaults to three directories below", component_args=hide_dirs),
-    "outdir_txt2img_samples": OptionInfo("outputs/txt2img-images", 'Output directory for txt2img images', component_args=hide_dirs),
-    "outdir_img2img_samples": OptionInfo("outputs/img2img-images", 'Output directory for img2img images', component_args=hide_dirs),
-    "outdir_extras_samples": OptionInfo("outputs/extras-images", 'Output directory for images from extras tab', component_args=hide_dirs),
+    "outdir_txt2img_samples": OptionInfo(util.truncate_path(os.path.join(default_output_dir, 'txt2img-images')), 'Output directory for txt2img images', component_args=hide_dirs),
+    "outdir_img2img_samples": OptionInfo(util.truncate_path(os.path.join(default_output_dir, 'img2img-images')), 'Output directory for img2img images', component_args=hide_dirs),
+    "outdir_extras_samples": OptionInfo(util.truncate_path(os.path.join(default_output_dir, 'extras-images')), 'Output directory for images from extras tab', component_args=hide_dirs),
     "outdir_grids": OptionInfo("", "Output directory for grids; if empty, defaults to two directories below", component_args=hide_dirs),
-    "outdir_txt2img_grids": OptionInfo("outputs/txt2img-grids", 'Output directory for txt2img grids', component_args=hide_dirs),
-    "outdir_img2img_grids": OptionInfo("outputs/img2img-grids", 'Output directory for img2img grids', component_args=hide_dirs),
-    "outdir_save": OptionInfo("log/images", "Directory for saving images using the Save button", component_args=hide_dirs),
-    "outdir_init_images": OptionInfo("outputs/init-images", "Directory for saving init images when using img2img", component_args=hide_dirs),
+    "outdir_txt2img_grids": OptionInfo(util.truncate_path(os.path.join(default_output_dir, 'txt2img-grids')), 'Output directory for txt2img grids', component_args=hide_dirs),
+    "outdir_img2img_grids": OptionInfo(util.truncate_path(os.path.join(default_output_dir, 'img2img-grids')), 'Output directory for img2img grids', component_args=hide_dirs),
+    "outdir_save": OptionInfo(util.truncate_path(os.path.join(data_path, 'log', 'images')), "Directory for saving images using the Save button", component_args=hide_dirs),
+    "outdir_init_images": OptionInfo(util.truncate_path(os.path.join(default_output_dir, 'init-images')), "Directory for saving init images when using img2img", component_args=hide_dirs),
 }))
 
-options_templates.update(options_section(('saving-to-dirs', "Saving to a directory"), {
+options_templates.update(options_section(('saving-to-dirs', "Saving to a directory", "saving"), {
     "save_to_dirs": OptionInfo(True, "Save images to a subdirectory"),
     "grid_save_to_dirs": OptionInfo(True, "Save grids to a subdirectory"),
     "use_save_to_dirs_for_ui": OptionInfo(False, "When using \"Save\" button, save images to a subdirectory"),
@@ -84,40 +96,63 @@
     "directories_max_prompt_words": OptionInfo(8, "Max prompt words for [prompt_words] pattern", gr.Slider, {"minimum": 1, "maximum": 20, "step": 1, **hide_dirs}),
 }))
 
-options_templates.update(options_section(('upscaling', "Upscaling"), {
+options_templates.update(options_section(('upscaling', "Upscaling", "postprocessing"), {
     "ESRGAN_tile": OptionInfo(192, "Tile size for ESRGAN upscalers.", gr.Slider, {"minimum": 0, "maximum": 512, "step": 16}).info("0 = no tiling"),
     "ESRGAN_tile_overlap": OptionInfo(8, "Tile overlap for ESRGAN upscalers.", gr.Slider, {"minimum": 0, "maximum": 48, "step": 1}).info("Low values = visible seam"),
     "realesrgan_enabled_models": OptionInfo(["R-ESRGAN 4x+", "R-ESRGAN 4x+ Anime6B"], "Select which Real-ESRGAN models to show in the web UI.", gr.CheckboxGroup, lambda: {"choices": shared_items.realesrgan_models_names()}),
+    "dat_enabled_models": OptionInfo(["DAT x2", "DAT x3", "DAT x4"], "Select which DAT models to show in the web UI.", gr.CheckboxGroup, lambda: {"choices": shared_items.dat_models_names()}),
+    "DAT_tile": OptionInfo(192, "Tile size for DAT upscalers.", gr.Slider, {"minimum": 0, "maximum": 512, "step": 16}).info("0 = no tiling"),
+    "DAT_tile_overlap": OptionInfo(8, "Tile overlap for DAT upscalers.", gr.Slider, {"minimum": 0, "maximum": 48, "step": 1}).info("Low values = visible seam"),
     "upscaler_for_img2img": OptionInfo(None, "Upscaler for img2img", gr.Dropdown, lambda: {"choices": [x.name for x in shared.sd_upscalers]}),
+    "set_scale_by_when_changing_upscaler": OptionInfo(False, "Automatically set the Scale by factor based on the name of the selected Upscaler."),
 }))
 
-options_templates.update(options_section(('face-restoration', "Face restoration"), {
+options_templates.update(options_section(('face-restoration', "Face restoration", "postprocessing"), {
     "face_restoration": OptionInfo(False, "Restore faces", infotext='Face restoration').info("will use a third-party model on generation result to reconstruct faces"),
     "face_restoration_model": OptionInfo("CodeFormer", "Face restoration model", gr.Radio, lambda: {"choices": [x.name() for x in shared.face_restorers]}),
     "code_former_weight": OptionInfo(0.5, "CodeFormer weight", gr.Slider, {"minimum": 0, "maximum": 1, "step": 0.01}).info("0 = maximum effect; 1 = minimum effect"),
     "face_restoration_unload": OptionInfo(False, "Move face restoration model from VRAM into RAM after processing"),
 }))
 
-options_templates.update(options_section(('system', "System"), {
+options_templates.update(options_section(('system', "System", "system"), {
     "auto_launch_browser": OptionInfo("Local", "Automatically open webui in browser on startup", gr.Radio, lambda: {"choices": ["Disable", "Local", "Remote"]}),
+    "enable_console_prompts": OptionInfo(shared.cmd_opts.enable_console_prompts, "Print prompts to console when generating with txt2img and img2img."),
     "show_warnings": OptionInfo(False, "Show warnings in console.").needs_reload_ui(),
     "show_gradio_deprecation_warnings": OptionInfo(True, "Show gradio deprecation warnings in console.").needs_reload_ui(),
     "memmon_poll_rate": OptionInfo(8, "VRAM usage polls per second during generation.", gr.Slider, {"minimum": 0, "maximum": 40, "step": 1}).info("0 = disable"),
     "samples_log_stdout": OptionInfo(False, "Always print all generation info to standard output"),
     "multiple_tqdm": OptionInfo(True, "Add a second progress bar to the console that shows progress for an entire job."),
+    "enable_upscale_progressbar": OptionInfo(True, "Show a progress bar in the console for tiled upscaling."),
     "print_hypernet_extra": OptionInfo(False, "Print extra hypernetwork information to console."),
     "list_hidden_files": OptionInfo(True, "Load models/files in hidden directories").info("directory is hidden if its name starts with \".\""),
     "disable_mmap_load_safetensors": OptionInfo(False, "Disable memmapping for loading .safetensors files.").info("fixes very slow loading speed in some cases"),
     "hide_ldm_prints": OptionInfo(True, "Prevent Stability-AI's ldm/sgm modules from printing noise to console."),
+    "dump_stacks_on_signal": OptionInfo(False, "Print stack traces before exiting the program with ctrl+c."),
 }))
 
-options_templates.update(options_section(('API', "API"), {
+options_templates.update(options_section(('profiler', "Profiler", "system"), {
+    "profiling_explanation": OptionHTML("""
+Those settings allow you to enable torch profiler when generating pictures.
+Profiling allows you to see which code uses how much of computer's resources during generation.
+Each generation writes its own profile to one file, overwriting previous.
+The file can be viewed in <a href="chrome:tracing">Chrome</a>, or on a <a href="https://ui.perfetto.dev/">Perfetto</a> web site.
+Warning: writing profile can take a lot of time, up to 30 seconds, and the file itelf can be around 500MB in size.
+"""),
+    "profiling_enable": OptionInfo(False, "Enable profiling"),
+    "profiling_activities": OptionInfo(["CPU"], "Activities", gr.CheckboxGroup, {"choices": ["CPU", "CUDA"]}),
+    "profiling_record_shapes": OptionInfo(True, "Record shapes"),
+    "profiling_profile_memory": OptionInfo(True, "Profile memory"),
+    "profiling_with_stack": OptionInfo(True, "Include python stack"),
+    "profiling_filename": OptionInfo("trace.json", "Profile filename"),
+}))
+
+options_templates.update(options_section(('API', "API", "system"), {
     "api_enable_requests": OptionInfo(True, "Allow http:// and https:// URLs for input images in API", restrict_api=True),
     "api_forbid_local_requests": OptionInfo(True, "Forbid URLs to local resources", restrict_api=True),
     "api_useragent": OptionInfo("", "User agent for requests", restrict_api=True),
 }))
 
-options_templates.update(options_section(('training', "Training"), {
+options_templates.update(options_section(('training', "Training", "training"), {
     "unload_models_when_training": OptionInfo(False, "Move VAE and CLIP to RAM when training if possible. Saves VRAM."),
     "pin_memory": OptionInfo(False, "Turn on pin_memory for DataLoader. Makes training slightly faster but can increase memory usage."),
     "save_optimizer_state": OptionInfo(False, "Saves Optimizer state as separate *.optim file. Training of embedding or HN can be resumed with the matching optim file."),
@@ -132,16 +167,17 @@
     "training_tensorboard_flush_every": OptionInfo(120, "How often, in seconds, to flush the pending tensorboard events and summaries to disk."),
 }))
 
-options_templates.update(options_section(('sd', "Stable Diffusion"), {
-    "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": shared_items.list_checkpoint_tiles()}, refresh=shared_items.refresh_checkpoints, infotext='Model hash'),
+options_templates.update(options_section(('sd', "Stable Diffusion", "sd"), {
+    "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": shared_items.list_checkpoint_tiles(shared.opts.sd_checkpoint_dropdown_use_short)}, refresh=shared_items.refresh_checkpoints, infotext='Model hash'),
     "sd_checkpoints_limit": OptionInfo(1, "Maximum number of checkpoints loaded at the same time", gr.Slider, {"minimum": 1, "maximum": 10, "step": 1}),
     "sd_checkpoints_keep_in_cpu": OptionInfo(True, "Only keep one model on device").info("will keep models other than the currently used one in RAM rather than VRAM"),
     "sd_checkpoint_cache": OptionInfo(0, "Checkpoints to cache in RAM", gr.Slider, {"minimum": 0, "maximum": 10, "step": 1}).info("obsolete; set to 0 and use the two settings above instead"),
     "sd_unet": OptionInfo("Automatic", "SD Unet", gr.Dropdown, lambda: {"choices": shared_items.sd_unet_items()}, refresh=shared_items.refresh_unet_list).info("choose Unet model: Automatic = use one with same filename as checkpoint; None = use Unet from checkpoint"),
     "enable_quantization": OptionInfo(False, "Enable quantization in K samplers for sharper and cleaner results. This may change existing seeds").needs_reload_ui(),
-    "enable_emphasis": OptionInfo(True, "Enable emphasis").info("use (text) to make model pay more attention to text and [text] to make it pay less attention"),
+    "emphasis": OptionInfo("Original", "Emphasis mode", gr.Radio, lambda: {"choices": [x.name for x in sd_emphasis.options]}, infotext="Emphasis").info("makes it possible to make model to pay (more:1.1) or (less:0.9) attention to text when you use the syntax in prompt; " + sd_emphasis.get_options_descriptions()),
     "enable_batch_seeds": OptionInfo(True, "Make K-diffusion samplers produce same images in a batch as when making a single image"),
     "comma_padding_backtrack": OptionInfo(20, "Prompt word wrap length limit", gr.Slider, {"minimum": 0, "maximum": 74, "step": 1}).info("in tokens - for texts shorter than specified, if they don't fit into 75 token limit, move them to the next 75 token chunk"),
+    "sdxl_clip_l_skip": OptionInfo(False, "Clip skip SDXL", gr.Checkbox).info("Enable Clip skip for the secondary clip model in sdxl. Has no effect on SD 1.5 or SD 2.0/2.1."),
     "CLIP_stop_at_last_layers": OptionInfo(1, "Clip skip", gr.Slider, {"minimum": 1, "maximum": 12, "step": 1}, infotext="Clip skip").link("wiki", "https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Features#clip-skip").info("ignore last layers of CLIP network; 1 ignores none, 2 ignores one layer"),
     "upcast_attn": OptionInfo(False, "Upcast cross attention layer to float32"),
     "randn_source": OptionInfo("GPU", "Random number generator source.", gr.Radio, {"choices": ["GPU", "CPU", "NV"]}, infotext="RNG").info("changes seeds drastically; use CPU to produce the same picture across different videocard vendors; use NV to produce same picture as on NVidia videocards"),
@@ -149,14 +185,18 @@
     "hires_fix_refiner_pass": OptionInfo("second pass", "Hires fix: which pass to enable refiner for", gr.Radio, {"choices": ["first pass", "second pass", "both passes"]}, infotext="Hires refiner"),
 }))
 
-options_templates.update(options_section(('sdxl', "Stable Diffusion XL"), {
+options_templates.update(options_section(('sdxl', "Stable Diffusion XL", "sd"), {
     "sdxl_crop_top": OptionInfo(0, "crop top coordinate"),
     "sdxl_crop_left": OptionInfo(0, "crop left coordinate"),
     "sdxl_refiner_low_aesthetic_score": OptionInfo(2.5, "SDXL low aesthetic score", gr.Number).info("used for refiner model negative prompt"),
     "sdxl_refiner_high_aesthetic_score": OptionInfo(6.0, "SDXL high aesthetic score", gr.Number).info("used for refiner model prompt"),
 }))
 
-options_templates.update(options_section(('vae', "VAE"), {
+options_templates.update(options_section(('sd3', "Stable Diffusion 3", "sd"), {
+    "sd3_enable_t5": OptionInfo(False, "Enable T5").info("load T5 text encoder; increases VRAM use by a lot, potentially improving quality of generation; requires model reload to apply"),
+}))
+
+options_templates.update(options_section(('vae', "VAE", "sd"), {
     "sd_vae_explanation": OptionHTML("""
 <abbr title='Variational autoencoder'>VAE</abbr> is a neural network that transforms a standard <abbr title='red/green/blue'>RGB</abbr>
 image into latent space representation and back. Latent space representation is what stable diffusion is working on during sampling
@@ -166,12 +206,13 @@
     "sd_vae_checkpoint_cache": OptionInfo(0, "VAE Checkpoints to cache in RAM", gr.Slider, {"minimum": 0, "maximum": 10, "step": 1}),
     "sd_vae": OptionInfo("Automatic", "SD VAE", gr.Dropdown, lambda: {"choices": shared_items.sd_vae_items()}, refresh=shared_items.refresh_vae_list, infotext='VAE').info("choose VAE model: Automatic = use one with same filename as checkpoint; None = use VAE from checkpoint"),
     "sd_vae_overrides_per_model_preferences": OptionInfo(True, "Selected VAE overrides per-model preferences").info("you can set per-model VAE either by editing user metadata for checkpoints, or by making the VAE have same name as checkpoint"),
+    "auto_vae_precision_bfloat16": OptionInfo(False, "Automatically convert VAE to bfloat16").info("triggers when a tensor with NaNs is produced in VAE; disabling the option in this case will result in a black square image; if enabled, overrides the option below"),
     "auto_vae_precision": OptionInfo(True, "Automatically revert VAE to 32-bit floats").info("triggers when a tensor with NaNs is produced in VAE; disabling the option in this case will result in a black square image"),
     "sd_vae_encode_method": OptionInfo("Full", "VAE type for encode", gr.Radio, {"choices": ["Full", "TAESD"]}, infotext='VAE Encoder').info("method to encode image to latent (use in img2img, hires-fix or inpaint mask)"),
     "sd_vae_decode_method": OptionInfo("Full", "VAE type for decode", gr.Radio, {"choices": ["Full", "TAESD"]}, infotext='VAE Decoder').info("method to decode latent to image"),
 }))
 
-options_templates.update(options_section(('img2img', "img2img"), {
+options_templates.update(options_section(('img2img', "img2img", "sd"), {
     "inpainting_mask_weight": OptionInfo(1.0, "Inpainting conditioning mask strength", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}, infotext='Conditional mask weight'),
     "initial_noise_multiplier": OptionInfo(1.0, "Noise multiplier for img2img", gr.Slider, {"minimum": 0.0, "maximum": 1.5, "step": 0.001}, infotext='Noise multiplier'),
     "img2img_extra_noise": OptionInfo(0.0, "Extra noise multiplier for img2img and hires fix", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}, infotext='Extra noise').info("0 = disabled (default); should be lower than denoising strength"),
@@ -184,27 +225,35 @@
     "img2img_inpaint_sketch_default_brush_color": OptionInfo("#ffffff", "Inpaint sketch initial brush color", ui_components.FormColorPicker, {}).info("default brush color of img2img inpaint sketch").needs_reload_ui(),
     "return_mask": OptionInfo(False, "For inpainting, include the greyscale mask in results for web"),
     "return_mask_composite": OptionInfo(False, "For inpainting, include masked composite in results for web"),
+    "img2img_batch_show_results_limit": OptionInfo(32, "Show the first N batch img2img results in UI", gr.Slider, {"minimum": -1, "maximum": 1000, "step": 1}).info('0: disable, -1: show all images. Too many images can cause lag'),
+    "overlay_inpaint": OptionInfo(True, "Overlay original for inpaint").info("when inpainting, overlay the original image over the areas that weren't inpainted."),
 }))
 
-options_templates.update(options_section(('optimizations', "Optimizations"), {
+options_templates.update(options_section(('optimizations', "Optimizations", "sd"), {
     "cross_attention_optimization": OptionInfo("Automatic", "Cross attention optimization", gr.Dropdown, lambda: {"choices": shared_items.cross_attention_optimizations()}),
-    "s_min_uncond": OptionInfo(0.0, "Negative Guidance minimum sigma", gr.Slider, {"minimum": 0.0, "maximum": 15.0, "step": 0.01}).link("PR", "https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/9177").info("skip negative prompt for some steps when the image is almost ready; 0=disable, higher=faster"),
+    "s_min_uncond": OptionInfo(0.0, "Negative Guidance minimum sigma", gr.Slider, {"minimum": 0.0, "maximum": 15.0, "step": 0.01}, infotext='NGMS').link("PR", "https://github.com/AUTOMATIC1111/stablediffusion-webui/pull/9177").info("skip negative prompt for some steps when the image is almost ready; 0=disable, higher=faster"),
+    "s_min_uncond_all": OptionInfo(False, "Negative Guidance minimum sigma all steps", infotext='NGMS all steps').info("By default, NGMS above skips every other step; this makes it skip all steps"),
     "token_merging_ratio": OptionInfo(0.0, "Token merging ratio", gr.Slider, {"minimum": 0.0, "maximum": 0.9, "step": 0.1}, infotext='Token merging ratio').link("PR", "https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/9256").info("0=disable, higher=faster"),
     "token_merging_ratio_img2img": OptionInfo(0.0, "Token merging ratio for img2img", gr.Slider, {"minimum": 0.0, "maximum": 0.9, "step": 0.1}).info("only applies if non-zero and overrides above"),
     "token_merging_ratio_hr": OptionInfo(0.0, "Token merging ratio for high-res pass", gr.Slider, {"minimum": 0.0, "maximum": 0.9, "step": 0.1}, infotext='Token merging ratio hr').info("only applies if non-zero and overrides above"),
-    "pad_cond_uncond": OptionInfo(False, "Pad prompt/negative prompt to be same length", infotext='Pad conds').info("improves performance when prompt and negative prompt have different lengths; changes seeds"),
+    "pad_cond_uncond": OptionInfo(False, "Pad prompt/negative prompt", infotext='Pad conds').info("improves performance when prompt and negative prompt have different lengths; changes seeds"),
+    "pad_cond_uncond_v0": OptionInfo(False, "Pad prompt/negative prompt (v0)", infotext='Pad conds v0').info("alternative implementation for the above; used prior to 1.6.0 for DDIM sampler; overrides the above if set; WARNING: truncates negative prompt if it's too long; changes seeds"),
     "persistent_cond_cache": OptionInfo(True, "Persistent cond cache").info("do not recalculate conds from prompts if prompts have not changed since previous calculation"),
-    "batch_cond_uncond": OptionInfo(True, "Batch cond/uncond").info("do both conditional and unconditional denoising in one batch; uses a bit more VRAM during sampling, but improves speed; previously this was controlled by --always-batch-cond-uncond comandline argument"),
+    "batch_cond_uncond": OptionInfo(True, "Batch cond/uncond").info("do both conditional and unconditional denoising in one batch; uses a bit more VRAM during sampling, but improves speed; previously this was controlled by --always-batch-cond-uncond commandline argument"),
+    "fp8_storage": OptionInfo("Disable", "FP8 weight", gr.Radio, {"choices": ["Disable", "Enable for SDXL", "Enable"]}).info("Use FP8 to store Linear/Conv layers' weight. Require pytorch>=2.1.0."),
+    "cache_fp16_weight": OptionInfo(False, "Cache FP16 weight for LoRA").info("Cache fp16 weight when enabling FP8, will increase the quality of LoRA. Use more system ram."),
 }))
 
-options_templates.update(options_section(('compatibility', "Compatibility"), {
+options_templates.update(options_section(('compatibility', "Compatibility", "sd"), {
+    "auto_backcompat": OptionInfo(True, "Automatic backward compatibility").info("automatically enable options for backwards compatibility when importing generation parameters from infotext that has program version."),
     "use_old_emphasis_implementation": OptionInfo(False, "Use old emphasis implementation. Can be useful to reproduce old seeds."),
     "use_old_karras_scheduler_sigmas": OptionInfo(False, "Use old karras scheduler sigmas (0.1 to 10)."),
     "no_dpmpp_sde_batch_determinism": OptionInfo(False, "Do not make DPM++ SDE deterministic across different batch sizes."),
     "use_old_hires_fix_width_height": OptionInfo(False, "For hires fix, use width/height sliders to set final resolution rather than first pass (disables Upscale by, Resize width/height to)."),
-    "dont_fix_second_order_samplers_schedule": OptionInfo(False, "Do not fix prompt schedule for second order samplers."),
     "hires_fix_use_firstpass_conds": OptionInfo(False, "For hires fix, calculate conds of second pass using extra networks of first pass."),
     "use_old_scheduling": OptionInfo(False, "Use old prompt editing timelines.", infotext="Old prompt editing timelines").info("For [red:green:N]; old: If N < 1, it's a fraction of steps (and hires fix uses range from 0 to 1), if N >= 1, it's an absolute number of steps; new: If N has a decimal point in it, it's a fraction of steps (and hires fix uses range from 1 to 2), othewrwise it's an absolute number of steps"),
+    "use_downcasted_alpha_bar": OptionInfo(False, "Downcast model alphas_cumprod to fp16 before sampling. For reproducing old seeds.", infotext="Downcast alphas_cumprod"),
+    "refiner_switch_by_sample_steps": OptionInfo(False, "Switch to refiner by sampling steps instead of model timesteps. Old behavior for refiner.", infotext="Refiner switch by sampling steps")
 }))
 
 options_templates.update(options_section(('interrogate', "Interrogate"), {
@@ -222,14 +271,21 @@
     "deepbooru_filter_tags": OptionInfo("", "deepbooru: filter out those tags").info("separate by comma"),
 }))
 
-options_templates.update(options_section(('extra_networks', "Extra Networks"), {
+options_templates.update(options_section(('extra_networks', "Extra Networks", "sd"), {
     "extra_networks_show_hidden_directories": OptionInfo(True, "Show hidden directories").info("directory is hidden if its name starts with \".\"."),
+    "extra_networks_dir_button_function": OptionInfo(False, "Add a '/' to the beginning of directory buttons").info("Buttons will display the contents of the selected directory without acting as a search filter."),
     "extra_networks_hidden_models": OptionInfo("When searched", "Show cards for models in hidden directories", gr.Radio, {"choices": ["Always", "When searched", "Never"]}).info('"When searched" option will only show the item when the search string has 4 characters or more'),
     "extra_networks_default_multiplier": OptionInfo(1.0, "Default multiplier for extra networks", gr.Slider, {"minimum": 0.0, "maximum": 2.0, "step": 0.01}),
     "extra_networks_card_width": OptionInfo(0, "Card width for Extra Networks").info("in pixels"),
     "extra_networks_card_height": OptionInfo(0, "Card height for Extra Networks").info("in pixels"),
     "extra_networks_card_text_scale": OptionInfo(1.0, "Card text scale", gr.Slider, {"minimum": 0.0, "maximum": 2.0, "step": 0.01}).info("1 = original size"),
     "extra_networks_card_show_desc": OptionInfo(True, "Show description on card"),
+    "extra_networks_card_description_is_html": OptionInfo(False, "Treat card description as HTML"),
+    "extra_networks_card_order_field": OptionInfo("Path", "Default order field for Extra Networks cards", gr.Dropdown, {"choices": ['Path', 'Name', 'Date Created', 'Date Modified']}).needs_reload_ui(),
+    "extra_networks_card_order": OptionInfo("Ascending", "Default order for Extra Networks cards", gr.Dropdown, {"choices": ['Ascending', 'Descending']}).needs_reload_ui(),
+    "extra_networks_tree_view_style": OptionInfo("Dirs", "Extra Networks directory view style", gr.Radio, {"choices": ["Tree", "Dirs"]}).needs_reload_ui(),
+    "extra_networks_tree_view_default_enabled": OptionInfo(True, "Show the Extra Networks directory view by default").needs_reload_ui(),
+    "extra_networks_tree_view_default_width": OptionInfo(180, "Default width for the Extra Networks directory tree view", gr.Number).needs_reload_ui(),
     "extra_networks_add_text_separator": OptionInfo(" ", "Extra networks separator").info("extra text to add before <...> when adding extra network to prompt"),
     "ui_extra_networks_tab_reorder": OptionInfo("", "Extra networks tab order").needs_reload_ui(),
     "textual_inversion_print_at_load": OptionInfo(False, "Print a list of Textual Inversion embeddings when loading model"),
@@ -237,42 +293,73 @@
     "sd_hypernetwork": OptionInfo("None", "Add hypernetwork to prompt", gr.Dropdown, lambda: {"choices": ["None", *shared.hypernetworks]}, refresh=shared_items.reload_hypernetworks),
 }))
 
-options_templates.update(options_section(('ui', "User interface"), {
-    "localization": OptionInfo("None", "Localization", gr.Dropdown, lambda: {"choices": ["None"] + list(localization.localizations.keys())}, refresh=lambda: localization.list_localizations(cmd_opts.localizations_dir)).needs_reload_ui(),
-    "gradio_theme": OptionInfo("Default", "Gradio theme", ui_components.DropdownEditable, lambda: {"choices": ["Default"] + shared_gradio_themes.gradio_hf_hub_themes}).info("you can also manually enter any of themes from the <a href='https://huggingface.co/spaces/gradio/theme-gallery'>gallery</a>.").needs_reload_ui(),
-    "gradio_themes_cache": OptionInfo(True, "Cache gradio themes locally").info("disable to update the selected Gradio theme"),
-    "gallery_height": OptionInfo("", "Gallery height", gr.Textbox).info("an be any valid CSS value").needs_reload_ui(),
-    "return_grid": OptionInfo(True, "Show grid in results for web"),
-    "do_not_show_images": OptionInfo(False, "Do not show any images in results for web"),
-    "send_seed": OptionInfo(True, "Send seed when sending prompt or image to other interface"),
-    "send_size": OptionInfo(True, "Send size when sending prompt or image to another interface"),
-    "js_modal_lightbox": OptionInfo(True, "Enable full page image viewer"),
-    "js_modal_lightbox_initially_zoomed": OptionInfo(True, "Show images zoomed in by default in full page image viewer"),
-    "js_modal_lightbox_gamepad": OptionInfo(False, "Navigate image viewer with gamepad"),
-    "js_modal_lightbox_gamepad_repeat": OptionInfo(250, "Gamepad repeat period, in milliseconds"),
-    "show_progress_in_title": OptionInfo(True, "Show generation progress in window title."),
+options_templates.update(options_section(('ui_prompt_editing', "Prompt editing", "ui"), {
+    "keyedit_precision_attention": OptionInfo(0.1, "Precision for (attention:1.1) when editing the prompt with Ctrl+up/down", gr.Slider, {"minimum": 0.01, "maximum": 0.2, "step": 0.001}),
+    "keyedit_precision_extra": OptionInfo(0.05, "Precision for <extra networks:0.9> when editing the prompt with Ctrl+up/down", gr.Slider, {"minimum": 0.01, "maximum": 0.2, "step": 0.001}),
+    "keyedit_delimiters": OptionInfo(r".,\/!?%^*;:{}=`~() ", "Word delimiters when editing the prompt with Ctrl+up/down"),
+    "keyedit_delimiters_whitespace": OptionInfo(["Tab", "Carriage Return", "Line Feed"], "Ctrl+up/down whitespace delimiters", gr.CheckboxGroup, lambda: {"choices": ["Tab", "Carriage Return", "Line Feed"]}),
+    "keyedit_move": OptionInfo(True, "Alt+left/right moves prompt elements"),
+    "disable_token_counters": OptionInfo(False, "Disable prompt token counters"),
+    "include_styles_into_token_counters": OptionInfo(True, "Count tokens of enabled styles").info("When calculating how many tokens the prompt has, also consider tokens added by enabled styles."),
+}))
+
+options_templates.update(options_section(('ui_gallery', "Gallery", "ui"), {
+    "return_grid": OptionInfo(True, "Show grid in gallery"),
+    "do_not_show_images": OptionInfo(False, "Do not show any images in gallery"),
+    "js_modal_lightbox": OptionInfo(True, "Full page image viewer: enable"),
+    "js_modal_lightbox_initially_zoomed": OptionInfo(True, "Full page image viewer: show images zoomed in by default"),
+    "js_modal_lightbox_gamepad": OptionInfo(False, "Full page image viewer: navigate with gamepad"),
+    "js_modal_lightbox_gamepad_repeat": OptionInfo(250, "Full page image viewer: gamepad repeat period").info("in milliseconds"),
+    "sd_webui_modal_lightbox_icon_opacity": OptionInfo(1, "Full page image viewer: control icon unfocused opacity", gr.Slider, {"minimum": 0.0, "maximum": 1, "step": 0.01}, onchange=shared.reload_gradio_theme).info('for mouse only').needs_reload_ui(),
+    "sd_webui_modal_lightbox_toolbar_opacity": OptionInfo(0.9, "Full page image viewer: tool bar opacity", gr.Slider, {"minimum": 0.0, "maximum": 1, "step": 0.01}, onchange=shared.reload_gradio_theme).info('for mouse only').needs_reload_ui(),
+    "gallery_height": OptionInfo("", "Gallery height", gr.Textbox).info("can be any valid CSS value, for example 768px or 20em").needs_reload_ui(),
+    "open_dir_button_choice": OptionInfo("Subdirectory", "What directory the [📂] button opens", gr.Radio, {"choices": ["Output Root", "Subdirectory", "Subdirectory (even temp dir)"]}),
+}))
+
+options_templates.update(options_section(('ui_alternatives', "UI alternatives", "ui"), {
+    "compact_prompt_box": OptionInfo(False, "Compact prompt layout").info("puts prompt and negative prompt inside the Generate tab, leaving more vertical space for the image on the right").needs_reload_ui(),
     "samplers_in_dropdown": OptionInfo(True, "Use dropdown for sampler selection instead of radio group").needs_reload_ui(),
     "dimensions_and_batch_together": OptionInfo(True, "Show Width/Height and Batch sliders in same row").needs_reload_ui(),
-    "keyedit_precision_attention": OptionInfo(0.1, "Ctrl+up/down precision when editing (attention:1.1)", gr.Slider, {"minimum": 0.01, "maximum": 0.2, "step": 0.001}),
-    "keyedit_precision_extra": OptionInfo(0.05, "Ctrl+up/down precision when editing <extra networks:0.9>", gr.Slider, {"minimum": 0.01, "maximum": 0.2, "step": 0.001}),
-    "keyedit_delimiters": OptionInfo(".,\\/!?%^*;:{}=`~()", "Ctrl+up/down word delimiters"),
-    "keyedit_move": OptionInfo(True, "Alt+left/right moves prompt elements"),
+    "sd_checkpoint_dropdown_use_short": OptionInfo(False, "Checkpoint dropdown: use filenames without paths").info("models in subdirectories like photo/sd15.ckpt will be listed as just sd15.ckpt"),
+    "hires_fix_show_sampler": OptionInfo(False, "Hires fix: show hires checkpoint and sampler selection").needs_reload_ui(),
+    "hires_fix_show_prompts": OptionInfo(False, "Hires fix: show hires prompt and negative prompt").needs_reload_ui(),
+    "txt2img_settings_accordion": OptionInfo(False, "Settings in txt2img hidden under Accordion").needs_reload_ui(),
+    "img2img_settings_accordion": OptionInfo(False, "Settings in img2img hidden under Accordion").needs_reload_ui(),
+    "interrupt_after_current": OptionInfo(True, "Don't Interrupt in the middle").info("when using Interrupt button, if generating more than one image, stop after the generation of an image has finished, instead of immediately"),
+}))
+
+options_templates.update(options_section(('ui', "User interface", "ui"), {
+    "localization": OptionInfo("None", "Localization", gr.Dropdown, lambda: {"choices": ["None"] + list(localization.localizations.keys())}, refresh=lambda: localization.list_localizations(cmd_opts.localizations_dir)).needs_reload_ui(),
     "quicksettings_list": OptionInfo(["sd_model_checkpoint"], "Quicksettings list", ui_components.DropdownMulti, lambda: {"choices": list(shared.opts.data_labels.keys())}).js("info", "settingsHintsShowQuicksettings").info("setting entries that appear at the top of page rather than in settings tab").needs_reload_ui(),
     "ui_tab_order": OptionInfo([], "UI tab order", ui_components.DropdownMulti, lambda: {"choices": list(shared.tab_names)}).needs_reload_ui(),
     "hidden_tabs": OptionInfo([], "Hidden UI tabs", ui_components.DropdownMulti, lambda: {"choices": list(shared.tab_names)}).needs_reload_ui(),
-    "ui_reorder_list": OptionInfo([], "txt2img/img2img UI item order", ui_components.DropdownMulti, lambda: {"choices": list(shared_items.ui_reorder_categories())}).info("selected items appear first").needs_reload_ui(),
-    "hires_fix_show_sampler": OptionInfo(False, "Hires fix: show hires checkpoint and sampler selection").needs_reload_ui(),
-    "hires_fix_show_prompts": OptionInfo(False, "Hires fix: show hires prompt and negative prompt").needs_reload_ui(),
-    "disable_token_counters": OptionInfo(False, "Disable prompt token counters").needs_reload_ui(),
+    "ui_reorder_list": OptionInfo([], "UI item order for txt2img/img2img tabs", ui_components.DropdownMulti, lambda: {"choices": list(shared_items.ui_reorder_categories())}).info("selected items appear first").needs_reload_ui(),
+    "gradio_theme": OptionInfo("Default", "Gradio theme", ui_components.DropdownEditable, lambda: {"choices": ["Default"] + shared_gradio_themes.gradio_hf_hub_themes}).info("you can also manually enter any of themes from the <a href='https://huggingface.co/spaces/gradio/theme-gallery'>gallery</a>.").needs_reload_ui(),
+    "gradio_themes_cache": OptionInfo(True, "Cache gradio themes locally").info("disable to update the selected Gradio theme"),
+    "show_progress_in_title": OptionInfo(True, "Show generation progress in window title."),
+    "send_seed": OptionInfo(True, "Send seed when sending prompt or image to other interface"),
+    "send_size": OptionInfo(True, "Send size when sending prompt or image to another interface"),
+    "enable_reloading_ui_scripts": OptionInfo(False, "Reload UI scripts when using Reload UI option").info("useful for developing: if you make changes to UI scripts code, it is applied when the UI is reloded."),
+
 }))
 
 
-options_templates.update(options_section(('infotext', "Infotext"), {
-    "add_model_hash_to_info": OptionInfo(True, "Add model hash to generation information"),
-    "add_model_name_to_info": OptionInfo(True, "Add model name to generation information"),
-    "add_user_name_to_info": OptionInfo(False, "Add user name to generation information when authenticated"),
-    "add_version_to_infotext": OptionInfo(True, "Add program version to generation information"),
+options_templates.update(options_section(('infotext', "Infotext", "ui"), {
+    "infotext_explanation": OptionHTML("""
+Infotext is what this software calls the text that contains generation parameters and can be used to generate the same picture again.
+It is displayed in UI below the image. To use infotext, paste it into the prompt and click the ↙️ paste button.
+"""),
+    "enable_pnginfo": OptionInfo(True, "Write infotext to metadata of the generated image"),
+    "save_txt": OptionInfo(False, "Create a text file with infotext next to every generated image"),
+
+    "add_model_name_to_info": OptionInfo(True, "Add model name to infotext"),
+    "add_model_hash_to_info": OptionInfo(True, "Add model hash to infotext"),
+    "add_vae_name_to_info": OptionInfo(True, "Add VAE name to infotext"),
+    "add_vae_hash_to_info": OptionInfo(True, "Add VAE hash to infotext"),
+    "add_user_name_to_info": OptionInfo(False, "Add user name to infotext when authenticated"),
+    "add_version_to_infotext": OptionInfo(True, "Add program version to infotext"),
     "disable_weights_auto_swap": OptionInfo(True, "Disregard checkpoint information from pasted infotext").info("when reading generation parameters from text into UI"),
+    "infotext_skip_pasting": OptionInfo([], "Disregard fields from pasted infotext", ui_components.DropdownMulti, lambda: {"choices": shared_items.get_infotext_names()}),
     "infotext_styles": OptionInfo("Apply if any", "Infer styles from prompts of pasted infotext", gr.Radio, {"choices": ["Ignore", "Apply", "Discard", "Apply if any"]}).info("when reading generation parameters from text into UI)").html("""<ul style='margin-left: 1.5em'>
 <li>Ignore: keep prompt and styles dropdown as it is.</li>
 <li>Apply: remove style text from prompt, always replace styles dropdown value with found styles (even if none are found).</li>
@@ -282,7 +369,7 @@
 
 }))
 
-options_templates.update(options_section(('ui', "Live previews"), {
+options_templates.update(options_section(('ui', "Live previews", "ui"), {
     "show_progressbar": OptionInfo(True, "Show progressbar"),
     "live_previews_enable": OptionInfo(True, "Show live previews of the created image"),
     "live_previews_image_format": OptionInfo("png", "Live preview file format", gr.Radio, {"choices": ["jpeg", "png", "webp"]}),
@@ -293,9 +380,11 @@
     "live_preview_content": OptionInfo("Prompt", "Live preview subject", gr.Radio, {"choices": ["Combined", "Prompt", "Negative prompt"]}),
     "live_preview_refresh_period": OptionInfo(1000, "Progressbar and preview update period").info("in milliseconds"),
     "live_preview_fast_interrupt": OptionInfo(False, "Return image with chosen live preview method on interrupt").info("makes interrupts faster"),
+    "js_live_preview_in_modal_lightbox": OptionInfo(False, "Show Live preview in full page image viewer"),
+    "prevent_screen_sleep_during_generation": OptionInfo(True, "Prevent screen sleep during generation"),
 }))
 
-options_templates.update(options_section(('sampler-params', "Sampler parameters"), {
+options_templates.update(options_section(('sampler-params', "Sampler parameters", "sd"), {
     "hide_samplers": OptionInfo([], "Hide samplers in user interface", gr.CheckboxGroup, lambda: {"choices": [x.name for x in shared_items.list_samplers()]}).needs_reload_ui(),
     "eta_ddim": OptionInfo(0.0, "Eta for DDIM", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}, infotext='Eta DDIM').info("noise multiplier; higher = more unpredictable results"),
     "eta_ancestral": OptionInfo(1.0, "Eta for k-diffusion samplers", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}, infotext='Eta').info("noise multiplier; currently only applies to ancestral samplers (i.e. Euler a) and SDE samplers"),
@@ -304,23 +393,28 @@
     's_tmin':  OptionInfo(0.0, "sigma tmin",  gr.Slider, {"minimum": 0.0, "maximum": 10.0, "step": 0.01}, infotext='Sigma tmin').info('enable stochasticity; start value of the sigma range; only applies to Euler, Heun, and DPM2'),
     's_tmax':  OptionInfo(0.0, "sigma tmax",  gr.Slider, {"minimum": 0.0, "maximum": 999.0, "step": 0.01}, infotext='Sigma tmax').info("0 = inf; end value of the sigma range; only applies to Euler, Heun, and DPM2"),
     's_noise': OptionInfo(1.0, "sigma noise", gr.Slider, {"minimum": 0.0, "maximum": 1.1, "step": 0.001}, infotext='Sigma noise').info('amount of additional noise to counteract loss of detail during sampling'),
-    'k_sched_type':  OptionInfo("Automatic", "Scheduler type", gr.Dropdown, {"choices": ["Automatic", "karras", "exponential", "polyexponential"]}, infotext='Schedule type').info("lets you override the noise schedule for k-diffusion samplers; choosing Automatic disables the three parameters below"),
-    'sigma_min': OptionInfo(0.0, "sigma min", gr.Number, infotext='Schedule max sigma').info("0 = default (~0.03); minimum noise strength for k-diffusion noise scheduler"),
-    'sigma_max': OptionInfo(0.0, "sigma max", gr.Number, infotext='Schedule min sigma').info("0 = default (~14.6); maximum noise strength for k-diffusion noise scheduler"),
+    'sigma_min': OptionInfo(0.0, "sigma min", gr.Number, infotext='Schedule min sigma').info("0 = default (~0.03); minimum noise strength for k-diffusion noise scheduler"),
+    'sigma_max': OptionInfo(0.0, "sigma max", gr.Number, infotext='Schedule max sigma').info("0 = default (~14.6); maximum noise strength for k-diffusion noise scheduler"),
     'rho':  OptionInfo(0.0, "rho", gr.Number, infotext='Schedule rho').info("0 = default (7 for karras, 1 for polyexponential); higher values result in a steeper noise schedule (decreases faster)"),
     'eta_noise_seed_delta': OptionInfo(0, "Eta noise seed delta", gr.Number, {"precision": 0}, infotext='ENSD').info("ENSD; does not improve anything, just produces different results for ancestral samplers - only useful for reproducing images"),
     'always_discard_next_to_last_sigma': OptionInfo(False, "Always discard next-to-last sigma", infotext='Discard penultimate sigma').link("PR", "https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/6044"),
-    'sgm_noise_multiplier': OptionInfo(False, "SGM noise multiplier", infotext='SGM noise multplier').link("PR", "https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12818").info("Match initial noise to official SDXL implementation - only useful for reproducing images"),
+    'sgm_noise_multiplier': OptionInfo(False, "SGM noise multiplier", infotext='SGM noise multiplier').link("PR", "https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/12818").info("Match initial noise to official SDXL implementation - only useful for reproducing images"),
     'uni_pc_variant': OptionInfo("bh1", "UniPC variant", gr.Radio, {"choices": ["bh1", "bh2", "vary_coeff"]}, infotext='UniPC variant'),
     'uni_pc_skip_type': OptionInfo("time_uniform", "UniPC skip type", gr.Radio, {"choices": ["time_uniform", "time_quadratic", "logSNR"]}, infotext='UniPC skip type'),
     'uni_pc_order': OptionInfo(3, "UniPC order", gr.Slider, {"minimum": 1, "maximum": 50, "step": 1}, infotext='UniPC order').info("must be < sampling steps"),
     'uni_pc_lower_order_final': OptionInfo(True, "UniPC lower order final", infotext='UniPC lower order final'),
+    'sd_noise_schedule': OptionInfo("Default", "Noise schedule for sampling", gr.Radio, {"choices": ["Default", "Zero Terminal SNR"]}, infotext="Noise Schedule").info("for use with zero terminal SNR trained models"),
+    'skip_early_cond': OptionInfo(0.0, "Ignore negative prompt during early sampling", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}, infotext="Skip Early CFG").info("disables CFG on a proportion of steps at the beginning of generation; 0=skip none; 1=skip all; can both improve sample diversity/quality and speed up sampling"),
+    'beta_dist_alpha': OptionInfo(0.6, "Beta scheduler - alpha", gr.Slider, {"minimum": 0.01, "maximum": 1.0, "step": 0.01}, infotext='Beta scheduler alpha').info('Default = 0.6; the alpha parameter of the beta distribution used in Beta sampling'),
+    'beta_dist_beta': OptionInfo(0.6, "Beta scheduler - beta", gr.Slider, {"minimum": 0.01, "maximum": 1.0, "step": 0.01}, infotext='Beta scheduler beta').info('Default = 0.6; the beta parameter of the beta distribution used in Beta sampling'),
 }))
 
-options_templates.update(options_section(('postprocessing', "Postprocessing"), {
+options_templates.update(options_section(('postprocessing', "Postprocessing", "postprocessing"), {
     'postprocessing_enable_in_main_ui': OptionInfo([], "Enable postprocessing operations in txt2img and img2img tabs", ui_components.DropdownMulti, lambda: {"choices": [x.name for x in shared_items.postprocessing_scripts()]}),
+    'postprocessing_disable_in_extras': OptionInfo([], "Disable postprocessing operations in extras tab", ui_components.DropdownMulti, lambda: {"choices": [x.name for x in shared_items.postprocessing_scripts()]}),
     'postprocessing_operation_order': OptionInfo([], "Postprocessing operation order", ui_components.DropdownMulti, lambda: {"choices": [x.name for x in shared_items.postprocessing_scripts()]}),
     'upscaling_max_images_in_cache': OptionInfo(5, "Maximum number of images in upscaling cache", gr.Slider, {"minimum": 0, "maximum": 10, "step": 1}),
+    'postprocessing_existing_caption_action': OptionInfo("Ignore", "Action for existing captions", gr.Radio, {"choices": ["Ignore", "Keep", "Prepend", "Append"]}).info("when generating captions using postprocessing; Ignore = use generated; Keep = use original; Prepend/Append = combine both"),
 }))
 
 options_templates.update(options_section((None, "Hidden options"), {
@@ -329,4 +423,3 @@
     "restore_config_state_file": OptionInfo("", "Config state file to restore from, under 'config-states/' folder"),
     "sd_checkpoint_hash": OptionInfo("", "SHA256 hash of the current checkpoint"),
 }))
-
diff --git a/modules/shared_state.py b/modules/shared_state.py
index d272ee5bc2c..4cd53af6271 100644
--- a/modules/shared_state.py
+++ b/modules/shared_state.py
@@ -12,6 +12,7 @@
 class State:
     skipped = False
     interrupted = False
+    stopping_generation = False
     job = ""
     job_no = 0
     job_count = 0
@@ -79,6 +80,10 @@ def interrupt(self):
         self.interrupted = True
         log.info("Received interrupt request")
 
+    def stop_generating(self):
+        self.stopping_generation = True
+        log.info("Received stop generating request")
+
     def nextjob(self):
         if shared.opts.live_previews_enable and shared.opts.show_progress_every_n_steps == -1:
             self.do_set_current_image()
@@ -91,6 +96,7 @@ def dict(self):
         obj = {
             "skipped": self.skipped,
             "interrupted": self.interrupted,
+            "stopping_generation": self.stopping_generation,
             "job": self.job,
             "job_count": self.job_count,
             "job_timestamp": self.job_timestamp,
@@ -103,6 +109,7 @@ def dict(self):
 
     def begin(self, job: str = "(unknown)"):
         self.sampling_step = 0
+        self.time_start = time.time()
         self.job_count = -1
         self.processing_has_refined_job_count = False
         self.job_no = 0
@@ -113,8 +120,8 @@ def begin(self, job: str = "(unknown)"):
         self.id_live_preview = 0
         self.skipped = False
         self.interrupted = False
+        self.stopping_generation = False
         self.textinfo = None
-        self.time_start = time.time()
         self.job = job
         devices.torch_gc()
         log.info("Starting job %s", job)
@@ -150,10 +157,12 @@ def do_set_current_image(self):
             self.current_image_sampling_step = self.sampling_step
 
         except Exception:
-            # when switching models during genration, VAE would be on CPU, so creating an image will fail.
+            # when switching models during generation, VAE would be on CPU, so creating an image will fail.
             # we silently ignore this error
             errors.record_exception()
 
     def assign_current_image(self, image):
+        if shared.opts.live_previews_image_format == 'jpeg' and image.mode in ('RGBA', 'P'):
+            image = image.convert('RGB')
         self.current_image = image
         self.id_live_preview += 1
diff --git a/modules/styles.py b/modules/styles.py
index 0740fe1b1c0..25f22d3dd49 100644
--- a/modules/styles.py
+++ b/modules/styles.py
@@ -1,15 +1,17 @@
+from __future__ import annotations
+from pathlib import Path
+from modules import errors
 import csv
 import os
-import os.path
-import re
 import typing
 import shutil
 
 
 class PromptStyle(typing.NamedTuple):
     name: str
-    prompt: str
-    negative_prompt: str
+    prompt: str | None
+    negative_prompt: str | None
+    path: str | None = None
 
 
 def merge_prompts(style_prompt: str, prompt: str) -> str:
@@ -29,14 +31,19 @@ def apply_styles_to_prompt(prompt, styles):
     return prompt
 
 
-re_spaces = re.compile("  +")
+def extract_style_text_from_prompt(style_text, prompt):
+    """This function extracts the text from a given prompt based on a provided style text. It checks if the style text contains the placeholder {prompt} or if it appears at the end of the prompt. If a match is found, it returns True along with the extracted text. Otherwise, it returns False and the original prompt.
 
+    extract_style_text_from_prompt("masterpiece", "1girl, art by greg, masterpiece") outputs (True, "1girl, art by greg")
+    extract_style_text_from_prompt("masterpiece, {prompt}", "masterpiece, 1girl, art by greg") outputs (True, "1girl, art by greg")
+    extract_style_text_from_prompt("masterpiece, {prompt}", "exquisite, 1girl, art by greg") outputs (False, "exquisite, 1girl, art by greg")
+    """
+
+    stripped_prompt = prompt.strip()
+    stripped_style_text = style_text.strip()
 
-def extract_style_text_from_prompt(style_text, prompt):
-    stripped_prompt = re.sub(re_spaces, " ", prompt.strip())
-    stripped_style_text = re.sub(re_spaces, " ", style_text.strip())
     if "{prompt}" in stripped_style_text:
-        left, right = stripped_style_text.split("{prompt}", 2)
+        left, _, right = stripped_style_text.partition("{prompt}")
         if stripped_prompt.startswith(left) and stripped_prompt.endswith(right):
             prompt = stripped_prompt[len(left):len(stripped_prompt)-len(right)]
             return True, prompt
@@ -52,7 +59,12 @@ def extract_style_text_from_prompt(style_text, prompt):
     return False, prompt
 
 
-def extract_style_from_prompts(style: PromptStyle, prompt, negative_prompt):
+def extract_original_prompts(style: PromptStyle, prompt, negative_prompt):
+    """
+    Takes a style and compares it to the prompt and negative prompt. If the style
+    matches, returns True plus the prompt and negative prompt with the style text
+    removed. Otherwise, returns False with the original prompt and negative prompt.
+    """
     if not style.prompt and not style.negative_prompt:
         return False, prompt, negative_prompt
 
@@ -68,26 +80,91 @@ def extract_style_from_prompts(style: PromptStyle, prompt, negative_prompt):
 
 
 class StyleDatabase:
-    def __init__(self, path: str):
-        self.no_style = PromptStyle("None", "", "")
+    def __init__(self, paths: list[str | Path]):
+        self.no_style = PromptStyle("None", "", "", None)
         self.styles = {}
-        self.path = path
+        self.paths = paths
+        self.all_styles_files: list[Path] = []
+
+        folder, file = os.path.split(self.paths[0])
+        if '*' in file or '?' in file:
+            # if the first path is a wildcard pattern, find the first match else use "folder/styles.csv" as the default path
+            self.default_path = next(Path(folder).glob(file), Path(os.path.join(folder, 'styles.csv')))
+            self.paths.insert(0, self.default_path)
+        else:
+            self.default_path = Path(self.paths[0])
+
+        self.prompt_fields = [field for field in PromptStyle._fields if field != "path"]
 
         self.reload()
 
     def reload(self):
+        """
+        Clears the style database and reloads the styles from the CSV file(s)
+        matching the path used to initialize the database.
+        """
         self.styles.clear()
 
-        if not os.path.exists(self.path):
-            return
-
-        with open(self.path, "r", encoding="utf-8-sig", newline='') as file:
-            reader = csv.DictReader(file, skipinitialspace=True)
-            for row in reader:
-                # Support loading old CSV format with "name, text"-columns
-                prompt = row["prompt"] if "prompt" in row else row["text"]
-                negative_prompt = row.get("negative_prompt", "")
-                self.styles[row["name"]] = PromptStyle(row["name"], prompt, negative_prompt)
+        # scans for all styles files
+        all_styles_files = []
+        for pattern in self.paths:
+            folder, file = os.path.split(pattern)
+            if '*' in file or '?' in file:
+                found_files = Path(folder).glob(file)
+                [all_styles_files.append(file) for file in found_files]
+            else:
+                # if os.path.exists(pattern):
+                all_styles_files.append(Path(pattern))
+
+        # Remove any duplicate entries
+        seen = set()
+        self.all_styles_files = [s for s in all_styles_files if not (s in seen or seen.add(s))]
+
+        for styles_file in self.all_styles_files:
+            if len(all_styles_files) > 1:
+                # add divider when more than styles file
+                # '---------------- STYLES ----------------'
+                divider = f' {styles_file.stem.upper()} '.center(40, '-')
+                self.styles[divider] = PromptStyle(f"{divider}", None, None, "do_not_save")
+            if styles_file.is_file():
+                self.load_from_csv(styles_file)
+
+    def load_from_csv(self, path: str | Path):
+        try:
+            with open(path, "r", encoding="utf-8-sig", newline="") as file:
+                reader = csv.DictReader(file, skipinitialspace=True)
+                for row in reader:
+                    # Ignore empty rows or rows starting with a comment
+                    if not row or row["name"].startswith("#"):
+                        continue
+                    # Support loading old CSV format with "name, text"-columns
+                    prompt = row["prompt"] if "prompt" in row else row["text"]
+                    negative_prompt = row.get("negative_prompt", "")
+                    # Add style to database
+                    self.styles[row["name"]] = PromptStyle(
+                        row["name"], prompt, negative_prompt, str(path)
+                    )
+        except Exception:
+            errors.report(f'Error loading styles from {path}: ', exc_info=True)
+
+    def get_style_paths(self) -> set:
+        """Returns a set of all distinct paths of files that styles are loaded from."""
+        # Update any styles without a path to the default path
+        for style in list(self.styles.values()):
+            if not style.path:
+                self.styles[style.name] = style._replace(path=str(self.default_path))
+
+        # Create a list of all distinct paths, including the default path
+        style_paths = set()
+        style_paths.add(str(self.default_path))
+        for _, style in self.styles.items():
+            if style.path:
+                style_paths.add(style.path)
+
+        # Remove any paths for styles that are just list dividers
+        style_paths.discard("do_not_save")
+
+        return style_paths
 
     def get_style_prompts(self, styles):
         return [self.styles.get(x, self.no_style).prompt for x in styles]
@@ -96,20 +173,39 @@ def get_negative_style_prompts(self, styles):
         return [self.styles.get(x, self.no_style).negative_prompt for x in styles]
 
     def apply_styles_to_prompt(self, prompt, styles):
-        return apply_styles_to_prompt(prompt, [self.styles.get(x, self.no_style).prompt for x in styles])
+        return apply_styles_to_prompt(
+            prompt, [self.styles.get(x, self.no_style).prompt for x in styles]
+        )
 
     def apply_negative_styles_to_prompt(self, prompt, styles):
-        return apply_styles_to_prompt(prompt, [self.styles.get(x, self.no_style).negative_prompt for x in styles])
-
-    def save_styles(self, path: str) -> None:
-        # Always keep a backup file around
-        if os.path.exists(path):
-            shutil.copy(path, f"{path}.bak")
-
-        with open(path, "w", encoding="utf-8-sig", newline='') as file:
-            writer = csv.DictWriter(file, fieldnames=PromptStyle._fields)
-            writer.writeheader()
-            writer.writerows(style._asdict() for k, style in self.styles.items())
+        return apply_styles_to_prompt(
+            prompt, [self.styles.get(x, self.no_style).negative_prompt for x in styles]
+        )
+
+    def save_styles(self, path: str = None) -> None:
+        # The path argument is deprecated, but kept for backwards compatibility
+
+        style_paths = self.get_style_paths()
+
+        csv_names = [os.path.split(path)[1].lower() for path in style_paths]
+
+        for style_path in style_paths:
+            # Always keep a backup file around
+            if os.path.exists(style_path):
+                shutil.copy(style_path, f"{style_path}.bak")
+
+            # Write the styles to the CSV file
+            with open(style_path, "w", encoding="utf-8-sig", newline="") as file:
+                writer = csv.DictWriter(file, fieldnames=self.prompt_fields)
+                writer.writeheader()
+                for style in (s for s in self.styles.values() if s.path == style_path):
+                    # Skip style list dividers, e.g. "STYLES.CSV"
+                    if style.name.lower().strip("# ") in csv_names:
+                        continue
+                    # Write style fields, ignoring the path field
+                    writer.writerow(
+                        {k: v for k, v in style._asdict().items() if k != "path"}
+                    )
 
     def extract_styles_from_prompt(self, prompt, negative_prompt):
         extracted = []
@@ -120,7 +216,9 @@ def extract_styles_from_prompt(self, prompt, negative_prompt):
             found_style = None
 
             for style in applicable_styles:
-                is_match, new_prompt, new_neg_prompt = extract_style_from_prompts(style, prompt, negative_prompt)
+                is_match, new_prompt, new_neg_prompt = extract_original_prompts(
+                    style, prompt, negative_prompt
+                )
                 if is_match:
                     found_style = style
                     prompt = new_prompt
diff --git a/modules/sub_quadratic_attention.py b/modules/sub_quadratic_attention.py
index ae4ee4bbec0..4cb561ef207 100644
--- a/modules/sub_quadratic_attention.py
+++ b/modules/sub_quadratic_attention.py
@@ -15,7 +15,7 @@
 from torch import Tensor
 from torch.utils.checkpoint import checkpoint
 import math
-from typing import Optional, NamedTuple, List
+from typing import Optional, NamedTuple
 
 
 def narrow_trunc(
@@ -97,7 +97,7 @@ def chunk_scanner(chunk_idx: int) -> AttnChunk:
         )
         return summarize_chunk(query, key_chunk, value_chunk)
 
-    chunks: List[AttnChunk] = [
+    chunks: list[AttnChunk] = [
         chunk_scanner(chunk) for chunk in torch.arange(0, k_tokens, kv_chunk_size)
     ]
     acc_chunk = AttnChunk(*map(torch.stack, zip(*chunks)))
diff --git a/modules/sysinfo.py b/modules/sysinfo.py
index 2db7551dcf0..e9a83d74e03 100644
--- a/modules/sysinfo.py
+++ b/modules/sysinfo.py
@@ -1,16 +1,13 @@
 import json
 import os
 import sys
-import traceback
-
+import subprocess
 import platform
 import hashlib
-import pkg_resources
-import psutil
 import re
+from pathlib import Path
 
-import launch
-from modules import paths_internal, timer, shared, extensions, errors
+from modules import paths_internal, timer, shared_cmd_options, errors, launch_utils
 
 checksum_token = "DontStealMyGamePlz__WINNERS_DONT_USE_DRUGS__DONT_COPY_THAT_FLOPPY"
 environment_whitelist = {
@@ -25,13 +22,13 @@
     "XFORMERS_PACKAGE",
     "CLIP_PACKAGE",
     "OPENCLIP_PACKAGE",
+    "ASSETS_REPO",
     "STABLE_DIFFUSION_REPO",
     "K_DIFFUSION_REPO",
-    "CODEFORMER_REPO",
     "BLIP_REPO",
+    "ASSETS_COMMIT_HASH",
     "STABLE_DIFFUSION_COMMIT_HASH",
     "K_DIFFUSION_COMMIT_HASH",
-    "CODEFORMER_COMMIT_HASH",
     "BLIP_COMMIT_HASH",
     "COMMANDLINE_ARGS",
     "IGNORE_CMD_ARGS_ERRORS",
@@ -70,55 +67,66 @@ def check(x):
     return h.hexdigest() == m.group(1)
 
 
-def get_dict():
-    ram = psutil.virtual_memory()
+def get_cpu_info():
+    cpu_info = {"model": platform.processor()}
+    try:
+        import psutil
+        cpu_info["count logical"] = psutil.cpu_count(logical=True)
+        cpu_info["count physical"] = psutil.cpu_count(logical=False)
+    except Exception as e:
+        cpu_info["error"] = str(e)
+    return cpu_info
+
+
+def get_ram_info():
+    try:
+        import psutil
+        ram = psutil.virtual_memory()
+        return {x: pretty_bytes(getattr(ram, x, 0)) for x in ["total", "used", "free", "active", "inactive", "buffers", "cached", "shared"] if getattr(ram, x, 0) != 0}
+    except Exception as e:
+        return str(e)
 
+
+def get_packages():
+    try:
+        return subprocess.check_output([sys.executable, '-m', 'pip', 'freeze', '--all']).decode("utf8").splitlines()
+    except Exception as pip_error:
+        try:
+            import importlib.metadata
+            packages = importlib.metadata.distributions()
+            return sorted([f"{package.metadata['Name']}=={package.version}" for package in packages])
+        except Exception as e2:
+            return {'error pip': pip_error, 'error importlib': str(e2)}
+
+
+def get_dict():
+    config = get_config()
     res = {
         "Platform": platform.platform(),
         "Python": platform.python_version(),
-        "Version": launch.git_tag(),
-        "Commit": launch.commit_hash(),
+        "Version": launch_utils.git_tag(),
+        "Commit": launch_utils.commit_hash(),
+        "Git status": git_status(paths_internal.script_path),
         "Script path": paths_internal.script_path,
         "Data path": paths_internal.data_path,
         "Extensions dir": paths_internal.extensions_dir,
         "Checksum": checksum_token,
         "Commandline": get_argv(),
         "Torch env info": get_torch_sysinfo(),
-        "Exceptions": get_exceptions(),
-        "CPU": {
-            "model": platform.processor(),
-            "count logical": psutil.cpu_count(logical=True),
-            "count physical": psutil.cpu_count(logical=False),
-        },
-        "RAM": {
-            x: pretty_bytes(getattr(ram, x, 0)) for x in ["total", "used", "free", "active", "inactive", "buffers", "cached", "shared"] if getattr(ram, x, 0) != 0
-        },
-        "Extensions": get_extensions(enabled=True),
-        "Inactive extensions": get_extensions(enabled=False),
+        "Exceptions": errors.get_exceptions(),
+        "CPU": get_cpu_info(),
+        "RAM": get_ram_info(),
+        "Extensions": get_extensions(enabled=True, fallback_disabled_extensions=config.get('disabled_extensions', [])),
+        "Inactive extensions": get_extensions(enabled=False, fallback_disabled_extensions=config.get('disabled_extensions', [])),
         "Environment": get_environment(),
-        "Config": get_config(),
+        "Config": config,
         "Startup": timer.startup_record,
-        "Packages": sorted([f"{pkg.key}=={pkg.version}" for pkg in pkg_resources.working_set]),
+        "Packages": get_packages(),
     }
 
     return res
 
 
-def format_traceback(tb):
-    return [[f"{x.filename}, line {x.lineno}, {x.name}", x.line] for x in traceback.extract_tb(tb)]
-
-
-def format_exception(e, tb):
-    return {"exception": str(e), "traceback": format_traceback(tb)}
-
-
-def get_exceptions():
-    try:
-        return list(reversed(errors.exception_records))
-    except Exception as e:
-        return str(e)
-
-
 def get_environment():
     return {k: os.environ[k] for k in sorted(os.environ) if k in environment_whitelist}
 
@@ -127,11 +135,11 @@ def get_argv():
     res = []
 
     for v in sys.argv:
-        if shared.cmd_opts.gradio_auth and shared.cmd_opts.gradio_auth == v:
+        if shared_cmd_options.cmd_opts.gradio_auth and shared_cmd_options.cmd_opts.gradio_auth == v:
             res.append("<hidden>")
             continue
 
-        if shared.cmd_opts.api_auth and shared.cmd_opts.api_auth == v:
+        if shared_cmd_options.cmd_opts.api_auth and shared_cmd_options.cmd_opts.api_auth == v:
             res.append("<hidden>")
             continue
 
@@ -139,6 +147,7 @@ def get_argv():
 
     return res
 
+
 re_newline = re.compile(r"\r*\n")
 
 
@@ -152,25 +161,55 @@ def get_torch_sysinfo():
         return str(e)
 
 
-def get_extensions(*, enabled):
+def run_git(path, *args):
+    try:
+        return subprocess.check_output([launch_utils.git, '-C', path, *args], shell=False, encoding='utf8').strip()
+    except Exception as e:
+        return str(e)
 
+
+def git_status(path):
+    if (Path(path) / '.git').is_dir():
+        return run_git(paths_internal.script_path, 'status')
+
+
+def get_info_from_repo_path(path: Path):
+    is_repo = (path / '.git').is_dir()
+    return {
+        'name': path.name,
+        'path': str(path),
+        'commit': run_git(path, 'rev-parse', 'HEAD') if is_repo else None,
+        'branch': run_git(path, 'branch', '--show-current') if is_repo else None,
+        'remote': run_git(path, 'remote', 'get-url', 'origin') if is_repo else None,
+    }
+
+
+def get_extensions(*, enabled, fallback_disabled_extensions=None):
     try:
-        def to_json(x: extensions.Extension):
-            return {
-                "name": x.name,
-                "path": x.path,
-                "version": x.version,
-                "branch": x.branch,
-                "remote": x.remote,
-            }
-
-        return [to_json(x) for x in extensions.extensions if not x.is_builtin and x.enabled == enabled]
+        from modules import extensions
+        if extensions.extensions:
+            def to_json(x: extensions.Extension):
+                return {
+                    "name": x.name,
+                    "path": x.path,
+                    "commit": x.commit_hash,
+                    "branch": x.branch,
+                    "remote": x.remote,
+                }
+            return [to_json(x) for x in extensions.extensions if not x.is_builtin and x.enabled == enabled]
+        else:
+            return [get_info_from_repo_path(d) for d in Path(paths_internal.extensions_dir).iterdir() if d.is_dir() and enabled != (str(d.name) in fallback_disabled_extensions)]
     except Exception as e:
         return str(e)
 
 
 def get_config():
     try:
+        from modules import shared
         return shared.opts.data
-    except Exception as e:
-        return str(e)
+    except Exception as _:
+        try:
+            with open(shared_cmd_options.cmd_opts.ui_settings_file, 'r') as f:
+                return json.load(f)
+        except Exception as e:
+            return str(e)
diff --git a/modules/textual_inversion/autocrop.py b/modules/textual_inversion/autocrop.py
index 1675e39a54b..ca858ef4c4a 100644
--- a/modules/textual_inversion/autocrop.py
+++ b/modules/textual_inversion/autocrop.py
@@ -3,6 +3,8 @@
 import os
 import numpy as np
 from PIL import ImageDraw
+from modules import paths_internal
+from pkg_resources import parse_version
 
 GREEN = "#0F0"
 BLUE = "#00F"
@@ -25,7 +27,6 @@ def crop_image(im, settings):
         elif is_portrait(settings.crop_width, settings.crop_height):
             scale_by = settings.crop_height / im.height
 
-
     im = im.resize((int(im.width * scale_by), int(im.height * scale_by)))
     im_debug = im.copy()
 
@@ -64,11 +65,12 @@ def crop_image(im, settings):
         rect[3] -= 1
         d.rectangle(rect, outline=GREEN)
         results.append(im_debug)
-        if settings.destop_view_image:
+        if settings.desktop_view_image:
             im_debug.show()
 
     return results
 
+
 def focal_point(im, settings):
     corner_points = image_corner_points(im, settings) if settings.corner_points_weight > 0 else []
     entropy_points = image_entropy_points(im, settings) if settings.entropy_points_weight > 0 else []
@@ -78,118 +80,120 @@ def focal_point(im, settings):
 
     weight_pref_total = 0
     if corner_points:
-      weight_pref_total += settings.corner_points_weight
+        weight_pref_total += settings.corner_points_weight
     if entropy_points:
-      weight_pref_total += settings.entropy_points_weight
+        weight_pref_total += settings.entropy_points_weight
     if face_points:
-      weight_pref_total += settings.face_points_weight
+        weight_pref_total += settings.face_points_weight
 
     corner_centroid = None
     if corner_points:
-      corner_centroid = centroid(corner_points)
-      corner_centroid.weight = settings.corner_points_weight / weight_pref_total
-      pois.append(corner_centroid)
+        corner_centroid = centroid(corner_points)
+        corner_centroid.weight = settings.corner_points_weight / weight_pref_total
+        pois.append(corner_centroid)
 
     entropy_centroid = None
     if entropy_points:
-      entropy_centroid = centroid(entropy_points)
-      entropy_centroid.weight = settings.entropy_points_weight / weight_pref_total
-      pois.append(entropy_centroid)
+        entropy_centroid = centroid(entropy_points)
+        entropy_centroid.weight = settings.entropy_points_weight / weight_pref_total
+        pois.append(entropy_centroid)
 
     face_centroid = None
     if face_points:
-      face_centroid = centroid(face_points)
-      face_centroid.weight = settings.face_points_weight / weight_pref_total
-      pois.append(face_centroid)
+        face_centroid = centroid(face_points)
+        face_centroid.weight = settings.face_points_weight / weight_pref_total
+        pois.append(face_centroid)
 
     average_point = poi_average(pois, settings)
 
     if settings.annotate_image:
-      d = ImageDraw.Draw(im)
-      max_size = min(im.width, im.height) * 0.07
-      if corner_centroid is not None:
-        color = BLUE
-        box = corner_centroid.bounding(max_size * corner_centroid.weight)
-        d.text((box[0], box[1]-15), f"Edge: {corner_centroid.weight:.02f}", fill=color)
-        d.ellipse(box, outline=color)
-        if len(corner_points) > 1:
-          for f in corner_points:
-            d.rectangle(f.bounding(4), outline=color)
-      if entropy_centroid is not None:
-        color = "#ff0"
-        box = entropy_centroid.bounding(max_size * entropy_centroid.weight)
-        d.text((box[0], box[1]-15), f"Entropy: {entropy_centroid.weight:.02f}", fill=color)
-        d.ellipse(box, outline=color)
-        if len(entropy_points) > 1:
-          for f in entropy_points:
-            d.rectangle(f.bounding(4), outline=color)
-      if face_centroid is not None:
-        color = RED
-        box = face_centroid.bounding(max_size * face_centroid.weight)
-        d.text((box[0], box[1]-15), f"Face: {face_centroid.weight:.02f}", fill=color)
-        d.ellipse(box, outline=color)
-        if len(face_points) > 1:
-          for f in face_points:
-            d.rectangle(f.bounding(4), outline=color)
-
-      d.ellipse(average_point.bounding(max_size), outline=GREEN)
+        d = ImageDraw.Draw(im)
+        max_size = min(im.width, im.height) * 0.07
+        if corner_centroid is not None:
+            color = BLUE
+            box = corner_centroid.bounding(max_size * corner_centroid.weight)
+            d.text((box[0], box[1] - 15), f"Edge: {corner_centroid.weight:.02f}", fill=color)
+            d.ellipse(box, outline=color)
+            if len(corner_points) > 1:
+                for f in corner_points:
+                    d.rectangle(f.bounding(4), outline=color)
+        if entropy_centroid is not None:
+            color = "#ff0"
+            box = entropy_centroid.bounding(max_size * entropy_centroid.weight)
+            d.text((box[0], box[1] - 15), f"Entropy: {entropy_centroid.weight:.02f}", fill=color)
+            d.ellipse(box, outline=color)
+            if len(entropy_points) > 1:
+                for f in entropy_points:
+                    d.rectangle(f.bounding(4), outline=color)
+        if face_centroid is not None:
+            color = RED
+            box = face_centroid.bounding(max_size * face_centroid.weight)
+            d.text((box[0], box[1] - 15), f"Face: {face_centroid.weight:.02f}", fill=color)
+            d.ellipse(box, outline=color)
+            if len(face_points) > 1:
+                for f in face_points:
+                    d.rectangle(f.bounding(4), outline=color)
+
+        d.ellipse(average_point.bounding(max_size), outline=GREEN)
 
     return average_point
 
 
 def image_face_points(im, settings):
     if settings.dnn_model_path is not None:
-      detector = cv2.FaceDetectorYN.create(
-          settings.dnn_model_path,
-          "",
-          (im.width, im.height),
-          0.9, # score threshold
-          0.3, # nms threshold
-          5000 # keep top k before nms
-      )
-      faces = detector.detect(np.array(im))
-      results = []
-      if faces[1] is not None:
-        for face in faces[1]:
-          x = face[0]
-          y = face[1]
-          w = face[2]
-          h = face[3]
-          results.append(
-            PointOfInterest(
-              int(x + (w * 0.5)), # face focus left/right is center
-              int(y + (h * 0.33)), # face focus up/down is close to the top of the head
-              size = w,
-              weight = 1/len(faces[1])
-            )
-          )
-      return results
+        detector = cv2.FaceDetectorYN.create(
+            settings.dnn_model_path,
+            "",
+            (im.width, im.height),
+            0.9,  # score threshold
+            0.3,  # nms threshold
+            5000  # keep top k before nms
+        )
+        faces = detector.detect(np.array(im))
+        results = []
+        if faces[1] is not None:
+            for face in faces[1]:
+                x = face[0]
+                y = face[1]
+                w = face[2]
+                h = face[3]
+                results.append(
+                    PointOfInterest(
+                        int(x + (w * 0.5)),  # face focus left/right is center
+                        int(y + (h * 0.33)),  # face focus up/down is close to the top of the head
+                        size=w,
+                        weight=1 / len(faces[1])
+                    )
+                )
+        return results
     else:
-      np_im = np.array(im)
-      gray = cv2.cvtColor(np_im, cv2.COLOR_BGR2GRAY)
-
-      tries = [
-        [ f'{cv2.data.haarcascades}haarcascade_eye.xml', 0.01 ],
-        [ f'{cv2.data.haarcascades}haarcascade_frontalface_default.xml', 0.05 ],
-        [ f'{cv2.data.haarcascades}haarcascade_profileface.xml', 0.05 ],
-        [ f'{cv2.data.haarcascades}haarcascade_frontalface_alt.xml', 0.05 ],
-        [ f'{cv2.data.haarcascades}haarcascade_frontalface_alt2.xml', 0.05 ],
-        [ f'{cv2.data.haarcascades}haarcascade_frontalface_alt_tree.xml', 0.05 ],
-        [ f'{cv2.data.haarcascades}haarcascade_eye_tree_eyeglasses.xml', 0.05 ],
-        [ f'{cv2.data.haarcascades}haarcascade_upperbody.xml', 0.05 ]
-      ]
-      for t in tries:
-        classifier = cv2.CascadeClassifier(t[0])
-        minsize = int(min(im.width, im.height) * t[1]) # at least N percent of the smallest side
-        try:
-          faces = classifier.detectMultiScale(gray, scaleFactor=1.1,
-            minNeighbors=7, minSize=(minsize, minsize), flags=cv2.CASCADE_SCALE_IMAGE)
-        except Exception:
-          continue
-
-        if faces:
-          rects = [[f[0], f[1], f[0] + f[2], f[1] + f[3]] for f in faces]
-          return [PointOfInterest((r[0] +r[2]) // 2, (r[1] + r[3]) // 2, size=abs(r[0]-r[2]), weight=1/len(rects)) for r in rects]
+        np_im = np.array(im)
+        gray = cv2.cvtColor(np_im, cv2.COLOR_BGR2GRAY)
+
+        tries = [
+            [f'{cv2.data.haarcascades}haarcascade_eye.xml', 0.01],
+            [f'{cv2.data.haarcascades}haarcascade_frontalface_default.xml', 0.05],
+            [f'{cv2.data.haarcascades}haarcascade_profileface.xml', 0.05],
+            [f'{cv2.data.haarcascades}haarcascade_frontalface_alt.xml', 0.05],
+            [f'{cv2.data.haarcascades}haarcascade_frontalface_alt2.xml', 0.05],
+            [f'{cv2.data.haarcascades}haarcascade_frontalface_alt_tree.xml', 0.05],
+            [f'{cv2.data.haarcascades}haarcascade_eye_tree_eyeglasses.xml', 0.05],
+            [f'{cv2.data.haarcascades}haarcascade_upperbody.xml', 0.05]
+        ]
+        for t in tries:
+            classifier = cv2.CascadeClassifier(t[0])
+            minsize = int(min(im.width, im.height) * t[1])  # at least N percent of the smallest side
+            try:
+                faces = classifier.detectMultiScale(gray, scaleFactor=1.1,
+                                                    minNeighbors=7, minSize=(minsize, minsize),
+                                                    flags=cv2.CASCADE_SCALE_IMAGE)
+            except Exception:
+                continue
+
+            if faces:
+                rects = [[f[0], f[1], f[0] + f[2], f[1] + f[3]] for f in faces]
+                return [PointOfInterest((r[0] + r[2]) // 2, (r[1] + r[3]) // 2, size=abs(r[0] - r[2]),
+                                        weight=1 / len(rects)) for r in rects]
     return []
 
 
@@ -198,7 +202,7 @@ def image_corner_points(im, settings):
 
     # naive attempt at preventing focal points from collecting at watermarks near the bottom
     gd = ImageDraw.Draw(grayscale)
-    gd.rectangle([0, im.height*.9, im.width, im.height], fill="#999")
+    gd.rectangle([0, im.height * .9, im.width, im.height], fill="#999")
 
     np_im = np.array(grayscale)
 
@@ -206,7 +210,7 @@ def image_corner_points(im, settings):
         np_im,
         maxCorners=100,
         qualityLevel=0.04,
-        minDistance=min(grayscale.width, grayscale.height)*0.06,
+        minDistance=min(grayscale.width, grayscale.height) * 0.06,
         useHarrisDetector=False,
     )
 
@@ -215,8 +219,8 @@ def image_corner_points(im, settings):
 
     focal_points = []
     for point in points:
-      x, y = point.ravel()
-      focal_points.append(PointOfInterest(x, y, size=4, weight=1/len(points)))
+        x, y = point.ravel()
+        focal_points.append(PointOfInterest(x, y, size=4, weight=1 / len(points)))
 
     return focal_points
 
@@ -225,13 +229,13 @@ def image_entropy_points(im, settings):
     landscape = im.height < im.width
     portrait = im.height > im.width
     if landscape:
-      move_idx = [0, 2]
-      move_max = im.size[0]
+        move_idx = [0, 2]
+        move_max = im.size[0]
     elif portrait:
-      move_idx = [1, 3]
-      move_max = im.size[1]
+        move_idx = [1, 3]
+        move_max = im.size[1]
     else:
-      return []
+        return []
 
     e_max = 0
     crop_current = [0, 0, settings.crop_width, settings.crop_height]
@@ -241,14 +245,14 @@ def image_entropy_points(im, settings):
         e = image_entropy(crop)
 
         if (e > e_max):
-          e_max = e
-          crop_best = list(crop_current)
+            e_max = e
+            crop_best = list(crop_current)
 
         crop_current[move_idx[0]] += 4
         crop_current[move_idx[1]] += 4
 
-    x_mid = int(crop_best[0] + settings.crop_width/2)
-    y_mid = int(crop_best[1] + settings.crop_height/2)
+    x_mid = int(crop_best[0] + settings.crop_width / 2)
+    y_mid = int(crop_best[1] + settings.crop_height / 2)
 
     return [PointOfInterest(x_mid, y_mid, size=25, weight=1.0)]
 
@@ -294,22 +298,23 @@ def is_square(w, h):
     return w == h
 
 
-def download_and_cache_models(dirname):
-    download_url = 'https://github.com/opencv/opencv_zoo/blob/91fb0290f50896f38a0ab1e558b74b16bc009428/models/face_detection_yunet/face_detection_yunet_2022mar.onnx?raw=true'
-    model_file_name = 'face_detection_yunet.onnx'
+model_dir_opencv = os.path.join(paths_internal.models_path, 'opencv')
+if parse_version(cv2.__version__) >= parse_version('4.8'):
+    model_file_path = os.path.join(model_dir_opencv, 'face_detection_yunet_2023mar.onnx')
+    model_url = 'https://github.com/opencv/opencv_zoo/blob/b6e370b10f641879a87890d44e42173077154a05/models/face_detection_yunet/face_detection_yunet_2023mar.onnx?raw=true'
+else:
+    model_file_path = os.path.join(model_dir_opencv, 'face_detection_yunet.onnx')
+    model_url = 'https://github.com/opencv/opencv_zoo/blob/91fb0290f50896f38a0ab1e558b74b16bc009428/models/face_detection_yunet/face_detection_yunet_2022mar.onnx?raw=true'
 
-    os.makedirs(dirname, exist_ok=True)
 
-    cache_file = os.path.join(dirname, model_file_name)
-    if not os.path.exists(cache_file):
-        print(f"downloading face detection model from '{download_url}' to '{cache_file}'")
-        response = requests.get(download_url)
-        with open(cache_file, "wb") as f:
+def download_and_cache_models():
+    if not os.path.exists(model_file_path):
+        os.makedirs(model_dir_opencv, exist_ok=True)
+        print(f"downloading face detection model from '{model_url}' to '{model_file_path}'")
+        response = requests.get(model_url)
+        with open(model_file_path, "wb") as f:
             f.write(response.content)
-
-    if os.path.exists(cache_file):
-        return cache_file
-    return None
+    return model_file_path
 
 
 class PointOfInterest:
@@ -336,5 +341,5 @@ def __init__(self, crop_width=512, crop_height=512, corner_points_weight=0.5, en
         self.entropy_points_weight = entropy_points_weight
         self.face_points_weight = face_points_weight
         self.annotate_image = annotate_image
-        self.destop_view_image = False
+        self.desktop_view_image = False
         self.dnn_model_path = dnn_model_path
diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py
index 7ee05061545..71c032df76d 100644
--- a/modules/textual_inversion/dataset.py
+++ b/modules/textual_inversion/dataset.py
@@ -2,7 +2,6 @@
 import numpy as np
 import PIL
 import torch
-from PIL import Image
 from torch.utils.data import Dataset, DataLoader, Sampler
 from torchvision import transforms
 from collections import defaultdict
@@ -10,7 +9,7 @@
 
 import random
 import tqdm
-from modules import devices, shared
+from modules import devices, shared, images
 import re
 
 from ldm.modules.distributions.distributions import DiagonalGaussianDistribution
@@ -61,7 +60,7 @@ def __init__(self, data_root, width, height, repeats, flip_p=0.5, placeholder_to
             if shared.state.interrupted:
                 raise Exception("interrupted")
             try:
-                image = Image.open(path)
+                image = images.read(path)
                 #Currently does not work for single color transparency
                 #We would need to read image.info['transparency'] for that
                 if use_weight and 'A' in image.getbands():
diff --git a/modules/textual_inversion/image_embedding.py b/modules/textual_inversion/image_embedding.py
index 81cff7bf17c..eac0f9760ab 100644
--- a/modules/textual_inversion/image_embedding.py
+++ b/modules/textual_inversion/image_embedding.py
@@ -1,12 +1,16 @@
 import base64
 import json
+import os.path
 import warnings
+import logging
 
 import numpy as np
 import zlib
 from PIL import Image, ImageDraw
 import torch
 
+logger = logging.getLogger(__name__)
+
 
 class EmbeddingEncoder(json.JSONEncoder):
     def default(self, obj):
@@ -43,7 +47,7 @@ def lcg(m=2**32, a=1664525, c=1013904223, seed=0):
 
 def xor_block(block):
     g = lcg()
-    randblock = np.array([next(g) for _ in range(np.product(block.shape))]).astype(np.uint8).reshape(block.shape)
+    randblock = np.array([next(g) for _ in range(np.prod(block.shape))]).astype(np.uint8).reshape(block.shape)
     return np.bitwise_xor(block.astype(np.uint8), randblock & 0x0F)
 
 
@@ -114,7 +118,7 @@ def extract_image_data_embed(image):
     outarr = crop_black(np.array(image.convert('RGB').getdata()).reshape(image.size[1], image.size[0], d).astype(np.uint8)) & 0x0F
     black_cols = np.where(np.sum(outarr, axis=(0, 2)) == 0)
     if black_cols[0].shape[0] < 2:
-        print('No Image data blocks found.')
+        logger.debug(f'{os.path.basename(getattr(image, "filename", "unknown image file"))}: no embedded information found.')
         return None
 
     data_block_lower = outarr[:, :black_cols[0].min(), :].astype(np.uint8)
@@ -193,11 +197,11 @@ def caption_image_overlay(srcimage, title, footerLeft, footerMid, footerRight, t
 
     embedded_image = insert_image_data_embed(cap_image, test_embed)
 
-    retrived_embed = extract_image_data_embed(embedded_image)
+    retrieved_embed = extract_image_data_embed(embedded_image)
 
-    assert str(retrived_embed) == str(test_embed)
+    assert str(retrieved_embed) == str(test_embed)
 
-    embedded_image2 = insert_image_data_embed(cap_image, retrived_embed)
+    embedded_image2 = insert_image_data_embed(cap_image, retrieved_embed)
 
     assert embedded_image == embedded_image2
 
diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py
deleted file mode 100644
index dbd856bd859..00000000000
--- a/modules/textual_inversion/preprocess.py
+++ /dev/null
@@ -1,232 +0,0 @@
-import os
-from PIL import Image, ImageOps
-import math
-import tqdm
-
-from modules import paths, shared, images, deepbooru
-from modules.textual_inversion import autocrop
-
-
-def preprocess(id_task, process_src, process_dst, process_width, process_height, preprocess_txt_action, process_keep_original_size, process_flip, process_split, process_caption, process_caption_deepbooru=False, split_threshold=0.5, overlap_ratio=0.2, process_focal_crop=False, process_focal_crop_face_weight=0.9, process_focal_crop_entropy_weight=0.15, process_focal_crop_edges_weight=0.5, process_focal_crop_debug=False, process_multicrop=None, process_multicrop_mindim=None, process_multicrop_maxdim=None, process_multicrop_minarea=None, process_multicrop_maxarea=None, process_multicrop_objective=None, process_multicrop_threshold=None):
-    try:
-        if process_caption:
-            shared.interrogator.load()
-
-        if process_caption_deepbooru:
-            deepbooru.model.start()
-
-        preprocess_work(process_src, process_dst, process_width, process_height, preprocess_txt_action, process_keep_original_size, process_flip, process_split, process_caption, process_caption_deepbooru, split_threshold, overlap_ratio, process_focal_crop, process_focal_crop_face_weight, process_focal_crop_entropy_weight, process_focal_crop_edges_weight, process_focal_crop_debug, process_multicrop, process_multicrop_mindim, process_multicrop_maxdim, process_multicrop_minarea, process_multicrop_maxarea, process_multicrop_objective, process_multicrop_threshold)
-
-    finally:
-
-        if process_caption:
-            shared.interrogator.send_blip_to_ram()
-
-        if process_caption_deepbooru:
-            deepbooru.model.stop()
-
-
-def listfiles(dirname):
-    return os.listdir(dirname)
-
-
-class PreprocessParams:
-    src = None
-    dstdir = None
-    subindex = 0
-    flip = False
-    process_caption = False
-    process_caption_deepbooru = False
-    preprocess_txt_action = None
-
-
-def save_pic_with_caption(image, index, params: PreprocessParams, existing_caption=None):
-    caption = ""
-
-    if params.process_caption:
-        caption += shared.interrogator.generate_caption(image)
-
-    if params.process_caption_deepbooru:
-        if caption:
-            caption += ", "
-        caption += deepbooru.model.tag_multi(image)
-
-    filename_part = params.src
-    filename_part = os.path.splitext(filename_part)[0]
-    filename_part = os.path.basename(filename_part)
-
-    basename = f"{index:05}-{params.subindex}-{filename_part}"
-    image.save(os.path.join(params.dstdir, f"{basename}.png"))
-
-    if params.preprocess_txt_action == 'prepend' and existing_caption:
-        caption = f"{existing_caption} {caption}"
-    elif params.preprocess_txt_action == 'append' and existing_caption:
-        caption = f"{caption} {existing_caption}"
-    elif params.preprocess_txt_action == 'copy' and existing_caption:
-        caption = existing_caption
-
-    caption = caption.strip()
-
-    if caption:
-        with open(os.path.join(params.dstdir, f"{basename}.txt"), "w", encoding="utf8") as file:
-            file.write(caption)
-
-    params.subindex += 1
-
-
-def save_pic(image, index, params, existing_caption=None):
-    save_pic_with_caption(image, index, params, existing_caption=existing_caption)
-
-    if params.flip:
-        save_pic_with_caption(ImageOps.mirror(image), index, params, existing_caption=existing_caption)
-
-
-def split_pic(image, inverse_xy, width, height, overlap_ratio):
-    if inverse_xy:
-        from_w, from_h = image.height, image.width
-        to_w, to_h = height, width
-    else:
-        from_w, from_h = image.width, image.height
-        to_w, to_h = width, height
-    h = from_h * to_w // from_w
-    if inverse_xy:
-        image = image.resize((h, to_w))
-    else:
-        image = image.resize((to_w, h))
-
-    split_count = math.ceil((h - to_h * overlap_ratio) / (to_h * (1.0 - overlap_ratio)))
-    y_step = (h - to_h) / (split_count - 1)
-    for i in range(split_count):
-        y = int(y_step * i)
-        if inverse_xy:
-            splitted = image.crop((y, 0, y + to_h, to_w))
-        else:
-            splitted = image.crop((0, y, to_w, y + to_h))
-        yield splitted
-
-# not using torchvision.transforms.CenterCrop because it doesn't allow float regions
-def center_crop(image: Image, w: int, h: int):
-    iw, ih = image.size
-    if ih / h < iw / w:
-        sw = w * ih / h
-        box = (iw - sw) / 2, 0, iw - (iw - sw) / 2, ih
-    else:
-        sh = h * iw / w
-        box = 0, (ih - sh) / 2, iw, ih - (ih - sh) / 2
-    return image.resize((w, h), Image.Resampling.LANCZOS, box)
-
-
-def multicrop_pic(image: Image, mindim, maxdim, minarea, maxarea, objective, threshold):
-    iw, ih = image.size
-    err = lambda w, h: 1-(lambda x: x if x < 1 else 1/x)(iw/ih/(w/h))
-    wh = max(((w, h) for w in range(mindim, maxdim+1, 64) for h in range(mindim, maxdim+1, 64)
-        if minarea <= w * h <= maxarea and err(w, h) <= threshold),
-        key= lambda wh: (wh[0]*wh[1], -err(*wh))[::1 if objective=='Maximize area' else -1],
-        default=None
-    )
-    return wh and center_crop(image, *wh)
-
-
-def preprocess_work(process_src, process_dst, process_width, process_height, preprocess_txt_action, process_keep_original_size, process_flip, process_split, process_caption, process_caption_deepbooru=False, split_threshold=0.5, overlap_ratio=0.2, process_focal_crop=False, process_focal_crop_face_weight=0.9, process_focal_crop_entropy_weight=0.3, process_focal_crop_edges_weight=0.5, process_focal_crop_debug=False, process_multicrop=None, process_multicrop_mindim=None, process_multicrop_maxdim=None, process_multicrop_minarea=None, process_multicrop_maxarea=None, process_multicrop_objective=None, process_multicrop_threshold=None):
-    width = process_width
-    height = process_height
-    src = os.path.abspath(process_src)
-    dst = os.path.abspath(process_dst)
-    split_threshold = max(0.0, min(1.0, split_threshold))
-    overlap_ratio = max(0.0, min(0.9, overlap_ratio))
-
-    assert src != dst, 'same directory specified as source and destination'
-
-    os.makedirs(dst, exist_ok=True)
-
-    files = listfiles(src)
-
-    shared.state.job = "preprocess"
-    shared.state.textinfo = "Preprocessing..."
-    shared.state.job_count = len(files)
-
-    params = PreprocessParams()
-    params.dstdir = dst
-    params.flip = process_flip
-    params.process_caption = process_caption
-    params.process_caption_deepbooru = process_caption_deepbooru
-    params.preprocess_txt_action = preprocess_txt_action
-
-    pbar = tqdm.tqdm(files)
-    for index, imagefile in enumerate(pbar):
-        params.subindex = 0
-        filename = os.path.join(src, imagefile)
-        try:
-            img = Image.open(filename)
-            img = ImageOps.exif_transpose(img)
-            img = img.convert("RGB")
-        except Exception:
-            continue
-
-        description = f"Preprocessing [Image {index}/{len(files)}]"
-        pbar.set_description(description)
-        shared.state.textinfo = description
-
-        params.src = filename
-
-        existing_caption = None
-        existing_caption_filename = f"{os.path.splitext(filename)[0]}.txt"
-        if os.path.exists(existing_caption_filename):
-            with open(existing_caption_filename, 'r', encoding="utf8") as file:
-                existing_caption = file.read()
-
-        if shared.state.interrupted:
-            break
-
-        if img.height > img.width:
-            ratio = (img.width * height) / (img.height * width)
-            inverse_xy = False
-        else:
-            ratio = (img.height * width) / (img.width * height)
-            inverse_xy = True
-
-        process_default_resize = True
-
-        if process_split and ratio < 1.0 and ratio <= split_threshold:
-            for splitted in split_pic(img, inverse_xy, width, height, overlap_ratio):
-                save_pic(splitted, index, params, existing_caption=existing_caption)
-            process_default_resize = False
-
-        if process_focal_crop and img.height != img.width:
-
-            dnn_model_path = None
-            try:
-                dnn_model_path = autocrop.download_and_cache_models(os.path.join(paths.models_path, "opencv"))
-            except Exception as e:
-                print("Unable to load face detection model for auto crop selection. Falling back to lower quality haar method.", e)
-
-            autocrop_settings = autocrop.Settings(
-                crop_width = width,
-                crop_height = height,
-                face_points_weight = process_focal_crop_face_weight,
-                entropy_points_weight = process_focal_crop_entropy_weight,
-                corner_points_weight = process_focal_crop_edges_weight,
-                annotate_image = process_focal_crop_debug,
-                dnn_model_path = dnn_model_path,
-            )
-            for focal in autocrop.crop_image(img, autocrop_settings):
-                save_pic(focal, index, params, existing_caption=existing_caption)
-            process_default_resize = False
-
-        if process_multicrop:
-            cropped = multicrop_pic(img, process_multicrop_mindim, process_multicrop_maxdim, process_multicrop_minarea, process_multicrop_maxarea, process_multicrop_objective, process_multicrop_threshold)
-            if cropped is not None:
-                save_pic(cropped, index, params, existing_caption=existing_caption)
-            else:
-                print(f"skipped {img.width}x{img.height} image {filename} (can't find suitable size within error threshold)")
-            process_default_resize = False
-
-        if process_keep_original_size:
-            save_pic(img, index, params, existing_caption=existing_caption)
-            process_default_resize = False
-
-        if process_default_resize:
-            img = images.resize_image(1, img, width, height)
-            save_pic(img, index, params, existing_caption=existing_caption)
-
-        shared.state.nextjob()
diff --git a/modules/textual_inversion/logging.py b/modules/textual_inversion/saving_settings.py
similarity index 100%
rename from modules/textual_inversion/logging.py
rename to modules/textual_inversion/saving_settings.py
diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py
index aa79dc09843..dc7833e9394 100644
--- a/modules/textual_inversion/textual_inversion.py
+++ b/modules/textual_inversion/textual_inversion.py
@@ -11,14 +11,13 @@
 
 import numpy as np
 from PIL import Image, PngImagePlugin
-from torch.utils.tensorboard import SummaryWriter
 
 from modules import shared, devices, sd_hijack, sd_models, images, sd_samplers, sd_hijack_checkpoint, errors, hashes
 import modules.textual_inversion.dataset
 from modules.textual_inversion.learn_schedule import LearnRateScheduler
 
 from modules.textual_inversion.image_embedding import embedding_to_b64, embedding_from_b64, insert_image_data_embed, extract_image_data_embed, caption_image_overlay
-from modules.textual_inversion.logging import save_settings_to_file
+from modules.textual_inversion.saving_settings import save_settings_to_file
 
 
 TextualInversionTemplate = namedtuple("TextualInversionTemplate", ["name", "path"])
@@ -151,6 +150,7 @@ def register_embedding_by_name(self, embedding, model, name):
         return embedding
 
     def get_expected_shape(self):
+        devices.torch_npu_set_device()
         vec = shared.sd_model.cond_stage_model.encode_embedding_init_text(",", 1)
         return vec.shape[1]
 
@@ -172,7 +172,7 @@ def load_from_file(self, path, filename):
                 if data:
                     name = data.get('name', name)
                 else:
-                    # if data is None, means this is not an embeding, just a preview image
+                    # if data is None, means this is not an embedding, just a preview image
                     return
         elif ext in ['.BIN', '.PT']:
             data = torch.load(path, map_location="cpu")
@@ -181,45 +181,16 @@ def load_from_file(self, path, filename):
         else:
             return
 
+        if data is not None:
+            embedding = create_embedding_from_data(data, name, filename=filename, filepath=path)
 
-        # textual inversion embeddings
-        if 'string_to_param' in data:
-            param_dict = data['string_to_param']
-            param_dict = getattr(param_dict, '_parameters', param_dict)  # fix for torch 1.12.1 loading saved file from torch 1.11
-            assert len(param_dict) == 1, 'embedding file has multiple terms in it'
-            emb = next(iter(param_dict.items()))[1]
-            vec = emb.detach().to(devices.device, dtype=torch.float32)
-            shape = vec.shape[-1]
-            vectors = vec.shape[0]
-        elif type(data) == dict and 'clip_g' in data and 'clip_l' in data:  # SDXL embedding
-            vec = {k: v.detach().to(devices.device, dtype=torch.float32) for k, v in data.items()}
-            shape = data['clip_g'].shape[-1] + data['clip_l'].shape[-1]
-            vectors = data['clip_g'].shape[0]
-        elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor: # diffuser concepts
-            assert len(data.keys()) == 1, 'embedding file has multiple terms in it'
-
-            emb = next(iter(data.values()))
-            if len(emb.shape) == 1:
-                emb = emb.unsqueeze(0)
-            vec = emb.detach().to(devices.device, dtype=torch.float32)
-            shape = vec.shape[-1]
-            vectors = vec.shape[0]
-        else:
-            raise Exception(f"Couldn't identify {filename} as neither textual inversion embedding nor diffuser concept.")
-
-        embedding = Embedding(vec, name)
-        embedding.step = data.get('step', None)
-        embedding.sd_checkpoint = data.get('sd_checkpoint', None)
-        embedding.sd_checkpoint_name = data.get('sd_checkpoint_name', None)
-        embedding.vectors = vectors
-        embedding.shape = shape
-        embedding.filename = path
-        embedding.set_hash(hashes.sha256(embedding.filename, "textual_inversion/" + name) or '')
-
-        if self.expected_shape == -1 or self.expected_shape == embedding.shape:
-            self.register_embedding(embedding, shared.sd_model)
+            if self.expected_shape == -1 or self.expected_shape == embedding.shape:
+                self.register_embedding(embedding, shared.sd_model)
+            else:
+                self.skipped_embeddings[name] = embedding
         else:
-            self.skipped_embeddings[name] = embedding
+            print(f"Unable to load Textual inversion embedding due to data issue: '{name}'.")
+
 
     def load_from_dir(self, embdir):
         if not os.path.isdir(embdir.path):
@@ -313,6 +284,45 @@ def create_embedding(name, num_vectors_per_token, overwrite_old, init_text='*'):
     return fn
 
 
+def create_embedding_from_data(data, name, filename='unknown embedding file', filepath=None):
+    if 'string_to_param' in data:  # textual inversion embeddings
+        param_dict = data['string_to_param']
+        param_dict = getattr(param_dict, '_parameters', param_dict)  # fix for torch 1.12.1 loading saved file from torch 1.11
+        assert len(param_dict) == 1, 'embedding file has multiple terms in it'
+        emb = next(iter(param_dict.items()))[1]
+        vec = emb.detach().to(devices.device, dtype=torch.float32)
+        shape = vec.shape[-1]
+        vectors = vec.shape[0]
+    elif type(data) == dict and 'clip_g' in data and 'clip_l' in data:  # SDXL embedding
+        vec = {k: v.detach().to(devices.device, dtype=torch.float32) for k, v in data.items()}
+        shape = data['clip_g'].shape[-1] + data['clip_l'].shape[-1]
+        vectors = data['clip_g'].shape[0]
+    elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor:  # diffuser concepts
+        assert len(data.keys()) == 1, 'embedding file has multiple terms in it'
+
+        emb = next(iter(data.values()))
+        if len(emb.shape) == 1:
+            emb = emb.unsqueeze(0)
+        vec = emb.detach().to(devices.device, dtype=torch.float32)
+        shape = vec.shape[-1]
+        vectors = vec.shape[0]
+    else:
+        raise Exception(f"Couldn't identify {filename} as neither textual inversion embedding nor diffuser concept.")
+
+    embedding = Embedding(vec, name)
+    embedding.step = data.get('step', None)
+    embedding.sd_checkpoint = data.get('sd_checkpoint', None)
+    embedding.sd_checkpoint_name = data.get('sd_checkpoint_name', None)
+    embedding.vectors = vectors
+    embedding.shape = shape
+
+    if filepath:
+        embedding.filename = filepath
+        embedding.set_hash(hashes.sha256(filepath, "textual_inversion/" + name) or '')
+
+    return embedding
+
+
 def write_loss(log_directory, filename, step, epoch_len, values):
     if shared.opts.training_write_csv_every == 0:
         return
@@ -338,6 +348,7 @@ def write_loss(log_directory, filename, step, epoch_len, values):
         })
 
 def tensorboard_setup(log_directory):
+    from torch.utils.tensorboard import SummaryWriter
     os.makedirs(os.path.join(log_directory, "tensorboard"), exist_ok=True)
     return SummaryWriter(
             log_dir=os.path.join(log_directory, "tensorboard"),
@@ -386,7 +397,7 @@ def validate_train_inputs(model_name, learn_rate, batch_size, gradient_step, dat
         assert log_directory, "Log directory is empty"
 
 
-def train_embedding(id_task, embedding_name, learn_rate, batch_size, gradient_step, data_root, log_directory, training_width, training_height, varsize, steps, clip_grad_mode, clip_grad_value, shuffle_tags, tag_drop_out, latent_sampling_method, use_weight, create_image_every, save_embedding_every, template_filename, save_image_with_stored_embedding, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height):
+def train_embedding(id_task, embedding_name, learn_rate, batch_size, gradient_step, data_root, log_directory, training_width, training_height, varsize, steps, clip_grad_mode, clip_grad_value, shuffle_tags, tag_drop_out, latent_sampling_method, use_weight, create_image_every, save_embedding_every, template_filename, save_image_with_stored_embedding, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_name, preview_cfg_scale, preview_seed, preview_width, preview_height):
     from modules import processing
 
     save_embedding_every = save_embedding_every or 0
@@ -442,8 +453,12 @@ def train_embedding(id_task, embedding_name, learn_rate, batch_size, gradient_st
     shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..."
     old_parallel_processing_allowed = shared.parallel_processing_allowed
 
+    tensorboard_writer = None
     if shared.opts.training_enable_tensorboard:
-        tensorboard_writer = tensorboard_setup(log_directory)
+        try:
+            tensorboard_writer = tensorboard_setup(log_directory)
+        except ImportError:
+            errors.report("Error initializing tensorboard", exc_info=True)
 
     pin_memory = shared.opts.pin_memory
 
@@ -590,7 +605,7 @@ def train_embedding(id_task, embedding_name, learn_rate, batch_size, gradient_st
                         p.prompt = preview_prompt
                         p.negative_prompt = preview_negative_prompt
                         p.steps = preview_steps
-                        p.sampler_name = sd_samplers.samplers[preview_sampler_index].name
+                        p.sampler_name = sd_samplers.samplers_map[preview_sampler_name.lower()]
                         p.cfg_scale = preview_cfg_scale
                         p.seed = preview_seed
                         p.width = preview_width
@@ -616,7 +631,7 @@ def train_embedding(id_task, embedding_name, learn_rate, batch_size, gradient_st
                         last_saved_image, last_text_info = images.save_image(image, images_dir, "", p.seed, p.prompt, shared.opts.samples_format, processed.infotexts[0], p=p, forced_filename=forced_filename, save_to_dirs=False)
                         last_saved_image += f", prompt: {preview_text}"
 
-                        if shared.opts.training_enable_tensorboard and shared.opts.training_tensorboard_save_images:
+                        if tensorboard_writer and shared.opts.training_tensorboard_save_images:
                             tensorboard_add_image(tensorboard_writer, f"Validation at epoch {epoch_num}", image, embedding.step)
 
                     if save_image_with_stored_embedding and os.path.exists(last_saved_file) and embedding_yet_to_be_embedded:
diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py
index 35c4feeff45..f149ad1f0f2 100644
--- a/modules/textual_inversion/ui.py
+++ b/modules/textual_inversion/ui.py
@@ -3,7 +3,6 @@
 import gradio as gr
 
 import modules.textual_inversion.textual_inversion
-import modules.textual_inversion.preprocess
 from modules import sd_hijack, shared
 
 
@@ -15,12 +14,6 @@ def create_embedding(name, initialization_text, nvpt, overwrite_old):
     return gr.Dropdown.update(choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())), f"Created: {filename}", ""
 
 
-def preprocess(*args):
-    modules.textual_inversion.preprocess.preprocess(*args)
-
-    return f"Preprocessing {'interrupted' if shared.state.interrupted else 'finished'}.", ""
-
-
 def train_embedding(*args):
 
     assert not shared.cmd_opts.lowvram, 'Training models with lowvram not possible'
diff --git a/modules/torch_utils.py b/modules/torch_utils.py
new file mode 100644
index 00000000000..5ea3da094c5
--- /dev/null
+++ b/modules/torch_utils.py
@@ -0,0 +1,25 @@
+from __future__ import annotations
+
+import torch.nn
+import torch
+
+
+def get_param(model) -> torch.nn.Parameter:
+    """
+    Find the first parameter in a model or module.
+    """
+    if hasattr(model, "model") and hasattr(model.model, "parameters"):
+        # Unpeel a model descriptor to get at the actual Torch module.
+        model = model.model
+
+    for param in model.parameters():
+        return param
+
+    raise ValueError(f"No parameters found in model {model!r}")
+
+
+def float64(t: torch.Tensor):
+    """return torch.float64 if device is not mps or xpu, else return torch.float32"""
+    if t.device.type in ['mps', 'xpu']:
+        return torch.float32
+    return torch.float64
diff --git a/modules/txt2img.py b/modules/txt2img.py
index 1ee592ad944..6f20253aecf 100644
--- a/modules/txt2img.py
+++ b/modules/txt2img.py
@@ -1,17 +1,22 @@
+import json
 from contextlib import closing
 
 import modules.scripts
-from modules import processing
-from modules.generation_parameters_copypaste import create_override_settings_dict
-from modules.shared import opts, cmd_opts
+from modules import processing, infotext_utils
+from modules.infotext_utils import create_override_settings_dict, parse_generation_parameters
+from modules.shared import opts
 import modules.shared as shared
 from modules.ui import plaintext_to_html
+from PIL import Image
 import gradio as gr
 
 
-def txt2img(id_task: str, prompt: str, negative_prompt: str, prompt_styles, steps: int, sampler_name: str, n_iter: int, batch_size: int, cfg_scale: float, height: int, width: int, enable_hr: bool, denoising_strength: float, hr_scale: float, hr_upscaler: str, hr_second_pass_steps: int, hr_resize_x: int, hr_resize_y: int, hr_checkpoint_name: str, hr_sampler_name: str, hr_prompt: str, hr_negative_prompt, override_settings_texts, request: gr.Request, *args):
+def txt2img_create_processing(id_task: str, request: gr.Request, prompt: str, negative_prompt: str, prompt_styles, n_iter: int, batch_size: int, cfg_scale: float, height: int, width: int, enable_hr: bool, denoising_strength: float, hr_scale: float, hr_upscaler: str, hr_second_pass_steps: int, hr_resize_x: int, hr_resize_y: int, hr_checkpoint_name: str, hr_sampler_name: str, hr_scheduler: str, hr_prompt: str, hr_negative_prompt, override_settings_texts, *args, force_enable_hr=False):
     override_settings = create_override_settings_dict(override_settings_texts)
 
+    if force_enable_hr:
+        enable_hr = True
+
     p = processing.StableDiffusionProcessingTxt2Img(
         sd_model=shared.sd_model,
         outpath_samples=opts.outdir_samples or opts.outdir_txt2img_samples,
@@ -19,15 +24,13 @@ def txt2img(id_task: str, prompt: str, negative_prompt: str, prompt_styles, step
         prompt=prompt,
         styles=prompt_styles,
         negative_prompt=negative_prompt,
-        sampler_name=sampler_name,
         batch_size=batch_size,
         n_iter=n_iter,
-        steps=steps,
         cfg_scale=cfg_scale,
         width=width,
         height=height,
         enable_hr=enable_hr,
-        denoising_strength=denoising_strength if enable_hr else None,
+        denoising_strength=denoising_strength,
         hr_scale=hr_scale,
         hr_upscaler=hr_upscaler,
         hr_second_pass_steps=hr_second_pass_steps,
@@ -35,6 +38,7 @@ def txt2img(id_task: str, prompt: str, negative_prompt: str, prompt_styles, step
         hr_resize_y=hr_resize_y,
         hr_checkpoint_name=None if hr_checkpoint_name == 'Use same checkpoint' else hr_checkpoint_name,
         hr_sampler_name=None if hr_sampler_name == 'Use same sampler' else hr_sampler_name,
+        hr_scheduler=None if hr_scheduler == 'Use same scheduler' else hr_scheduler,
         hr_prompt=hr_prompt,
         hr_negative_prompt=hr_negative_prompt,
         override_settings=override_settings,
@@ -45,11 +49,61 @@ def txt2img(id_task: str, prompt: str, negative_prompt: str, prompt_styles, step
 
     p.user = request.username
 
-    if cmd_opts.enable_console_prompts:
+    if shared.opts.enable_console_prompts:
         print(f"\ntxt2img: {prompt}", file=shared.progress_print_out)
 
+    return p
+
+
+def txt2img_upscale(id_task: str, request: gr.Request, gallery, gallery_index, generation_info, *args):
+    assert len(gallery) > 0, 'No image to upscale'
+    assert 0 <= gallery_index < len(gallery), f'Bad image index: {gallery_index}'
+
+    p = txt2img_create_processing(id_task, request, *args, force_enable_hr=True)
+    p.batch_size = 1
+    p.n_iter = 1
+    # txt2img_upscale attribute that signifies this is called by txt2img_upscale
+    p.txt2img_upscale = True
+
+    geninfo = json.loads(generation_info)
+
+    image_info = gallery[gallery_index] if 0 <= gallery_index < len(gallery) else gallery[0]
+    p.firstpass_image = infotext_utils.image_from_url_text(image_info)
+
+    parameters = parse_generation_parameters(geninfo.get('infotexts')[gallery_index], [])
+    p.seed = parameters.get('Seed', -1)
+    p.subseed = parameters.get('Variation seed', -1)
+
+    p.override_settings['save_images_before_highres_fix'] = False
+
+    with closing(p):
+        processed = modules.scripts.scripts_txt2img.run(p, *p.script_args)
+
+        if processed is None:
+            processed = processing.process_images(p)
+
+    shared.total_tqdm.clear()
+
+    new_gallery = []
+    for i, image in enumerate(gallery):
+        if i == gallery_index:
+            geninfo["infotexts"][gallery_index: gallery_index+1] = processed.infotexts
+            new_gallery.extend(processed.images)
+        else:
+            fake_image = Image.new(mode="RGB", size=(1, 1))
+            fake_image.already_saved_as = image["name"].rsplit('?', 1)[0]
+            new_gallery.append(fake_image)
+
+    geninfo["infotexts"][gallery_index] = processed.info
+
+    return new_gallery, json.dumps(geninfo), plaintext_to_html(processed.info), plaintext_to_html(processed.comments, classname="comments")
+
+
+def txt2img(id_task: str, request: gr.Request, *args):
+    p = txt2img_create_processing(id_task, request, *args)
+
     with closing(p):
-        processed = modules.scripts.scripts_txt2img.run(p, *args)
+        processed = modules.scripts.scripts_txt2img.run(p, *p.script_args)
 
         if processed is None:
             processed = processing.process_images(p)
diff --git a/modules/ui.py b/modules/ui.py
index 579bab9800c..f48638f69ad 100644
--- a/modules/ui.py
+++ b/modules/ui.py
@@ -4,15 +4,16 @@
 import sys
 from functools import reduce
 import warnings
+from contextlib import ExitStack
 
 import gradio as gr
 import gradio.utils
 import numpy as np
 from PIL import Image, PngImagePlugin  # noqa: F401
-from modules.call_queue import wrap_gradio_gpu_call, wrap_queued_call, wrap_gradio_call
+from modules.call_queue import wrap_gradio_gpu_call, wrap_queued_call, wrap_gradio_call, wrap_gradio_call_no_job # noqa: F401
 
-from modules import gradio_extensons  # noqa: F401
-from modules import sd_hijack, sd_models, script_callbacks, ui_extensions, deepbooru, extra_networks, ui_common, ui_postprocessing, progress, ui_loadsave, shared_items, ui_settings, timer, sysinfo, ui_checkpoint_merger, ui_prompt_styles, scripts, sd_samplers, processing, ui_extra_networks
+from modules import gradio_extensons, sd_schedulers  # noqa: F401
+from modules import sd_hijack, sd_models, script_callbacks, ui_extensions, deepbooru, extra_networks, ui_common, ui_postprocessing, progress, ui_loadsave, shared_items, ui_settings, timer, sysinfo, ui_checkpoint_merger, scripts, sd_samplers, processing, ui_extra_networks, ui_toprow, launch_utils
 from modules.ui_components import FormRow, FormGroup, ToolButton, FormHTML, InputAccordion, ResizeHandleRow
 from modules.paths import script_path
 from modules.ui_common import create_refresh_button
@@ -20,15 +21,14 @@
 
 from modules.shared import opts, cmd_opts
 
-import modules.generation_parameters_copypaste as parameters_copypaste
+import modules.infotext_utils as parameters_copypaste
 import modules.hypernetworks.ui as hypernetworks_ui
 import modules.textual_inversion.ui as textual_inversion_ui
 import modules.textual_inversion.textual_inversion as textual_inversion
 import modules.shared as shared
-import modules.images
 from modules import prompt_parser
 from modules.sd_hijack import model_hijack
-from modules.generation_parameters_copypaste import image_from_url_text
+from modules.infotext_utils import image_from_url_text, PasteField
 
 create_setting_component = ui_settings.create_setting_component
 
@@ -38,9 +38,11 @@
 # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the browser will not show any UI
 mimetypes.init()
 mimetypes.add_type('application/javascript', '.js')
+mimetypes.add_type('application/javascript', '.mjs')
 
 # Likewise, add explicit content-type header for certain missing image types
 mimetypes.add_type('image/webp', '.webp')
+mimetypes.add_type('image/avif', '.avif')
 
 if not cmd_opts.share and not cmd_opts.listen:
     # fix gradio phoning home
@@ -151,11 +153,26 @@ def connect_clear_prompt(button):
     )
 
 
-def update_token_counter(text, steps):
+def update_token_counter(text, steps, styles, *, is_positive=True):
+    params = script_callbacks.BeforeTokenCounterParams(text, steps, styles, is_positive=is_positive)
+    script_callbacks.before_token_counter_callback(params)
+    text = params.prompt
+    steps = params.steps
+    styles = params.styles
+    is_positive = params.is_positive
+
+    if shared.opts.include_styles_into_token_counters:
+        apply_styles = shared.prompt_styles.apply_styles_to_prompt if is_positive else shared.prompt_styles.apply_negative_styles_to_prompt
+        text = apply_styles(text, styles)
+
     try:
         text, _ = extra_networks.parse_prompt(text)
 
-        _, prompt_flat_list, _ = prompt_parser.get_multicond_prompt_list([text])
+        if is_positive:
+            _, prompt_flat_list, _ = prompt_parser.get_multicond_prompt_list([text])
+        else:
+            prompt_flat_list = [text]
+
         prompt_schedules = prompt_parser.get_learned_conditioning_prompt_schedules(prompt_flat_list, steps)
 
     except Exception:
@@ -169,76 +186,8 @@ def update_token_counter(text, steps):
     return f"<span class='gr-box gr-text-input'>{token_count}/{max_length}</span>"
 
 
-class Toprow:
-    """Creates a top row UI with prompts, generate button, styles, extra little buttons for things, and enables some functionality related to their operation"""
-
-    def __init__(self, is_img2img):
-        id_part = "img2img" if is_img2img else "txt2img"
-        self.id_part = id_part
-
-        with gr.Row(elem_id=f"{id_part}_toprow", variant="compact"):
-            with gr.Column(elem_id=f"{id_part}_prompt_container", scale=6):
-                with gr.Row():
-                    with gr.Column(scale=80):
-                        with gr.Row():
-                            self.prompt = gr.Textbox(label="Prompt", elem_id=f"{id_part}_prompt", show_label=False, lines=3, placeholder="Prompt (press Ctrl+Enter or Alt+Enter to generate)", elem_classes=["prompt"])
-                            self.prompt_img = gr.File(label="", elem_id=f"{id_part}_prompt_image", file_count="single", type="binary", visible=False)
-
-                with gr.Row():
-                    with gr.Column(scale=80):
-                        with gr.Row():
-                            self.negative_prompt = gr.Textbox(label="Negative prompt", elem_id=f"{id_part}_neg_prompt", show_label=False, lines=3, placeholder="Negative prompt (press Ctrl+Enter or Alt+Enter to generate)", elem_classes=["prompt"])
-
-            self.button_interrogate = None
-            self.button_deepbooru = None
-            if is_img2img:
-                with gr.Column(scale=1, elem_classes="interrogate-col"):
-                    self.button_interrogate = gr.Button('Interrogate\nCLIP', elem_id="interrogate")
-                    self.button_deepbooru = gr.Button('Interrogate\nDeepBooru', elem_id="deepbooru")
-
-            with gr.Column(scale=1, elem_id=f"{id_part}_actions_column"):
-                with gr.Row(elem_id=f"{id_part}_generate_box", elem_classes="generate-box"):
-                    self.interrupt = gr.Button('Interrupt', elem_id=f"{id_part}_interrupt", elem_classes="generate-box-interrupt")
-                    self.skip = gr.Button('Skip', elem_id=f"{id_part}_skip", elem_classes="generate-box-skip")
-                    self.submit = gr.Button('Generate', elem_id=f"{id_part}_generate", variant='primary')
-
-                    self.skip.click(
-                        fn=lambda: shared.state.skip(),
-                        inputs=[],
-                        outputs=[],
-                    )
-
-                    self.interrupt.click(
-                        fn=lambda: shared.state.interrupt(),
-                        inputs=[],
-                        outputs=[],
-                    )
-
-                with gr.Row(elem_id=f"{id_part}_tools"):
-                    self.paste = ToolButton(value=paste_symbol, elem_id="paste")
-                    self.clear_prompt_button = ToolButton(value=clear_prompt_symbol, elem_id=f"{id_part}_clear_prompt")
-                    self.restore_progress_button = ToolButton(value=restore_progress_symbol, elem_id=f"{id_part}_restore_progress", visible=False)
-
-                    self.token_counter = gr.HTML(value="<span>0/75</span>", elem_id=f"{id_part}_token_counter", elem_classes=["token-counter"])
-                    self.token_button = gr.Button(visible=False, elem_id=f"{id_part}_token_button")
-                    self.negative_token_counter = gr.HTML(value="<span>0/75</span>", elem_id=f"{id_part}_negative_token_counter", elem_classes=["token-counter"])
-                    self.negative_token_button = gr.Button(visible=False, elem_id=f"{id_part}_negative_token_button")
-
-                    self.clear_prompt_button.click(
-                        fn=lambda *x: x,
-                        _js="confirm_clear_prompt",
-                        inputs=[self.prompt, self.negative_prompt],
-                        outputs=[self.prompt, self.negative_prompt],
-                    )
-
-                self.ui_styles = ui_prompt_styles.UiPromptStyles(id_part, self.prompt, self.negative_prompt)
-
-        self.prompt_img.change(
-            fn=modules.images.image_data,
-            inputs=[self.prompt_img],
-            outputs=[self.prompt, self.prompt_img],
-            show_progress=False,
-        )
+def update_negative_prompt_token_counter(*args):
+    return update_token_counter(*args, is_positive=False)
 
 
 def setup_progressbar(*args, **kwargs):
@@ -278,21 +227,8 @@ def apply_setting(key, value):
     return getattr(opts, key)
 
 
-def create_output_panel(tabname, outdir):
-    return ui_common.create_output_panel(tabname, outdir)
-
-
-def create_sampler_and_steps_selection(choices, tabname):
-    if opts.samplers_in_dropdown:
-        with FormRow(elem_id=f"sampler_selection_{tabname}"):
-            sampler_name = gr.Dropdown(label='Sampling method', elem_id=f"{tabname}_sampling", choices=choices, value=choices[0])
-            steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{tabname}_steps", label="Sampling steps", value=20)
-    else:
-        with FormGroup(elem_id=f"sampler_selection_{tabname}"):
-            steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{tabname}_steps", label="Sampling steps", value=20)
-            sampler_name = gr.Radio(label='Sampling method', elem_id=f"{tabname}_sampling", choices=choices, value=choices[0])
-
-    return steps, sampler_name
+def create_output_panel(tabname, outdir, toprow=None):
+    return ui_common.create_output_panel(tabname, outdir, toprow)
 
 
 def ordered_ui_categories():
@@ -322,24 +258,31 @@ def create_ui():
 
     parameters_copypaste.reset()
 
+    settings = ui_settings.UiSettings()
+    settings.register_settings()
+
     scripts.scripts_current = scripts.scripts_txt2img
     scripts.scripts_txt2img.initialize_scripts(is_img2img=False)
 
     with gr.Blocks(analytics_enabled=False) as txt2img_interface:
-        toprow = Toprow(is_img2img=False)
+        toprow = ui_toprow.Toprow(is_img2img=False, is_compact=shared.opts.compact_prompt_box)
 
         dummy_component = gr.Label(visible=False)
 
-        extra_tabs = gr.Tabs(elem_id="txt2img_extra_tabs")
+        extra_tabs = gr.Tabs(elem_id="txt2img_extra_tabs", elem_classes=["extra-networks"])
         extra_tabs.__enter__()
 
         with gr.Tab("Generation", id="txt2img_generation") as txt2img_generation_tab, ResizeHandleRow(equal_height=False):
-            with gr.Column(variant='compact', elem_id="txt2img_settings"):
+            with ExitStack() as stack:
+                if shared.opts.txt2img_settings_accordion:
+                    stack.enter_context(gr.Accordion("Open for Settings", open=False))
+                stack.enter_context(gr.Column(variant='compact', elem_id="txt2img_settings"))
+
                 scripts.scripts_txt2img.prepare_ui()
 
                 for category in ordered_ui_categories():
-                    if category == "sampler":
-                        steps, sampler_name = create_sampler_and_steps_selection(sd_samplers.visible_sampler_names(), "txt2img")
+                    if category == "prompt":
+                        toprow.create_inline_toprow_prompts()
 
                     elif category == "dimensions":
                         with FormRow():
@@ -348,7 +291,7 @@ def create_ui():
                                 height = gr.Slider(minimum=64, maximum=2048, step=8, label="Height", value=512, elem_id="txt2img_height")
 
                             with gr.Column(elem_id="txt2img_dimensions_row", scale=1, elem_classes="dimensions-tools"):
-                                res_switch_btn = ToolButton(value=switch_values_symbol, elem_id="txt2img_res_switch_btn", label="Switch dims")
+                                res_switch_btn = ToolButton(value=switch_values_symbol, elem_id="txt2img_res_switch_btn", tooltip="Switch width/height")
 
                             if opts.dimensions_and_batch_together:
                                 with gr.Column(elem_id="txt2img_column_batch"):
@@ -381,10 +324,11 @@ def create_ui():
 
                                 with FormRow(elem_id="txt2img_hires_fix_row3", variant="compact", visible=opts.hires_fix_show_sampler) as hr_sampler_container:
 
-                                    hr_checkpoint_name = gr.Dropdown(label='Hires checkpoint', elem_id="hr_checkpoint", choices=["Use same checkpoint"] + modules.sd_models.checkpoint_tiles(use_short=True), value="Use same checkpoint")
+                                    hr_checkpoint_name = gr.Dropdown(label='Checkpoint', elem_id="hr_checkpoint", choices=["Use same checkpoint"] + modules.sd_models.checkpoint_tiles(use_short=True), value="Use same checkpoint")
                                     create_refresh_button(hr_checkpoint_name, modules.sd_models.list_models, lambda: {"choices": ["Use same checkpoint"] + modules.sd_models.checkpoint_tiles(use_short=True)}, "hr_checkpoint_refresh")
 
                                     hr_sampler_name = gr.Dropdown(label='Hires sampling method', elem_id="hr_sampler", choices=["Use same sampler"] + sd_samplers.visible_sampler_names(), value="Use same sampler")
+                                    hr_scheduler = gr.Dropdown(label='Hires schedule type', elem_id="hr_scheduler", choices=["Use same scheduler"] + [x.label for x in sd_schedulers.schedulers], value="Use same scheduler")
 
                                 with FormRow(elem_id="txt2img_hires_fix_row4", variant="compact", visible=opts.hires_fix_show_prompts) as hr_prompts_container:
                                     with gr.Column(scale=80):
@@ -432,50 +376,59 @@ def create_ui():
                     show_progress=False,
                 )
 
-            txt2img_gallery, generation_info, html_info, html_log = create_output_panel("txt2img", opts.outdir_txt2img_samples)
+            output_panel = create_output_panel("txt2img", opts.outdir_txt2img_samples, toprow)
+
+            txt2img_inputs = [
+                dummy_component,
+                toprow.prompt,
+                toprow.negative_prompt,
+                toprow.ui_styles.dropdown,
+                batch_count,
+                batch_size,
+                cfg_scale,
+                height,
+                width,
+                enable_hr,
+                denoising_strength,
+                hr_scale,
+                hr_upscaler,
+                hr_second_pass_steps,
+                hr_resize_x,
+                hr_resize_y,
+                hr_checkpoint_name,
+                hr_sampler_name,
+                hr_scheduler,
+                hr_prompt,
+                hr_negative_prompt,
+                override_settings,
+            ] + custom_inputs
+
+            txt2img_outputs = [
+                output_panel.gallery,
+                output_panel.generation_info,
+                output_panel.infotext,
+                output_panel.html_log,
+            ]
 
             txt2img_args = dict(
                 fn=wrap_gradio_gpu_call(modules.txt2img.txt2img, extra_outputs=[None, '', '']),
                 _js="submit",
-                inputs=[
-                    dummy_component,
-                    toprow.prompt,
-                    toprow.negative_prompt,
-                    toprow.ui_styles.dropdown,
-                    steps,
-                    sampler_name,
-                    batch_count,
-                    batch_size,
-                    cfg_scale,
-                    height,
-                    width,
-                    enable_hr,
-                    denoising_strength,
-                    hr_scale,
-                    hr_upscaler,
-                    hr_second_pass_steps,
-                    hr_resize_x,
-                    hr_resize_y,
-                    hr_checkpoint_name,
-                    hr_sampler_name,
-                    hr_prompt,
-                    hr_negative_prompt,
-                    override_settings,
-
-                ] + custom_inputs,
-
-                outputs=[
-                    txt2img_gallery,
-                    generation_info,
-                    html_info,
-                    html_log,
-                ],
+                inputs=txt2img_inputs,
+                outputs=txt2img_outputs,
                 show_progress=False,
             )
 
             toprow.prompt.submit(**txt2img_args)
             toprow.submit.click(**txt2img_args)
 
+            output_panel.button_upscale.click(
+                fn=wrap_gradio_gpu_call(modules.txt2img.txt2img_upscale, extra_outputs=[None, '', '']),
+                _js="submit_txt2img_upscale",
+                inputs=txt2img_inputs[0:1] + [output_panel.gallery, dummy_component, output_panel.generation_info] + txt2img_inputs[1:],
+                outputs=txt2img_outputs,
+                show_progress=False,
+            )
+
             res_switch_btn.click(fn=None, _js="function(){switchWidthHeight('txt2img')}", inputs=None, outputs=None, show_progress=False)
 
             toprow.restore_progress_button.click(
@@ -483,37 +436,36 @@ def create_ui():
                 _js="restoreProgressTxt2img",
                 inputs=[dummy_component],
                 outputs=[
-                    txt2img_gallery,
-                    generation_info,
-                    html_info,
-                    html_log,
+                    output_panel.gallery,
+                    output_panel.generation_info,
+                    output_panel.infotext,
+                    output_panel.html_log,
                 ],
                 show_progress=False,
             )
 
             txt2img_paste_fields = [
-                (toprow.prompt, "Prompt"),
-                (toprow.negative_prompt, "Negative prompt"),
-                (steps, "Steps"),
-                (sampler_name, "Sampler"),
-                (cfg_scale, "CFG scale"),
-                (width, "Size-1"),
-                (height, "Size-2"),
-                (batch_size, "Batch size"),
-                (toprow.ui_styles.dropdown, lambda d: d["Styles array"] if isinstance(d.get("Styles array"), list) else gr.update()),
-                (denoising_strength, "Denoising strength"),
-                (enable_hr, lambda d: "Denoising strength" in d and ("Hires upscale" in d or "Hires upscaler" in d or "Hires resize-1" in d)),
-                (hr_scale, "Hires upscale"),
-                (hr_upscaler, "Hires upscaler"),
-                (hr_second_pass_steps, "Hires steps"),
-                (hr_resize_x, "Hires resize-1"),
-                (hr_resize_y, "Hires resize-2"),
-                (hr_checkpoint_name, "Hires checkpoint"),
-                (hr_sampler_name, "Hires sampler"),
-                (hr_sampler_container, lambda d: gr.update(visible=True) if d.get("Hires sampler", "Use same sampler") != "Use same sampler" or d.get("Hires checkpoint", "Use same checkpoint") != "Use same checkpoint" else gr.update()),
-                (hr_prompt, "Hires prompt"),
-                (hr_negative_prompt, "Hires negative prompt"),
-                (hr_prompts_container, lambda d: gr.update(visible=True) if d.get("Hires prompt", "") != "" or d.get("Hires negative prompt", "") != "" else gr.update()),
+                PasteField(toprow.prompt, "Prompt", api="prompt"),
+                PasteField(toprow.negative_prompt, "Negative prompt", api="negative_prompt"),
+                PasteField(cfg_scale, "CFG scale", api="cfg_scale"),
+                PasteField(width, "Size-1", api="width"),
+                PasteField(height, "Size-2", api="height"),
+                PasteField(batch_size, "Batch size", api="batch_size"),
+                PasteField(toprow.ui_styles.dropdown, lambda d: d["Styles array"] if isinstance(d.get("Styles array"), list) else gr.update(), api="styles"),
+                PasteField(denoising_strength, "Denoising strength", api="denoising_strength"),
+                PasteField(enable_hr, lambda d: "Denoising strength" in d and ("Hires upscale" in d or "Hires upscaler" in d or "Hires resize-1" in d), api="enable_hr"),
+                PasteField(hr_scale, "Hires upscale", api="hr_scale"),
+                PasteField(hr_upscaler, "Hires upscaler", api="hr_upscaler"),
+                PasteField(hr_second_pass_steps, "Hires steps", api="hr_second_pass_steps"),
+                PasteField(hr_resize_x, "Hires resize-1", api="hr_resize_x"),
+                PasteField(hr_resize_y, "Hires resize-2", api="hr_resize_y"),
+                PasteField(hr_checkpoint_name, "Hires checkpoint", api="hr_checkpoint_name"),
+                PasteField(hr_sampler_name, sd_samplers.get_hr_sampler_from_infotext, api="hr_sampler_name"),
+                PasteField(hr_scheduler, sd_samplers.get_hr_scheduler_from_infotext, api="hr_scheduler"),
+                PasteField(hr_sampler_container, lambda d: gr.update(visible=True) if d.get("Hires sampler", "Use same sampler") != "Use same sampler" or d.get("Hires checkpoint", "Use same checkpoint") != "Use same checkpoint" or d.get("Hires schedule type", "Use same scheduler") != "Use same scheduler" else gr.update()),
+                PasteField(hr_prompt, "Hires prompt", api="hr_prompt"),
+                PasteField(hr_negative_prompt, "Hires negative prompt", api="hr_negative_prompt"),
+                PasteField(hr_prompts_container, lambda d: gr.update(visible=True) if d.get("Hires prompt", "") != "" or d.get("Hires negative prompt", "") != "" else gr.update()),
                 *scripts.scripts_txt2img.infotext_fields
             ]
             parameters_copypaste.add_paste_fields("txt2img", None, txt2img_paste_fields, override_settings)
@@ -521,22 +473,26 @@ def create_ui():
                 paste_button=toprow.paste, tabname="txt2img", source_text_component=toprow.prompt, source_image_component=None,
             ))
 
+            steps = scripts.scripts_txt2img.script('Sampler').steps
+
             txt2img_preview_params = [
                 toprow.prompt,
                 toprow.negative_prompt,
                 steps,
-                sampler_name,
+                scripts.scripts_txt2img.script('Sampler').sampler_name,
                 cfg_scale,
                 scripts.scripts_txt2img.script('Seed').seed,
                 width,
                 height,
             ]
 
-            toprow.token_button.click(fn=wrap_queued_call(update_token_counter), inputs=[toprow.prompt, steps], outputs=[toprow.token_counter])
-            toprow.negative_token_button.click(fn=wrap_queued_call(update_token_counter), inputs=[toprow.negative_prompt, steps], outputs=[toprow.negative_token_counter])
+            toprow.ui_styles.dropdown.change(fn=wrap_queued_call(update_token_counter), inputs=[toprow.prompt, steps, toprow.ui_styles.dropdown], outputs=[toprow.token_counter])
+            toprow.ui_styles.dropdown.change(fn=wrap_queued_call(update_negative_prompt_token_counter), inputs=[toprow.negative_prompt, steps, toprow.ui_styles.dropdown], outputs=[toprow.negative_token_counter])
+            toprow.token_button.click(fn=wrap_queued_call(update_token_counter), inputs=[toprow.prompt, steps, toprow.ui_styles.dropdown], outputs=[toprow.token_counter])
+            toprow.negative_token_button.click(fn=wrap_queued_call(update_negative_prompt_token_counter), inputs=[toprow.negative_prompt, steps, toprow.ui_styles.dropdown], outputs=[toprow.negative_token_counter])
 
         extra_networks_ui = ui_extra_networks.create_ui(txt2img_interface, [txt2img_generation_tab], 'txt2img')
-        ui_extra_networks.setup_ui(extra_networks_ui, txt2img_gallery)
+        ui_extra_networks.setup_ui(extra_networks_ui, output_panel.gallery)
 
         extra_tabs.__exit__()
 
@@ -544,13 +500,17 @@ def create_ui():
     scripts.scripts_img2img.initialize_scripts(is_img2img=True)
 
     with gr.Blocks(analytics_enabled=False) as img2img_interface:
-        toprow = Toprow(is_img2img=True)
+        toprow = ui_toprow.Toprow(is_img2img=True, is_compact=shared.opts.compact_prompt_box)
 
-        extra_tabs = gr.Tabs(elem_id="img2img_extra_tabs")
+        extra_tabs = gr.Tabs(elem_id="img2img_extra_tabs", elem_classes=["extra-networks"])
         extra_tabs.__enter__()
 
         with gr.Tab("Generation", id="img2img_generation") as img2img_generation_tab, ResizeHandleRow(equal_height=False):
-            with gr.Column(variant='compact', elem_id="img2img_settings"):
+            with ExitStack() as stack:
+                if shared.opts.img2img_settings_accordion:
+                    stack.enter_context(gr.Accordion("Open for Settings", open=False))
+                stack.enter_context(gr.Column(variant='compact', elem_id="img2img_settings"))
+
                 copy_image_buttons = []
                 copy_image_destinations = {}
 
@@ -567,104 +527,112 @@ def add_copy_image_controls(tab_name, elem):
                             button = gr.Button(title)
                             copy_image_buttons.append((button, name, elem))
 
-                with gr.Tabs(elem_id="mode_img2img"):
-                    img2img_selected_tab = gr.State(0)
-
-                    with gr.TabItem('img2img', id='img2img', elem_id="img2img_img2img_tab") as tab_img2img:
-                        init_img = gr.Image(label="Image for img2img", elem_id="img2img_image", show_label=False, source="upload", interactive=True, type="pil", tool="editor", image_mode="RGBA", height=opts.img2img_editor_height)
-                        add_copy_image_controls('img2img', init_img)
-
-                    with gr.TabItem('Sketch', id='img2img_sketch', elem_id="img2img_img2img_sketch_tab") as tab_sketch:
-                        sketch = gr.Image(label="Image for img2img", elem_id="img2img_sketch", show_label=False, source="upload", interactive=True, type="pil", tool="color-sketch", image_mode="RGB", height=opts.img2img_editor_height, brush_color=opts.img2img_sketch_default_brush_color)
-                        add_copy_image_controls('sketch', sketch)
-
-                    with gr.TabItem('Inpaint', id='inpaint', elem_id="img2img_inpaint_tab") as tab_inpaint:
-                        init_img_with_mask = gr.Image(label="Image for inpainting with mask", show_label=False, elem_id="img2maskimg", source="upload", interactive=True, type="pil", tool="sketch", image_mode="RGBA", height=opts.img2img_editor_height, brush_color=opts.img2img_inpaint_mask_brush_color)
-                        add_copy_image_controls('inpaint', init_img_with_mask)
-
-                    with gr.TabItem('Inpaint sketch', id='inpaint_sketch', elem_id="img2img_inpaint_sketch_tab") as tab_inpaint_color:
-                        inpaint_color_sketch = gr.Image(label="Color sketch inpainting", show_label=False, elem_id="inpaint_sketch", source="upload", interactive=True, type="pil", tool="color-sketch", image_mode="RGB", height=opts.img2img_editor_height, brush_color=opts.img2img_inpaint_sketch_default_brush_color)
-                        inpaint_color_sketch_orig = gr.State(None)
-                        add_copy_image_controls('inpaint_sketch', inpaint_color_sketch)
-
-                        def update_orig(image, state):
-                            if image is not None:
-                                same_size = state is not None and state.size == image.size
-                                has_exact_match = np.any(np.all(np.array(image) == np.array(state), axis=-1))
-                                edited = same_size and has_exact_match
-                                return image if not edited or state is None else state
-
-                        inpaint_color_sketch.change(update_orig, [inpaint_color_sketch, inpaint_color_sketch_orig], inpaint_color_sketch_orig)
-
-                    with gr.TabItem('Inpaint upload', id='inpaint_upload', elem_id="img2img_inpaint_upload_tab") as tab_inpaint_upload:
-                        init_img_inpaint = gr.Image(label="Image for img2img", show_label=False, source="upload", interactive=True, type="pil", elem_id="img_inpaint_base")
-                        init_mask_inpaint = gr.Image(label="Mask", source="upload", interactive=True, type="pil", image_mode="RGBA", elem_id="img_inpaint_mask")
-
-                    with gr.TabItem('Batch', id='batch', elem_id="img2img_batch_tab") as tab_batch:
-                        hidden = '<br>Disabled when launched with --hide-ui-dir-config.' if shared.cmd_opts.hide_ui_dir_config else ''
-                        gr.HTML(
-                            "<p style='padding-bottom: 1em;' class=\"text-gray-500\">Process images in a directory on the same machine where the server is running." +
-                            "<br>Use an empty output directory to save pictures normally instead of writing to the output directory." +
-                            f"<br>Add inpaint batch mask directory to enable inpaint batch processing."
-                            f"{hidden}</p>"
-                        )
-                        img2img_batch_input_dir = gr.Textbox(label="Input directory", **shared.hide_dirs, elem_id="img2img_batch_input_dir")
-                        img2img_batch_output_dir = gr.Textbox(label="Output directory", **shared.hide_dirs, elem_id="img2img_batch_output_dir")
-                        img2img_batch_inpaint_mask_dir = gr.Textbox(label="Inpaint batch mask directory (required for inpaint batch processing only)", **shared.hide_dirs, elem_id="img2img_batch_inpaint_mask_dir")
-                        with gr.Accordion("PNG info", open=False):
-                            img2img_batch_use_png_info = gr.Checkbox(label="Append png info to prompts", **shared.hide_dirs, elem_id="img2img_batch_use_png_info")
-                            img2img_batch_png_info_dir = gr.Textbox(label="PNG info directory", **shared.hide_dirs, placeholder="Leave empty to use input directory", elem_id="img2img_batch_png_info_dir")
-                            img2img_batch_png_info_props = gr.CheckboxGroup(["Prompt", "Negative prompt", "Seed", "CFG scale", "Sampler", "Steps"], label="Parameters to take from png info", info="Prompts from png info will be appended to prompts set in ui.")
-
-                    img2img_tabs = [tab_img2img, tab_sketch, tab_inpaint, tab_inpaint_color, tab_inpaint_upload, tab_batch]
-
-                    for i, tab in enumerate(img2img_tabs):
-                        tab.select(fn=lambda tabnum=i: tabnum, inputs=[], outputs=[img2img_selected_tab])
-
-                def copy_image(img):
-                    if isinstance(img, dict) and 'image' in img:
-                        return img['image']
-
-                    return img
-
-                for button, name, elem in copy_image_buttons:
-                    button.click(
-                        fn=copy_image,
-                        inputs=[elem],
-                        outputs=[copy_image_destinations[name]],
-                    )
-                    button.click(
-                        fn=lambda: None,
-                        _js=f"switch_to_{name.replace(' ', '_')}",
-                        inputs=[],
-                        outputs=[],
-                    )
-
-                with FormRow():
-                    resize_mode = gr.Radio(label="Resize mode", elem_id="resize_mode", choices=["Just resize", "Crop and resize", "Resize and fill", "Just resize (latent upscale)"], type="index", value="Just resize")
-
                 scripts.scripts_img2img.prepare_ui()
 
                 for category in ordered_ui_categories():
-                    if category == "sampler":
-                        steps, sampler_name = create_sampler_and_steps_selection(sd_samplers.visible_sampler_names(), "img2img")
+                    if category == "prompt":
+                        toprow.create_inline_toprow_prompts()
+
+                    if category == "image":
+                        with gr.Tabs(elem_id="mode_img2img"):
+                            img2img_selected_tab = gr.Number(value=0, visible=False)
+
+                            with gr.TabItem('img2img', id='img2img', elem_id="img2img_img2img_tab") as tab_img2img:
+                                init_img = gr.Image(label="Image for img2img", elem_id="img2img_image", show_label=False, source="upload", interactive=True, type="pil", tool="editor", image_mode="RGBA", height=opts.img2img_editor_height)
+                                add_copy_image_controls('img2img', init_img)
+
+                            with gr.TabItem('Sketch', id='img2img_sketch', elem_id="img2img_img2img_sketch_tab") as tab_sketch:
+                                sketch = gr.Image(label="Image for img2img", elem_id="img2img_sketch", show_label=False, source="upload", interactive=True, type="pil", tool="color-sketch", image_mode="RGB", height=opts.img2img_editor_height, brush_color=opts.img2img_sketch_default_brush_color)
+                                add_copy_image_controls('sketch', sketch)
+
+                            with gr.TabItem('Inpaint', id='inpaint', elem_id="img2img_inpaint_tab") as tab_inpaint:
+                                init_img_with_mask = gr.Image(label="Image for inpainting with mask", show_label=False, elem_id="img2maskimg", source="upload", interactive=True, type="pil", tool="sketch", image_mode="RGBA", height=opts.img2img_editor_height, brush_color=opts.img2img_inpaint_mask_brush_color)
+                                add_copy_image_controls('inpaint', init_img_with_mask)
+
+                            with gr.TabItem('Inpaint sketch', id='inpaint_sketch', elem_id="img2img_inpaint_sketch_tab") as tab_inpaint_color:
+                                inpaint_color_sketch = gr.Image(label="Color sketch inpainting", show_label=False, elem_id="inpaint_sketch", source="upload", interactive=True, type="pil", tool="color-sketch", image_mode="RGB", height=opts.img2img_editor_height, brush_color=opts.img2img_inpaint_sketch_default_brush_color)
+                                inpaint_color_sketch_orig = gr.State(None)
+                                add_copy_image_controls('inpaint_sketch', inpaint_color_sketch)
+
+                                def update_orig(image, state):
+                                    if image is not None:
+                                        same_size = state is not None and state.size == image.size
+                                        has_exact_match = np.any(np.all(np.array(image) == np.array(state), axis=-1))
+                                        edited = same_size and has_exact_match
+                                        return image if not edited or state is None else state
+
+                                inpaint_color_sketch.change(update_orig, [inpaint_color_sketch, inpaint_color_sketch_orig], inpaint_color_sketch_orig)
+
+                            with gr.TabItem('Inpaint upload', id='inpaint_upload', elem_id="img2img_inpaint_upload_tab") as tab_inpaint_upload:
+                                init_img_inpaint = gr.Image(label="Image for img2img", show_label=False, source="upload", interactive=True, type="pil", elem_id="img_inpaint_base")
+                                init_mask_inpaint = gr.Image(label="Mask", source="upload", interactive=True, type="pil", image_mode="RGBA", elem_id="img_inpaint_mask")
+
+                            with gr.TabItem('Batch', id='batch', elem_id="img2img_batch_tab") as tab_batch:
+                                with gr.Tabs(elem_id="img2img_batch_source"):
+                                    img2img_batch_source_type = gr.Textbox(visible=False, value="upload")
+                                    with gr.TabItem('Upload', id='batch_upload', elem_id="img2img_batch_upload_tab") as tab_batch_upload:
+                                        img2img_batch_upload = gr.Files(label="Files", interactive=True, elem_id="img2img_batch_upload")
+                                    with gr.TabItem('From directory', id='batch_from_dir', elem_id="img2img_batch_from_dir_tab") as tab_batch_from_dir:
+                                        hidden = '<br>Disabled when launched with --hide-ui-dir-config.' if shared.cmd_opts.hide_ui_dir_config else ''
+                                        gr.HTML(
+                                            "<p style='padding-bottom: 1em;' class=\"text-gray-500\">Process images in a directory on the same machine where the server is running." +
+                                            "<br>Use an empty output directory to save pictures normally instead of writing to the output directory." +
+                                            f"<br>Add inpaint batch mask directory to enable inpaint batch processing."
+                                            f"{hidden}</p>"
+                                        )
+                                        img2img_batch_input_dir = gr.Textbox(label="Input directory", **shared.hide_dirs, elem_id="img2img_batch_input_dir")
+                                        img2img_batch_output_dir = gr.Textbox(label="Output directory", **shared.hide_dirs, elem_id="img2img_batch_output_dir")
+                                        img2img_batch_inpaint_mask_dir = gr.Textbox(label="Inpaint batch mask directory (required for inpaint batch processing only)", **shared.hide_dirs, elem_id="img2img_batch_inpaint_mask_dir")
+                                tab_batch_upload.select(fn=lambda: "upload", inputs=[], outputs=[img2img_batch_source_type])
+                                tab_batch_from_dir.select(fn=lambda: "from dir", inputs=[], outputs=[img2img_batch_source_type])
+                                with gr.Accordion("PNG info", open=False):
+                                    img2img_batch_use_png_info = gr.Checkbox(label="Append png info to prompts", elem_id="img2img_batch_use_png_info")
+                                    img2img_batch_png_info_dir = gr.Textbox(label="PNG info directory", **shared.hide_dirs, placeholder="Leave empty to use input directory", elem_id="img2img_batch_png_info_dir")
+                                    img2img_batch_png_info_props = gr.CheckboxGroup(["Prompt", "Negative prompt", "Seed", "CFG scale", "Sampler", "Steps", "Model hash"], label="Parameters to take from png info", info="Prompts from png info will be appended to prompts set in ui.")
+
+                            img2img_tabs = [tab_img2img, tab_sketch, tab_inpaint, tab_inpaint_color, tab_inpaint_upload, tab_batch]
+
+                            for i, tab in enumerate(img2img_tabs):
+                                tab.select(fn=lambda tabnum=i: tabnum, inputs=[], outputs=[img2img_selected_tab])
+
+                        def copy_image(img):
+                            if isinstance(img, dict) and 'image' in img:
+                                return img['image']
+
+                            return img
+
+                        for button, name, elem in copy_image_buttons:
+                            button.click(
+                                fn=copy_image,
+                                inputs=[elem],
+                                outputs=[copy_image_destinations[name]],
+                            )
+                            button.click(
+                                fn=lambda: None,
+                                _js=f"switch_to_{name.replace(' ', '_')}",
+                                inputs=[],
+                                outputs=[],
+                            )
+
+                        with FormRow():
+                            resize_mode = gr.Radio(label="Resize mode", elem_id="resize_mode", choices=["Just resize", "Crop and resize", "Resize and fill", "Just resize (latent upscale)"], type="index", value="Just resize")
 
                     elif category == "dimensions":
                         with FormRow():
                             with gr.Column(elem_id="img2img_column_size", scale=4):
-                                selected_scale_tab = gr.State(value=0)
+                                selected_scale_tab = gr.Number(value=0, visible=False)
 
-                                with gr.Tabs():
-                                    with gr.Tab(label="Resize to", elem_id="img2img_tab_resize_to") as tab_scale_to:
+                                with gr.Tabs(elem_id="img2img_tabs_resize"):
+                                    with gr.Tab(label="Resize to", id="to", elem_id="img2img_tab_resize_to") as tab_scale_to:
                                         with FormRow():
                                             with gr.Column(elem_id="img2img_column_size", scale=4):
                                                 width = gr.Slider(minimum=64, maximum=2048, step=8, label="Width", value=512, elem_id="img2img_width")
                                                 height = gr.Slider(minimum=64, maximum=2048, step=8, label="Height", value=512, elem_id="img2img_height")
                                             with gr.Column(elem_id="img2img_dimensions_row", scale=1, elem_classes="dimensions-tools"):
-                                                res_switch_btn = ToolButton(value=switch_values_symbol, elem_id="img2img_res_switch_btn")
-                                                detect_image_size_btn = ToolButton(value=detect_image_size_symbol, elem_id="img2img_detect_image_size_btn")
+                                                res_switch_btn = ToolButton(value=switch_values_symbol, elem_id="img2img_res_switch_btn", tooltip="Switch width/height")
+                                                detect_image_size_btn = ToolButton(value=detect_image_size_symbol, elem_id="img2img_detect_image_size_btn", tooltip="Auto detect size from img2img")
 
-                                    with gr.Tab(label="Resize by", elem_id="img2img_tab_resize_by") as tab_scale_by:
+                                    with gr.Tab(label="Resize by", id="by", elem_id="img2img_tab_resize_by") as tab_scale_by:
                                         scale_by = gr.Slider(minimum=0.05, maximum=4.0, step=0.05, label="Scale", value=1.0, elem_id="img2img_scale")
 
                                         with FormRow():
@@ -683,12 +651,6 @@ def copy_image(img):
                                     scale_by.release(**on_change_args)
                                     button_update_resize_to.click(**on_change_args)
 
-                                    # the code below is meant to update the resolution label after the image in the image selection UI has changed.
-                                    # as it is now the event keeps firing continuously for inpaint edits, which ruins the page with constant requests.
-                                    # I assume this must be a gradio bug and for now we'll just do it for non-inpaint inputs.
-                                    for component in [init_img, sketch]:
-                                        component.change(fn=lambda: None, _js="updateImg2imgResizeToTextAfterChangingImage", inputs=[], outputs=[], show_progress=False)
-
                             tab_scale_to.select(fn=lambda: 0, inputs=[], outputs=[selected_scale_tab])
                             tab_scale_by.select(fn=lambda: 1, inputs=[], outputs=[selected_scale_tab])
 
@@ -746,20 +708,26 @@ def copy_image(img):
                                 with gr.Column(scale=4):
                                     inpaint_full_res_padding = gr.Slider(label='Only masked padding, pixels', minimum=0, maximum=256, step=4, value=32, elem_id="img2img_inpaint_full_res_padding")
 
-                            def select_img2img_tab(tab):
-                                return gr.update(visible=tab in [2, 3, 4]), gr.update(visible=tab == 3),
-
-                            for i, elem in enumerate(img2img_tabs):
-                                elem.select(
-                                    fn=lambda tab=i: select_img2img_tab(tab),
-                                    inputs=[],
-                                    outputs=[inpaint_controls, mask_alpha],
-                                )
-
                     if category not in {"accordions"}:
                         scripts.scripts_img2img.setup_ui_for_section(category)
 
-            img2img_gallery, generation_info, html_info, html_log = create_output_panel("img2img", opts.outdir_img2img_samples)
+            # the code below is meant to update the resolution label after the image in the image selection UI has changed.
+            # as it is now the event keeps firing continuously for inpaint edits, which ruins the page with constant requests.
+            # I assume this must be a gradio bug and for now we'll just do it for non-inpaint inputs.
+            for component in [init_img, sketch]:
+                component.change(fn=lambda: None, _js="updateImg2imgResizeToTextAfterChangingImage", inputs=[], outputs=[], show_progress=False)
+
+            def select_img2img_tab(tab):
+                return gr.update(visible=tab in [2, 3, 4]), gr.update(visible=tab == 3),
+
+            for i, elem in enumerate(img2img_tabs):
+                elem.select(
+                    fn=lambda tab=i: select_img2img_tab(tab),
+                    inputs=[],
+                    outputs=[inpaint_controls, mask_alpha],
+                )
+
+            output_panel = create_output_panel("img2img", opts.outdir_img2img_samples, toprow)
 
             img2img_args = dict(
                 fn=wrap_gradio_gpu_call(modules.img2img.img2img, extra_outputs=[None, '', '']),
@@ -777,8 +745,6 @@ def select_img2img_tab(tab):
                     inpaint_color_sketch_orig,
                     init_img_inpaint,
                     init_mask_inpaint,
-                    steps,
-                    sampler_name,
                     mask_blur,
                     mask_alpha,
                     inpainting_fill,
@@ -802,12 +768,14 @@ def select_img2img_tab(tab):
                     img2img_batch_use_png_info,
                     img2img_batch_png_info_props,
                     img2img_batch_png_info_dir,
+                    img2img_batch_source_type,
+                    img2img_batch_upload,
                 ] + custom_inputs,
                 outputs=[
-                    img2img_gallery,
-                    generation_info,
-                    html_info,
-                    html_log,
+                    output_panel.gallery,
+                    output_panel.generation_info,
+                    output_panel.infotext,
+                    output_panel.html_log,
                 ],
                 show_progress=False,
             )
@@ -845,10 +813,10 @@ def select_img2img_tab(tab):
                 _js="restoreProgressImg2img",
                 inputs=[dummy_component],
                 outputs=[
-                    img2img_gallery,
-                    generation_info,
-                    html_info,
-                    html_log,
+                    output_panel.gallery,
+                    output_panel.generation_info,
+                    output_panel.infotext,
+                    output_panel.html_log,
                 ],
                 show_progress=False,
             )
@@ -863,14 +831,16 @@ def select_img2img_tab(tab):
                 **interrogate_args,
             )
 
-            toprow.token_button.click(fn=update_token_counter, inputs=[toprow.prompt, steps], outputs=[toprow.token_counter])
-            toprow.negative_token_button.click(fn=wrap_queued_call(update_token_counter), inputs=[toprow.negative_prompt, steps], outputs=[toprow.negative_token_counter])
+            steps = scripts.scripts_img2img.script('Sampler').steps
+
+            toprow.ui_styles.dropdown.change(fn=wrap_queued_call(update_token_counter), inputs=[toprow.prompt, steps, toprow.ui_styles.dropdown], outputs=[toprow.token_counter])
+            toprow.ui_styles.dropdown.change(fn=wrap_queued_call(update_negative_prompt_token_counter), inputs=[toprow.negative_prompt, steps, toprow.ui_styles.dropdown], outputs=[toprow.negative_token_counter])
+            toprow.token_button.click(fn=update_token_counter, inputs=[toprow.prompt, steps, toprow.ui_styles.dropdown], outputs=[toprow.token_counter])
+            toprow.negative_token_button.click(fn=wrap_queued_call(update_negative_prompt_token_counter), inputs=[toprow.negative_prompt, steps, toprow.ui_styles.dropdown], outputs=[toprow.negative_token_counter])
 
             img2img_paste_fields = [
                 (toprow.prompt, "Prompt"),
                 (toprow.negative_prompt, "Negative prompt"),
-                (steps, "Steps"),
-                (sampler_name, "Sampler"),
                 (cfg_scale, "CFG scale"),
                 (image_cfg_scale, "Image CFG scale"),
                 (width, "Size-1"),
@@ -879,6 +849,10 @@ def select_img2img_tab(tab):
                 (toprow.ui_styles.dropdown, lambda d: d["Styles array"] if isinstance(d.get("Styles array"), list) else gr.update()),
                 (denoising_strength, "Denoising strength"),
                 (mask_blur, "Mask blur"),
+                (inpainting_mask_invert, 'Mask mode'),
+                (inpainting_fill, 'Masked content'),
+                (inpaint_full_res, 'Inpaint area'),
+                (inpaint_full_res_padding, 'Masked area padding'),
                 *scripts.scripts_img2img.infotext_fields
             ]
             parameters_copypaste.add_paste_fields("img2img", init_img, img2img_paste_fields, override_settings)
@@ -888,7 +862,7 @@ def select_img2img_tab(tab):
             ))
 
         extra_networks_ui_img2img = ui_extra_networks.create_ui(img2img_interface, [img2img_generation_tab], 'img2img')
-        ui_extra_networks.setup_ui(extra_networks_ui_img2img, img2img_gallery)
+        ui_extra_networks.setup_ui(extra_networks_ui_img2img, output_panel.gallery)
 
         extra_tabs.__exit__()
 
@@ -898,7 +872,7 @@ def select_img2img_tab(tab):
         ui_postprocessing.create_ui()
 
     with gr.Blocks(analytics_enabled=False) as pnginfo_interface:
-        with gr.Row(equal_height=False):
+        with ResizeHandleRow(equal_height=False):
             with gr.Column(variant='panel'):
                 image = gr.Image(elem_id="pnginfo_image", label="Source", source="upload", interactive=True, type="pil")
 
@@ -915,7 +889,7 @@ def select_img2img_tab(tab):
                     ))
 
         image.change(
-            fn=wrap_gradio_call(modules.extras.run_pnginfo),
+            fn=wrap_gradio_call_no_job(modules.extras.run_pnginfo),
             inputs=[image],
             outputs=[html, generation_info, html2],
         )
@@ -926,7 +900,7 @@ def select_img2img_tab(tab):
         with gr.Row(equal_height=False):
             gr.HTML(value="<p style='margin-bottom: 0.7em'>See <b><a href=\"https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Textual-Inversion\">wiki</a></b> for detailed explanation.</p>")
 
-        with gr.Row(variant="compact", equal_height=False):
+        with ResizeHandleRow(variant="compact", equal_height=False):
             with gr.Tabs(elem_id="train_tabs"):
 
                 with gr.Tab(label="Create embedding", id="create_embedding"):
@@ -960,71 +934,6 @@ def select_img2img_tab(tab):
                         with gr.Column():
                             create_hypernetwork = gr.Button(value="Create hypernetwork", variant='primary', elem_id="train_create_hypernetwork")
 
-                with gr.Tab(label="Preprocess images", id="preprocess_images"):
-                    process_src = gr.Textbox(label='Source directory', elem_id="train_process_src")
-                    process_dst = gr.Textbox(label='Destination directory', elem_id="train_process_dst")
-                    process_width = gr.Slider(minimum=64, maximum=2048, step=8, label="Width", value=512, elem_id="train_process_width")
-                    process_height = gr.Slider(minimum=64, maximum=2048, step=8, label="Height", value=512, elem_id="train_process_height")
-                    preprocess_txt_action = gr.Dropdown(label='Existing Caption txt Action', value="ignore", choices=["ignore", "copy", "prepend", "append"], elem_id="train_preprocess_txt_action")
-
-                    with gr.Row():
-                        process_keep_original_size = gr.Checkbox(label='Keep original size', elem_id="train_process_keep_original_size")
-                        process_flip = gr.Checkbox(label='Create flipped copies', elem_id="train_process_flip")
-                        process_split = gr.Checkbox(label='Split oversized images', elem_id="train_process_split")
-                        process_focal_crop = gr.Checkbox(label='Auto focal point crop', elem_id="train_process_focal_crop")
-                        process_multicrop = gr.Checkbox(label='Auto-sized crop', elem_id="train_process_multicrop")
-                        process_caption = gr.Checkbox(label='Use BLIP for caption', elem_id="train_process_caption")
-                        process_caption_deepbooru = gr.Checkbox(label='Use deepbooru for caption', visible=True, elem_id="train_process_caption_deepbooru")
-
-                    with gr.Row(visible=False) as process_split_extra_row:
-                        process_split_threshold = gr.Slider(label='Split image threshold', value=0.5, minimum=0.0, maximum=1.0, step=0.05, elem_id="train_process_split_threshold")
-                        process_overlap_ratio = gr.Slider(label='Split image overlap ratio', value=0.2, minimum=0.0, maximum=0.9, step=0.05, elem_id="train_process_overlap_ratio")
-
-                    with gr.Row(visible=False) as process_focal_crop_row:
-                        process_focal_crop_face_weight = gr.Slider(label='Focal point face weight', value=0.9, minimum=0.0, maximum=1.0, step=0.05, elem_id="train_process_focal_crop_face_weight")
-                        process_focal_crop_entropy_weight = gr.Slider(label='Focal point entropy weight', value=0.15, minimum=0.0, maximum=1.0, step=0.05, elem_id="train_process_focal_crop_entropy_weight")
-                        process_focal_crop_edges_weight = gr.Slider(label='Focal point edges weight', value=0.5, minimum=0.0, maximum=1.0, step=0.05, elem_id="train_process_focal_crop_edges_weight")
-                        process_focal_crop_debug = gr.Checkbox(label='Create debug image', elem_id="train_process_focal_crop_debug")
-
-                    with gr.Column(visible=False) as process_multicrop_col:
-                        gr.Markdown('Each image is center-cropped with an automatically chosen width and height.')
-                        with gr.Row():
-                            process_multicrop_mindim = gr.Slider(minimum=64, maximum=2048, step=8, label="Dimension lower bound", value=384, elem_id="train_process_multicrop_mindim")
-                            process_multicrop_maxdim = gr.Slider(minimum=64, maximum=2048, step=8, label="Dimension upper bound", value=768, elem_id="train_process_multicrop_maxdim")
-                        with gr.Row():
-                            process_multicrop_minarea = gr.Slider(minimum=64*64, maximum=2048*2048, step=1, label="Area lower bound", value=64*64, elem_id="train_process_multicrop_minarea")
-                            process_multicrop_maxarea = gr.Slider(minimum=64*64, maximum=2048*2048, step=1, label="Area upper bound", value=640*640, elem_id="train_process_multicrop_maxarea")
-                        with gr.Row():
-                            process_multicrop_objective = gr.Radio(["Maximize area", "Minimize error"], value="Maximize area", label="Resizing objective", elem_id="train_process_multicrop_objective")
-                            process_multicrop_threshold = gr.Slider(minimum=0, maximum=1, step=0.01, label="Error threshold", value=0.1, elem_id="train_process_multicrop_threshold")
-
-                    with gr.Row():
-                        with gr.Column(scale=3):
-                            gr.HTML(value="")
-
-                        with gr.Column():
-                            with gr.Row():
-                                interrupt_preprocessing = gr.Button("Interrupt", elem_id="train_interrupt_preprocessing")
-                            run_preprocess = gr.Button(value="Preprocess", variant='primary', elem_id="train_run_preprocess")
-
-                    process_split.change(
-                        fn=lambda show: gr_show(show),
-                        inputs=[process_split],
-                        outputs=[process_split_extra_row],
-                    )
-
-                    process_focal_crop.change(
-                        fn=lambda show: gr_show(show),
-                        inputs=[process_focal_crop],
-                        outputs=[process_focal_crop_row],
-                    )
-
-                    process_multicrop.change(
-                        fn=lambda show: gr_show(show),
-                        inputs=[process_multicrop],
-                        outputs=[process_multicrop_col],
-                    )
-
                 def get_textual_inversion_template_names():
                     return sorted(textual_inversion.textual_inversion_templates)
 
@@ -1125,42 +1034,6 @@ def get_textual_inversion_template_names():
             ]
         )
 
-        run_preprocess.click(
-            fn=wrap_gradio_gpu_call(textual_inversion_ui.preprocess, extra_outputs=[gr.update()]),
-            _js="start_training_textual_inversion",
-            inputs=[
-                dummy_component,
-                process_src,
-                process_dst,
-                process_width,
-                process_height,
-                preprocess_txt_action,
-                process_keep_original_size,
-                process_flip,
-                process_split,
-                process_caption,
-                process_caption_deepbooru,
-                process_split_threshold,
-                process_overlap_ratio,
-                process_focal_crop,
-                process_focal_crop_face_weight,
-                process_focal_crop_entropy_weight,
-                process_focal_crop_edges_weight,
-                process_focal_crop_debug,
-                process_multicrop,
-                process_multicrop_mindim,
-                process_multicrop_maxdim,
-                process_multicrop_minarea,
-                process_multicrop_maxarea,
-                process_multicrop_objective,
-                process_multicrop_threshold,
-            ],
-            outputs=[
-                ti_output,
-                ti_outcome,
-            ],
-        )
-
         train_embedding.click(
             fn=wrap_gradio_gpu_call(textual_inversion_ui.train_embedding, extra_outputs=[gr.update()]),
             _js="start_training_textual_inversion",
@@ -1234,15 +1107,9 @@ def get_textual_inversion_template_names():
             outputs=[],
         )
 
-        interrupt_preprocessing.click(
-            fn=lambda: shared.state.interrupt(),
-            inputs=[],
-            outputs=[],
-        )
-
     loadsave = ui_loadsave.UiLoadsave(cmd_opts.ui_config_file)
+    ui_settings_from_file = loadsave.ui_settings.copy()
 
-    settings = ui_settings.UiSettings()
     settings.create_ui(loadsave, dummy_component)
 
     interfaces = [
@@ -1286,7 +1153,7 @@ def get_textual_inversion_template_names():
 
             loadsave.setup_ui()
 
-        if os.path.exists(os.path.join(script_path, "notification.mp3")):
+        if os.path.exists(os.path.join(script_path, "notification.mp3")) and shared.opts.notification_audio:
             gr.Audio(interactive=False, value=os.path.join(script_path, "notification.mp3"), elem_id="audio_notification", visible=False)
 
         footer = shared.html("footer.html")
@@ -1301,7 +1168,8 @@ def get_textual_inversion_template_names():
 
         modelmerger_ui.setup_ui(dummy_component=dummy_component, sd_model_checkpoint_component=settings.component_dict['sd_model_checkpoint'])
 
-    loadsave.dump_defaults()
+    if ui_settings_from_file != loadsave.ui_settings:
+        loadsave.dump_defaults()
     demo.ui_loadsave = loadsave
 
     return demo
@@ -1338,7 +1206,6 @@ def versions_html():
 
 def setup_ui_api(app):
     from pydantic import BaseModel, Field
-    from typing import List
 
     class QuicksettingsHint(BaseModel):
         name: str = Field(title="Name of the quicksettings field")
@@ -1347,7 +1214,7 @@ class QuicksettingsHint(BaseModel):
     def quicksettings_hint():
         return [QuicksettingsHint(name=k, label=v.label) for k, v in opts.data_labels.items()]
 
-    app.add_api_route("/internal/quicksettings-hint", quicksettings_hint, methods=["GET"], response_model=List[QuicksettingsHint])
+    app.add_api_route("/internal/quicksettings-hint", quicksettings_hint, methods=["GET"], response_model=list[QuicksettingsHint])
 
     app.add_api_route("/internal/ping", lambda: {}, methods=["GET"])
 
@@ -1357,10 +1224,12 @@ def download_sysinfo(attachment=False):
         from fastapi.responses import PlainTextResponse
 
         text = sysinfo.get()
-        filename = f"sysinfo-{datetime.datetime.utcnow().strftime('%Y-%m-%d-%H-%M')}.txt"
+        filename = f"sysinfo-{datetime.datetime.utcnow().strftime('%Y-%m-%d-%H-%M')}.json"
 
         return PlainTextResponse(text, headers={'Content-Disposition': f'{"attachment" if attachment else "inline"}; filename="{filename}"'})
 
     app.add_api_route("/internal/sysinfo", download_sysinfo, methods=["GET"])
     app.add_api_route("/internal/sysinfo-download", lambda: download_sysinfo(attachment=True), methods=["GET"])
 
+    import fastapi.staticfiles
+    app.mount("/webui-assets", fastapi.staticfiles.StaticFiles(directory=launch_utils.repo_dir('stable-diffusion-webui-assets')), name="webui-assets")
diff --git a/modules/ui_common.py b/modules/ui_common.py
index 84a7d7f2756..395bb3b61ee 100644
--- a/modules/ui_common.py
+++ b/modules/ui_common.py
@@ -1,17 +1,17 @@
+import csv
+import dataclasses
 import json
 import html
 import os
-import platform
-import sys
+from contextlib import nullcontext
 
 import gradio as gr
-import subprocess as sp
 
-from modules import call_queue, shared
-from modules.generation_parameters_copypaste import image_from_url_text
+from modules import call_queue, shared, ui_tempdir, util
+from modules.infotext_utils import image_from_url_text
 import modules.images
 from modules.ui_components import ToolButton
-import modules.generation_parameters_copypaste as parameters_copypaste
+import modules.infotext_utils as parameters_copypaste
 
 folder_symbol = '\U0001f4c2'  # 📂
 refresh_symbol = '\U0001f504'  # 🔄
@@ -35,12 +35,38 @@ def plaintext_to_html(text, classname=None):
     return f"<p class='{classname}'>{content}</p>" if classname else f"<p>{content}</p>"
 
 
+def update_logfile(logfile_path, fields):
+    """Update a logfile from old format to new format to maintain CSV integrity."""
+    with open(logfile_path, "r", encoding="utf8", newline="") as file:
+        reader = csv.reader(file)
+        rows = list(reader)
+
+    # blank file: leave it as is
+    if not rows:
+        return
+
+    # file is already synced, do nothing
+    if len(rows[0]) == len(fields):
+        return
+
+    rows[0] = fields
+
+    # append new fields to each row as empty values
+    for row in rows[1:]:
+        while len(row) < len(fields):
+            row.append("")
+
+    with open(logfile_path, "w", encoding="utf8", newline="") as file:
+        writer = csv.writer(file)
+        writer.writerows(rows)
+
+
 def save_files(js_data, images, do_make_zip, index):
-    import csv
     filenames = []
     fullfns = []
+    parsed_infotexts = []
 
-    #quick dictionary to class object conversion. Its necessary due apply_filename_pattern requiring it
+    # quick dictionary to class object conversion. Its necessary due apply_filename_pattern requiring it
     class MyObject:
         def __init__(self, d=None):
             if d is not None:
@@ -48,35 +74,56 @@ def __init__(self, d=None):
                     setattr(self, key, value)
 
     data = json.loads(js_data)
-
     p = MyObject(data)
+
     path = shared.opts.outdir_save
     save_to_dirs = shared.opts.use_save_to_dirs_for_ui
     extension: str = shared.opts.samples_format
     start_index = 0
-    only_one = False
 
     if index > -1 and shared.opts.save_selected_only and (index >= data["index_of_first_image"]):  # ensures we are looking at a specific non-grid picture, and we have save_selected_only
-        only_one = True
         images = [images[index]]
         start_index = index
 
     os.makedirs(shared.opts.outdir_save, exist_ok=True)
 
-    with open(os.path.join(shared.opts.outdir_save, "log.csv"), "a", encoding="utf8", newline='') as file:
-        at_start = file.tell() == 0
-        writer = csv.writer(file)
-        if at_start:
-            writer.writerow(["prompt", "seed", "width", "height", "sampler", "cfgs", "steps", "filename", "negative_prompt"])
+    fields = [
+        "prompt",
+        "seed",
+        "width",
+        "height",
+        "sampler",
+        "cfgs",
+        "steps",
+        "filename",
+        "negative_prompt",
+        "sd_model_name",
+        "sd_model_hash",
+    ]
+    logfile_path = os.path.join(shared.opts.outdir_save, "log.csv")
+
+    # NOTE: ensure csv integrity when fields are added by
+    # updating headers and padding with delimiters where needed
+    if shared.opts.save_write_log_csv and os.path.exists(logfile_path):
+        update_logfile(logfile_path, fields)
+
+    with (open(logfile_path, "a", encoding="utf8", newline='') if shared.opts.save_write_log_csv else nullcontext()) as file:
+        if file:
+            at_start = file.tell() == 0
+            writer = csv.writer(file)
+            if at_start:
+                writer.writerow(fields)
 
         for image_index, filedata in enumerate(images, start_index):
             image = image_from_url_text(filedata)
 
             is_grid = image_index < p.index_of_first_image
-            i = 0 if is_grid else (image_index - p.index_of_first_image)
 
             p.batch_index = image_index-1
-            fullfn, txt_fullfn = modules.images.save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs)
+
+            parameters = parameters_copypaste.parse_generation_parameters(data["infotexts"][image_index], [])
+            parsed_infotexts.append(parameters)
+            fullfn, txt_fullfn = modules.images.save_image(image, path, "", seed=parameters['Seed'], prompt=parameters['Prompt'], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs)
 
             filename = os.path.relpath(fullfn, path)
             filenames.append(filename)
@@ -85,12 +132,13 @@ def __init__(self, d=None):
                 filenames.append(os.path.basename(txt_fullfn))
                 fullfns.append(txt_fullfn)
 
-        writer.writerow([data["prompt"], data["seed"], data["width"], data["height"], data["sampler_name"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]])
+        if file:
+            writer.writerow([parsed_infotexts[0]['Prompt'], parsed_infotexts[0]['Seed'], data["width"], data["height"], data["sampler_name"], data["cfg_scale"], data["steps"], filenames[0], parsed_infotexts[0]['Negative prompt'], data["sd_model_name"], data["sd_model_hash"]])
 
     # Make Zip
     if do_make_zip:
-        zip_fileseed = p.all_seeds[index-1] if only_one else p.all_seeds[0]
-        namegen = modules.images.FilenameGenerator(p, zip_fileseed, p.all_prompts[0], image, True)
+        p.all_seeds = [parameters['Seed'] for parameters in parsed_infotexts]
+        namegen = modules.images.FilenameGenerator(p, parsed_infotexts[0]['Seed'], parsed_infotexts[0]['Prompt'], image, True)
         zip_filename = namegen.apply(shared.opts.grid_zip_filename_pattern or "[datetime]_[[model_name]]_[seed]-[seed_last]")
         zip_filepath = os.path.join(path, f"{zip_filename}.zip")
 
@@ -104,38 +152,40 @@ def __init__(self, d=None):
     return gr.File.update(value=fullfns, visible=True), plaintext_to_html(f"Saved: {filenames[0]}")
 
 
-def create_output_panel(tabname, outdir):
+@dataclasses.dataclass
+class OutputPanel:
+    gallery = None
+    generation_info = None
+    infotext = None
+    html_log = None
+    button_upscale = None
 
-    def open_folder(f):
-        if not os.path.exists(f):
-            print(f'Folder "{f}" does not exist. After you create an image, the folder will be created.')
-            return
-        elif not os.path.isdir(f):
-            print(f"""
-WARNING
-An open_folder request was made with an argument that is not a folder.
-This could be an error or a malicious attempt to run code on your computer.
-Requested path was: {f}
-""", file=sys.stderr)
+
+def create_output_panel(tabname, outdir, toprow=None):
+    res = OutputPanel()
+
+    def open_folder(f, images=None, index=None):
+        if shared.cmd_opts.hide_ui_dir_config:
             return
 
-        if not shared.cmd_opts.hide_ui_dir_config:
-            path = os.path.normpath(f)
-            if platform.system() == "Windows":
-                os.startfile(path)
-            elif platform.system() == "Darwin":
-                sp.Popen(["open", path])
-            elif "microsoft-standard-WSL2" in platform.uname().release:
-                sp.Popen(["wsl-open", path])
-            else:
-                sp.Popen(["xdg-open", path])
+        try:
+            if 'Sub' in shared.opts.open_dir_button_choice:
+                image_dir = os.path.split(images[index]["name"].rsplit('?', 1)[0])[0]
+                if 'temp' in shared.opts.open_dir_button_choice or not ui_tempdir.is_gradio_temp_path(image_dir):
+                    f = image_dir
+        except Exception:
+            pass
+
+        util.open_folder(f)
 
-    with gr.Column(variant='panel', elem_id=f"{tabname}_results"):
-        with gr.Group(elem_id=f"{tabname}_gallery_container"):
-            result_gallery = gr.Gallery(label='Output', show_label=False, elem_id=f"{tabname}_gallery", columns=4, preview=True, height=shared.opts.gallery_height or None)
+    with gr.Column(elem_id=f"{tabname}_results"):
+        if toprow:
+            toprow.create_inline_toprow_image()
+
+        with gr.Column(variant='panel', elem_id=f"{tabname}_results_panel"):
+            with gr.Group(elem_id=f"{tabname}_gallery_container"):
+                res.gallery = gr.Gallery(label='Output', show_label=False, elem_id=f"{tabname}_gallery", columns=4, preview=True, height=shared.opts.gallery_height or None)
 
-        generation_info = None
-        with gr.Column():
             with gr.Row(elem_id=f"image_buttons_{tabname}", elem_classes="image-buttons"):
                 open_folder_button = ToolButton(folder_symbol, elem_id=f'{tabname}_open_folder', visible=not shared.cmd_opts.hide_ui_dir_config, tooltip="Open images output directory.")
 
@@ -149,9 +199,16 @@ def open_folder(f):
                     'extras': ToolButton('📐', elem_id=f'{tabname}_send_to_extras', tooltip="Send image and generation parameters to extras tab.")
                 }
 
+                if tabname == 'txt2img':
+                    res.button_upscale = ToolButton('✨', elem_id=f'{tabname}_upscale', tooltip="Create an upscaled version of the current image using hires fix settings.")
+
             open_folder_button.click(
-                fn=lambda: open_folder(shared.opts.outdir_samples or outdir),
-                inputs=[],
+                fn=lambda images, index: open_folder(shared.opts.outdir_samples or outdir, images, index),
+                _js="(y, w) => [y, selected_gallery_index()]",
+                inputs=[
+                    res.gallery,
+                    open_folder_button,  # placeholder for index
+                ],
                 outputs=[],
             )
 
@@ -159,55 +216,55 @@ def open_folder(f):
                 download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False, elem_id=f'download_files_{tabname}')
 
                 with gr.Group():
-                    html_info = gr.HTML(elem_id=f'html_info_{tabname}', elem_classes="infotext")
-                    html_log = gr.HTML(elem_id=f'html_log_{tabname}', elem_classes="html-log")
+                    res.infotext = gr.HTML(elem_id=f'html_info_{tabname}', elem_classes="infotext")
+                    res.html_log = gr.HTML(elem_id=f'html_log_{tabname}', elem_classes="html-log")
 
-                    generation_info = gr.Textbox(visible=False, elem_id=f'generation_info_{tabname}')
+                    res.generation_info = gr.Textbox(visible=False, elem_id=f'generation_info_{tabname}')
                     if tabname == 'txt2img' or tabname == 'img2img':
                         generation_info_button = gr.Button(visible=False, elem_id=f"{tabname}_generation_info_button")
                         generation_info_button.click(
                             fn=update_generation_info,
                             _js="function(x, y, z){ return [x, y, selected_gallery_index()] }",
-                            inputs=[generation_info, html_info, html_info],
-                            outputs=[html_info, html_info],
+                            inputs=[res.generation_info, res.infotext, res.infotext],
+                            outputs=[res.infotext, res.infotext],
                             show_progress=False,
                         )
 
                     save.click(
-                        fn=call_queue.wrap_gradio_call(save_files),
+                        fn=call_queue.wrap_gradio_call_no_job(save_files),
                         _js="(x, y, z, w) => [x, y, false, selected_gallery_index()]",
                         inputs=[
-                            generation_info,
-                            result_gallery,
-                            html_info,
-                            html_info,
+                            res.generation_info,
+                            res.gallery,
+                            res.infotext,
+                            res.infotext,
                         ],
                         outputs=[
                             download_files,
-                            html_log,
+                            res.html_log,
                         ],
                         show_progress=False,
                     )
 
                     save_zip.click(
-                        fn=call_queue.wrap_gradio_call(save_files),
+                        fn=call_queue.wrap_gradio_call_no_job(save_files),
                         _js="(x, y, z, w) => [x, y, true, selected_gallery_index()]",
                         inputs=[
-                            generation_info,
-                            result_gallery,
-                            html_info,
-                            html_info,
+                            res.generation_info,
+                            res.gallery,
+                            res.infotext,
+                            res.infotext,
                         ],
                         outputs=[
                             download_files,
-                            html_log,
+                            res.html_log,
                         ]
                     )
 
             else:
-                html_info_x = gr.HTML(elem_id=f'html_info_x_{tabname}')
-                html_info = gr.HTML(elem_id=f'html_info_{tabname}', elem_classes="infotext")
-                html_log = gr.HTML(elem_id=f'html_log_{tabname}')
+                res.generation_info = gr.HTML(elem_id=f'html_info_x_{tabname}')
+                res.infotext = gr.HTML(elem_id=f'html_info_{tabname}', elem_classes="infotext")
+                res.html_log = gr.HTML(elem_id=f'html_log_{tabname}')
 
             paste_field_names = []
             if tabname == "txt2img":
@@ -217,11 +274,11 @@ def open_folder(f):
 
             for paste_tabname, paste_button in buttons.items():
                 parameters_copypaste.register_paste_params_button(parameters_copypaste.ParamBinding(
-                    paste_button=paste_button, tabname=paste_tabname, source_tabname="txt2img" if tabname == "txt2img" else None, source_image_component=result_gallery,
+                    paste_button=paste_button, tabname=paste_tabname, source_tabname="txt2img" if tabname == "txt2img" else None, source_image_component=res.gallery,
                     paste_field_names=paste_field_names
                 ))
 
-            return result_gallery, generation_info if tabname != "extras" else html_info_x, html_info, html_log
+    return res
 
 
 def create_refresh_button(refresh_component, refresh_method, refreshed_args, elem_id):
diff --git a/modules/ui_components.py b/modules/ui_components.py
index 55979f62629..9cf67722a3d 100644
--- a/modules/ui_components.py
+++ b/modules/ui_components.py
@@ -88,7 +88,7 @@ def get_block_name(self):
 class InputAccordion(gr.Checkbox):
     """A gr.Accordion that can be used as an input - returns True if open, False if closed.
 
-    Actaully just a hidden checkbox, but creates an accordion that follows and is followed by the state of the checkbox.
+    Actually just a hidden checkbox, but creates an accordion that follows and is followed by the state of the checkbox.
     """
 
     global_index = 0
diff --git a/modules/ui_extensions.py b/modules/ui_extensions.py
index 2e8c1d6d21d..23aff709627 100644
--- a/modules/ui_extensions.py
+++ b/modules/ui_extensions.py
@@ -58,14 +58,15 @@ def apply_and_restart(disable_list, update_list, disable_all):
 
 def save_config_state(name):
     current_config_state = config_states.get_config()
-    if not name:
-        name = "Config"
+
+    name = os.path.basename(name or "Config")
+
     current_config_state["name"] = name
     timestamp = datetime.now().strftime('%Y_%m_%d-%H_%M_%S')
     filename = os.path.join(config_states_dir, f"{timestamp}_{name}.json")
     print(f"Saving backup of webui/extension state to {filename}.")
     with open(filename, "w", encoding="utf-8") as f:
-        json.dump(current_config_state, f, indent=4)
+        json.dump(current_config_state, f, indent=4, ensure_ascii=False)
     config_states.list_config_states()
     new_value = next(iter(config_states.all_config_states.keys()), "Current")
     new_choices = ["Current"] + list(config_states.all_config_states.keys())
@@ -197,7 +198,7 @@ def update_config_states_table(state_name):
         config_state = config_states.all_config_states[state_name]
 
     config_name = config_state.get("name", "Config")
-    created_date = time.asctime(time.gmtime(config_state["created_at"]))
+    created_date = datetime.fromtimestamp(config_state["created_at"]).strftime('%Y-%m-%d %H:%M:%S')
     filepath = config_state.get("filepath", "<unknown>")
 
     try:
@@ -335,6 +336,11 @@ def normalize_git_url(url):
     return url
 
 
+def get_extension_dirname_from_url(url):
+    *parts, last_part = url.split('/')
+    return normalize_git_url(last_part)
+
+
 def install_extension_from_url(dirname, url, branch_name=None):
     check_access()
 
@@ -346,10 +352,7 @@ def install_extension_from_url(dirname, url, branch_name=None):
     assert url, 'No URL specified'
 
     if dirname is None or dirname == "":
-        *parts, last_part = url.split('/')
-        last_part = normalize_git_url(last_part)
-
-        dirname = last_part
+        dirname = get_extension_dirname_from_url(url)
 
     target_dir = os.path.join(extensions.extensions_dir, dirname)
     assert not os.path.exists(target_dir), f'Extension directory already exists: {target_dir}'
@@ -378,7 +381,7 @@ def install_extension_from_url(dirname, url, branch_name=None):
         except OSError as err:
             if err.errno == errno.EXDEV:
                 # Cross device link, typical in docker or when tmp/ and extensions/ are on different file systems
-                # Since we can't use a rename, do the slower but more versitile shutil.move()
+                # Since we can't use a rename, do the slower but more versatile shutil.move()
                 shutil.move(tmpdir, target_dir)
             else:
                 # Something else, not enough free space, permissions, etc.  rethrow it so that it gets handled.
@@ -393,15 +396,15 @@ def install_extension_from_url(dirname, url, branch_name=None):
         shutil.rmtree(tmpdir, True)
 
 
-def install_extension_from_index(url, hide_tags, sort_column, filter_text):
+def install_extension_from_index(url, selected_tags, showing_type, filtering_type, sort_column, filter_text):
     ext_table, message = install_extension_from_url(None, url)
 
-    code, _ = refresh_available_extensions_from_data(hide_tags, sort_column, filter_text)
+    code, _ = refresh_available_extensions_from_data(selected_tags, showing_type, filtering_type, sort_column, filter_text)
 
     return code, ext_table, message, ''
 
 
-def refresh_available_extensions(url, hide_tags, sort_column):
+def refresh_available_extensions(url, selected_tags, showing_type, filtering_type, sort_column):
     global available_extensions
 
     import urllib.request
@@ -410,19 +413,19 @@ def refresh_available_extensions(url, hide_tags, sort_column):
 
     available_extensions = json.loads(text)
 
-    code, tags = refresh_available_extensions_from_data(hide_tags, sort_column)
+    code, tags = refresh_available_extensions_from_data(selected_tags, showing_type, filtering_type, sort_column)
 
     return url, code, gr.CheckboxGroup.update(choices=tags), '', ''
 
 
-def refresh_available_extensions_for_tags(hide_tags, sort_column, filter_text):
-    code, _ = refresh_available_extensions_from_data(hide_tags, sort_column, filter_text)
+def refresh_available_extensions_for_tags(selected_tags, showing_type, filtering_type, sort_column, filter_text):
+    code, _ = refresh_available_extensions_from_data(selected_tags, showing_type, filtering_type, sort_column, filter_text)
 
     return code, ''
 
 
-def search_extensions(filter_text, hide_tags, sort_column):
-    code, _ = refresh_available_extensions_from_data(hide_tags, sort_column, filter_text)
+def search_extensions(filter_text, selected_tags, showing_type, filtering_type, sort_column):
+    code, _ = refresh_available_extensions_from_data(selected_tags, showing_type, filtering_type, sort_column, filter_text)
 
     return code, ''
 
@@ -447,12 +450,13 @@ def get_date(info: dict, key):
         return ''
 
 
-def refresh_available_extensions_from_data(hide_tags, sort_column, filter_text=""):
+def refresh_available_extensions_from_data(selected_tags, showing_type, filtering_type, sort_column, filter_text=""):
     extlist = available_extensions["extensions"]
-    installed_extension_urls = {normalize_git_url(extension.remote): extension.name for extension in extensions.extensions}
+    installed_extensions = {extension.name for extension in extensions.extensions}
+    installed_extension_urls = {normalize_git_url(extension.remote) for extension in extensions.extensions if extension.remote is not None}
 
     tags = available_extensions.get("tags", {})
-    tags_to_hide = set(hide_tags)
+    selected_tags = set(selected_tags)
     hidden = 0
 
     code = f"""<!-- {time.time()} -->
@@ -482,12 +486,22 @@ def refresh_available_extensions_from_data(hide_tags, sort_column, filter_text="
         if url is None:
             continue
 
-        existing = installed_extension_urls.get(normalize_git_url(url), None)
+        existing = get_extension_dirname_from_url(url) in installed_extensions or normalize_git_url(url) in installed_extension_urls
         extension_tags = extension_tags + ["installed"] if existing else extension_tags
 
-        if any(x for x in extension_tags if x in tags_to_hide):
-            hidden += 1
-            continue
+        if len(selected_tags) > 0:
+            matched_tags = [x for x in extension_tags if x in selected_tags]
+            if filtering_type == 'or':
+                need_hide = len(matched_tags) > 0
+            else:
+                need_hide = len(matched_tags) == len(selected_tags)
+
+            if showing_type == 'show':
+                need_hide = not need_hide
+
+            if need_hide:
+                hidden += 1
+                continue
 
         if filter_text and filter_text.strip():
             if filter_text.lower() not in html.escape(name).lower() and filter_text.lower() not in html.escape(description).lower():
@@ -545,6 +559,7 @@ def create_ui():
                     extensions_disable_all = gr.Radio(label="Disable all extensions", choices=["none", "extra", "all"], value=shared.opts.disable_all_extensions, elem_id="extensions_disable_all")
                     extensions_disabled_list = gr.Text(elem_id="extensions_disabled_list", visible=False, container=False)
                     extensions_update_list = gr.Text(elem_id="extensions_update_list", visible=False, container=False)
+                    refresh = gr.Button(value='Refresh', variant="compact")
 
                 html = ""
 
@@ -563,7 +578,8 @@ def create_ui():
                 with gr.Row(elem_classes="progress-container"):
                     extensions_table = gr.HTML('Loading...', elem_id="extensions_installed_html")
 
-                ui.load(fn=extension_table, inputs=[], outputs=[extensions_table])
+                ui.load(fn=extension_table, inputs=[], outputs=[extensions_table], show_progress=False)
+                refresh.click(fn=extension_table, inputs=[], outputs=[extensions_table], show_progress=False)
 
                 apply.click(
                     fn=apply_and_restart,
@@ -588,8 +604,12 @@ def create_ui():
                     install_extension_button = gr.Button(elem_id="install_extension_button", visible=False)
 
                 with gr.Row():
-                    hide_tags = gr.CheckboxGroup(value=["ads", "localization", "installed"], label="Hide extensions with tags", choices=["script", "ads", "localization", "installed"])
-                    sort_column = gr.Radio(value="newest first", label="Order", choices=["newest first", "oldest first", "a-z", "z-a", "internal order",'update time', 'create time', "stars"], type="index")
+                    selected_tags = gr.CheckboxGroup(value=["ads", "localization", "installed"], label="Extension tags", choices=["script", "ads", "localization", "installed"], elem_classes=['compact-checkbox-group'])
+                    sort_column = gr.Radio(value="newest first", label="Order", choices=["newest first", "oldest first", "a-z", "z-a", "internal order",'update time', 'create time', "stars"], type="index", elem_classes=['compact-checkbox-group'])
+
+                with gr.Row():
+                    showing_type = gr.Radio(value="hide", label="Showing type", choices=["hide", "show"], elem_classes=['compact-checkbox-group'])
+                    filtering_type = gr.Radio(value="or", label="Filtering type", choices=["or", "and"], elem_classes=['compact-checkbox-group'])
 
                 with gr.Row():
                     search_extensions_text = gr.Text(label="Search", container=False)
@@ -599,31 +619,43 @@ def create_ui():
 
                 refresh_available_extensions_button.click(
                     fn=modules.ui.wrap_gradio_call(refresh_available_extensions, extra_outputs=[gr.update(), gr.update(), gr.update(), gr.update()]),
-                    inputs=[available_extensions_index, hide_tags, sort_column],
-                    outputs=[available_extensions_index, available_extensions_table, hide_tags, search_extensions_text, install_result],
+                    inputs=[available_extensions_index, selected_tags, showing_type, filtering_type, sort_column],
+                    outputs=[available_extensions_index, available_extensions_table, selected_tags, search_extensions_text, install_result],
                 )
 
                 install_extension_button.click(
-                    fn=modules.ui.wrap_gradio_call(install_extension_from_index, extra_outputs=[gr.update(), gr.update()]),
-                    inputs=[extension_to_install, hide_tags, sort_column, search_extensions_text],
+                    fn=modules.ui.wrap_gradio_call_no_job(install_extension_from_index, extra_outputs=[gr.update(), gr.update()]),
+                    inputs=[extension_to_install, selected_tags, showing_type, filtering_type, sort_column, search_extensions_text],
                     outputs=[available_extensions_table, extensions_table, install_result],
                 )
 
                 search_extensions_text.change(
-                    fn=modules.ui.wrap_gradio_call(search_extensions, extra_outputs=[gr.update()]),
-                    inputs=[search_extensions_text, hide_tags, sort_column],
+                    fn=modules.ui.wrap_gradio_call_no_job(search_extensions, extra_outputs=[gr.update()]),
+                    inputs=[search_extensions_text, selected_tags, showing_type, filtering_type, sort_column],
                     outputs=[available_extensions_table, install_result],
                 )
 
-                hide_tags.change(
-                    fn=modules.ui.wrap_gradio_call(refresh_available_extensions_for_tags, extra_outputs=[gr.update()]),
-                    inputs=[hide_tags, sort_column, search_extensions_text],
+                selected_tags.change(
+                    fn=modules.ui.wrap_gradio_call_no_job(refresh_available_extensions_for_tags, extra_outputs=[gr.update()]),
+                    inputs=[selected_tags, showing_type, filtering_type, sort_column, search_extensions_text],
+                    outputs=[available_extensions_table, install_result]
+                )
+
+                showing_type.change(
+                    fn=modules.ui.wrap_gradio_call_no_job(refresh_available_extensions_for_tags, extra_outputs=[gr.update()]),
+                    inputs=[selected_tags, showing_type, filtering_type, sort_column, search_extensions_text],
+                    outputs=[available_extensions_table, install_result]
+                )
+
+                filtering_type.change(
+                    fn=modules.ui.wrap_gradio_call_no_job(refresh_available_extensions_for_tags, extra_outputs=[gr.update()]),
+                    inputs=[selected_tags, showing_type, filtering_type, sort_column, search_extensions_text],
                     outputs=[available_extensions_table, install_result]
                 )
 
                 sort_column.change(
-                    fn=modules.ui.wrap_gradio_call(refresh_available_extensions_for_tags, extra_outputs=[gr.update()]),
-                    inputs=[hide_tags, sort_column, search_extensions_text],
+                    fn=modules.ui.wrap_gradio_call_no_job(refresh_available_extensions_for_tags, extra_outputs=[gr.update()]),
+                    inputs=[selected_tags, showing_type, filtering_type, sort_column, search_extensions_text],
                     outputs=[available_extensions_table, install_result]
                 )
 
@@ -635,7 +667,7 @@ def create_ui():
                 install_result = gr.HTML(elem_id="extension_install_result")
 
                 install_button.click(
-                    fn=modules.ui.wrap_gradio_call(lambda *args: [gr.update(), *install_extension_from_url(*args)], extra_outputs=[gr.update(), gr.update()]),
+                    fn=modules.ui.wrap_gradio_call_no_job(lambda *args: [gr.update(), *install_extension_from_url(*args)], extra_outputs=[gr.update(), gr.update()]),
                     inputs=[install_dirname, install_url, install_branch],
                     outputs=[install_url, extensions_table, install_result],
                 )
diff --git a/modules/ui_extra_networks.py b/modules/ui_extra_networks.py
index 063bd7b80e6..6e9ec164552 100644
--- a/modules/ui_extra_networks.py
+++ b/modules/ui_extra_networks.py
@@ -1,20 +1,90 @@
+import functools
 import os.path
 import urllib.parse
+from base64 import b64decode
+from io import BytesIO
 from pathlib import Path
+from typing import Optional, Union
+from dataclasses import dataclass
 
-from modules import shared, ui_extra_networks_user_metadata, errors, extra_networks
+from modules import shared, ui_extra_networks_user_metadata, errors, extra_networks, util
 from modules.images import read_info_from_image, save_image_with_geninfo
 import gradio as gr
 import json
 import html
 from fastapi.exceptions import HTTPException
+from PIL import Image
 
-from modules.generation_parameters_copypaste import image_from_url_text
-from modules.ui_components import ToolButton
+from modules.infotext_utils import image_from_url_text
 
 extra_pages = []
 allowed_dirs = set()
-
+default_allowed_preview_extensions = ["png", "jpg", "jpeg", "webp", "gif"]
+
+@functools.cache
+def allowed_preview_extensions_with_extra(extra_extensions=None):
+    return set(default_allowed_preview_extensions) | set(extra_extensions or [])
+
+
+def allowed_preview_extensions():
+    return allowed_preview_extensions_with_extra((shared.opts.samples_format, ))
+
+
+@dataclass
+class ExtraNetworksItem:
+    """Wrapper for dictionaries representing ExtraNetworks items."""
+    item: dict
+
+
+def get_tree(paths: Union[str, list[str]], items: dict[str, ExtraNetworksItem]) -> dict:
+    """Recursively builds a directory tree.
+
+    Args:
+        paths: Path or list of paths to directories. These paths are treated as roots from which
+            the tree will be built.
+        items: A dictionary associating filepaths to an ExtraNetworksItem instance.
+
+    Returns:
+        The result directory tree.
+    """
+    if isinstance(paths, (str,)):
+        paths = [paths]
+
+    def _get_tree(_paths: list[str], _root: str):
+        _res = {}
+        for path in _paths:
+            relpath = os.path.relpath(path, _root)
+            if os.path.isdir(path):
+                dir_items = os.listdir(path)
+                # Ignore empty directories.
+                if not dir_items:
+                    continue
+                dir_tree = _get_tree([os.path.join(path, x) for x in dir_items], _root)
+                # We only want to store non-empty folders in the tree.
+                if dir_tree:
+                    _res[relpath] = dir_tree
+            else:
+                if path not in items:
+                    continue
+                # Add the ExtraNetworksItem to the result.
+                _res[relpath] = items[path]
+        return _res
+
+    res = {}
+    # Handle each root directory separately.
+    # Each root WILL have a key/value at the root of the result dict though
+    # the value can be an empty dict if the directory is empty. We want these
+    # placeholders for empty dirs so we can inform the user later.
+    for path in paths:
+        root = os.path.dirname(path)
+        relpath = os.path.relpath(path, root)
+        # Wrap the path in a list since that is what the `_get_tree` expects.
+        res[relpath] = _get_tree([path], root)
+        if res[relpath]:
+            # We need to pull the inner path out one for these root dirs.
+            res[relpath] = res[relpath][relpath]
+
+    return res
 
 def register_page(page):
     """registers extra networks page for the UI; recommend doing it in on_before_ui() callback for extensions"""
@@ -33,14 +103,39 @@ def fetch_file(filename: str = ""):
     if not any(Path(x).absolute() in Path(filename).absolute().parents for x in allowed_dirs):
         raise ValueError(f"File cannot be fetched: {filename}. Must be in one of directories registered by extra pages.")
 
-    ext = os.path.splitext(filename)[1].lower()
-    if ext not in (".png", ".jpg", ".jpeg", ".webp", ".gif"):
-        raise ValueError(f"File cannot be fetched: {filename}. Only png, jpg, webp, and gif.")
+    ext = os.path.splitext(filename)[1].lower()[1:]
+    if ext not in allowed_preview_extensions():
+        raise ValueError(f"File cannot be fetched: {filename}. Extensions allowed: {allowed_preview_extensions()}.")
 
     # would profit from returning 304
     return FileResponse(filename, headers={"Accept-Ranges": "bytes"})
 
 
+def fetch_cover_images(page: str = "", item: str = "", index: int = 0):
+    from starlette.responses import Response
+
+    page = next(iter([x for x in extra_pages if x.name == page]), None)
+    if page is None:
+        raise HTTPException(status_code=404, detail="File not found")
+
+    metadata = page.metadata.get(item)
+    if metadata is None:
+        raise HTTPException(status_code=404, detail="File not found")
+
+    cover_images = json.loads(metadata.get('ssmd_cover_images', {}))
+    image = cover_images[index] if index < len(cover_images) else None
+    if not image:
+        raise HTTPException(status_code=404, detail="File not found")
+
+    try:
+        image = Image.open(BytesIO(b64decode(image)))
+        buffer = BytesIO()
+        image.save(buffer, format=image.format)
+        return Response(content=buffer.getvalue(), media_type=image.get_format_mimetype())
+    except Exception as err:
+        raise ValueError(f"File cannot be fetched: {item}. Failed to load cover image.") from err
+
+
 def get_metadata(page: str = "", item: str = ""):
     from starlette.responses import JSONResponse
 
@@ -52,6 +147,8 @@ def get_metadata(page: str = "", item: str = ""):
     if metadata is None:
         return JSONResponse({})
 
+    metadata = {i:metadata[i] for i in metadata if i != 'ssmd_cover_images'}  # those are cover images, and they are too big to display in UI as text
+
     return JSONResponse({"metadata": json.dumps(metadata, indent=4, ensure_ascii=False)})
 
 
@@ -67,14 +164,15 @@ def get_single_card(page: str = "", tabname: str = "", name: str = ""):
         errors.display(e, "creating item for extra network")
         item = page.items.get(name)
 
-    page.read_user_metadata(item)
-    item_html = page.create_html_for_item(item, tabname)
+    page.read_user_metadata(item, use_cache=False)
+    item_html = page.create_item_html(tabname, item, shared.html("extra-networks-card.html"))
 
     return JSONResponse({"html": item_html})
 
 
 def add_pages_to_demo(app):
     app.add_api_route("/sd_extra_networks/thumb", fetch_file, methods=["GET"])
+    app.add_api_route("/sd_extra_networks/cover-images", fetch_cover_images, methods=["GET"])
     app.add_api_route("/sd_extra_networks/metadata", get_metadata, methods=["GET"])
     app.add_api_route("/sd_extra_networks/get-single-card", get_single_card, methods=["GET"])
 
@@ -89,18 +187,29 @@ class ExtraNetworksPage:
     def __init__(self, title):
         self.title = title
         self.name = title.lower()
-        self.id_page = self.name.replace(" ", "_")
-        self.card_page = shared.html("extra-networks-card.html")
+        # This is the actual name of the extra networks tab (not txt2img/img2img).
+        self.extra_networks_tabname = self.name.replace(" ", "_")
+        self.allow_prompt = True
         self.allow_negative_prompt = False
         self.metadata = {}
         self.items = {}
+        self.lister = util.MassFileLister()
+        # HTML Templates
+        self.pane_tpl = shared.html("extra-networks-pane.html")
+        self.pane_content_tree_tpl = shared.html("extra-networks-pane-tree.html")
+        self.pane_content_dirs_tpl = shared.html("extra-networks-pane-dirs.html")
+        self.card_tpl = shared.html("extra-networks-card.html")
+        self.btn_tree_tpl = shared.html("extra-networks-tree-button.html")
+        self.btn_copy_path_tpl = shared.html("extra-networks-copy-path-button.html")
+        self.btn_metadata_tpl = shared.html("extra-networks-metadata-button.html")
+        self.btn_edit_item_tpl = shared.html("extra-networks-edit-item-button.html")
 
     def refresh(self):
         pass
 
-    def read_user_metadata(self, item):
+    def read_user_metadata(self, item, use_cache=True):
         filename = item.get("filename", None)
-        metadata = extra_networks.get_user_metadata(filename)
+        metadata = extra_networks.get_user_metadata(filename, lister=self.lister if use_cache else None)
 
         desc = metadata.get("description", None)
         if desc is not None:
@@ -110,23 +219,297 @@ def read_user_metadata(self, item):
 
     def link_preview(self, filename):
         quoted_filename = urllib.parse.quote(filename.replace('\\', '/'))
-        mtime = os.path.getmtime(filename)
+        mtime, _ = self.lister.mctime(filename)
         return f"./sd_extra_networks/thumb?filename={quoted_filename}&mtime={mtime}"
 
     def search_terms_from_path(self, filename, possible_directories=None):
         abspath = os.path.abspath(filename)
-
         for parentdir in (possible_directories if possible_directories is not None else self.allowed_directories_for_previews()):
-            parentdir = os.path.abspath(parentdir)
+            parentdir = os.path.dirname(os.path.abspath(parentdir))
             if abspath.startswith(parentdir):
-                return abspath[len(parentdir):].replace('\\', '/')
+                return os.path.relpath(abspath, parentdir)
 
         return ""
 
-    def create_html(self, tabname):
-        items_html = ''
+    def create_item_html(
+        self,
+        tabname: str,
+        item: dict,
+        template: Optional[str] = None,
+    ) -> Union[str, dict]:
+        """Generates HTML for a single ExtraNetworks Item.
+
+        Args:
+            tabname: The name of the active tab.
+            item: Dictionary containing item information.
+            template: Optional template string to use.
+
+        Returns:
+            If a template is passed: HTML string generated for this item.
+                Can be empty if the item is not meant to be shown.
+            If no template is passed: A dictionary containing the generated item's attributes.
+        """
+        preview = item.get("preview", None)
+        style_height = f"height: {shared.opts.extra_networks_card_height}px;" if shared.opts.extra_networks_card_height else ''
+        style_width = f"width: {shared.opts.extra_networks_card_width}px;" if shared.opts.extra_networks_card_width else ''
+        style_font_size = f"font-size: {shared.opts.extra_networks_card_text_scale*100}%;"
+        card_style = style_height + style_width + style_font_size
+        background_image = f'<img src="{html.escape(preview)}" class="preview" loading="lazy">' if preview else ''
 
-        self.metadata = {}
+        onclick = item.get("onclick", None)
+        if onclick is None:
+            # Don't quote prompt/neg_prompt since they are stored as js strings already.
+            onclick_js_tpl = "cardClicked('{tabname}', {prompt}, {neg_prompt}, {allow_neg});"
+            onclick = onclick_js_tpl.format(
+                **{
+                    "tabname": tabname,
+                    "prompt": item["prompt"],
+                    "neg_prompt": item.get("negative_prompt", "''"),
+                    "allow_neg": str(self.allow_negative_prompt).lower(),
+                }
+            )
+            onclick = html.escape(onclick)
+
+        btn_copy_path = self.btn_copy_path_tpl.format(**{"filename": item["filename"]})
+        btn_metadata = ""
+        metadata = item.get("metadata")
+        if metadata:
+            btn_metadata = self.btn_metadata_tpl.format(
+                **{
+                    "extra_networks_tabname": self.extra_networks_tabname,
+                }
+            )
+        btn_edit_item = self.btn_edit_item_tpl.format(
+            **{
+                "tabname": tabname,
+                "extra_networks_tabname": self.extra_networks_tabname,
+            }
+        )
+
+        local_path = ""
+        filename = item.get("filename", "")
+        for reldir in self.allowed_directories_for_previews():
+            absdir = os.path.abspath(reldir)
+
+            if filename.startswith(absdir):
+                local_path = filename[len(absdir):]
+
+        # if this is true, the item must not be shown in the default view, and must instead only be
+        # shown when searching for it
+        if shared.opts.extra_networks_hidden_models == "Always":
+            search_only = False
+        else:
+            search_only = "/." in local_path or "\\." in local_path
+
+        if search_only and shared.opts.extra_networks_hidden_models == "Never":
+            return ""
+
+        sort_keys = " ".join(
+            [
+                f'data-sort-{k}="{html.escape(str(v))}"'
+                for k, v in item.get("sort_keys", {}).items()
+            ]
+        ).strip()
+
+        search_terms_html = ""
+        search_term_template = "<span class='hidden {class}'>{search_term}</span>"
+        for search_term in item.get("search_terms", []):
+            search_terms_html += search_term_template.format(
+                **{
+                    "class": f"search_terms{' search_only' if search_only else ''}",
+                    "search_term": search_term,
+                }
+            )
+
+        description = (item.get("description", "") or "" if shared.opts.extra_networks_card_show_desc else "")
+        if not shared.opts.extra_networks_card_description_is_html:
+            description = html.escape(description)
+
+        # Some items here might not be used depending on HTML template used.
+        args = {
+            "background_image": background_image,
+            "card_clicked": onclick,
+            "copy_path_button": btn_copy_path,
+            "description": description,
+            "edit_button": btn_edit_item,
+            "local_preview": quote_js(item["local_preview"]),
+            "metadata_button": btn_metadata,
+            "name": html.escape(item["name"]),
+            "prompt": item.get("prompt", None),
+            "save_card_preview": html.escape(f"return saveCardPreview(event, '{tabname}', '{item['local_preview']}');"),
+            "search_only": " search_only" if search_only else "",
+            "search_terms": search_terms_html,
+            "sort_keys": sort_keys,
+            "style": card_style,
+            "tabname": tabname,
+            "extra_networks_tabname": self.extra_networks_tabname,
+        }
+
+        if template:
+            return template.format(**args)
+        else:
+            return args
+
+    def create_tree_dir_item_html(
+        self,
+        tabname: str,
+        dir_path: str,
+        content: Optional[str] = None,
+    ) -> Optional[str]:
+        """Generates HTML for a directory item in the tree.
+
+        The generated HTML is of the format:
+        ```html
+        <li class="tree-list-item tree-list-item--has-subitem">
+            <div class="tree-list-content tree-list-content-dir"></div>
+            <ul class="tree-list tree-list--subgroup">
+                {content}
+            </ul>
+        </li>
+        ```
+
+        Args:
+            tabname: The name of the active tab.
+            dir_path: Path to the directory for this item.
+            content: Optional HTML string that will be wrapped by this <ul>.
+
+        Returns:
+            HTML formatted string.
+        """
+        if not content:
+            return None
+
+        btn = self.btn_tree_tpl.format(
+            **{
+                "search_terms": "",
+                "subclass": "tree-list-content-dir",
+                "tabname": tabname,
+                "extra_networks_tabname": self.extra_networks_tabname,
+                "onclick_extra": "",
+                "data_path": dir_path,
+                "data_hash": "",
+                "action_list_item_action_leading": "<i class='tree-list-item-action-chevron'></i>",
+                "action_list_item_visual_leading": "🗀",
+                "action_list_item_label": os.path.basename(dir_path),
+                "action_list_item_visual_trailing": "",
+                "action_list_item_action_trailing": "",
+            }
+        )
+        ul = f"<ul class='tree-list tree-list--subgroup' hidden>{content}</ul>"
+        return (
+            "<li class='tree-list-item tree-list-item--has-subitem' data-tree-entry-type='dir'>"
+            f"{btn}{ul}"
+            "</li>"
+        )
+
+    def create_tree_file_item_html(self, tabname: str, file_path: str, item: dict) -> str:
+        """Generates HTML for a file item in the tree.
+
+        The generated HTML is of the format:
+        ```html
+        <li class="tree-list-item tree-list-item--subitem">
+            <span data-filterable-item-text hidden></span>
+            <div class="tree-list-content tree-list-content-file"></div>
+        </li>
+        ```
+
+        Args:
+            tabname: The name of the active tab.
+            file_path: The path to the file for this item.
+            item: Dictionary containing the item information.
+
+        Returns:
+            HTML formatted string.
+        """
+        item_html_args = self.create_item_html(tabname, item)
+        action_buttons = "".join(
+            [
+                item_html_args["copy_path_button"],
+                item_html_args["metadata_button"],
+                item_html_args["edit_button"],
+            ]
+        )
+        action_buttons = f"<div class=\"button-row\">{action_buttons}</div>"
+        btn = self.btn_tree_tpl.format(
+            **{
+                "search_terms": "",
+                "subclass": "tree-list-content-file",
+                "tabname": tabname,
+                "extra_networks_tabname": self.extra_networks_tabname,
+                "onclick_extra": item_html_args["card_clicked"],
+                "data_path": file_path,
+                "data_hash": item["shorthash"],
+                "action_list_item_action_leading": "<i class='tree-list-item-action-chevron'></i>",
+                "action_list_item_visual_leading": "🗎",
+                "action_list_item_label": item["name"],
+                "action_list_item_visual_trailing": "",
+                "action_list_item_action_trailing": action_buttons,
+            }
+        )
+        return (
+            "<li class='tree-list-item tree-list-item--subitem' data-tree-entry-type='file'>"
+            f"{btn}"
+            "</li>"
+        )
+
+    def create_tree_view_html(self, tabname: str) -> str:
+        """Generates HTML for displaying folders in a tree view.
+
+        Args:
+            tabname: The name of the active tab.
+
+        Returns:
+            HTML string generated for this tree view.
+        """
+        res = ""
+
+        # Setup the tree dictionary.
+        roots = self.allowed_directories_for_previews()
+        tree_items = {v["filename"]: ExtraNetworksItem(v) for v in self.items.values()}
+        tree = get_tree([os.path.abspath(x) for x in roots], items=tree_items)
+
+        if not tree:
+            return res
+
+        def _build_tree(data: Optional[dict[str, ExtraNetworksItem]] = None) -> Optional[str]:
+            """Recursively builds HTML for a tree.
+
+            Args:
+                data: Dictionary representing a directory tree. Can be NoneType.
+                    Data keys should be absolute paths from the root and values
+                    should be subdirectory trees or an ExtraNetworksItem.
+
+            Returns:
+                If data is not None: HTML string
+                Else: None
+            """
+            if not data:
+                return None
+
+            # Lists for storing <li> items html for directories and files separately.
+            _dir_li = []
+            _file_li = []
+
+            for k, v in sorted(data.items(), key=lambda x: shared.natural_sort_key(x[0])):
+                if isinstance(v, (ExtraNetworksItem,)):
+                    _file_li.append(self.create_tree_file_item_html(tabname, k, v.item))
+                else:
+                    _dir_li.append(self.create_tree_dir_item_html(tabname, k, _build_tree(v)))
+
+            # Directories should always be displayed before files so we order them here.
+            return "".join(_dir_li) + "".join(_file_li)
+
+        # Add each root directory to the tree.
+        for k, v in sorted(tree.items(), key=lambda x: shared.natural_sort_key(x[0])):
+            item_html = self.create_tree_dir_item_html(tabname, k, _build_tree(v))
+            # Only add non-empty entries to the tree.
+            if item_html is not None:
+                res += item_html
+
+        return f"<ul class='tree-list tree-list--tree'>{res}</ul>"
+
+    def create_dirs_view_html(self, tabname: str) -> str:
+        """Generates HTML for displaying folders."""
 
         subdirs = {}
         for parentdir in [os.path.abspath(x) for x in self.allowed_directories_for_previews()]:
@@ -137,15 +520,20 @@ def create_html(self, tabname):
                     if not os.path.isdir(x):
                         continue
 
-                    subdir = os.path.abspath(x)[len(parentdir):].replace("\\", "/")
-                    while subdir.startswith("/"):
-                        subdir = subdir[1:]
+                    subdir = os.path.abspath(x)[len(parentdir):]
+
+                    if shared.opts.extra_networks_dir_button_function:
+                        if not subdir.startswith(os.path.sep):
+                            subdir = os.path.sep + subdir
+                    else:
+                        while subdir.startswith(os.path.sep):
+                            subdir = subdir[1:]
 
                     is_empty = len(os.listdir(x)) == 0
-                    if not is_empty and not subdir.endswith("/"):
-                        subdir = subdir + "/"
+                    if not is_empty and not subdir.endswith(os.path.sep):
+                        subdir = subdir + os.path.sep
 
-                    if ("/." in subdir or subdir.startswith(".")) and not shared.opts.extra_networks_show_hidden_directories:
+                    if (os.path.sep + "." in subdir or subdir.startswith(".")) and not shared.opts.extra_networks_show_hidden_directories:
                         continue
 
                     subdirs[subdir] = 1
@@ -154,118 +542,106 @@ def create_html(self, tabname):
             subdirs = {"": 1, **subdirs}
 
         subdirs_html = "".join([f"""
-<button class='lg secondary gradio-button custom-button{" search-all" if subdir=="" else ""}' onclick='extraNetworksSearchButton("{tabname}_extra_search", event)'>
-{html.escape(subdir if subdir!="" else "all")}
-</button>
-""" for subdir in subdirs])
+        <button class='lg secondary gradio-button custom-button{" search-all" if subdir == "" else ""}' onclick='extraNetworksSearchButton("{tabname}", "{self.extra_networks_tabname}", event)'>
+        {html.escape(subdir if subdir != "" else "all")}
+        </button>
+        """ for subdir in subdirs])
 
-        self.items = {x["name"]: x for x in self.list_items()}
-        for item in self.items.values():
-            metadata = item.get("metadata")
-            if metadata:
-                self.metadata[item["name"]] = metadata
+        return subdirs_html
 
-            if "user_metadata" not in item:
-                self.read_user_metadata(item)
+    def create_card_view_html(self, tabname: str, *, none_message) -> str:
+        """Generates HTML for the network Card View section for a tab.
 
-            items_html += self.create_html_for_item(item, tabname)
+        This HTML goes into the `extra-networks-pane.html` <div> with
+        `id='{tabname}_{extra_networks_tabname}_cards`.
 
-        if items_html == '':
-            dirs = "".join([f"<li>{x}</li>" for x in self.allowed_directories_for_previews()])
-            items_html = shared.html("extra-networks-no-cards.html").format(dirs=dirs)
+        Args:
+            tabname: The name of the active tab.
+            none_message: HTML text to show when there are no cards.
 
-        self_name_id = self.name.replace(" ", "_")
+        Returns:
+            HTML formatted string.
+        """
+        res = []
+        for item in self.items.values():
+            res.append(self.create_item_html(tabname, item, self.card_tpl))
 
-        res = f"""
-<div id='{tabname}_{self_name_id}_subdirs' class='extra-network-subdirs extra-network-subdirs-cards'>
-{subdirs_html}
-</div>
-<div id='{tabname}_{self_name_id}_cards' class='extra-network-cards'>
-{items_html}
-</div>
-"""
+        if not res:
+            dirs = "".join([f"<li>{x}</li>" for x in self.allowed_directories_for_previews()])
+            res = [none_message or shared.html("extra-networks-no-cards.html").format(dirs=dirs)]
 
-        return res
+        return "".join(res)
 
-    def create_item(self, name, index=None):
-        raise NotImplementedError()
+    def create_html(self, tabname, *, empty=False):
+        """Generates an HTML string for the current pane.
 
-    def list_items(self):
-        raise NotImplementedError()
+        The generated HTML uses `extra-networks-pane.html` as a template.
 
-    def allowed_directories_for_previews(self):
-        return []
+        Args:
+            tabname: The name of the active tab.
+            empty: create an empty HTML page with no items
 
-    def create_html_for_item(self, item, tabname):
+        Returns:
+            HTML formatted string.
         """
-        Create HTML for card item in tab tabname; can return empty string if the item is not meant to be shown.
-        """
-
-        preview = item.get("preview", None)
-
-        onclick = item.get("onclick", None)
-        if onclick is None:
-            onclick = '"' + html.escape(f"""return cardClicked({quote_js(tabname)}, {item["prompt"]}, {"true" if self.allow_negative_prompt else "false"})""") + '"'
+        self.lister.reset()
+        self.metadata = {}
 
-        height = f"height: {shared.opts.extra_networks_card_height}px;" if shared.opts.extra_networks_card_height else ''
-        width = f"width: {shared.opts.extra_networks_card_width}px;" if shared.opts.extra_networks_card_width else ''
-        background_image = f'<img src="{html.escape(preview)}" class="preview" loading="lazy">' if preview else ''
-        metadata_button = ""
-        metadata = item.get("metadata")
-        if metadata:
-            metadata_button = f"<div class='metadata-button card-button' title='Show internal metadata' onclick='extraNetworksRequestMetadata(event, {quote_js(self.name)}, {quote_js(item['name'])})'></div>"
+        items_list = [] if empty else self.list_items()
+        self.items = {x["name"]: x for x in items_list}
 
-        edit_button = f"<div class='edit-button card-button' title='Edit metadata' onclick='extraNetworksEditUserMetadata(event, {quote_js(tabname)}, {quote_js(self.id_page)}, {quote_js(item['name'])})'></div>"
+        # Populate the instance metadata for each item.
+        for item in self.items.values():
+            metadata = item.get("metadata")
+            if metadata:
+                self.metadata[item["name"]] = metadata
 
-        local_path = ""
-        filename = item.get("filename", "")
-        for reldir in self.allowed_directories_for_previews():
-            absdir = os.path.abspath(reldir)
+            if "user_metadata" not in item:
+                self.read_user_metadata(item)
 
-            if filename.startswith(absdir):
-                local_path = filename[len(absdir):]
+        show_tree = shared.opts.extra_networks_tree_view_default_enabled
+
+        page_params = {
+            "tabname": tabname,
+            "extra_networks_tabname": self.extra_networks_tabname,
+            "data_sortdir": shared.opts.extra_networks_card_order,
+            "sort_path_active": ' extra-network-control--enabled' if shared.opts.extra_networks_card_order_field == 'Path' else '',
+            "sort_name_active": ' extra-network-control--enabled' if shared.opts.extra_networks_card_order_field == 'Name' else '',
+            "sort_date_created_active": ' extra-network-control--enabled' if shared.opts.extra_networks_card_order_field == 'Date Created' else '',
+            "sort_date_modified_active": ' extra-network-control--enabled' if shared.opts.extra_networks_card_order_field == 'Date Modified' else '',
+            "tree_view_btn_extra_class": "extra-network-control--enabled" if show_tree else "",
+            "items_html": self.create_card_view_html(tabname, none_message="Loading..." if empty else None),
+            "extra_networks_tree_view_default_width": shared.opts.extra_networks_tree_view_default_width,
+            "tree_view_div_default_display_class": "" if show_tree else "extra-network-dirs-hidden",
+        }
 
-        # if this is true, the item must not be shown in the default view, and must instead only be
-        # shown when searching for it
-        if shared.opts.extra_networks_hidden_models == "Always":
-            search_only = False
+        if shared.opts.extra_networks_tree_view_style == "Tree":
+            pane_content = self.pane_content_tree_tpl.format(**page_params, tree_html=self.create_tree_view_html(tabname))
         else:
-            search_only = "/." in local_path or "\\." in local_path
+            pane_content = self.pane_content_dirs_tpl.format(**page_params, dirs_html=self.create_dirs_view_html(tabname))
 
-        if search_only and shared.opts.extra_networks_hidden_models == "Never":
-            return ""
+        return self.pane_tpl.format(**page_params, pane_content=pane_content)
 
-        sort_keys = " ".join([html.escape(f'data-sort-{k}={v}') for k, v in item.get("sort_keys", {}).items()]).strip()
+    def create_item(self, name, index=None):
+        raise NotImplementedError()
 
-        args = {
-            "background_image": background_image,
-            "style": f"'display: none; {height}{width}; font-size: {shared.opts.extra_networks_card_text_scale*100}%'",
-            "prompt": item.get("prompt", None),
-            "tabname": quote_js(tabname),
-            "local_preview": quote_js(item["local_preview"]),
-            "name": html.escape(item["name"]),
-            "description": (item.get("description") or "" if shared.opts.extra_networks_card_show_desc else ""),
-            "card_clicked": onclick,
-            "save_card_preview": '"' + html.escape(f"""return saveCardPreview(event, {quote_js(tabname)}, {quote_js(item["local_preview"])})""") + '"',
-            "search_term": item.get("search_term", ""),
-            "metadata_button": metadata_button,
-            "edit_button": edit_button,
-            "search_only": " search_only" if search_only else "",
-            "sort_keys": sort_keys,
-        }
+    def list_items(self):
+        raise NotImplementedError()
 
-        return self.card_page.format(**args)
+    def allowed_directories_for_previews(self):
+        return []
 
     def get_sort_keys(self, path):
         """
         List of default keys used for sorting in the UI.
         """
         pth = Path(path)
-        stat = pth.stat()
+        mtime, ctime = self.lister.mctime(path)
         return {
-            "date_created": int(stat.st_ctime or 0),
-            "date_modified": int(stat.st_mtime or 0),
+            "date_created": int(mtime),
+            "date_modified": int(ctime),
             "name": pth.name.lower(),
+            "path": str(pth).lower(),
         }
 
     def find_preview(self, path):
@@ -273,23 +649,33 @@ def find_preview(self, path):
         Find a preview PNG for a given path (without extension) and call link_preview on it.
         """
 
-        preview_extensions = ["png", "jpg", "jpeg", "webp"]
-        if shared.opts.samples_format not in preview_extensions:
-            preview_extensions.append(shared.opts.samples_format)
-
-        potential_files = sum([[path + "." + ext, path + ".preview." + ext] for ext in preview_extensions], [])
+        potential_files = sum([[f"{path}.{ext}", f"{path}.preview.{ext}"] for ext in allowed_preview_extensions()], [])
 
         for file in potential_files:
-            if os.path.isfile(file):
+            if self.lister.exists(file):
                 return self.link_preview(file)
 
         return None
 
+    def find_embedded_preview(self, path, name, metadata):
+        """
+        Find if embedded preview exists in safetensors metadata and return endpoint for it.
+        """
+
+        file = f"{path}.safetensors"
+        if self.lister.exists(file) and 'ssmd_cover_images' in metadata and len(list(filter(None, json.loads(metadata['ssmd_cover_images'])))) > 0:
+            return f"./sd_extra_networks/cover-images?page={self.extra_networks_tabname}&item={name}"
+
+        return None
+
     def find_description(self, path):
         """
         Find and read a description file for a given path (without extension).
         """
         for file in [f"{path}.txt", f"{path}.description.txt"]:
+            if not self.lister.exists(file):
+                continue
+
             try:
                 with open(file, "r", encoding="utf-8", errors="replace") as f:
                     return f.read()
@@ -347,8 +733,6 @@ def tab_name_score(name):
 
 
 def create_ui(interface: gr.Blocks, unrelated_tabs, tabname):
-    from modules.ui import switch_values_symbol
-
     ui = ExtraNetworksUi()
     ui.pages = []
     ui.pages_contents = []
@@ -359,50 +743,51 @@ def create_ui(interface: gr.Blocks, unrelated_tabs, tabname):
     related_tabs = []
 
     for page in ui.stored_extra_pages:
-        with gr.Tab(page.title, id=page.id_page) as tab:
-            elem_id = f"{tabname}_{page.id_page}_cards_html"
-            page_elem = gr.HTML('Loading...', elem_id=elem_id)
-            ui.pages.append(page_elem)
-
-            page_elem.change(fn=lambda: None, _js='function(){applyExtraNetworkFilter(' + quote_js(tabname) + '); return []}', inputs=[], outputs=[])
+        with gr.Tab(page.title, elem_id=f"{tabname}_{page.extra_networks_tabname}", elem_classes=["extra-page"]) as tab:
+            with gr.Column(elem_id=f"{tabname}_{page.extra_networks_tabname}_prompts", elem_classes=["extra-page-prompts"]):
+                pass
 
+            elem_id = f"{tabname}_{page.extra_networks_tabname}_cards_html"
+            page_elem = gr.HTML(page.create_html(tabname, empty=True), elem_id=elem_id)
+            ui.pages.append(page_elem)
             editor = page.create_user_metadata_editor(ui, tabname)
             editor.create_ui()
             ui.user_metadata_editors.append(editor)
-
             related_tabs.append(tab)
 
-    edit_search = gr.Textbox('', show_label=False, elem_id=tabname+"_extra_search", elem_classes="search", placeholder="Search...", visible=False, interactive=True)
-    dropdown_sort = gr.Dropdown(choices=['Default Sort', 'Date Created', 'Date Modified', 'Name'], value='Default Sort', elem_id=tabname+"_extra_sort", elem_classes="sort", multiselect=False, visible=False, show_label=False, interactive=True, label=tabname+"_extra_sort_order")
-    button_sortorder = ToolButton(switch_values_symbol, elem_id=tabname+"_extra_sortorder", elem_classes="sortorder", visible=False)
-    button_refresh = gr.Button('Refresh', elem_id=tabname+"_extra_refresh", visible=False)
-    checkbox_show_dirs = gr.Checkbox(True, label='Show dirs', elem_id=tabname+"_extra_show_dirs", elem_classes="show-dirs", visible=False)
-
-    ui.button_save_preview = gr.Button('Save preview', elem_id=tabname+"_save_preview", visible=False)
-    ui.preview_target_filename = gr.Textbox('Preview save filename', elem_id=tabname+"_preview_filename", visible=False)
+    ui.button_save_preview = gr.Button('Save preview', elem_id=f"{tabname}_save_preview", visible=False)
+    ui.preview_target_filename = gr.Textbox('Preview save filename', elem_id=f"{tabname}_preview_filename", visible=False)
 
     for tab in unrelated_tabs:
-        tab.select(fn=lambda: [gr.update(visible=False) for _ in range(5)], inputs=[], outputs=[edit_search, dropdown_sort, button_sortorder, button_refresh, checkbox_show_dirs], show_progress=False)
-
-    for tab in related_tabs:
-        tab.select(fn=lambda: [gr.update(visible=True) for _ in range(5)], inputs=[], outputs=[edit_search, dropdown_sort, button_sortorder, button_refresh, checkbox_show_dirs], show_progress=False)
+        tab.select(fn=None, _js=f"function(){{extraNetworksUnrelatedTabSelected('{tabname}');}}", inputs=[], outputs=[], show_progress=False)
+
+    for page, tab in zip(ui.stored_extra_pages, related_tabs):
+        jscode = (
+            "function(){{"
+            f"extraNetworksTabSelected('{tabname}', '{tabname}_{page.extra_networks_tabname}_prompts', {str(page.allow_prompt).lower()}, {str(page.allow_negative_prompt).lower()}, '{tabname}_{page.extra_networks_tabname}');"
+            f"applyExtraNetworkFilter('{tabname}_{page.extra_networks_tabname}');"
+            "}}"
+        )
+        tab.select(fn=None, _js=jscode, inputs=[], outputs=[], show_progress=False)
+
+        def refresh():
+            for pg in ui.stored_extra_pages:
+                pg.refresh()
+            create_html()
+            return ui.pages_contents
+
+        button_refresh = gr.Button("Refresh", elem_id=f"{tabname}_{page.extra_networks_tabname}_extra_refresh_internal", visible=False)
+        button_refresh.click(fn=refresh, inputs=[], outputs=ui.pages).then(fn=lambda: None, _js="function(){ " + f"applyExtraNetworkFilter('{tabname}_{page.extra_networks_tabname}');" + " }").then(fn=lambda: None, _js='setupAllResizeHandles')
+
+    def create_html():
+        ui.pages_contents = [pg.create_html(ui.tabname) for pg in ui.stored_extra_pages]
 
     def pages_html():
         if not ui.pages_contents:
-            return refresh()
-
+            create_html()
         return ui.pages_contents
 
-    def refresh():
-        for pg in ui.stored_extra_pages:
-            pg.refresh()
-
-        ui.pages_contents = [pg.create_html(ui.tabname) for pg in ui.stored_extra_pages]
-
-        return ui.pages_contents
-
-    interface.load(fn=pages_html, inputs=[], outputs=[*ui.pages])
-    button_refresh.click(fn=refresh, inputs=[], outputs=ui.pages)
+    interface.load(fn=pages_html, inputs=[], outputs=ui.pages).then(fn=lambda: None, _js='setupAllResizeHandles')
 
     return ui
 
@@ -451,5 +836,3 @@ def save_preview(index, images, filename):
 
     for editor in ui.user_metadata_editors:
         editor.setup_ui(gallery)
-
-
diff --git a/modules/ui_extra_networks_checkpoints.py b/modules/ui_extra_networks_checkpoints.py
index ca6c26076f9..d69d144dba4 100644
--- a/modules/ui_extra_networks_checkpoints.py
+++ b/modules/ui_extra_networks_checkpoints.py
@@ -2,7 +2,6 @@
 import os
 
 from modules import shared, ui_extra_networks, sd_models
-from modules.ui_extra_networks import quote_js
 from modules.ui_extra_networks_checkpoints_user_metadata import CheckpointUserMetadataEditor
 
 
@@ -10,29 +9,40 @@ class ExtraNetworksPageCheckpoints(ui_extra_networks.ExtraNetworksPage):
     def __init__(self):
         super().__init__('Checkpoints')
 
+        self.allow_prompt = False
+
     def refresh(self):
         shared.refresh_checkpoints()
 
     def create_item(self, name, index=None, enable_filter=True):
         checkpoint: sd_models.CheckpointInfo = sd_models.checkpoint_aliases.get(name)
+        if checkpoint is None:
+            return
+
         path, ext = os.path.splitext(checkpoint.filename)
+        search_terms = [self.search_terms_from_path(checkpoint.filename)]
+        if checkpoint.sha256:
+            search_terms.append(checkpoint.sha256)
         return {
             "name": checkpoint.name_for_extra,
             "filename": checkpoint.filename,
             "shorthash": checkpoint.shorthash,
             "preview": self.find_preview(path),
             "description": self.find_description(path),
-            "search_term": self.search_terms_from_path(checkpoint.filename) + " " + (checkpoint.sha256 or ""),
-            "onclick": '"' + html.escape(f"""return selectCheckpoint({quote_js(name)})""") + '"',
+            "search_terms": search_terms,
+            "onclick": html.escape(f"return selectCheckpoint({ui_extra_networks.quote_js(name)})"),
             "local_preview": f"{path}.{shared.opts.samples_format}",
             "metadata": checkpoint.metadata,
             "sort_keys": {'default': index, **self.get_sort_keys(checkpoint.filename)},
         }
 
     def list_items(self):
+        # instantiate a list to protect against concurrent modification
         names = list(sd_models.checkpoints_list)
         for index, name in enumerate(names):
-            yield self.create_item(name, index)
+            item = self.create_item(name, index)
+            if item is not None:
+                yield item
 
     def allowed_directories_for_previews(self):
         return [v for v in [shared.cmd_opts.ckpt_dir, sd_models.model_path] if v is not None]
diff --git a/modules/ui_extra_networks_hypernets.py b/modules/ui_extra_networks_hypernets.py
index 4cedf085196..2fb4bd190a1 100644
--- a/modules/ui_extra_networks_hypernets.py
+++ b/modules/ui_extra_networks_hypernets.py
@@ -13,26 +13,35 @@ def refresh(self):
         shared.reload_hypernetworks()
 
     def create_item(self, name, index=None, enable_filter=True):
-        full_path = shared.hypernetworks[name]
+        full_path = shared.hypernetworks.get(name)
+        if full_path is None:
+            return
+
         path, ext = os.path.splitext(full_path)
         sha256 = sha256_from_cache(full_path, f'hypernet/{name}')
         shorthash = sha256[0:10] if sha256 else None
-
+        search_terms = [self.search_terms_from_path(path)]
+        if sha256:
+            search_terms.append(sha256)
         return {
             "name": name,
             "filename": full_path,
             "shorthash": shorthash,
             "preview": self.find_preview(path),
             "description": self.find_description(path),
-            "search_term": self.search_terms_from_path(path) + " " + (sha256 or ""),
+            "search_terms": search_terms,
             "prompt": quote_js(f"<hypernet:{name}:") + " + opts.extra_networks_default_multiplier + " + quote_js(">"),
             "local_preview": f"{path}.preview.{shared.opts.samples_format}",
             "sort_keys": {'default': index, **self.get_sort_keys(path + ext)},
         }
 
     def list_items(self):
-        for index, name in enumerate(shared.hypernetworks):
-            yield self.create_item(name, index)
+        # instantiate a list to protect against concurrent modification
+        names = list(shared.hypernetworks)
+        for index, name in enumerate(names):
+            item = self.create_item(name, index)
+            if item is not None:
+                yield item
 
     def allowed_directories_for_previews(self):
         return [shared.cmd_opts.hypernetwork_dir]
diff --git a/modules/ui_extra_networks_textual_inversion.py b/modules/ui_extra_networks_textual_inversion.py
index 55ef0ea7b54..deb7cb8733b 100644
--- a/modules/ui_extra_networks_textual_inversion.py
+++ b/modules/ui_extra_networks_textual_inversion.py
@@ -14,23 +14,32 @@ def refresh(self):
 
     def create_item(self, name, index=None, enable_filter=True):
         embedding = sd_hijack.model_hijack.embedding_db.word_embeddings.get(name)
+        if embedding is None:
+            return
 
         path, ext = os.path.splitext(embedding.filename)
+        search_terms = [self.search_terms_from_path(embedding.filename)]
+        if embedding.hash:
+            search_terms.append(embedding.hash)
         return {
             "name": name,
             "filename": embedding.filename,
             "shorthash": embedding.shorthash,
             "preview": self.find_preview(path),
             "description": self.find_description(path),
-            "search_term": self.search_terms_from_path(embedding.filename) + " " + (embedding.hash or ""),
+            "search_terms": search_terms,
             "prompt": quote_js(embedding.name),
             "local_preview": f"{path}.preview.{shared.opts.samples_format}",
             "sort_keys": {'default': index, **self.get_sort_keys(embedding.filename)},
         }
 
     def list_items(self):
-        for index, name in enumerate(sd_hijack.model_hijack.embedding_db.word_embeddings):
-            yield self.create_item(name, index)
+        # instantiate a list to protect against concurrent modification
+        names = list(sd_hijack.model_hijack.embedding_db.word_embeddings)
+        for index, name in enumerate(names):
+            item = self.create_item(name, index)
+            if item is not None:
+                yield item
 
     def allowed_directories_for_previews(self):
         return list(sd_hijack.model_hijack.embedding_db.embedding_dirs)
diff --git a/modules/ui_extra_networks_user_metadata.py b/modules/ui_extra_networks_user_metadata.py
index bfec140cc73..3a07db10542 100644
--- a/modules/ui_extra_networks_user_metadata.py
+++ b/modules/ui_extra_networks_user_metadata.py
@@ -5,7 +5,7 @@
 
 import gradio as gr
 
-from modules import generation_parameters_copypaste, images, sysinfo, errors, ui_extra_networks
+from modules import infotext_utils, images, sysinfo, errors, ui_extra_networks
 
 
 class UserMetadataEditor:
@@ -14,7 +14,7 @@ def __init__(self, ui, tabname, page):
         self.ui = ui
         self.tabname = tabname
         self.page = page
-        self.id_part = f"{self.tabname}_{self.page.id_page}_edit_user_metadata"
+        self.id_part = f"{self.tabname}_{self.page.extra_networks_tabname}_edit_user_metadata"
 
         self.box = None
 
@@ -133,8 +133,10 @@ def write_user_metadata(self, name, metadata):
         filename = item.get("filename", None)
         basename, ext = os.path.splitext(filename)
 
-        with open(basename + '.json', "w", encoding="utf8") as file:
-            json.dump(metadata, file, indent=4)
+        metadata_path = basename + '.json'
+        with open(metadata_path, "w", encoding="utf8") as file:
+            json.dump(metadata, file, indent=4, ensure_ascii=False)
+        self.page.lister.update_file_entry(metadata_path)
 
     def save_user_metadata(self, name, desc, notes):
         user_metadata = self.get_user_metadata(name)
@@ -181,17 +183,18 @@ def save_preview(self, index, gallery, name):
         index = len(gallery) - 1 if index >= len(gallery) else index
 
         img_info = gallery[index if index >= 0 else 0]
-        image = generation_parameters_copypaste.image_from_url_text(img_info)
+        image = infotext_utils.image_from_url_text(img_info)
         geninfo, items = images.read_info_from_image(image)
 
         images.save_image_with_geninfo(image, geninfo, item["local_preview"])
-
+        self.page.lister.update_file_entry(item["local_preview"])
+        item['preview'] = self.page.find_preview(item["local_preview"])
         return self.get_card_html(name), ''
 
     def setup_ui(self, gallery):
         self.button_replace_preview.click(
             fn=self.save_preview,
-            _js="function(x, y, z){return [selected_gallery_index(), y, z]}",
+            _js=f"function(x, y, z){{return [selected_gallery_index_id('{self.tabname + '_gallery_container'}'), y, z]}}",
             inputs=[self.edit_name_input, gallery, self.edit_name_input],
             outputs=[self.html_preview, self.html_status]
         ).then(
@@ -200,6 +203,3 @@ def setup_ui(self, gallery):
             inputs=[self.edit_name_input],
             outputs=[]
         )
-
-
-
diff --git a/modules/ui_gradio_extensions.py b/modules/ui_gradio_extensions.py
index b824b113732..ed57c1e9896 100644
--- a/modules/ui_gradio_extensions.py
+++ b/modules/ui_gradio_extensions.py
@@ -1,17 +1,12 @@
 import os
 import gradio as gr
 
-from modules import localization, shared, scripts
+from modules import localization, shared, scripts, util
 from modules.paths import script_path, data_path
 
 
 def webpath(fn):
-    if fn.startswith(script_path):
-        web_path = os.path.relpath(fn, script_path).replace('\\', '/')
-    else:
-        web_path = os.path.abspath(fn)
-
-    return f'file={web_path}?{os.path.getmtime(fn)}'
+    return f'file={util.truncate_path(fn)}?{os.path.getmtime(fn)}'
 
 
 def javascript_html():
@@ -40,13 +35,16 @@ def stylesheet(fn):
         return f'<link rel="stylesheet" property="stylesheet" href="{webpath(fn)}">'
 
     for cssfile in scripts.list_files_with_name("style.css"):
-        if not os.path.isfile(cssfile):
-            continue
-
         head += stylesheet(cssfile)
 
-    if os.path.exists(os.path.join(data_path, "user.css")):
-        head += stylesheet(os.path.join(data_path, "user.css"))
+    user_css = os.path.join(data_path, "user.css")
+    if os.path.exists(user_css):
+        head += stylesheet(user_css)
+
+    from modules.shared_gradio_themes import resolve_var
+    light = resolve_var('background_fill_primary')
+    dark = resolve_var('background_fill_primary_dark')
+    head += f'<style>html {{ background-color: {light}; }} @media (prefers-color-scheme: dark) {{ html {{background-color:  {dark}; }} }}</style>'
 
     return head
 
@@ -57,7 +55,7 @@ def reload_javascript():
 
     def template_response(*args, **kwargs):
         res = shared.GradioTemplateResponseOriginal(*args, **kwargs)
-        res.body = res.body.replace(b'</head>', f'{js}</head>'.encode("utf8"))
+        res.body = res.body.replace(b'</head>', f'{js}<meta name="referrer" content="no-referrer"/></head>'.encode("utf8"))
         res.body = res.body.replace(b'</body>', f'{css}</body>'.encode("utf8"))
         res.init_headers()
         return res
diff --git a/modules/ui_loadsave.py b/modules/ui_loadsave.py
index ec8fa8e89e3..0cc1ab82af4 100644
--- a/modules/ui_loadsave.py
+++ b/modules/ui_loadsave.py
@@ -4,7 +4,7 @@
 import gradio as gr
 
 from modules import errors
-from modules.ui_components import ToolButton
+from modules.ui_components import ToolButton, InputAccordion
 
 
 def radio_choices(comp):  # gradio 3.41 changes choices from list of values to list of pairs
@@ -26,14 +26,13 @@ def __init__(self, filename):
         self.ui_defaults_review = None
 
         try:
-            if os.path.exists(self.filename):
-                self.ui_settings = self.read_from_file()
+            self.ui_settings = self.read_from_file()
+        except FileNotFoundError:
+            pass
         except Exception as e:
             self.error_loading = True
             errors.display(e, "loading settings")
 
-
-
     def add_component(self, path, x):
         """adds component to the registry of tracked components"""
 
@@ -43,20 +42,24 @@ def apply_field(obj, field, condition=None, init_field=None):
             key = f"{path}/{field}"
 
             if getattr(obj, 'custom_script_source', None) is not None:
-              key = f"customscript/{obj.custom_script_source}/{key}"
+                key = f"customscript/{obj.custom_script_source}/{key}"
 
             if getattr(obj, 'do_not_save_to_config', False):
                 return
 
             saved_value = self.ui_settings.get(key, None)
+
+            if isinstance(obj, gr.Accordion) and isinstance(x, InputAccordion) and field == 'value':
+                field = 'open'
+
             if saved_value is None:
                 self.ui_settings[key] = getattr(obj, field)
             elif condition and not condition(saved_value):
                 pass
             else:
-                if isinstance(x, gr.Textbox) and field == 'value':  # due to an undesirable behavior of gr.Textbox, if you give it an int value instead of str, everything dies
+                if isinstance(obj, gr.Textbox) and field == 'value':  # due to an undesirable behavior of gr.Textbox, if you give it an int value instead of str, everything dies
                     saved_value = str(saved_value)
-                elif isinstance(x, gr.Number) and field == 'value':
+                elif isinstance(obj, gr.Number) and field == 'value':
                     try:
                         saved_value = float(saved_value)
                     except ValueError:
@@ -67,7 +70,7 @@ def apply_field(obj, field, condition=None, init_field=None):
                     init_field(saved_value)
 
             if field == 'value' and key not in self.component_mapping:
-                self.component_mapping[key] = x
+                self.component_mapping[key] = obj
 
         if type(x) in [gr.Slider, gr.Radio, gr.Checkbox, gr.Textbox, gr.Number, gr.Dropdown, ToolButton, gr.Button] and x.visible:
             apply_field(x, 'visible')
@@ -100,6 +103,14 @@ def check_dropdown(val):
 
             apply_field(x, 'value', check_dropdown, getattr(x, 'init_field', None))
 
+        if type(x) == InputAccordion:
+            if hasattr(x, 'custom_script_source'):
+                x.accordion.custom_script_source = x.custom_script_source
+            if x.accordion.visible:
+                apply_field(x.accordion, 'visible')
+            apply_field(x, 'value')
+            apply_field(x.accordion, 'value')
+
         def check_tab_id(tab_id):
             tab_items = list(filter(lambda e: isinstance(e, gr.TabItem), x.children))
             if type(tab_id) == str:
@@ -133,10 +144,10 @@ def read_from_file(self):
 
     def write_to_file(self, current_ui_settings):
         with open(self.filename, "w", encoding="utf8") as file:
-            json.dump(current_ui_settings, file, indent=4)
+            json.dump(current_ui_settings, file, indent=4, ensure_ascii=False)
 
     def dump_defaults(self):
-        """saves default values to a file unless tjhe file is present and there was an error loading default values at start"""
+        """saves default values to a file unless the file is present and there was an error loading default values at start"""
 
         if self.error_loading and os.path.exists(self.filename):
             return
diff --git a/modules/ui_postprocessing.py b/modules/ui_postprocessing.py
index 802e1ce71a1..dc08350d126 100644
--- a/modules/ui_postprocessing.py
+++ b/modules/ui_postprocessing.py
@@ -1,16 +1,18 @@
 import gradio as gr
-from modules import scripts, shared, ui_common, postprocessing, call_queue
-import modules.generation_parameters_copypaste as parameters_copypaste
+from modules import scripts, shared, ui_common, postprocessing, call_queue, ui_toprow
+import modules.infotext_utils as parameters_copypaste
+from modules.ui_components import ResizeHandleRow
 
 
 def create_ui():
-    tab_index = gr.State(value=0)
+    dummy_component = gr.Label(visible=False)
+    tab_index = gr.Number(value=0, visible=False)
 
-    with gr.Row(equal_height=False, variant='compact'):
+    with ResizeHandleRow(equal_height=False, variant='compact'):
         with gr.Column(variant='compact'):
             with gr.Tabs(elem_id="mode_extras"):
                 with gr.TabItem('Single Image', id="single_image", elem_id="extras_single_tab") as tab_single:
-                    extras_image = gr.Image(label="Source", source="upload", interactive=True, type="pil", elem_id="extras_image")
+                    extras_image = gr.Image(label="Source", source="upload", interactive=True, type="pil", elem_id="extras_image", image_mode="RGBA")
 
                 with gr.TabItem('Batch Process', id="batch_process", elem_id="extras_batch_process_tab") as tab_batch:
                     image_batch = gr.Files(label="Batch Process", interactive=True, elem_id="extras_image_batch")
@@ -20,20 +22,24 @@ def create_ui():
                     extras_batch_output_dir = gr.Textbox(label="Output directory", **shared.hide_dirs, placeholder="Leave blank to save images to the default path.", elem_id="extras_batch_output_dir")
                     show_extras_results = gr.Checkbox(label='Show result images', value=True, elem_id="extras_show_extras_results")
 
-            submit = gr.Button('Generate', elem_id="extras_generate", variant='primary')
-
             script_inputs = scripts.scripts_postproc.setup_ui()
 
         with gr.Column():
-            result_images, html_info_x, html_info, html_log = ui_common.create_output_panel("extras", shared.opts.outdir_extras_samples)
+            toprow = ui_toprow.Toprow(is_compact=True, is_img2img=False, id_part="extras")
+            toprow.create_inline_toprow_image()
+            submit = toprow.submit
+
+            output_panel = ui_common.create_output_panel("extras", shared.opts.outdir_extras_samples)
 
     tab_single.select(fn=lambda: 0, inputs=[], outputs=[tab_index])
     tab_batch.select(fn=lambda: 1, inputs=[], outputs=[tab_index])
     tab_batch_dir.select(fn=lambda: 2, inputs=[], outputs=[tab_index])
 
     submit.click(
-        fn=call_queue.wrap_gradio_gpu_call(postprocessing.run_postprocessing, extra_outputs=[None, '']),
+        fn=call_queue.wrap_gradio_gpu_call(postprocessing.run_postprocessing_webui, extra_outputs=[None, '']),
+        _js="submit_extras",
         inputs=[
+            dummy_component,
             tab_index,
             extras_image,
             image_batch,
@@ -43,10 +49,11 @@ def create_ui():
             *script_inputs
         ],
         outputs=[
-            result_images,
-            html_info_x,
-            html_info,
-        ]
+            output_panel.gallery,
+            output_panel.generation_info,
+            output_panel.html_log,
+        ],
+        show_progress=False,
     )
 
     parameters_copypaste.add_paste_fields("extras", extras_image, None)
diff --git a/modules/ui_prompt_styles.py b/modules/ui_prompt_styles.py
index 85eb3a6417e..f71b40c419b 100644
--- a/modules/ui_prompt_styles.py
+++ b/modules/ui_prompt_styles.py
@@ -4,6 +4,7 @@
 
 styles_edit_symbol = '\U0001f58c\uFE0F'  # 🖌️
 styles_materialize_symbol = '\U0001f4cb'  # 📋
+styles_copy_symbol = '\U0001f4dd'  # 📝
 
 
 def select_style(name):
@@ -21,9 +22,12 @@ def save_style(name, prompt, negative_prompt):
     if not name:
         return gr.update(visible=False)
 
-    style = styles.PromptStyle(name, prompt, negative_prompt)
+    existing_style = shared.prompt_styles.styles.get(name)
+    path = existing_style.path if existing_style is not None else None
+
+    style = styles.PromptStyle(name, prompt, negative_prompt, path)
     shared.prompt_styles.styles[style.name] = style
-    shared.prompt_styles.save_styles(shared.styles_filename)
+    shared.prompt_styles.save_styles()
 
     return gr.update(visible=True)
 
@@ -33,7 +37,7 @@ def delete_style(name):
         return
 
     shared.prompt_styles.styles.pop(name, None)
-    shared.prompt_styles.save_styles(shared.styles_filename)
+    shared.prompt_styles.save_styles()
 
     return '', '', ''
 
@@ -52,6 +56,8 @@ def refresh_styles():
 class UiPromptStyles:
     def __init__(self, tabname, main_ui_prompt, main_ui_negative_prompt):
         self.tabname = tabname
+        self.main_ui_prompt = main_ui_prompt
+        self.main_ui_negative_prompt = main_ui_negative_prompt
 
         with gr.Row(elem_id=f"{tabname}_styles_row"):
             self.dropdown = gr.Dropdown(label="Styles", show_label=False, elem_id=f"{tabname}_styles", choices=list(shared.prompt_styles.styles), value=[], multiselect=True, tooltip="Styles")
@@ -61,13 +67,14 @@ def __init__(self, tabname, main_ui_prompt, main_ui_negative_prompt):
             with gr.Row():
                 self.selection = gr.Dropdown(label="Styles", elem_id=f"{tabname}_styles_edit_select", choices=list(shared.prompt_styles.styles), value=[], allow_custom_value=True, info="Styles allow you to add custom text to prompt. Use the {prompt} token in style text, and it will be replaced with user's prompt when applying style. Otherwise, style's text will be added to the end of the prompt.")
                 ui_common.create_refresh_button([self.dropdown, self.selection], shared.prompt_styles.reload, lambda: {"choices": list(shared.prompt_styles.styles)}, f"refresh_{tabname}_styles")
-                self.materialize = ui_components.ToolButton(value=styles_materialize_symbol, elem_id=f"{tabname}_style_apply", tooltip="Apply all selected styles from the style selction dropdown in main UI to the prompt.")
+                self.materialize = ui_components.ToolButton(value=styles_materialize_symbol, elem_id=f"{tabname}_style_apply_dialog", tooltip="Apply all selected styles from the style selection dropdown in main UI to the prompt.")
+                self.copy = ui_components.ToolButton(value=styles_copy_symbol, elem_id=f"{tabname}_style_copy", tooltip="Copy main UI prompt to style.")
 
             with gr.Row():
-                self.prompt = gr.Textbox(label="Prompt", show_label=True, elem_id=f"{tabname}_edit_style_prompt", lines=3)
+                self.prompt = gr.Textbox(label="Prompt", show_label=True, elem_id=f"{tabname}_edit_style_prompt", lines=3, elem_classes=["prompt"])
 
             with gr.Row():
-                self.neg_prompt = gr.Textbox(label="Negative prompt", show_label=True, elem_id=f"{tabname}_edit_style_neg_prompt", lines=3)
+                self.neg_prompt = gr.Textbox(label="Negative prompt", show_label=True, elem_id=f"{tabname}_edit_style_neg_prompt", lines=3, elem_classes=["prompt"])
 
             with gr.Row():
                 self.save = gr.Button('Save', variant='primary', elem_id=f'{tabname}_edit_style_save', visible=False)
@@ -96,15 +103,21 @@ def __init__(self, tabname, main_ui_prompt, main_ui_negative_prompt):
             show_progress=False,
         ).then(refresh_styles, outputs=[self.dropdown, self.selection], show_progress=False)
 
-        self.materialize.click(
-            fn=materialize_styles,
-            inputs=[main_ui_prompt, main_ui_negative_prompt, self.dropdown],
-            outputs=[main_ui_prompt, main_ui_negative_prompt, self.dropdown],
+        self.setup_apply_button(self.materialize)
+
+        self.copy.click(
+            fn=lambda p, n: (p, n),
+            inputs=[main_ui_prompt, main_ui_negative_prompt],
+            outputs=[self.prompt, self.neg_prompt],
             show_progress=False,
-        ).then(fn=None, _js="function(){update_"+tabname+"_tokens(); closePopup();}", show_progress=False)
+        )
 
         ui_common.setup_dialog(button_show=edit_button, dialog=styles_dialog, button_close=self.close)
 
-
-
-
+    def setup_apply_button(self, button):
+        button.click(
+            fn=materialize_styles,
+            inputs=[self.main_ui_prompt, self.main_ui_negative_prompt, self.dropdown],
+            outputs=[self.main_ui_prompt, self.main_ui_negative_prompt, self.dropdown],
+            show_progress=False,
+        ).then(fn=None, _js="function(){update_"+self.tabname+"_tokens(); closePopup();}", show_progress=False)
diff --git a/modules/ui_settings.py b/modules/ui_settings.py
index 8ff9c074718..e53ad50f8f4 100644
--- a/modules/ui_settings.py
+++ b/modules/ui_settings.py
@@ -1,10 +1,12 @@
 import gradio as gr
 
-from modules import ui_common, shared, script_callbacks, scripts, sd_models, sysinfo
-from modules.call_queue import wrap_gradio_call
+from modules import ui_common, shared, script_callbacks, scripts, sd_models, sysinfo, timer, shared_items
+from modules.call_queue import wrap_gradio_call_no_job
+from modules.options import options_section
 from modules.shared import opts
 from modules.ui_components import FormRow
 from modules.ui_gradio_extensions import reload_javascript
+from concurrent.futures import ThreadPoolExecutor, as_completed
 
 
 def get_value_for_setting(key):
@@ -63,6 +65,9 @@ class UiSettings:
     quicksettings_list = None
     quicksettings_names = None
     text_settings = None
+    show_all_pages = None
+    show_one_page = None
+    search_input = None
 
     def run_settings(self, *args):
         changed = []
@@ -94,6 +99,9 @@ def run_settings_single(self, value, key):
 
         return get_value_for_setting(key), opts.dumpjson()
 
+    def register_settings(self):
+        script_callbacks.ui_settings_callback()
+
     def create_ui(self, loadsave, dummy_component):
         self.components = []
         self.component_dict = {}
@@ -101,7 +109,11 @@ def create_ui(self, loadsave, dummy_component):
 
         shared.settings_components = self.component_dict
 
-        script_callbacks.ui_settings_callback()
+        # we add this as late as possible so that scripts have already registered their callbacks
+        opts.data_labels.update(options_section(('callbacks', "Callbacks", "system"), {
+            **shared_items.callbacks_order_settings(),
+        }))
+
         opts.reorder()
 
         with gr.Blocks(analytics_enabled=False) as settings_interface:
@@ -135,7 +147,7 @@ def create_ui(self, loadsave, dummy_component):
                         gr.Group()
                         current_tab = gr.TabItem(elem_id=f"settings_{elem_id}", label=text)
                         current_tab.__enter__()
-                        current_row = gr.Column(variant='compact')
+                        current_row = gr.Column(elem_id=f"column_settings_{elem_id}", variant='compact')
                         current_row.__enter__()
 
                         previous_section = item.section
@@ -173,26 +185,43 @@ def create_ui(self, loadsave, dummy_component):
                     download_localization = gr.Button(value='Download localization template', elem_id="download_localization")
                     reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary', elem_id="settings_reload_script_bodies")
                     with gr.Row():
-                        unload_sd_model = gr.Button(value='Unload SD checkpoint to free VRAM', elem_id="sett_unload_sd_model")
-                        reload_sd_model = gr.Button(value='Reload the last SD checkpoint back into VRAM', elem_id="sett_reload_sd_model")
+                        unload_sd_model = gr.Button(value='Unload SD checkpoint to RAM', elem_id="sett_unload_sd_model")
+                        reload_sd_model = gr.Button(value='Load SD checkpoint to VRAM from RAM', elem_id="sett_reload_sd_model")
+                    with gr.Row():
+                        calculate_all_checkpoint_hash = gr.Button(value='Calculate hash for all checkpoint', elem_id="calculate_all_checkpoint_hash")
+                        calculate_all_checkpoint_hash_threads = gr.Number(value=1, label="Number of parallel calculations", elem_id="calculate_all_checkpoint_hash_threads", precision=0, minimum=1)
 
                 with gr.TabItem("Licenses", id="licenses", elem_id="settings_tab_licenses"):
                     gr.HTML(shared.html("licenses.html"), elem_id="licenses")
 
-                gr.Button(value="Show all pages", elem_id="settings_show_all_pages")
+                self.show_all_pages = gr.Button(value="Show all pages", elem_id="settings_show_all_pages")
+                self.show_one_page = gr.Button(value="Show only one page", elem_id="settings_show_one_page", visible=False)
+                self.show_one_page.click(lambda: None)
+
+                self.search_input = gr.Textbox(value="", elem_id="settings_search", max_lines=1, placeholder="Search...", show_label=False)
 
                 self.text_settings = gr.Textbox(elem_id="settings_json", value=lambda: opts.dumpjson(), visible=False)
 
+            def call_func_and_return_text(func, text):
+                def handler():
+                    t = timer.Timer()
+                    func()
+                    t.record(text)
+
+                    return f'{text} in {t.total:.1f}s'
+
+                return handler
+
             unload_sd_model.click(
-                fn=sd_models.unload_model_weights,
+                fn=call_func_and_return_text(sd_models.unload_model_weights, 'Unloaded the checkpoint'),
                 inputs=[],
-                outputs=[]
+                outputs=[self.result]
             )
 
             reload_sd_model.click(
-                fn=sd_models.reload_model_weights,
+                fn=call_func_and_return_text(lambda: sd_models.send_model_to_device(shared.sd_model), 'Loaded the checkpoint'),
                 inputs=[],
-                outputs=[]
+                outputs=[self.result]
             )
 
             request_notifications.click(
@@ -241,6 +270,21 @@ def check_file(x):
                 outputs=[sysinfo_check_output],
             )
 
+            def calculate_all_checkpoint_hash_fn(max_thread):
+                checkpoints_list = sd_models.checkpoints_list.values()
+                with ThreadPoolExecutor(max_workers=max_thread) as executor:
+                    futures = [executor.submit(checkpoint.calculate_shorthash) for checkpoint in checkpoints_list]
+                    completed = 0
+                    for _ in as_completed(futures):
+                        completed += 1
+                        print(f"{completed} / {len(checkpoints_list)} ")
+                    print("Finish calculating hash for all checkpoints")
+
+            calculate_all_checkpoint_hash.click(
+                fn=calculate_all_checkpoint_hash_fn,
+                inputs=[calculate_all_checkpoint_hash_threads],
+            )
+
         self.interface = settings_interface
 
     def add_quicksettings(self):
@@ -251,7 +295,7 @@ def add_quicksettings(self):
 
     def add_functionality(self, demo):
         self.submit.click(
-            fn=wrap_gradio_call(lambda *args: self.run_settings(*args), extra_outputs=[gr.update()]),
+            fn=wrap_gradio_call_no_job(lambda *args: self.run_settings(*args), extra_outputs=[gr.update()]),
             inputs=self.components,
             outputs=[self.text_settings, self.result],
         )
@@ -294,3 +338,8 @@ def get_settings_values():
             outputs=[self.component_dict[k] for k in component_keys],
             queue=False,
         )
+
+    def search(self, text):
+        print(text)
+
+        return [gr.update(visible=text in (comp.label or "")) for comp in self.components]
diff --git a/modules/ui_tempdir.py b/modules/ui_tempdir.py
index 85015db56b5..ecd6bdec355 100644
--- a/modules/ui_tempdir.py
+++ b/modules/ui_tempdir.py
@@ -35,12 +35,9 @@ def save_pil_to_file(self, pil_image, dir=None, format="png"):
     already_saved_as = getattr(pil_image, 'already_saved_as', None)
     if already_saved_as and os.path.isfile(already_saved_as):
         register_tmp_file(shared.demo, already_saved_as)
-        filename = already_saved_as
-
-        if not shared.opts.save_images_add_number:
-            filename += f'?{os.path.getmtime(already_saved_as)}'
-
-        return filename
+        filename_with_mtime = f'{already_saved_as}?{os.path.getmtime(already_saved_as)}'
+        register_tmp_file(shared.demo, filename_with_mtime)
+        return filename_with_mtime
 
     if shared.opts.temp_dir != "":
         dir = shared.opts.temp_dir
@@ -86,3 +83,18 @@ def cleanup_tmpdr():
 
             filename = os.path.join(root, name)
             os.remove(filename)
+
+
+def is_gradio_temp_path(path):
+    """
+    Check if the path is a temp dir used by gradio
+    """
+    path = Path(path)
+    if shared.opts.temp_dir and path.is_relative_to(shared.opts.temp_dir):
+        return True
+    if gradio_temp_dir := os.environ.get("GRADIO_TEMP_DIR"):
+        if path.is_relative_to(gradio_temp_dir):
+            return True
+    if path.is_relative_to(Path(tempfile.gettempdir()) / "gradio"):
+        return True
+    return False
diff --git a/modules/ui_toprow.py b/modules/ui_toprow.py
new file mode 100644
index 00000000000..dc3c3aa3837
--- /dev/null
+++ b/modules/ui_toprow.py
@@ -0,0 +1,144 @@
+import gradio as gr
+
+from modules import shared, ui_prompt_styles
+import modules.images
+
+from modules.ui_components import ToolButton
+
+
+class Toprow:
+    """Creates a top row UI with prompts, generate button, styles, extra little buttons for things, and enables some functionality related to their operation"""
+
+    prompt = None
+    prompt_img = None
+    negative_prompt = None
+
+    button_interrogate = None
+    button_deepbooru = None
+
+    interrupt = None
+    interrupting = None
+    skip = None
+    submit = None
+
+    paste = None
+    clear_prompt_button = None
+    apply_styles = None
+    restore_progress_button = None
+
+    token_counter = None
+    token_button = None
+    negative_token_counter = None
+    negative_token_button = None
+
+    ui_styles = None
+
+    submit_box = None
+
+    def __init__(self, is_img2img, is_compact=False, id_part=None):
+        if id_part is None:
+            id_part = "img2img" if is_img2img else "txt2img"
+
+        self.id_part = id_part
+        self.is_img2img = is_img2img
+        self.is_compact = is_compact
+
+        if not is_compact:
+            with gr.Row(elem_id=f"{id_part}_toprow", variant="compact"):
+                self.create_classic_toprow()
+        else:
+            self.create_submit_box()
+
+    def create_classic_toprow(self):
+        self.create_prompts()
+
+        with gr.Column(scale=1, elem_id=f"{self.id_part}_actions_column"):
+            self.create_submit_box()
+
+            self.create_tools_row()
+
+            self.create_styles_ui()
+
+    def create_inline_toprow_prompts(self):
+        if not self.is_compact:
+            return
+
+        self.create_prompts()
+
+        with gr.Row(elem_classes=["toprow-compact-stylerow"]):
+            with gr.Column(elem_classes=["toprow-compact-tools"]):
+                self.create_tools_row()
+            with gr.Column():
+                self.create_styles_ui()
+
+    def create_inline_toprow_image(self):
+        if not self.is_compact:
+            return
+
+        self.submit_box.render()
+
+    def create_prompts(self):
+        with gr.Column(elem_id=f"{self.id_part}_prompt_container", elem_classes=["prompt-container-compact"] if self.is_compact else [], scale=6):
+            with gr.Row(elem_id=f"{self.id_part}_prompt_row", elem_classes=["prompt-row"]):
+                self.prompt = gr.Textbox(label="Prompt", elem_id=f"{self.id_part}_prompt", show_label=False, lines=3, placeholder="Prompt\n(Press Ctrl+Enter to generate, Alt+Enter to skip, Esc to interrupt)", elem_classes=["prompt"])
+                self.prompt_img = gr.File(label="", elem_id=f"{self.id_part}_prompt_image", file_count="single", type="binary", visible=False)
+
+            with gr.Row(elem_id=f"{self.id_part}_neg_prompt_row", elem_classes=["prompt-row"]):
+                self.negative_prompt = gr.Textbox(label="Negative prompt", elem_id=f"{self.id_part}_neg_prompt", show_label=False, lines=3, placeholder="Negative prompt\n(Press Ctrl+Enter to generate, Alt+Enter to skip, Esc to interrupt)", elem_classes=["prompt"])
+
+        self.prompt_img.change(
+            fn=modules.images.image_data,
+            inputs=[self.prompt_img],
+            outputs=[self.prompt, self.prompt_img],
+            show_progress=False,
+        )
+
+    def create_submit_box(self):
+        with gr.Row(elem_id=f"{self.id_part}_generate_box", elem_classes=["generate-box"] + (["generate-box-compact"] if self.is_compact else []), render=not self.is_compact) as submit_box:
+            self.submit_box = submit_box
+
+            self.interrupt = gr.Button('Interrupt', elem_id=f"{self.id_part}_interrupt", elem_classes="generate-box-interrupt", tooltip="End generation immediately or after completing current batch")
+            self.skip = gr.Button('Skip', elem_id=f"{self.id_part}_skip", elem_classes="generate-box-skip", tooltip="Stop generation of current batch and continues onto next batch")
+            self.interrupting = gr.Button('Interrupting...', elem_id=f"{self.id_part}_interrupting", elem_classes="generate-box-interrupting", tooltip="Interrupting generation...")
+            self.submit = gr.Button('Generate', elem_id=f"{self.id_part}_generate", variant='primary', tooltip="Right click generate forever menu")
+
+            def interrupt_function():
+                if not shared.state.stopping_generation and shared.state.job_count > 1 and shared.opts.interrupt_after_current:
+                    shared.state.stop_generating()
+                    gr.Info("Generation will stop after finishing this image, click again to stop immediately.")
+                else:
+                    shared.state.interrupt()
+
+            self.skip.click(fn=shared.state.skip)
+            self.interrupt.click(fn=interrupt_function, _js='function(){ showSubmitInterruptingPlaceholder("' + self.id_part + '"); }')
+            self.interrupting.click(fn=interrupt_function)
+
+    def create_tools_row(self):
+        with gr.Row(elem_id=f"{self.id_part}_tools"):
+            from modules.ui import paste_symbol, clear_prompt_symbol, restore_progress_symbol
+
+            self.paste = ToolButton(value=paste_symbol, elem_id="paste", tooltip="Read generation parameters from prompt or last generation if prompt is empty into user interface.")
+            self.clear_prompt_button = ToolButton(value=clear_prompt_symbol, elem_id=f"{self.id_part}_clear_prompt", tooltip="Clear prompt")
+            self.apply_styles = ToolButton(value=ui_prompt_styles.styles_materialize_symbol, elem_id=f"{self.id_part}_style_apply", tooltip="Apply all selected styles to prompts.")
+
+            if self.is_img2img:
+                self.button_interrogate = ToolButton('📎', tooltip='Interrogate CLIP - use CLIP neural network to create a text describing the image, and put it into the prompt field', elem_id="interrogate")
+                self.button_deepbooru = ToolButton('📦', tooltip='Interrogate DeepBooru - use DeepBooru neural network to create a text describing the image, and put it into the prompt field', elem_id="deepbooru")
+
+            self.restore_progress_button = ToolButton(value=restore_progress_symbol, elem_id=f"{self.id_part}_restore_progress", visible=False, tooltip="Restore progress")
+
+            self.token_counter = gr.HTML(value="<span>0/75</span>", elem_id=f"{self.id_part}_token_counter", elem_classes=["token-counter"], visible=False)
+            self.token_button = gr.Button(visible=False, elem_id=f"{self.id_part}_token_button")
+            self.negative_token_counter = gr.HTML(value="<span>0/75</span>", elem_id=f"{self.id_part}_negative_token_counter", elem_classes=["token-counter"], visible=False)
+            self.negative_token_button = gr.Button(visible=False, elem_id=f"{self.id_part}_negative_token_button")
+
+            self.clear_prompt_button.click(
+                fn=lambda *x: x,
+                _js="confirm_clear_prompt",
+                inputs=[self.prompt, self.negative_prompt],
+                outputs=[self.prompt, self.negative_prompt],
+            )
+
+    def create_styles_ui(self):
+        self.ui_styles = ui_prompt_styles.UiPromptStyles(self.id_part, self.prompt, self.negative_prompt)
+        self.ui_styles.setup_apply_button(self.apply_styles)
diff --git a/modules/upscaler.py b/modules/upscaler.py
index e682bbaa26c..507881fede2 100644
--- a/modules/upscaler.py
+++ b/modules/upscaler.py
@@ -20,7 +20,7 @@ class Upscaler:
     filter = None
     model = None
     user_path = None
-    scalers: []
+    scalers: list
     tile = True
 
     def __init__(self, create_dirs=False):
@@ -56,7 +56,13 @@ def upscale(self, img: PIL.Image, scale, selected_model: str = None):
         dest_w = int((img.width * scale) // 8 * 8)
         dest_h = int((img.height * scale) // 8 * 8)
 
-        for _ in range(3):
+        for i in range(3):
+            if img.width >= dest_w and img.height >= dest_h and (i > 0 or scale != 1):
+                break
+
+            if shared.state.interrupted:
+                break
+
             shape = (img.width, img.height)
 
             img = self.do_upscale(img, selected_model)
@@ -64,9 +70,6 @@ def upscale(self, img: PIL.Image, scale, selected_model: str = None):
             if shape == (img.width, img.height):
                 break
 
-            if img.width >= dest_w and img.height >= dest_h:
-                break
-
         if img.width != dest_w or img.height != dest_h:
             img = img.resize((int(dest_w), int(dest_h)), resample=LANCZOS)
 
@@ -98,6 +101,9 @@ def __init__(self, name: str, path: str, upscaler: Upscaler = None, scale: int =
         self.scale = scale
         self.model = model
 
+    def __repr__(self):
+        return f"<UpscalerData name={self.name} path={self.data_path} scale={self.scale}>"
+
 
 class UpscalerNone(Upscaler):
     name = "None"
diff --git a/modules/upscaler_utils.py b/modules/upscaler_utils.py
new file mode 100644
index 00000000000..a8408f05bca
--- /dev/null
+++ b/modules/upscaler_utils.py
@@ -0,0 +1,190 @@
+import logging
+from typing import Callable
+
+import numpy as np
+import torch
+import tqdm
+from PIL import Image
+
+from modules import devices, images, shared, torch_utils
+
+logger = logging.getLogger(__name__)
+
+
+def pil_image_to_torch_bgr(img: Image.Image) -> torch.Tensor:
+    img = np.array(img.convert("RGB"))
+    img = img[:, :, ::-1]  # flip RGB to BGR
+    img = np.transpose(img, (2, 0, 1))  # HWC to CHW
+    img = np.ascontiguousarray(img) / 255  # Rescale to [0, 1]
+    return torch.from_numpy(img)
+
+
+def torch_bgr_to_pil_image(tensor: torch.Tensor) -> Image.Image:
+    if tensor.ndim == 4:
+        # If we're given a tensor with a batch dimension, squeeze it out
+        # (but only if it's a batch of size 1).
+        if tensor.shape[0] != 1:
+            raise ValueError(f"{tensor.shape} does not describe a BCHW tensor")
+        tensor = tensor.squeeze(0)
+    assert tensor.ndim == 3, f"{tensor.shape} does not describe a CHW tensor"
+    # TODO: is `tensor.float().cpu()...numpy()` the most efficient idiom?
+    arr = tensor.float().cpu().clamp_(0, 1).numpy()  # clamp
+    arr = 255.0 * np.moveaxis(arr, 0, 2)  # CHW to HWC, rescale
+    arr = arr.round().astype(np.uint8)
+    arr = arr[:, :, ::-1]  # flip BGR to RGB
+    return Image.fromarray(arr, "RGB")
+
+
+def upscale_pil_patch(model, img: Image.Image) -> Image.Image:
+    """
+    Upscale a given PIL image using the given model.
+    """
+    param = torch_utils.get_param(model)
+
+    with torch.inference_mode():
+        tensor = pil_image_to_torch_bgr(img).unsqueeze(0)  # add batch dimension
+        tensor = tensor.to(device=param.device, dtype=param.dtype)
+        with devices.without_autocast():
+            return torch_bgr_to_pil_image(model(tensor))
+
+
+def upscale_with_model(
+    model: Callable[[torch.Tensor], torch.Tensor],
+    img: Image.Image,
+    *,
+    tile_size: int,
+    tile_overlap: int = 0,
+    desc="tiled upscale",
+) -> Image.Image:
+    if tile_size <= 0:
+        logger.debug("Upscaling %s without tiling", img)
+        output = upscale_pil_patch(model, img)
+        logger.debug("=> %s", output)
+        return output
+
+    grid = images.split_grid(img, tile_size, tile_size, tile_overlap)
+    newtiles = []
+
+    with tqdm.tqdm(total=grid.tile_count, desc=desc, disable=not shared.opts.enable_upscale_progressbar) as p:
+        for y, h, row in grid.tiles:
+            newrow = []
+            for x, w, tile in row:
+                if shared.state.interrupted:
+                    return img
+                output = upscale_pil_patch(model, tile)
+                scale_factor = output.width // tile.width
+                newrow.append([x * scale_factor, w * scale_factor, output])
+                p.update(1)
+            newtiles.append([y * scale_factor, h * scale_factor, newrow])
+
+    newgrid = images.Grid(
+        newtiles,
+        tile_w=grid.tile_w * scale_factor,
+        tile_h=grid.tile_h * scale_factor,
+        image_w=grid.image_w * scale_factor,
+        image_h=grid.image_h * scale_factor,
+        overlap=grid.overlap * scale_factor,
+    )
+    return images.combine_grid(newgrid)
+
+
+def tiled_upscale_2(
+    img: torch.Tensor,
+    model,
+    *,
+    tile_size: int,
+    tile_overlap: int,
+    scale: int,
+    device: torch.device,
+    desc="Tiled upscale",
+):
+    # Alternative implementation of `upscale_with_model` originally used by
+    # SwinIR and ScuNET.  It differs from `upscale_with_model` in that tiling and
+    # weighting is done in PyTorch space, as opposed to `images.Grid` doing it in
+    # Pillow space without weighting.
+
+    b, c, h, w = img.size()
+    tile_size = min(tile_size, h, w)
+
+    if tile_size <= 0:
+        logger.debug("Upscaling %s without tiling", img.shape)
+        return model(img)
+
+    stride = tile_size - tile_overlap
+    h_idx_list = list(range(0, h - tile_size, stride)) + [h - tile_size]
+    w_idx_list = list(range(0, w - tile_size, stride)) + [w - tile_size]
+    result = torch.zeros(
+        b,
+        c,
+        h * scale,
+        w * scale,
+        device=device,
+        dtype=img.dtype,
+    )
+    weights = torch.zeros_like(result)
+    logger.debug("Upscaling %s to %s with tiles", img.shape, result.shape)
+    with tqdm.tqdm(total=len(h_idx_list) * len(w_idx_list), desc=desc, disable=not shared.opts.enable_upscale_progressbar) as pbar:
+        for h_idx in h_idx_list:
+            if shared.state.interrupted or shared.state.skipped:
+                break
+
+            for w_idx in w_idx_list:
+                if shared.state.interrupted or shared.state.skipped:
+                    break
+
+                # Only move this patch to the device if it's not already there.
+                in_patch = img[
+                    ...,
+                    h_idx : h_idx + tile_size,
+                    w_idx : w_idx + tile_size,
+                ].to(device=device)
+
+                out_patch = model(in_patch)
+
+                result[
+                    ...,
+                    h_idx * scale : (h_idx + tile_size) * scale,
+                    w_idx * scale : (w_idx + tile_size) * scale,
+                ].add_(out_patch)
+
+                out_patch_mask = torch.ones_like(out_patch)
+
+                weights[
+                    ...,
+                    h_idx * scale : (h_idx + tile_size) * scale,
+                    w_idx * scale : (w_idx + tile_size) * scale,
+                ].add_(out_patch_mask)
+
+                pbar.update(1)
+
+    output = result.div_(weights)
+
+    return output
+
+
+def upscale_2(
+    img: Image.Image,
+    model,
+    *,
+    tile_size: int,
+    tile_overlap: int,
+    scale: int,
+    desc: str,
+):
+    """
+    Convenience wrapper around `tiled_upscale_2` that handles PIL images.
+    """
+    param = torch_utils.get_param(model)
+    tensor = pil_image_to_torch_bgr(img).to(dtype=param.dtype).unsqueeze(0)  # add batch dimension
+
+    with torch.no_grad():
+        output = tiled_upscale_2(
+            tensor,
+            model,
+            tile_size=tile_size,
+            tile_overlap=tile_overlap,
+            scale=scale,
+            desc=desc,
+            device=param.device,
+        )
+    return torch_bgr_to_pil_image(output)
diff --git a/modules/util.py b/modules/util.py
index 60afc0670c7..7911b0db72c 100644
--- a/modules/util.py
+++ b/modules/util.py
@@ -2,7 +2,7 @@
 import re
 
 from modules import shared
-from modules.paths_internal import script_path
+from modules.paths_internal import script_path, cwd
 
 
 def natural_sort_key(s, regex=re.compile('([0-9]+)')):
@@ -21,11 +21,11 @@ def html_path(filename):
 def html(filename):
     path = html_path(filename)
 
-    if os.path.exists(path):
+    try:
         with open(path, encoding="utf8") as file:
             return file.read()
-
-    return ""
+    except OSError:
+        return ""
 
 
 def walk_files(path, allowed_extensions=None):
@@ -42,7 +42,7 @@ def walk_files(path, allowed_extensions=None):
         for filename in sorted(files, key=natural_sort_key):
             if allowed_extensions is not None:
                 _, ext = os.path.splitext(filename)
-                if ext not in allowed_extensions:
+                if ext.lower() not in allowed_extensions:
                     continue
 
             if not shared.opts.list_hidden_files and ("/." in root or "\\." in root):
@@ -56,3 +56,158 @@ def ldm_print(*args, **kwargs):
         return
 
     print(*args, **kwargs)
+
+
+def truncate_path(target_path, base_path=cwd):
+    abs_target, abs_base = os.path.abspath(target_path), os.path.abspath(base_path)
+    try:
+        if os.path.commonpath([abs_target, abs_base]) == abs_base:
+            return os.path.relpath(abs_target, abs_base)
+    except ValueError:
+        pass
+    return abs_target
+
+
+class MassFileListerCachedDir:
+    """A class that caches file metadata for a specific directory."""
+
+    def __init__(self, dirname):
+        self.files = None
+        self.files_cased = None
+        self.dirname = dirname
+
+        stats = ((x.name, x.stat(follow_symlinks=False)) for x in os.scandir(self.dirname))
+        files = [(n, s.st_mtime, s.st_ctime) for n, s in stats]
+        self.files = {x[0].lower(): x for x in files}
+        self.files_cased = {x[0]: x for x in files}
+
+    def update_entry(self, filename):
+        """Add a file to the cache"""
+        file_path = os.path.join(self.dirname, filename)
+        try:
+            stat = os.stat(file_path)
+            entry = (filename, stat.st_mtime, stat.st_ctime)
+            self.files[filename.lower()] = entry
+            self.files_cased[filename] = entry
+        except FileNotFoundError as e:
+            print(f'MassFileListerCachedDir.add_entry: "{file_path}" {e}')
+
+
+class MassFileLister:
+    """A class that provides a way to check for the existence and mtime/ctile of files without doing more than one stat call per file."""
+
+    def __init__(self):
+        self.cached_dirs = {}
+
+    def find(self, path):
+        """
+        Find the metadata for a file at the given path.
+
+        Returns:
+            tuple or None: A tuple of (name, mtime, ctime) if the file exists, or None if it does not.
+        """
+
+        dirname, filename = os.path.split(path)
+
+        cached_dir = self.cached_dirs.get(dirname)
+        if cached_dir is None:
+            cached_dir = MassFileListerCachedDir(dirname)
+            self.cached_dirs[dirname] = cached_dir
+
+        stats = cached_dir.files_cased.get(filename)
+        if stats is not None:
+            return stats
+
+        stats = cached_dir.files.get(filename.lower())
+        if stats is None:
+            return None
+
+        try:
+            os_stats = os.stat(path, follow_symlinks=False)
+            return filename, os_stats.st_mtime, os_stats.st_ctime
+        except Exception:
+            return None
+
+    def exists(self, path):
+        """Check if a file exists at the given path."""
+
+        return self.find(path) is not None
+
+    def mctime(self, path):
+        """
+        Get the modification and creation times for a file at the given path.
+
+        Returns:
+            tuple: A tuple of (mtime, ctime) if the file exists, or (0, 0) if it does not.
+        """
+
+        stats = self.find(path)
+        return (0, 0) if stats is None else stats[1:3]
+
+    def reset(self):
+        """Clear the cache of all directories."""
+        self.cached_dirs.clear()
+
+    def update_file_entry(self, path):
+        """Update the cache for a specific directory."""
+        dirname, filename = os.path.split(path)
+        if cached_dir := self.cached_dirs.get(dirname):
+            cached_dir.update_entry(filename)
+
+def topological_sort(dependencies):
+    """Accepts a dictionary mapping name to its dependencies, returns a list of names ordered according to dependencies.
+    Ignores errors relating to missing dependencies or circular dependencies
+    """
+
+    visited = {}
+    result = []
+
+    def inner(name):
+        visited[name] = True
+
+        for dep in dependencies.get(name, []):
+            if dep in dependencies and dep not in visited:
+                inner(dep)
+
+        result.append(name)
+
+    for depname in dependencies:
+        if depname not in visited:
+            inner(depname)
+
+    return result
+
+
+def open_folder(path):
+    """Open a folder in the file manager of the respect OS."""
+    # import at function level to avoid potential issues
+    import gradio as gr
+    import platform
+    import sys
+    import subprocess
+
+    if not os.path.exists(path):
+        msg = f'Folder "{path}" does not exist. after you save an image, the folder will be created.'
+        print(msg)
+        gr.Info(msg)
+        return
+    elif not os.path.isdir(path):
+        msg = f"""
+WARNING
+An open_folder request was made with an path that is not a folder.
+This could be an error or a malicious attempt to run code on your computer.
+Requested path was: {path}
+"""
+        print(msg, file=sys.stderr)
+        gr.Warning(msg)
+        return
+
+    path = os.path.normpath(path)
+    if platform.system() == "Windows":
+        os.startfile(path)
+    elif platform.system() == "Darwin":
+        subprocess.Popen(["open", path])
+    elif "microsoft-standard-WSL2" in platform.uname().release:
+        subprocess.Popen(["explorer.exe", subprocess.check_output(["wslpath", "-w", path])])
+    else:
+        subprocess.Popen(["xdg-open", path])
diff --git a/modules/xlmr.py b/modules/xlmr.py
index a407a3cade8..319771b7bf0 100644
--- a/modules/xlmr.py
+++ b/modules/xlmr.py
@@ -5,6 +5,9 @@
 from transformers import XLMRobertaModel,XLMRobertaTokenizer
 from typing import Optional
 
+from modules import torch_utils
+
+
 class BertSeriesConfig(BertConfig):
     def __init__(self, vocab_size=30522, hidden_size=768, num_hidden_layers=12, num_attention_heads=12, intermediate_size=3072, hidden_act="gelu", hidden_dropout_prob=0.1, attention_probs_dropout_prob=0.1, max_position_embeddings=512, type_vocab_size=2, initializer_range=0.02, layer_norm_eps=1e-12, pad_token_id=0, position_embedding_type="absolute", use_cache=True, classifier_dropout=None,project_dim=512, pooler_fn="average",learn_encoder=False,model_type='bert',**kwargs):
 
@@ -62,7 +65,7 @@ def __init__(self, config=None, **kargs):
         self.post_init()
 
     def encode(self,c):
-        device = next(self.parameters()).device
+        device = torch_utils.get_param(self).device
         text = self.tokenizer(c,
                         truncation=True,
                         max_length=77,
diff --git a/modules/xlmr_m18.py b/modules/xlmr_m18.py
new file mode 100644
index 00000000000..f60555049f5
--- /dev/null
+++ b/modules/xlmr_m18.py
@@ -0,0 +1,166 @@
+from transformers import BertPreTrainedModel,BertConfig
+import torch.nn as nn
+import torch
+from transformers.models.xlm_roberta.configuration_xlm_roberta import XLMRobertaConfig
+from transformers import XLMRobertaModel,XLMRobertaTokenizer
+from typing import Optional
+from modules import torch_utils
+
+
+class BertSeriesConfig(BertConfig):
+    def __init__(self, vocab_size=30522, hidden_size=768, num_hidden_layers=12, num_attention_heads=12, intermediate_size=3072, hidden_act="gelu", hidden_dropout_prob=0.1, attention_probs_dropout_prob=0.1, max_position_embeddings=512, type_vocab_size=2, initializer_range=0.02, layer_norm_eps=1e-12, pad_token_id=0, position_embedding_type="absolute", use_cache=True, classifier_dropout=None,project_dim=512, pooler_fn="average",learn_encoder=False,model_type='bert',**kwargs):
+
+        super().__init__(vocab_size, hidden_size, num_hidden_layers, num_attention_heads, intermediate_size, hidden_act, hidden_dropout_prob, attention_probs_dropout_prob, max_position_embeddings, type_vocab_size, initializer_range, layer_norm_eps, pad_token_id, position_embedding_type, use_cache, classifier_dropout, **kwargs)
+        self.project_dim = project_dim
+        self.pooler_fn = pooler_fn
+        self.learn_encoder = learn_encoder
+
+class RobertaSeriesConfig(XLMRobertaConfig):
+    def __init__(self, pad_token_id=1, bos_token_id=0, eos_token_id=2,project_dim=512,pooler_fn='cls',learn_encoder=False, **kwargs):
+        super().__init__(pad_token_id=pad_token_id, bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)
+        self.project_dim = project_dim
+        self.pooler_fn = pooler_fn
+        self.learn_encoder = learn_encoder
+
+
+class BertSeriesModelWithTransformation(BertPreTrainedModel):
+
+    _keys_to_ignore_on_load_unexpected = [r"pooler"]
+    _keys_to_ignore_on_load_missing = [r"position_ids", r"predictions.decoder.bias"]
+    config_class = BertSeriesConfig
+
+    def __init__(self, config=None, **kargs):
+        # modify initialization for autoloading
+        if config is None:
+            config = XLMRobertaConfig()
+            config.attention_probs_dropout_prob= 0.1
+            config.bos_token_id=0
+            config.eos_token_id=2
+            config.hidden_act='gelu'
+            config.hidden_dropout_prob=0.1
+            config.hidden_size=1024
+            config.initializer_range=0.02
+            config.intermediate_size=4096
+            config.layer_norm_eps=1e-05
+            config.max_position_embeddings=514
+
+            config.num_attention_heads=16
+            config.num_hidden_layers=24
+            config.output_past=True
+            config.pad_token_id=1
+            config.position_embedding_type= "absolute"
+
+            config.type_vocab_size= 1
+            config.use_cache=True
+            config.vocab_size= 250002
+            config.project_dim = 1024
+            config.learn_encoder = False
+        super().__init__(config)
+        self.roberta = XLMRobertaModel(config)
+        self.transformation = nn.Linear(config.hidden_size,config.project_dim)
+        # self.pre_LN=nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.tokenizer = XLMRobertaTokenizer.from_pretrained('xlm-roberta-large')
+        # self.pooler = lambda x: x[:,0]
+        # self.post_init()
+
+        self.has_pre_transformation = True
+        if self.has_pre_transformation:
+            self.transformation_pre = nn.Linear(config.hidden_size, config.project_dim)
+            self.pre_LN = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.post_init()
+
+    def encode(self,c):
+        device = torch_utils.get_param(self).device
+        text = self.tokenizer(c,
+                        truncation=True,
+                        max_length=77,
+                        return_length=False,
+                        return_overflowing_tokens=False,
+                        padding="max_length",
+                        return_tensors="pt")
+        text["input_ids"] = torch.tensor(text["input_ids"]).to(device)
+        text["attention_mask"] = torch.tensor(
+            text['attention_mask']).to(device)
+        features = self(**text)
+        return features['projection_state']
+
+    def forward(
+        self,
+        input_ids: Optional[torch.Tensor] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        token_type_ids: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.Tensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        inputs_embeds: Optional[torch.Tensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.Tensor] = None,
+        output_attentions: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+    ) :
+        r"""
+        """
+
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+
+        outputs = self.roberta(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            output_attentions=output_attentions,
+            output_hidden_states=True,
+            return_dict=return_dict,
+        )
+
+        # # last module outputs
+        # sequence_output = outputs[0]
+
+
+        # # project every module
+        # sequence_output_ln = self.pre_LN(sequence_output)
+
+        # # pooler
+        # pooler_output = self.pooler(sequence_output_ln)
+        # pooler_output = self.transformation(pooler_output)
+        # projection_state = self.transformation(outputs.last_hidden_state)
+
+        if self.has_pre_transformation:
+            sequence_output2 = outputs["hidden_states"][-2]
+            sequence_output2 = self.pre_LN(sequence_output2)
+            projection_state2 = self.transformation_pre(sequence_output2)
+
+            return {
+                "projection_state": projection_state2,
+                "last_hidden_state": outputs.last_hidden_state,
+                "hidden_states": outputs.hidden_states,
+                "attentions": outputs.attentions,
+            }
+        else:
+            projection_state = self.transformation(outputs.last_hidden_state)
+            return {
+                "projection_state": projection_state,
+                "last_hidden_state": outputs.last_hidden_state,
+                "hidden_states": outputs.hidden_states,
+                "attentions": outputs.attentions,
+            }
+
+
+        # return {
+        #     'pooler_output':pooler_output,
+        #     'last_hidden_state':outputs.last_hidden_state,
+        #     'hidden_states':outputs.hidden_states,
+        #     'attentions':outputs.attentions,
+        #     'projection_state':projection_state,
+        #     'sequence_out': sequence_output
+        # }
+
+
+class RobertaSeriesModelWithTransformation(BertSeriesModelWithTransformation):
+    base_model_prefix = 'roberta'
+    config_class= RobertaSeriesConfig
diff --git a/modules/xpu_specific.py b/modules/xpu_specific.py
new file mode 100644
index 00000000000..2971dbc3cf5
--- /dev/null
+++ b/modules/xpu_specific.py
@@ -0,0 +1,138 @@
+from modules import shared
+from modules.sd_hijack_utils import CondFunc
+
+has_ipex = False
+try:
+    import torch
+    import intel_extension_for_pytorch as ipex # noqa: F401
+    has_ipex = True
+except Exception:
+    pass
+
+
+def check_for_xpu():
+    return has_ipex and hasattr(torch, 'xpu') and torch.xpu.is_available()
+
+
+def get_xpu_device_string():
+    if shared.cmd_opts.device_id is not None:
+        return f"xpu:{shared.cmd_opts.device_id}"
+    return "xpu"
+
+
+def torch_xpu_gc():
+    with torch.xpu.device(get_xpu_device_string()):
+        torch.xpu.empty_cache()
+
+
+has_xpu = check_for_xpu()
+
+
+# Arc GPU cannot allocate a single block larger than 4GB: https://github.com/intel/compute-runtime/issues/627
+# Here we implement a slicing algorithm to split large batch size into smaller chunks,
+# so that SDPA of each chunk wouldn't require any allocation larger than ARC_SINGLE_ALLOCATION_LIMIT.
+# The heuristic limit (TOTAL_VRAM // 8) is tuned for Intel Arc A770 16G and Arc A750 8G,
+# which is the best trade-off between VRAM usage and performance.
+ARC_SINGLE_ALLOCATION_LIMIT = {}
+orig_sdp_attn_func = torch.nn.functional.scaled_dot_product_attention
+def torch_xpu_scaled_dot_product_attention(
+    query, key, value, attn_mask=None, dropout_p=0.0, is_causal=False, *args, **kwargs
+):
+    # cast to same dtype first
+    key = key.to(query.dtype)
+    value = value.to(query.dtype)
+    if attn_mask is not None and attn_mask.dtype != torch.bool:
+        attn_mask = attn_mask.to(query.dtype)
+
+    N = query.shape[:-2]  # Batch size
+    L = query.size(-2)  # Target sequence length
+    E = query.size(-1)  # Embedding dimension of the query and key
+    S = key.size(-2)  # Source sequence length
+    Ev = value.size(-1)  # Embedding dimension of the value
+
+    total_batch_size = torch.numel(torch.empty(N))
+    device_id = query.device.index
+    if device_id not in ARC_SINGLE_ALLOCATION_LIMIT:
+        ARC_SINGLE_ALLOCATION_LIMIT[device_id] = min(torch.xpu.get_device_properties(device_id).total_memory // 8, 4 * 1024 * 1024 * 1024)
+    batch_size_limit = max(1, ARC_SINGLE_ALLOCATION_LIMIT[device_id] // (L * S * query.element_size()))
+
+    if total_batch_size <= batch_size_limit:
+        return orig_sdp_attn_func(
+            query,
+            key,
+            value,
+            attn_mask,
+            dropout_p,
+            is_causal,
+            *args, **kwargs
+        )
+
+    query = torch.reshape(query, (-1, L, E))
+    key = torch.reshape(key, (-1, S, E))
+    value = torch.reshape(value, (-1, S, Ev))
+    if attn_mask is not None:
+        attn_mask = attn_mask.view(-1, L, S)
+    chunk_count = (total_batch_size + batch_size_limit - 1) // batch_size_limit
+    outputs = []
+    for i in range(chunk_count):
+        attn_mask_chunk = (
+            None
+            if attn_mask is None
+            else attn_mask[i * batch_size_limit : (i + 1) * batch_size_limit, :, :]
+        )
+        chunk_output = orig_sdp_attn_func(
+            query[i * batch_size_limit : (i + 1) * batch_size_limit, :, :],
+            key[i * batch_size_limit : (i + 1) * batch_size_limit, :, :],
+            value[i * batch_size_limit : (i + 1) * batch_size_limit, :, :],
+            attn_mask_chunk,
+            dropout_p,
+            is_causal,
+            *args, **kwargs
+        )
+        outputs.append(chunk_output)
+    result = torch.cat(outputs, dim=0)
+    return torch.reshape(result, (*N, L, Ev))
+
+
+def is_xpu_device(device: str | torch.device = None):
+    if device is None:
+        return False
+    if isinstance(device, str):
+        return device.startswith("xpu")
+    return device.type == "xpu"
+
+
+if has_xpu:
+    try:
+        # torch.Generator supports "xpu" device since 2.1
+        torch.Generator("xpu")
+    except RuntimeError:
+        # W/A for https://github.com/intel/intel-extension-for-pytorch/issues/452: torch.Generator API doesn't support XPU device (for torch < 2.1)
+        CondFunc('torch.Generator',
+            lambda orig_func, device=None: torch.xpu.Generator(device),
+            lambda orig_func, device=None: is_xpu_device(device))
+
+    # W/A for some OPs that could not handle different input dtypes
+    CondFunc('torch.nn.functional.layer_norm',
+        lambda orig_func, input, normalized_shape=None, weight=None, *args, **kwargs:
+        orig_func(input.to(weight.data.dtype), normalized_shape, weight, *args, **kwargs),
+        lambda orig_func, input, normalized_shape=None, weight=None, *args, **kwargs:
+        weight is not None and input.dtype != weight.data.dtype)
+    CondFunc('torch.nn.modules.GroupNorm.forward',
+        lambda orig_func, self, input: orig_func(self, input.to(self.weight.data.dtype)),
+        lambda orig_func, self, input: input.dtype != self.weight.data.dtype)
+    CondFunc('torch.nn.modules.linear.Linear.forward',
+        lambda orig_func, self, input: orig_func(self, input.to(self.weight.data.dtype)),
+        lambda orig_func, self, input: input.dtype != self.weight.data.dtype)
+    CondFunc('torch.nn.modules.conv.Conv2d.forward',
+        lambda orig_func, self, input: orig_func(self, input.to(self.weight.data.dtype)),
+        lambda orig_func, self, input: input.dtype != self.weight.data.dtype)
+    CondFunc('torch.bmm',
+        lambda orig_func, input, mat2, out=None: orig_func(input.to(mat2.dtype), mat2, out=out),
+        lambda orig_func, input, mat2, out=None: input.dtype != mat2.dtype)
+    CondFunc('torch.cat',
+        lambda orig_func, tensors, dim=0, out=None: orig_func([t.to(tensors[0].dtype) for t in tensors], dim=dim, out=out),
+        lambda orig_func, tensors, dim=0, out=None: not all(t.dtype == tensors[0].dtype for t in tensors))
+    CondFunc('torch.nn.functional.scaled_dot_product_attention',
+        lambda orig_func, *args, **kwargs: torch_xpu_scaled_dot_product_attention(*args, **kwargs),
+        lambda orig_func, query, *args, **kwargs: query.is_xpu)
diff --git a/pyproject.toml b/pyproject.toml
index 80541a8f353..10ebc84b35f 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -2,6 +2,8 @@
 
 target-version = "py39"
 
+[tool.ruff.lint]
+
 extend-select = [
   "B",
   "C",
@@ -16,6 +18,7 @@ exclude = [
 
 ignore = [
 	"E501", # Line too long
+	"E721", # Do not compare types, use `isinstance`
 	"E731", # Do not assign a `lambda` expression, use a `def`
 	
 	"I001", # Import block is un-sorted or un-formatted
@@ -24,10 +27,10 @@ ignore = [
 	"W605", # invalid escape sequence, messes with some docstrings
 ]
 
-[tool.ruff.per-file-ignores]
+[tool.ruff.lint.per-file-ignores]
 "webui.py" = ["E402"]  # Module level import not at top of file
 
-[tool.ruff.flake8-bugbear]
+[tool.ruff.lint.flake8-bugbear]
 # Allow default arguments like, e.g., `data: List[str] = fastapi.Query(None)`.
 extend-immutable-calls = ["fastapi.Depends", "fastapi.security.HTTPBasic"]
 
diff --git a/requirements.txt b/requirements.txt
index 80b438455ce..0d6bac600e1 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -2,12 +2,12 @@ GitPython
 Pillow
 accelerate
 
-basicsr
 blendmodes
 clean-fid
+diskcache
 einops
+facexlib
 fastapi>=0.90.1
-gfpgan
 gradio==3.41.2
 inflection
 jsonmerge
@@ -18,17 +18,17 @@ omegaconf
 open-clip-torch
 
 piexif
+protobuf==3.20.0
 psutil
 pytorch_lightning
-realesrgan
 requests
 resize-right
 
 safetensors
 scikit-image>=0.19
-timm
 tomesd
 torch
 torchdiffeq
 torchsde
 transformers==4.30.2
+pillow-avif-plugin==1.4.3
\ No newline at end of file
diff --git a/requirements_npu.txt b/requirements_npu.txt
new file mode 100644
index 00000000000..5e6a43646a0
--- /dev/null
+++ b/requirements_npu.txt
@@ -0,0 +1,4 @@
+cloudpickle
+decorator
+synr==0.5.0
+tornado
diff --git a/requirements_versions.txt b/requirements_versions.txt
index f8ae1f385ae..0306ce94fda 100644
--- a/requirements_versions.txt
+++ b/requirements_versions.txt
@@ -1,31 +1,35 @@
+setuptools==69.5.1  # temp fix for compatibility with some old packages
 GitPython==3.1.32
 Pillow==9.5.0
 accelerate==0.21.0
-basicsr==1.4.2
 blendmodes==2022
 clean-fid==0.1.35
+diskcache==5.6.3
 einops==0.4.1
+facexlib==0.3.0
 fastapi==0.94.0
-gfpgan==1.3.8
 gradio==3.41.2
 httpcore==0.15
 inflection==0.5.1
 jsonmerge==1.8.0
 kornia==0.6.7
 lark==1.1.2
-numpy==1.23.5
+numpy==1.26.2
 omegaconf==2.2.3
 open-clip-torch==2.20.0
 piexif==1.1.3
+protobuf==3.20.0
 psutil==5.9.5
 pytorch_lightning==1.9.4
-realesrgan==0.3.0
 resize-right==0.0.2
-safetensors==0.3.1
+safetensors==0.4.2
 scikit-image==0.21.0
-timm==0.9.2
+spandrel==0.3.4
+spandrel-extra-arches==0.1.1
 tomesd==0.1.3
 torch
 torchdiffeq==0.2.3
-torchsde==0.2.5
+torchsde==0.2.6
 transformers==4.30.2
+httpx==0.24.1
+pillow-avif-plugin==1.4.3
diff --git a/script.js b/script.js
index 34cca7651dd..de1a9000d4f 100644
--- a/script.js
+++ b/script.js
@@ -29,6 +29,7 @@ var uiAfterUpdateCallbacks = [];
 var uiLoadedCallbacks = [];
 var uiTabChangeCallbacks = [];
 var optionsChangedCallbacks = [];
+var optionsAvailableCallbacks = [];
 var uiAfterUpdateTimeout = null;
 var uiCurrentTab = null;
 
@@ -77,6 +78,20 @@ function onOptionsChanged(callback) {
     optionsChangedCallbacks.push(callback);
 }
 
+/**
+ * Register callback to be called when the options (in opts global variable) are available.
+ * The callback receives no arguments.
+ * If you register the callback after the options are available, it's just immediately called.
+ */
+function onOptionsAvailable(callback) {
+    if (Object.keys(opts).length != 0) {
+        callback();
+        return;
+    }
+
+    optionsAvailableCallbacks.push(callback);
+}
+
 function executeCallbacks(queue, arg) {
     for (const callback of queue) {
         try {
@@ -121,22 +136,58 @@ document.addEventListener("DOMContentLoaded", function() {
 });
 
 /**
- * Add a ctrl+enter as a shortcut to start a generation
+ * Add keyboard shortcuts:
+ * Ctrl+Enter to start/restart a generation
+ * Alt/Option+Enter to skip a generation
+ * Esc to interrupt a generation
  */
 document.addEventListener('keydown', function(e) {
-    var handled = false;
-    if (e.key !== undefined) {
-        if ((e.key == "Enter" && (e.metaKey || e.ctrlKey || e.altKey))) handled = true;
-    } else if (e.keyCode !== undefined) {
-        if ((e.keyCode == 13 && (e.metaKey || e.ctrlKey || e.altKey))) handled = true;
-    }
-    if (handled) {
-        var button = get_uiCurrentTabContent().querySelector('button[id$=_generate]');
-        if (button) {
-            button.click();
+    const isEnter = e.key === 'Enter' || e.keyCode === 13;
+    const isCtrlKey = e.metaKey || e.ctrlKey;
+    const isAltKey = e.altKey;
+    const isEsc = e.key === 'Escape';
+
+    const generateButton = get_uiCurrentTabContent().querySelector('button[id$=_generate]');
+    const interruptButton = get_uiCurrentTabContent().querySelector('button[id$=_interrupt]');
+    const skipButton = get_uiCurrentTabContent().querySelector('button[id$=_skip]');
+
+    if (isCtrlKey && isEnter) {
+        if (interruptButton.style.display === 'block') {
+            interruptButton.click();
+            const callback = (mutationList) => {
+                for (const mutation of mutationList) {
+                    if (mutation.type === 'attributes' && mutation.attributeName === 'style') {
+                        if (interruptButton.style.display === 'none') {
+                            generateButton.click();
+                            observer.disconnect();
+                        }
+                    }
+                }
+            };
+            const observer = new MutationObserver(callback);
+            observer.observe(interruptButton, {attributes: true});
+        } else {
+            generateButton.click();
         }
         e.preventDefault();
     }
+
+    if (isAltKey && isEnter) {
+        skipButton.click();
+        e.preventDefault();
+    }
+
+    if (isEsc) {
+        const globalPopup = document.querySelector('.global-popup');
+        const lightboxModal = document.querySelector('#lightboxModal');
+        if (!globalPopup || globalPopup.style.display === 'none') {
+            if (document.activeElement === lightboxModal) return;
+            if (interruptButton.style.display === 'block') {
+                interruptButton.click();
+                e.preventDefault();
+            }
+        }
+    }
 });
 
 /**
diff --git a/scripts/loopback.py b/scripts/loopback.py
index 2d5feaf9b26..800ee882a16 100644
--- a/scripts/loopback.py
+++ b/scripts/loopback.py
@@ -95,7 +95,7 @@ def calculate_denoising_strength(loop):
                 processed = processing.process_images(p)
 
                 # Generation cancelled.
-                if state.interrupted:
+                if state.interrupted or state.stopping_generation:
                     break
 
                 if initial_seed is None:
@@ -122,8 +122,8 @@ def calculate_denoising_strength(loop):
 
             p.inpainting_fill = original_inpainting_fill
 
-            if state.interrupted:
-                    break
+            if state.interrupted or state.stopping_generation:
+                break
 
         if len(history) > 1:
             grid = images.image_grid(history, rows=1)
diff --git a/scripts/outpainting_mk_2.py b/scripts/outpainting_mk_2.py
index c98ab48098e..5df9dff9c48 100644
--- a/scripts/outpainting_mk_2.py
+++ b/scripts/outpainting_mk_2.py
@@ -102,7 +102,7 @@ def _get_masked_window_rgb(np_mask_grey, hardness=1.):
     shaped_noise_fft = _fft2(noise_rgb)
     shaped_noise_fft[:, :, :] = np.absolute(shaped_noise_fft[:, :, :]) ** 2 * (src_dist ** noise_q) * src_phase  # perform the actual shaping
 
-    brightness_variation = 0.  # color_variation # todo: temporarily tieing brightness variation to color variation for now
+    brightness_variation = 0.  # color_variation # todo: temporarily tying brightness variation to color variation for now
     contrast_adjusted_np_src = _np_src_image[:] * (brightness_variation + 1.) - brightness_variation * 2.
 
     # scikit-image is used for histogram matching, very convenient!
diff --git a/scripts/postprocessing_codeformer.py b/scripts/postprocessing_codeformer.py
index a7d80d40e2b..53a0cc44cde 100644
--- a/scripts/postprocessing_codeformer.py
+++ b/scripts/postprocessing_codeformer.py
@@ -1,31 +1,31 @@
 from PIL import Image
 import numpy as np
 
-from modules import scripts_postprocessing, codeformer_model
+from modules import scripts_postprocessing, codeformer_model, ui_components
 import gradio as gr
 
-from modules.ui_components import FormRow
-
 
 class ScriptPostprocessingCodeFormer(scripts_postprocessing.ScriptPostprocessing):
     name = "CodeFormer"
     order = 3000
 
     def ui(self):
-        with FormRow():
-            codeformer_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="CodeFormer visibility", value=0, elem_id="extras_codeformer_visibility")
-            codeformer_weight = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="CodeFormer weight (0 = maximum effect, 1 = minimum effect)", value=0, elem_id="extras_codeformer_weight")
+        with ui_components.InputAccordion(False, label="CodeFormer") as enable:
+            with gr.Row():
+                codeformer_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="Visibility", value=1.0, elem_id="extras_codeformer_visibility")
+                codeformer_weight = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="Weight (0 = maximum effect, 1 = minimum effect)", value=0, elem_id="extras_codeformer_weight")
 
         return {
+            "enable": enable,
             "codeformer_visibility": codeformer_visibility,
             "codeformer_weight": codeformer_weight,
         }
 
-    def process(self, pp: scripts_postprocessing.PostprocessedImage, codeformer_visibility, codeformer_weight):
-        if codeformer_visibility == 0:
+    def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, codeformer_visibility, codeformer_weight):
+        if codeformer_visibility == 0 or not enable:
             return
 
-        restored_img = codeformer_model.codeformer.restore(np.array(pp.image, dtype=np.uint8), w=codeformer_weight)
+        restored_img = codeformer_model.codeformer.restore(np.array(pp.image.convert("RGB"), dtype=np.uint8), w=codeformer_weight)
         res = Image.fromarray(restored_img)
 
         if codeformer_visibility < 1.0:
diff --git a/scripts/postprocessing_gfpgan.py b/scripts/postprocessing_gfpgan.py
index d854f3f7748..57e3623995c 100644
--- a/scripts/postprocessing_gfpgan.py
+++ b/scripts/postprocessing_gfpgan.py
@@ -1,29 +1,28 @@
 from PIL import Image
 import numpy as np
 
-from modules import scripts_postprocessing, gfpgan_model
+from modules import scripts_postprocessing, gfpgan_model, ui_components
 import gradio as gr
 
-from modules.ui_components import FormRow
-
 
 class ScriptPostprocessingGfpGan(scripts_postprocessing.ScriptPostprocessing):
     name = "GFPGAN"
     order = 2000
 
     def ui(self):
-        with FormRow():
-            gfpgan_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="GFPGAN visibility", value=0, elem_id="extras_gfpgan_visibility")
+        with ui_components.InputAccordion(False, label="GFPGAN") as enable:
+            gfpgan_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="Visibility", value=1.0, elem_id="extras_gfpgan_visibility")
 
         return {
+            "enable": enable,
             "gfpgan_visibility": gfpgan_visibility,
         }
 
-    def process(self, pp: scripts_postprocessing.PostprocessedImage, gfpgan_visibility):
-        if gfpgan_visibility == 0:
+    def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, gfpgan_visibility):
+        if gfpgan_visibility == 0 or not enable:
             return
 
-        restored_img = gfpgan_model.gfpgan_fix_faces(np.array(pp.image, dtype=np.uint8))
+        restored_img = gfpgan_model.gfpgan_fix_faces(np.array(pp.image.convert("RGB"), dtype=np.uint8))
         res = Image.fromarray(restored_img)
 
         if gfpgan_visibility < 1.0:
diff --git a/scripts/postprocessing_upscale.py b/scripts/postprocessing_upscale.py
index edb70ac01ca..2409fd2073e 100644
--- a/scripts/postprocessing_upscale.py
+++ b/scripts/postprocessing_upscale.py
@@ -1,51 +1,82 @@
+import re
+
 from PIL import Image
 import numpy as np
 
 from modules import scripts_postprocessing, shared
 import gradio as gr
 
-from modules.ui_components import FormRow, ToolButton
+from modules.ui_components import FormRow, ToolButton, InputAccordion
 from modules.ui import switch_values_symbol
 
 upscale_cache = {}
 
 
+def limit_size_by_one_dimention(w, h, limit):
+    if h > w and h > limit:
+        w = limit * w // h
+        h = limit
+    elif w > limit:
+        h = limit * h // w
+        w = limit
+
+    return int(w), int(h)
+
+
 class ScriptPostprocessingUpscale(scripts_postprocessing.ScriptPostprocessing):
     name = "Upscale"
     order = 1000
 
     def ui(self):
-        selected_tab = gr.State(value=0)
+        selected_tab = gr.Number(value=0, visible=False)
+
+        with InputAccordion(True, label="Upscale", elem_id="extras_upscale") as upscale_enabled:
+            with FormRow():
+                extras_upscaler_1 = gr.Dropdown(label='Upscaler 1', elem_id="extras_upscaler_1", choices=[x.name for x in shared.sd_upscalers], value=shared.sd_upscalers[0].name)
+
+            with FormRow():
+                extras_upscaler_2 = gr.Dropdown(label='Upscaler 2', elem_id="extras_upscaler_2", choices=[x.name for x in shared.sd_upscalers], value=shared.sd_upscalers[0].name)
+                extras_upscaler_2_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="Upscaler 2 visibility", value=0.0, elem_id="extras_upscaler_2_visibility")
 
-        with gr.Column():
             with FormRow():
                 with gr.Tabs(elem_id="extras_resize_mode"):
                     with gr.TabItem('Scale by', elem_id="extras_scale_by_tab") as tab_scale_by:
-                        upscaling_resize = gr.Slider(minimum=1.0, maximum=8.0, step=0.05, label="Resize", value=4, elem_id="extras_upscaling_resize")
+                        with gr.Row():
+                            with gr.Column(scale=4):
+                                upscaling_resize = gr.Slider(minimum=1.0, maximum=8.0, step=0.05, label="Resize", value=4, elem_id="extras_upscaling_resize")
+                            with gr.Column(scale=1, min_width=160):
+                                max_side_length = gr.Number(label="Max side length", value=0, elem_id="extras_upscale_max_side_length", tooltip="If any of two sides of the image ends up larger than specified, will downscale it to fit. 0 = no limit.", min_width=160, step=8, minimum=0)
 
                     with gr.TabItem('Scale to', elem_id="extras_scale_to_tab") as tab_scale_to:
                         with FormRow():
                             with gr.Column(elem_id="upscaling_column_size", scale=4):
-                                upscaling_resize_w = gr.Slider(minimum=64, maximum=2048, step=8, label="Width", value=512, elem_id="extras_upscaling_resize_w")
-                                upscaling_resize_h = gr.Slider(minimum=64, maximum=2048, step=8, label="Height", value=512, elem_id="extras_upscaling_resize_h")
+                                upscaling_resize_w = gr.Slider(minimum=64, maximum=8192, step=8, label="Width", value=512, elem_id="extras_upscaling_resize_w")
+                                upscaling_resize_h = gr.Slider(minimum=64, maximum=8192, step=8, label="Height", value=512, elem_id="extras_upscaling_resize_h")
                             with gr.Column(elem_id="upscaling_dimensions_row", scale=1, elem_classes="dimensions-tools"):
-                                upscaling_res_switch_btn = ToolButton(value=switch_values_symbol, elem_id="upscaling_res_switch_btn")
+                                upscaling_res_switch_btn = ToolButton(value=switch_values_symbol, elem_id="upscaling_res_switch_btn", tooltip="Switch width/height")
                                 upscaling_crop = gr.Checkbox(label='Crop to fit', value=True, elem_id="extras_upscaling_crop")
 
-            with FormRow():
-                extras_upscaler_1 = gr.Dropdown(label='Upscaler 1', elem_id="extras_upscaler_1", choices=[x.name for x in shared.sd_upscalers], value=shared.sd_upscalers[0].name)
+        def on_selected_upscale_method(upscale_method):
+            if not shared.opts.set_scale_by_when_changing_upscaler:
+                return gr.update()
 
-            with FormRow():
-                extras_upscaler_2 = gr.Dropdown(label='Upscaler 2', elem_id="extras_upscaler_2", choices=[x.name for x in shared.sd_upscalers], value=shared.sd_upscalers[0].name)
-                extras_upscaler_2_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="Upscaler 2 visibility", value=0.0, elem_id="extras_upscaler_2_visibility")
+            match = re.search(r'(\d)[xX]|[xX](\d)', upscale_method)
+            if not match:
+                return gr.update()
+
+            return gr.update(value=int(match.group(1) or match.group(2)))
 
         upscaling_res_switch_btn.click(lambda w, h: (h, w), inputs=[upscaling_resize_w, upscaling_resize_h], outputs=[upscaling_resize_w, upscaling_resize_h], show_progress=False)
         tab_scale_by.select(fn=lambda: 0, inputs=[], outputs=[selected_tab])
         tab_scale_to.select(fn=lambda: 1, inputs=[], outputs=[selected_tab])
 
+        extras_upscaler_1.change(on_selected_upscale_method, inputs=[extras_upscaler_1], outputs=[upscaling_resize], show_progress="hidden")
+
         return {
+            "upscale_enabled": upscale_enabled,
             "upscale_mode": selected_tab,
             "upscale_by": upscaling_resize,
+            "max_side_length": max_side_length,
             "upscale_to_width": upscaling_resize_w,
             "upscale_to_height": upscaling_resize_h,
             "upscale_crop": upscaling_crop,
@@ -54,12 +85,18 @@ def ui(self):
             "upscaler_2_visibility": extras_upscaler_2_visibility,
         }
 
-    def upscale(self, image, info, upscaler, upscale_mode, upscale_by,  upscale_to_width, upscale_to_height, upscale_crop):
+    def upscale(self, image, info, upscaler, upscale_mode, upscale_by, max_side_length, upscale_to_width, upscale_to_height, upscale_crop):
         if upscale_mode == 1:
             upscale_by = max(upscale_to_width/image.width, upscale_to_height/image.height)
             info["Postprocess upscale to"] = f"{upscale_to_width}x{upscale_to_height}"
         else:
             info["Postprocess upscale by"] = upscale_by
+            if max_side_length != 0 and max(*image.size)*upscale_by > max_side_length:
+                upscale_mode = 1
+                upscale_crop = False
+                upscale_to_width, upscale_to_height = limit_size_by_one_dimention(image.width*upscale_by, image.height*upscale_by, max_side_length)
+                upscale_by = max(upscale_to_width/image.width, upscale_to_height/image.height)
+                info["Max side length"] = max_side_length
 
         cache_key = (hash(np.array(image.getdata()).tobytes()), upscaler.name, upscale_mode, upscale_by,  upscale_to_width, upscale_to_height, upscale_crop)
         cached_image = upscale_cache.pop(cache_key, None)
@@ -81,7 +118,21 @@ def upscale(self, image, info, upscaler, upscale_mode, upscale_by,  upscale_to_w
 
         return image
 
-    def process(self, pp: scripts_postprocessing.PostprocessedImage, upscale_mode=1, upscale_by=2.0, upscale_to_width=None, upscale_to_height=None, upscale_crop=False, upscaler_1_name=None, upscaler_2_name=None, upscaler_2_visibility=0.0):
+    def process_firstpass(self, pp: scripts_postprocessing.PostprocessedImage, upscale_enabled=True, upscale_mode=1, upscale_by=2.0, max_side_length=0, upscale_to_width=None, upscale_to_height=None, upscale_crop=False, upscaler_1_name=None, upscaler_2_name=None, upscaler_2_visibility=0.0):
+        if upscale_mode == 1:
+            pp.shared.target_width = upscale_to_width
+            pp.shared.target_height = upscale_to_height
+        else:
+            pp.shared.target_width = int(pp.image.width * upscale_by)
+            pp.shared.target_height = int(pp.image.height * upscale_by)
+
+            pp.shared.target_width, pp.shared.target_height = limit_size_by_one_dimention(pp.shared.target_width, pp.shared.target_height, max_side_length)
+
+    def process(self, pp: scripts_postprocessing.PostprocessedImage, upscale_enabled=True, upscale_mode=1, upscale_by=2.0, max_side_length=0, upscale_to_width=None, upscale_to_height=None, upscale_crop=False, upscaler_1_name=None, upscaler_2_name=None, upscaler_2_visibility=0.0):
+        if not upscale_enabled:
+            return
+
+        upscaler_1_name = upscaler_1_name
         if upscaler_1_name == "None":
             upscaler_1_name = None
 
@@ -91,17 +142,20 @@ def process(self, pp: scripts_postprocessing.PostprocessedImage, upscale_mode=1,
         if not upscaler1:
             return
 
+        upscaler_2_name = upscaler_2_name
         if upscaler_2_name == "None":
             upscaler_2_name = None
 
         upscaler2 = next(iter([x for x in shared.sd_upscalers if x.name == upscaler_2_name and x.name != "None"]), None)
         assert upscaler2 or (upscaler_2_name is None), f'could not find upscaler named {upscaler_2_name}'
 
-        upscaled_image = self.upscale(pp.image, pp.info, upscaler1, upscale_mode, upscale_by, upscale_to_width, upscale_to_height, upscale_crop)
+        upscaled_image = self.upscale(pp.image, pp.info, upscaler1, upscale_mode, upscale_by, max_side_length, upscale_to_width, upscale_to_height, upscale_crop)
         pp.info["Postprocess upscaler"] = upscaler1.name
 
         if upscaler2 and upscaler_2_visibility > 0:
-            second_upscale = self.upscale(pp.image, pp.info, upscaler2, upscale_mode, upscale_by, upscale_to_width, upscale_to_height, upscale_crop)
+            second_upscale = self.upscale(pp.image, pp.info, upscaler2, upscale_mode, upscale_by, max_side_length, upscale_to_width, upscale_to_height, upscale_crop)
+            if upscaled_image.mode != second_upscale.mode:
+                second_upscale = second_upscale.convert(upscaled_image.mode)
             upscaled_image = Image.blend(upscaled_image, second_upscale, upscaler_2_visibility)
 
             pp.info["Postprocess upscaler 2"] = upscaler2.name
@@ -126,6 +180,10 @@ def ui(self):
             "upscaler_name": upscaler_name,
         }
 
+    def process_firstpass(self, pp: scripts_postprocessing.PostprocessedImage, upscale_by=2.0, upscaler_name=None):
+        pp.shared.target_width = int(pp.image.width * upscale_by)
+        pp.shared.target_height = int(pp.image.height * upscale_by)
+
     def process(self, pp: scripts_postprocessing.PostprocessedImage, upscale_by=2.0, upscaler_name=None):
         if upscaler_name is None or upscaler_name == "None":
             return
@@ -133,5 +191,5 @@ def process(self, pp: scripts_postprocessing.PostprocessedImage, upscale_by=2.0,
         upscaler1 = next(iter([x for x in shared.sd_upscalers if x.name == upscaler_name]), None)
         assert upscaler1, f'could not find upscaler named {upscaler_name}'
 
-        pp.image = self.upscale(pp.image, pp.info, upscaler1, 0, upscale_by, 0, 0, False)
+        pp.image = self.upscale(pp.image, pp.info, upscaler1, 0, upscale_by, 0, 0, 0, False)
         pp.info["Postprocess upscaler"] = upscaler1.name
diff --git a/scripts/prompts_from_file.py b/scripts/prompts_from_file.py
index 50320d553bd..a4a2f24dd25 100644
--- a/scripts/prompts_from_file.py
+++ b/scripts/prompts_from_file.py
@@ -5,11 +5,17 @@
 import modules.scripts as scripts
 import gradio as gr
 
-from modules import sd_samplers, errors
+from modules import sd_samplers, errors, sd_models
 from modules.processing import Processed, process_images
 from modules.shared import state
 
 
+def process_model_tag(tag):
+    info = sd_models.get_closet_checkpoint_match(tag)
+    assert info is not None, f'Unknown checkpoint: {tag}'
+    return info.name
+
+
 def process_string_tag(tag):
     return tag
 
@@ -27,7 +33,7 @@ def process_boolean_tag(tag):
 
 
 prompt_tags = {
-    "sd_model": None,
+    "sd_model": process_model_tag,
     "outpath_samples": process_string_tag,
     "outpath_grids": process_string_tag,
     "prompt_for_display": process_string_tag,
@@ -108,6 +114,7 @@ def title(self):
     def ui(self, is_img2img):
         checkbox_iterate = gr.Checkbox(label="Iterate seed every line", value=False, elem_id=self.elem_id("checkbox_iterate"))
         checkbox_iterate_batch = gr.Checkbox(label="Use same random seed for all lines", value=False, elem_id=self.elem_id("checkbox_iterate_batch"))
+        prompt_position = gr.Radio(["start", "end"], label="Insert prompts at the", elem_id=self.elem_id("prompt_position"), value="start")
 
         prompt_txt = gr.Textbox(label="List of prompt inputs", lines=1, elem_id=self.elem_id("prompt_txt"))
         file = gr.File(label="Upload prompt inputs", type='binary', elem_id=self.elem_id("file"))
@@ -118,9 +125,9 @@ def ui(self, is_img2img):
         # We don't shrink back to 1, because that causes the control to ignore [enter], and it may
         # be unclear to the user that shift-enter is needed.
         prompt_txt.change(lambda tb: gr.update(lines=7) if ("\n" in tb) else gr.update(lines=2), inputs=[prompt_txt], outputs=[prompt_txt], show_progress=False)
-        return [checkbox_iterate, checkbox_iterate_batch, prompt_txt]
+        return [checkbox_iterate, checkbox_iterate_batch, prompt_position, prompt_txt]
 
-    def run(self, p, checkbox_iterate, checkbox_iterate_batch, prompt_txt: str):
+    def run(self, p, checkbox_iterate, checkbox_iterate_batch, prompt_position, prompt_txt: str):
         lines = [x for x in (x.strip() for x in prompt_txt.splitlines()) if x]
 
         p.do_not_save_grid = True
@@ -156,7 +163,22 @@ def run(self, p, checkbox_iterate, checkbox_iterate_batch, prompt_txt: str):
 
             copy_p = copy.copy(p)
             for k, v in args.items():
-                setattr(copy_p, k, v)
+                if k == "sd_model":
+                    copy_p.override_settings['sd_model_checkpoint'] = v
+                else:
+                    setattr(copy_p, k, v)
+
+            if args.get("prompt") and p.prompt:
+                if prompt_position == "start":
+                    copy_p.prompt = args.get("prompt") + " " + p.prompt
+                else:
+                    copy_p.prompt = p.prompt + " " + args.get("prompt")
+
+            if args.get("negative_prompt") and p.negative_prompt:
+                if prompt_position == "start":
+                    copy_p.negative_prompt = args.get("negative_prompt") + " " + p.negative_prompt
+                else:
+                    copy_p.negative_prompt = p.negative_prompt + " " + args.get("negative_prompt")
 
             proc = process_images(copy_p)
             images += proc.images
diff --git a/scripts/xyz_grid.py b/scripts/xyz_grid.py
index 939d86053bd..6a42a04d9a3 100644
--- a/scripts/xyz_grid.py
+++ b/scripts/xyz_grid.py
@@ -11,7 +11,7 @@
 import modules.scripts as scripts
 import gradio as gr
 
-from modules import images, sd_samplers, processing, sd_models, sd_vae, sd_samplers_kdiffusion, errors
+from modules import images, sd_samplers, processing, sd_models, sd_vae, sd_schedulers, errors
 from modules.processing import process_images, Processed, StableDiffusionProcessingTxt2Img
 from modules.shared import opts, state
 import modules.shared as shared
@@ -45,7 +45,7 @@ def apply_prompt(p, x, xs):
 def apply_order(p, x, xs):
     token_order = []
 
-    # Initally grab the tokens from the prompt, so they can be replaced in order of earliest seen
+    # Initially grab the tokens from the prompt, so they can be replaced in order of earliest seen
     for token in x:
         token_order.append((p.prompt.find(token), token))
 
@@ -95,33 +95,38 @@ def confirm_checkpoints_or_none(p, xs):
             raise RuntimeError(f"Unknown checkpoint: {x}")
 
 
-def apply_clip_skip(p, x, xs):
-    opts.data["CLIP_stop_at_last_layers"] = x
+def confirm_range(min_val, max_val, axis_label):
+    """Generates a AxisOption.confirm() function that checks all values are within the specified range."""
 
+    def confirm_range_fun(p, xs):
+        for x in xs:
+            if not (max_val >= x >= min_val):
+                raise ValueError(f'{axis_label} value "{x}" out of range [{min_val}, {max_val}]')
+
+    return confirm_range_fun
 
-def apply_upscale_latent_space(p, x, xs):
-    if x.lower().strip() != '0':
-        opts.data["use_scale_latent_for_hires_fix"] = True
-    else:
-        opts.data["use_scale_latent_for_hires_fix"] = False
+
+def apply_size(p, x: str, xs) -> None:
+    try:
+        width, _, height = x.partition('x')
+        width = int(width.strip())
+        height = int(height.strip())
+        p.width = width
+        p.height = height
+    except ValueError:
+        print(f"Invalid size in XYZ plot: {x}")
 
 
 def find_vae(name: str):
-    if name.lower() in ['auto', 'automatic']:
-        return modules.sd_vae.unspecified
-    if name.lower() == 'none':
-        return None
-    else:
-        choices = [x for x in sorted(modules.sd_vae.vae_dict, key=lambda x: len(x)) if name.lower().strip() in x.lower()]
-        if len(choices) == 0:
-            print(f"No VAE found for {name}; using automatic")
-            return modules.sd_vae.unspecified
-        else:
-            return modules.sd_vae.vae_dict[choices[0]]
+    if (name := name.strip().lower()) in ('auto', 'automatic'):
+        return 'Automatic'
+    elif name == 'none':
+        return 'None'
+    return next((k for k in modules.sd_vae.vae_dict if k.lower() == name), print(f'No VAE found for {name}; using Automatic') or 'Automatic')
 
 
 def apply_vae(p, x, xs):
-    modules.sd_vae.reload_vae_weights(shared.sd_model, vae_file=find_vae(x))
+    p.override_settings['sd_vae'] = find_vae(x)
 
 
 def apply_styles(p: StableDiffusionProcessingTxt2Img, x: str, _):
@@ -129,7 +134,7 @@ def apply_styles(p: StableDiffusionProcessingTxt2Img, x: str, _):
 
 
 def apply_uni_pc_order(p, x, xs):
-    opts.data["uni_pc_order"] = min(x, p.steps - 1)
+    p.override_settings['uni_pc_order'] = min(x, p.steps - 1)
 
 
 def apply_face_restore(p, opt, x):
@@ -151,12 +156,14 @@ def fun(p, x, xs):
         if boolean:
             x = True if x.lower() == "true" else False
         p.override_settings[field] = x
+
     return fun
 
 
 def boolean_choice(reverse: bool = False):
     def choice():
         return ["False", "True"] if reverse else ["True", "False"]
+
     return choice
 
 
@@ -201,17 +208,18 @@ def list_to_csv_string(data_list):
 
 
 def csv_string_to_list_strip(data_str):
-    return list(map(str.strip, chain.from_iterable(csv.reader(StringIO(data_str)))))
+    return list(map(str.strip, chain.from_iterable(csv.reader(StringIO(data_str), skipinitialspace=True))))
 
 
 class AxisOption:
-    def __init__(self, label, type, apply, format_value=format_value_add_label, confirm=None, cost=0.0, choices=None):
+    def __init__(self, label, type, apply, format_value=format_value_add_label, confirm=None, cost=0.0, choices=None, prepare=None):
         self.label = label
         self.type = type
         self.apply = apply
         self.format_value = format_value
         self.confirm = confirm
         self.cost = cost
+        self.prepare = prepare
         self.choices = choices
 
 
@@ -247,18 +255,20 @@ def __init__(self, *args, **kwargs):
     AxisOption("Sigma min", float, apply_field("s_tmin")),
     AxisOption("Sigma max", float, apply_field("s_tmax")),
     AxisOption("Sigma noise", float, apply_field("s_noise")),
-    AxisOption("Schedule type", str, apply_override("k_sched_type"), choices=lambda: list(sd_samplers_kdiffusion.k_diffusion_scheduler)),
+    AxisOption("Schedule type", str, apply_field("scheduler"), choices=lambda: [x.label for x in sd_schedulers.schedulers]),
     AxisOption("Schedule min sigma", float, apply_override("sigma_min")),
     AxisOption("Schedule max sigma", float, apply_override("sigma_max")),
     AxisOption("Schedule rho", float, apply_override("rho")),
+    AxisOption("Beta schedule alpha", float, apply_override("beta_dist_alpha")),
+    AxisOption("Beta schedule beta", float, apply_override("beta_dist_beta")),
     AxisOption("Eta", float, apply_field("eta")),
-    AxisOption("Clip skip", int, apply_clip_skip),
+    AxisOption("Clip skip", int, apply_override('CLIP_stop_at_last_layers')),
     AxisOption("Denoising", float, apply_field("denoising_strength")),
     AxisOption("Initial noise multiplier", float, apply_field("initial_noise_multiplier")),
     AxisOption("Extra noise", float, apply_override("img2img_extra_noise")),
     AxisOptionTxt2Img("Hires upscaler", str, apply_field("hr_upscaler"), choices=lambda: [*shared.latent_upscale_modes, *[x.name for x in shared.sd_upscalers]]),
     AxisOptionImg2Img("Cond. Image Mask Weight", float, apply_field("inpainting_mask_weight")),
-    AxisOption("VAE", str, apply_vae, cost=0.7, choices=lambda: ['None'] + list(sd_vae.vae_dict)),
+    AxisOption("VAE", str, apply_vae, cost=0.7, choices=lambda: ['Automatic', 'None'] + list(sd_vae.vae_dict)),
     AxisOption("Styles", str, apply_styles, choices=lambda: list(shared.prompt_styles.styles)),
     AxisOption("UniPC Order", int, apply_uni_pc_order, cost=0.5),
     AxisOption("Face restore", str, apply_face_restore, format_value=format_value),
@@ -269,6 +279,8 @@ def __init__(self, *args, **kwargs):
     AxisOption("Refiner checkpoint", str, apply_field('refiner_checkpoint'), format_value=format_remove_path, confirm=confirm_checkpoints_or_none, cost=1.0, choices=lambda: ['None'] + sorted(sd_models.checkpoints_list, key=str.casefold)),
     AxisOption("Refiner switch at", float, apply_field('refiner_switch_at')),
     AxisOption("RNG source", str, apply_override("randn_source"), choices=lambda: ["GPU", "CPU", "NV"]),
+    AxisOption("FP8 mode", str, apply_override("fp8_storage"), cost=0.9, choices=lambda: ["Disable", "Enable for SDXL", "Enable"]),
+    AxisOption("Size", str, apply_size),
 ]
 
 
@@ -364,16 +376,17 @@ def index(ix, iy, iz):
         end_index = start_index + len(xs) * len(ys)
         grid = images.image_grid(processed_result.images[start_index:end_index], rows=len(ys))
         if draw_legend:
-            grid = images.draw_grid_annotations(grid, processed_result.images[start_index].size[0], processed_result.images[start_index].size[1], hor_texts, ver_texts, margin_size)
+            grid_max_w, grid_max_h = map(max, zip(*(img.size for img in processed_result.images[start_index:end_index])))
+            grid = images.draw_grid_annotations(grid, grid_max_w, grid_max_h, hor_texts, ver_texts, margin_size)
         processed_result.images.insert(i, grid)
         processed_result.all_prompts.insert(i, processed_result.all_prompts[start_index])
         processed_result.all_seeds.insert(i, processed_result.all_seeds[start_index])
         processed_result.infotexts.insert(i, processed_result.infotexts[start_index])
 
-    sub_grid_size = processed_result.images[0].size
     z_grid = images.image_grid(processed_result.images[:z_count], rows=1)
+    z_sub_grid_max_w, z_sub_grid_max_h = map(max, zip(*(img.size for img in processed_result.images[:z_count])))
     if draw_legend:
-        z_grid = images.draw_grid_annotations(z_grid, sub_grid_size[0], sub_grid_size[1], title_texts, [[images.GridAnnotation()]])
+        z_grid = images.draw_grid_annotations(z_grid, z_sub_grid_max_w, z_sub_grid_max_h, title_texts, [[images.GridAnnotation()]])
     processed_result.images.insert(0, z_grid)
     # TODO: Deeper aspects of the program rely on grid info being misaligned between metadata arrays, which is not ideal.
     # processed_result.all_prompts.insert(0, processed_result.all_prompts[0])
@@ -385,18 +398,12 @@ def index(ix, iy, iz):
 
 class SharedSettingsStackHelper(object):
     def __enter__(self):
-        self.CLIP_stop_at_last_layers = opts.CLIP_stop_at_last_layers
-        self.vae = opts.sd_vae
-        self.uni_pc_order = opts.uni_pc_order
+        pass
 
     def __exit__(self, exc_type, exc_value, tb):
-        opts.data["sd_vae"] = self.vae
-        opts.data["uni_pc_order"] = self.uni_pc_order
         modules.sd_models.reload_model_weights()
         modules.sd_vae.reload_vae_weights()
 
-        opts.data["CLIP_stop_at_last_layers"] = self.CLIP_stop_at_last_layers
-
 
 re_range = re.compile(r"\s*([+-]?\s*\d+)\s*-\s*([+-]?\s*\d+)(?:\s*\(([+-]\d+)\s*\))?\s*")
 re_range_float = re.compile(r"\s*([+-]?\s*\d+(?:.\d*)?)\s*-\s*([+-]?\s*\d+(?:.\d*)?)(?:\s*\(([+-]\d+(?:.\d*)?)\s*\))?\s*")
@@ -436,13 +443,16 @@ def ui(self, is_img2img):
             with gr.Column():
                 draw_legend = gr.Checkbox(label='Draw legend', value=True, elem_id=self.elem_id("draw_legend"))
                 no_fixed_seeds = gr.Checkbox(label='Keep -1 for seeds', value=False, elem_id=self.elem_id("no_fixed_seeds"))
+                with gr.Row():
+                    vary_seeds_x = gr.Checkbox(label='Vary seeds for X', value=False, min_width=80, elem_id=self.elem_id("vary_seeds_x"), tooltip="Use different seeds for images along X axis.")
+                    vary_seeds_y = gr.Checkbox(label='Vary seeds for Y', value=False, min_width=80, elem_id=self.elem_id("vary_seeds_y"), tooltip="Use different seeds for images along Y axis.")
+                    vary_seeds_z = gr.Checkbox(label='Vary seeds for Z', value=False, min_width=80, elem_id=self.elem_id("vary_seeds_z"), tooltip="Use different seeds for images along Z axis.")
             with gr.Column():
                 include_lone_images = gr.Checkbox(label='Include Sub Images', value=False, elem_id=self.elem_id("include_lone_images"))
                 include_sub_grids = gr.Checkbox(label='Include Sub Grids', value=False, elem_id=self.elem_id("include_sub_grids"))
+                csv_mode = gr.Checkbox(label='Use text inputs instead of dropdowns', value=False, elem_id=self.elem_id("csv_mode"))
             with gr.Column():
                 margin_size = gr.Slider(label="Grid margins (px)", minimum=0, maximum=500, value=0, step=2, elem_id=self.elem_id("margin_size"))
-            with gr.Column():
-                csv_mode = gr.Checkbox(label='Use text inputs instead of dropdowns', value=False, elem_id=self.elem_id("csv_mode"))
 
         with gr.Row(variant="compact", elem_id="swap_axes"):
             swap_xy_axes_button = gr.Button(value="Swap X/Y axes", elem_id="xy_grid_swap_axes_button")
@@ -474,6 +484,8 @@ def fill(axis_type, csv_mode):
         fill_z_button.click(fn=fill, inputs=[z_type, csv_mode], outputs=[z_values, z_values_dropdown])
 
         def select_axis(axis_type, axis_values, axis_values_dropdown, csv_mode):
+            axis_type = axis_type or 0  # if axle type is None set to 0
+
             choices = self.current_axis_options[axis_type].choices
             has_choices = choices is not None
 
@@ -521,9 +533,11 @@ def get_dropdown_update_from_params(axis, params):
             (z_values_dropdown, lambda params: get_dropdown_update_from_params("Z", params)),
         )
 
-        return [x_type, x_values, x_values_dropdown, y_type, y_values, y_values_dropdown, z_type, z_values, z_values_dropdown, draw_legend, include_lone_images, include_sub_grids, no_fixed_seeds, margin_size, csv_mode]
+        return [x_type, x_values, x_values_dropdown, y_type, y_values, y_values_dropdown, z_type, z_values, z_values_dropdown, draw_legend, include_lone_images, include_sub_grids, no_fixed_seeds, vary_seeds_x, vary_seeds_y, vary_seeds_z, margin_size, csv_mode]
+
+    def run(self, p, x_type, x_values, x_values_dropdown, y_type, y_values, y_values_dropdown, z_type, z_values, z_values_dropdown, draw_legend, include_lone_images, include_sub_grids, no_fixed_seeds, vary_seeds_x, vary_seeds_y, vary_seeds_z, margin_size, csv_mode):
+        x_type, y_type, z_type = x_type or 0, y_type or 0, z_type or 0  # if axle type is None set to 0
 
-    def run(self, p, x_type, x_values, x_values_dropdown, y_type, y_values, y_values_dropdown, z_type, z_values, z_values_dropdown, draw_legend, include_lone_images, include_sub_grids, no_fixed_seeds, margin_size, csv_mode):
         if not no_fixed_seeds:
             modules.processing.fix_seed(p)
 
@@ -536,6 +550,8 @@ def process_axis(opt, vals, vals_dropdown):
 
             if opt.choices is not None and not csv_mode:
                 valslist = vals_dropdown
+            elif opt.prepare is not None:
+                valslist = opt.prepare(vals)
             else:
                 valslist = csv_string_to_list_strip(vals)
 
@@ -543,11 +559,13 @@ def process_axis(opt, vals, vals_dropdown):
                 valslist_ext = []
 
                 for val in valslist:
+                    if val.strip() == '':
+                        continue
                     m = re_range.fullmatch(val)
                     mc = re_range_count.fullmatch(val)
                     if m is not None:
                         start = int(m.group(1))
-                        end = int(m.group(2))+1
+                        end = int(m.group(2)) + 1
                         step = int(m.group(3)) if m.group(3) is not None else 1
 
                         valslist_ext += list(range(start, end, step))
@@ -565,6 +583,8 @@ def process_axis(opt, vals, vals_dropdown):
                 valslist_ext = []
 
                 for val in valslist:
+                    if val.strip() == '':
+                        continue
                     m = re_range_float.fullmatch(val)
                     mc = re_range_count_float.fullmatch(val)
                     if m is not None:
@@ -685,7 +705,7 @@ def fix_axis_seeds(axis_opt, axis_list):
         grid_infotext = [None] * (1 + len(zs))
 
         def cell(x, y, z, ix, iy, iz):
-            if shared.state.interrupted:
+            if shared.state.interrupted or state.stopping_generation:
                 return Processed(p, [], p.seed, "")
 
             pc = copy(p)
@@ -694,6 +714,16 @@ def cell(x, y, z, ix, iy, iz):
             y_opt.apply(pc, y, ys)
             z_opt.apply(pc, z, zs)
 
+            xdim = len(xs) if vary_seeds_x else 1
+            ydim = len(ys) if vary_seeds_y else 1
+
+            if vary_seeds_x:
+                pc.seed += ix
+            if vary_seeds_y:
+                pc.seed += iy * xdim
+            if vary_seeds_z:
+                pc.seed += iz * xdim * ydim
+
             try:
                 res = process_images(pc)
             except Exception as e:
@@ -760,19 +790,21 @@ def cell(x, y, z, ix, iy, iz):
         z_count = len(zs)
 
         # Set the grid infotexts to the real ones with extra_generation_params (1 main grid + z_count sub-grids)
-        processed.infotexts[:1+z_count] = grid_infotext[:1+z_count]
+        processed.infotexts[:1 + z_count] = grid_infotext[:1 + z_count]
 
         if not include_lone_images:
             # Don't need sub-images anymore, drop from list:
-            processed.images = processed.images[:z_count+1]
+            processed.images = processed.images[:z_count + 1]
 
         if opts.grid_save:
             # Auto-save main and sub-grids:
             grid_count = z_count + 1 if z_count > 1 else 1
             for g in range(grid_count):
                 # TODO: See previous comment about intentional data misalignment.
-                adj_g = g-1 if g > 0 else g
+                adj_g = g - 1 if g > 0 else g
                 images.save_image(processed.images[g], p.outpath_grids, "xyz_grid", info=processed.infotexts[g], extension=opts.grid_format, prompt=processed.all_prompts[adj_g], seed=processed.all_seeds[adj_g], grid=True, p=processed)
+                if not include_sub_grids:  # if not include_sub_grids then skip saving after the first grid
+                    break
 
         if not include_sub_grids:
             # Done with sub-grids, drop all related information:
diff --git a/style.css b/style.css
index fb4e2f1f02e..64ef61bad46 100644
--- a/style.css
+++ b/style.css
@@ -1,6 +1,6 @@
 /* temporary fix to load default gradio font in frontend instead of backend */
 
-@import url('https://fonts.googleapis.com/css2?family=Source+Sans+Pro:wght@400;600&display=swap');
+@import url('webui-assets/css/sourcesanspro.css');
 
 
 /* temporary fix to hide gradio crop tool until it's fixed https://github.com/gradio-app/gradio/issues/3810 */
@@ -28,7 +28,7 @@ div.gradio-container{
 }
 
 .hidden{
-    display: none;
+    display: none !important;
 }
 
 .compact{
@@ -83,8 +83,10 @@ div.compact{
     white-space: nowrap;
 }
 
-.gradio-dropdown ul.options li.item {
-    padding: 0.05em 0;
+@media (pointer:fine) {
+    .gradio-dropdown ul.options li.item {
+        padding: 0.05em 0;
+    }
 }
 
 .gradio-dropdown ul.options li.item.selected {
@@ -202,6 +204,11 @@ div.block.gradio-accordion {
     padding: 8px 8px;
 }
 
+input[type="checkbox"].input-accordion-checkbox{
+    vertical-align: sub;
+    margin-right: 0.5em;
+}
+
 
 /* txt2img/img2img specific */
 
@@ -215,6 +222,10 @@ div.block.gradio-accordion {
     top: -0.75em;
 }
 
+.block.token-counter-visible{
+    display: block !important;
+}
+
 .block.token-counter span{
     background: var(--input-background-fill) !important;
     box-shadow: 0 0 0.0 0.3em rgba(192,192,192,0.15), inset 0 0 0.6em rgba(192,192,192,0.075);
@@ -268,7 +279,7 @@ div.block.gradio-accordion {
     display: inline-block;
 }
 
-.html-log .performance p.time, .performance p.vram, .performance p.time abbr, .performance p.vram abbr {
+.html-log .performance p.time, .performance p.vram, .performance p.profile, .performance p.time abbr, .performance p.vram abbr {
     margin-bottom: 0;
     color: var(--block-title-text-color);
 }
@@ -280,6 +291,10 @@ div.block.gradio-accordion {
     margin-left: auto;
 }
 
+.html-log .performance p.profile {
+    margin-left: 0.5em;
+}
+
 .html-log .performance .measurement{
     color: var(--body-text-color);
     font-weight: bold;
@@ -289,6 +304,13 @@ div.block.gradio-accordion {
     min-height: 4.5em;
 }
 
+#txt2img_generate, #img2img_generate {
+    min-height: 4.5em;
+}
+.generate-box-compact #txt2img_generate, .generate-box-compact #img2img_generate {
+    min-height: 3em;
+}
+
 @media screen and (min-width: 2500px) {
     #txt2img_gallery, #img2img_gallery {
         min-height: 768px;
@@ -317,17 +339,17 @@ div.block.gradio-accordion {
 .generate-box{
     position: relative;
 }
-.gradio-button.generate-box-skip, .gradio-button.generate-box-interrupt{
+.gradio-button.generate-box-skip, .gradio-button.generate-box-interrupt, .gradio-button.generate-box-interrupting{
     position: absolute;
     width: 50%;
     height: 100%;
     display: none;
     background: #b4c0cc;
 }
-.gradio-button.generate-box-skip:hover, .gradio-button.generate-box-interrupt:hover{
+.gradio-button.generate-box-skip:hover, .gradio-button.generate-box-interrupt:hover, .gradio-button.generate-box-interrupting:hover{
     background: #c2cfdb;
 }
-.gradio-button.generate-box-interrupt{
+.gradio-button.generate-box-interrupt, .gradio-button.generate-box-interrupting{
     left: 0;
     border-radius: 0.5rem 0 0 0.5rem;
 }
@@ -396,6 +418,15 @@ div#extras_scale_to_tab div.form{
     min-width: 0.5em;
 }
 
+div.toprow-compact-stylerow{
+    margin: 0.5em 0;
+}
+
+div.toprow-compact-tools{
+    min-width: fit-content !important;
+    max-width: fit-content;
+}
+
 /* settings */
 #quicksettings {
     align-items: end;
@@ -421,6 +452,7 @@ div#extras_scale_to_tab div.form{
 #settings > div{
     border: none;
     margin-left: 10em;
+    padding: 0 var(--spacing-xl);
 }
 
 #settings > div.tab-nav{
@@ -435,6 +467,16 @@ div#extras_scale_to_tab div.form{
     border: none;
     text-align: left;
     white-space: initial;
+    padding: 4px;
+}
+
+#settings > div.tab-nav .settings-category{
+    display: block;
+    margin: 1em 0 0.25em 0;
+    font-weight: bold;
+    text-decoration: underline;
+    cursor: default;
+    user-select: none;
 }
 
 #settings_result{
@@ -490,6 +532,10 @@ table.popup-table .link{
     opacity: 0.75;
 }
 
+.settings-comment .info ol{
+    margin: 0.4em 0 0.8em 1em;
+}
+
 #sysinfo_download a.sysinfo_big_link{
     font-size: 24pt;
 }
@@ -516,7 +562,8 @@ table.popup-table .link{
     height: 20px;
     background: #b4c0cc;
     border-radius: 3px !important;
-    top: -20px;
+    top: -14px;
+    left: 0px;
     width: 100%;
 }
 
@@ -581,7 +628,6 @@ table.popup-table .link{
     width: 100%;
     height: 100%;
     overflow: auto;
-    background-color: rgba(20, 20, 20, 0.95);
 }
 
 .global-popup *{
@@ -590,9 +636,6 @@ table.popup-table .link{
 
 .global-popup-close:before {
     content: "×";
-}
-
-.global-popup-close{
     position: fixed;
     right: 0.25em;
     top: 0;
@@ -601,10 +644,22 @@ table.popup-table .link{
     font-size: 32pt;
 }
 
+.global-popup-close{
+    position: fixed;
+    left: 0;
+    top: 0;
+    width: 100%;
+    height: 100%;
+    background-color: rgba(20, 20, 20, 0.95);
+}
+
 .global-popup-inner{
     display: inline-block;
     margin: auto;
     padding: 2em;
+    z-index: 1001;
+    max-height: 90%;
+    max-width: 90%;
 }
 
 /* fullpage image viewer */
@@ -636,7 +691,7 @@ table.popup-table .link{
     transition: 0.2s ease background-color;
 }
 .modalControls:hover {
-    background-color:rgba(0,0,0,0.9);
+    background-color:rgba(0,0,0, var(--sd-webui-modal-lightbox-toolbar-opacity));
 }
 .modalClose {
     margin-left: auto;
@@ -706,6 +761,22 @@ table.popup-table .link{
     display: none;
 }
 
+@media (pointer: fine) {
+    .modalPrev:hover,
+    .modalNext:hover,
+    .modalControls:hover ~ .modalPrev,
+    .modalControls:hover ~ .modalNext,
+    .modalControls:hover .cursor {
+        opacity: 1;
+    }
+
+    .modalPrev,
+    .modalNext,
+    .modalControls .cursor {
+        opacity: var(--sd-webui-modal-lightbox-icon-opacity);
+    }
+}
+
 /* context menu (ie for the generate button) */
 
 #context-menu{
@@ -713,9 +784,9 @@ table.popup-table .link{
     position:absolute;
     display:block;
     padding:0px 0;
-    border:2px solid #a55000;
+    border:2px solid var(--primary-800);
     border-radius:8px;
-    box-shadow:1px 1px 2px #CE6400;
+    box-shadow:1px 1px 2px var(--primary-500);
     width: 200px;
 }
 
@@ -732,7 +803,7 @@ table.popup-table .link{
 }
 
 .context-menu-items a:hover{
-    background: #a55000;
+    background: var(--primary-700);
 }
 
 
@@ -740,6 +811,8 @@ table.popup-table .link{
 
 #tab_extensions table{
     border-collapse: collapse;
+    overflow-x: auto;
+    display: block;
 }
 
 #tab_extensions table td, #tab_extensions table th{
@@ -787,6 +860,24 @@ table.popup-table .link{
     display: inline-block;
 }
 
+.compact-checkbox-group  div label {
+    padding: 0.1em 0.3em !important;
+}
+
+/* extensions tab table row hover highlight */
+
+#extensions tr:hover td,
+#config_state_extensions tr:hover td,
+#available_extensions tr:hover td {
+    background: rgba(0, 0, 0, 0.15);
+}
+
+.dark #extensions tr:hover td ,
+.dark #config_state_extensions tr:hover td ,
+.dark #available_extensions tr:hover td {
+    background: rgba(255, 255, 255, 0.15);
+}
+
 /* replace original footer with ours */
 
 footer {
@@ -808,31 +899,33 @@ footer {
 
 /* extra networks UI */
 
-.extra-network-cards{
-    height: calc(100vh - 24rem);
-    overflow: clip scroll;
-    resize: vertical;
-    min-height: 52rem;
+.extra-page > div.gap{
+    gap: 0;
 }
 
-.extra-networks > div.tab-nav{
-    min-height: 3.4rem;
+.extra-page-prompts{
+    margin-bottom: 0;
 }
 
-.extra-networks > div > [id *= '_extra_']{
-    margin: 0.3em;
+.extra-page-prompts.extra-page-prompts-active{
+    margin-bottom: 1em;
 }
 
-.extra-network-subdirs{
-    padding: 0.2em 0.35em;
+.extra-networks > div.tab-nav{
+    min-height: 2.7rem;
 }
 
-.extra-network-subdirs button{
-    margin: 0 0.15em;
+.extra-networks-controls-div{
+    align-self: center;
+    margin-left: auto;
 }
+
+.extra-networks > div > [id *= '_extra_']{
+    margin: 0.3em;
+}
+
 .extra-networks .tab-nav .search,
-.extra-networks .tab-nav .sort,
-.extra-networks .tab-nav .show-dirs
+.extra-networks .tab-nav .sort
 {
     margin: 0.3em;
     align-self: center;
@@ -853,53 +946,69 @@ footer {
     width: auto;
 }
 
-.extra-network-cards .nocards{
+.extra-network-pane .nocards{
     margin: 1.25em 0.5em 0.5em 0.5em;
 }
 
-.extra-network-cards .nocards h1{
+.extra-network-pane .nocards h1{
     font-size: 1.5em;
     margin-bottom: 1em;
 }
 
-.extra-network-cards .nocards li{
+.extra-network-pane .nocards li{
     margin-left: 0.5em;
 }
 
+.extra-network-pane .card .button-row{
+    display: inline-flex;
+    visibility: hidden;
+    color: white;
+}
 
-.extra-network-cards .card .button-row{
-    display: none;
+.extra-network-pane .card .button-row {
     position: absolute;
-    color: white;
     right: 0;
-    z-index: 1
+    z-index: 1;
 }
-.extra-network-cards .card:hover .button-row{
-    display: flex;
+
+.extra-network-pane .card:hover .button-row{
+    visibility: visible;
 }
 
-.extra-network-cards .card .card-button{
+.extra-network-pane .card-button{
     color: white;
 }
 
-.extra-network-cards .card .metadata-button:before{
+.extra-network-pane .copy-path-button::before {
+    content: "⎘";
+}
+
+.extra-network-pane .metadata-button::before{
     content: "🛈";
 }
 
-.extra-network-cards .card .edit-button:before{
+.extra-network-pane .edit-button::before{
     content: "🛠";
 }
 
-.extra-network-cards .card .card-button {
+.extra-network-pane .card-button {
+    width: 1.5em;
     text-shadow: 2px 2px 3px black;
+    color: white;
     padding: 0.25em 0.1em;
-    font-size: 200%;
-    width: 1.5em;
 }
-.extra-network-cards .card .card-button:hover{
+
+.extra-network-pane .card-button:hover{
     color: red;
 }
 
+.extra-network-pane .card .card-button {
+    font-size: 2rem;
+}
+
+.extra-network-pane .card-minimal .card-button {
+    font-size: 1rem;
+}
 
 .standalone-card-preview.card .preview{
     position: absolute;
@@ -908,7 +1017,7 @@ footer {
     height:100%;
 }
 
-.extra-network-cards .card, .standalone-card-preview.card{
+.extra-network-pane .card, .standalone-card-preview.card{
     display: inline-block;
     margin: 0.5rem;
     width: 16rem;
@@ -925,15 +1034,15 @@ footer {
     background-image: url('./file=html/card-no-preview.png')
 }
 
-.extra-network-cards .card:hover{
+.extra-network-pane .card:hover{
     box-shadow: 0 0 2px 0.3em rgba(0, 128, 255, 0.35);
 }
 
-.extra-network-cards .card .actions .additional{
+.extra-network-pane .card .actions .additional{
     display: none;
 }
 
-.extra-network-cards .card .actions{
+.extra-network-pane .card .actions{
     position: absolute;
     bottom: 0;
     left: 0;
@@ -944,45 +1053,45 @@ footer {
     text-shadow: 0 0 0.2em black;
 }
 
-.extra-network-cards .card .actions *{
+.extra-network-pane .card .actions *{
     color: white;
 }
 
-.extra-network-cards .card .actions .name{
+.extra-network-pane .card .actions .name{
     font-size: 1.7em;
     font-weight: bold;
     line-break: anywhere;
 }
 
-.extra-network-cards .card .actions .description {
+.extra-network-pane .card .actions .description {
     display: block;
     max-height: 3em;
     white-space: pre-wrap;
     line-height: 1.1;
 }
 
-.extra-network-cards .card .actions .description:hover {
+.extra-network-pane .card .actions .description:hover {
     max-height: none;
 }
 
-.extra-network-cards .card .actions:hover .additional{
+.extra-network-pane .card .actions:hover .additional{
     display: block;
 }
 
-.extra-network-cards .card ul{
+.extra-network-pane .card ul{
     margin: 0.25em 0 0.75em 0.25em;
     cursor: unset;
 }
 
-.extra-network-cards .card ul a{
+.extra-network-pane .card ul a{
     cursor: pointer;
 }
 
-.extra-network-cards .card ul a:hover{
+.extra-network-pane .card ul a:hover{
     color: red;
 }
 
-.extra-network-cards .card .preview{
+.extra-network-pane .card .preview{
     position: absolute;
     object-fit: cover;
     width: 100%;
@@ -1025,9 +1134,6 @@ div.block.gradio-box.edit-user-metadata {
     margin-top: 1.5em;
 }
 
-
-
-
 div.block.gradio-box.popup-dialog, .popup-dialog {
     width: 56em;
     background: var(--body-background-fill);
@@ -1102,3 +1208,460 @@ body.resizing .resize-handle {
     left: 7.5px;
     border-left: 1px dashed var(--border-color-primary);
 }
+
+/* ========================= */
+.extra-network-pane {
+    display: flex;
+    height: calc(100vh - 24rem);
+    resize: vertical;
+    min-height: 52rem;
+    flex-direction: column;
+    overflow: hidden;
+}
+
+.extra-network-pane .extra-network-pane-content-dirs {
+    display: flex;
+    flex: 1;
+    flex-direction: column;
+    overflow: hidden;
+}
+
+.extra-network-pane .extra-network-pane-content-tree {
+    display: flex;
+    flex: 1;
+    overflow: hidden;
+}
+
+.extra-network-dirs-hidden .extra-network-dirs{ display: none; }
+.extra-network-dirs-hidden .extra-network-tree{ display: none; }
+.extra-network-dirs-hidden .resize-handle { display: none; }
+.extra-network-dirs-hidden .resize-handle-row { display: flex !important; }
+
+.extra-network-pane .extra-network-tree {
+    flex: 1;
+    font-size: 1rem;
+    border: 1px solid var(--block-border-color);
+    overflow: clip auto !important;
+}
+
+.extra-network-pane .extra-network-cards {
+    flex: 3;
+    overflow: clip auto !important;
+    border: 1px solid var(--block-border-color);
+}
+
+.extra-network-pane .extra-network-tree .tree-list {
+    flex: 1;
+    display: flex;
+    flex-direction: column;
+    padding: 0;
+    width: 100%;
+    overflow: hidden;
+}
+
+
+.extra-network-pane .extra-network-cards::-webkit-scrollbar,
+.extra-network-pane .extra-network-tree::-webkit-scrollbar {
+    background-color: transparent;
+    width: 16px;
+}
+
+.extra-network-pane .extra-network-cards::-webkit-scrollbar-track,
+.extra-network-pane .extra-network-tree::-webkit-scrollbar-track {
+    background-color: transparent;
+    background-clip: content-box;
+}
+
+.extra-network-pane .extra-network-cards::-webkit-scrollbar-thumb,
+.extra-network-pane .extra-network-tree::-webkit-scrollbar-thumb {
+    background-color: var(--border-color-primary);
+    border-radius: 16px;
+    border: 4px solid var(--background-fill-primary);
+}
+
+.extra-network-pane .extra-network-cards::-webkit-scrollbar-button,
+.extra-network-pane .extra-network-tree::-webkit-scrollbar-button {
+    display: none;
+}
+
+.extra-network-control {
+    position: relative;
+    display: flex;
+    width: 100%;
+    padding: 0 !important;
+    margin-top: 0 !important;
+    margin-bottom: 0 !important;
+    font-size: 1rem;
+    text-align: left;
+    user-select: none;
+    background-color: transparent;
+    border: none;
+    transition: background 33.333ms linear;
+    grid-template-rows: min-content;
+    grid-template-columns: minmax(0, auto) repeat(4, min-content);
+    grid-gap: 0.1rem;
+    align-items: start;
+}
+
+.extra-network-control small{
+    color: var(--input-placeholder-color);
+    line-height: 2.2rem;
+    margin: 0 0.5rem 0 0.75rem;
+}
+
+.extra-network-tree .tree-list--tree {}
+
+/* Remove auto indentation from tree. Will be overridden later. */
+.extra-network-tree .tree-list--subgroup {
+    margin: 0 !important;
+    padding: 0 !important;
+    box-shadow: 0.5rem 0 0 var(--body-background-fill) inset,
+                0.7rem 0 0 var(--neutral-800) inset;
+}
+
+/* Set indentation for each depth of tree. */
+.extra-network-tree .tree-list--subgroup > .tree-list-item {
+    margin-left: 0.4rem !important;
+    padding-left: 0.4rem !important;
+}
+
+/* Styles for tree <li> elements. */
+.extra-network-tree .tree-list-item {
+    list-style: none;
+    position: relative;
+    background-color: transparent;
+}
+
+/* Directory <ul> visibility based on data-expanded attribute. */
+.extra-network-tree .tree-list-content+.tree-list--subgroup {
+    height: 0;
+    visibility: hidden;
+    opacity: 0;
+}
+
+.extra-network-tree .tree-list-content[data-expanded]+.tree-list--subgroup {
+    height: auto;
+    visibility: visible;
+    opacity: 1;
+}
+
+/* File <li> */
+.extra-network-tree .tree-list-item--subitem {
+    padding-top: 0 !important;
+    padding-bottom: 0 !important;
+    margin-top: 0 !important;
+    margin-bottom: 0 !important;
+}
+
+/* <li> containing <ul> */
+.extra-network-tree .tree-list-item--has-subitem {}
+
+/* BUTTON ELEMENTS */
+/* <button> */
+.extra-network-tree .tree-list-content {
+    position: relative;
+    display: grid;
+    width: 100%;
+    padding: 0 !important;
+    margin-top: 0 !important;
+    margin-bottom: 0 !important;
+    font-size: 1rem;
+    text-align: left;
+    user-select: none;
+    background-color: transparent;
+    border: none;
+    transition: background 33.333ms linear;
+    grid-template-rows: min-content;
+    grid-template-areas: "leading-action leading-visual label trailing-visual trailing-action";
+    grid-template-columns: min-content min-content minmax(0, auto) min-content min-content;
+    grid-gap: 0.1rem;
+    align-items: start;
+    flex-grow: 1;
+    flex-basis: 100%;
+}
+/* Buttons for directories. */
+.extra-network-tree .tree-list-content-dir {}   
+
+/* Buttons for files. */
+.extra-network-tree .tree-list-item--has-subitem .tree-list--subgroup > li:first-child {
+    padding-top: 0.5rem !important;
+}
+
+.dark .extra-network-tree div.tree-list-content:hover {
+    -webkit-transition: all 0.05s ease-in-out;
+	transition: all 0.05s ease-in-out;
+    background-color: var(--neutral-800);
+}
+
+.dark .extra-network-tree div.tree-list-content[data-selected] {
+    background-color: var(--neutral-700);
+}
+
+.extra-network-tree div.tree-list-content[data-selected] {
+    background-color: var(--neutral-300);
+}
+
+.extra-network-tree div.tree-list-content:hover {
+    -webkit-transition: all 0.05s ease-in-out;
+	transition: all 0.05s ease-in-out;
+    background-color: var(--neutral-200);
+}
+
+/* ==== CHEVRON ICON ACTIONS ==== */
+/* Define the animation for the arrow when it is clicked. */
+.extra-network-tree .tree-list-content-dir .tree-list-item-action-chevron {
+    -ms-transform: rotate(135deg);
+    -webkit-transform: rotate(135deg);
+    transform: rotate(135deg);
+    transition: transform 0.2s;
+}
+
+.extra-network-tree .tree-list-content-dir[data-expanded] .tree-list-item-action-chevron {
+    -ms-transform: rotate(225deg);
+    -webkit-transform: rotate(225deg);
+    transform: rotate(225deg);
+    transition: transform 0.2s;
+}
+
+.tree-list-item-action-chevron {
+    display: inline-flex;
+    /* Uses box shadow to generate a pseudo chevron `>` icon. */
+    padding: 0.3rem;
+    box-shadow: 0.1rem 0.1rem 0 0 var(--neutral-200) inset;
+    transform: rotate(135deg);
+}
+
+/* ==== SEARCH INPUT ACTIONS ==== */
+/* Add icon to left side of <input> */
+.extra-network-control .extra-network-control--search::before {
+    content: "🔎︎";
+    position: absolute;
+    margin: 0.5rem;
+    font-size: 1rem;
+    color: var(--input-placeholder-color);
+}
+
+.extra-network-control .extra-network-control--search {
+    display: inline-flex;
+    position: relative;
+}
+
+.extra-network-control .extra-network-control--search .extra-network-control--search-text {
+    border: 1px solid var(--button-secondary-border-color);
+    border-radius: 0.5rem;
+    color: var(--button-secondary-text-color);
+    background-color: transparent;
+    width: 100%;
+    padding-left: 2rem;
+    line-height: 1rem;
+}
+
+
+.extra-network-control .extra-network-control--search .extra-network-control--search-text::placeholder {
+    color: var(--input-placeholder-color);
+}
+
+
+/* <input> clear button (x on right side) styling */
+.extra-network-control .extra-network-control--search .extra-network-control--search-text::-webkit-search-cancel-button {
+    -webkit-appearance: none;
+    appearance: none;
+    cursor: pointer;
+    height: 1rem;
+    width: 1rem;
+    mask-image: url('data:image/svg+xml,<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="black" stroke-width="4" stroke-linecap="round" stroke-linejoin="round"><line x1="18" y1="6" x2="6" y2="18"></line><line x1="6" y1="6" x2="18" y2="18"></line></svg>');
+    mask-repeat: no-repeat;
+    mask-position: center center;
+    mask-size: 100%;
+    background-color: var(--input-placeholder-color);
+}
+
+/* ==== SORT ICON ACTIONS ==== */
+.extra-network-control .extra-network-control--sort {
+    padding: 0.25rem;
+    display: inline-flex;
+    cursor: pointer;
+    justify-self: center;
+    align-self: center;
+}
+
+.extra-network-control .extra-network-control--sort .extra-network-control--sort-icon {
+    height: 1.5rem;
+    width: 1.5rem;
+    mask-repeat: no-repeat;
+    mask-position: center center;
+    mask-size: 100%;
+    background-color: var(--input-placeholder-color);
+}
+
+.extra-network-control .extra-network-control--sort[data-sortkey="default"] .extra-network-control--sort-icon {
+    mask-image: url('data:image/svg+xml,<svg viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg"><g id="SVGRepo_bgCarrier" stroke-width="0"></g><g id="SVGRepo_tracerCarrier" stroke-linecap="round" stroke-linejoin="round"></g><g id="SVGRepo_iconCarrier"><path fill-rule="evenodd" clip-rule="evenodd" d="M1 5C1 3.34315 2.34315 2 4 2H8.43845C9.81505 2 11.015 2.93689 11.3489 4.27239L11.7808 6H13.5H20C21.6569 6 23 7.34315 23 9V11C23 11.5523 22.5523 12 22 12C21.4477 12 21 11.5523 21 11V9C21 8.44772 20.5523 8 20 8H13.5H11.7808H4C3.44772 8 3 8.44772 3 9V10V19C3 19.5523 3.44772 20 4 20H9C9.55228 20 10 20.4477 10 21C10 21.5523 9.55228 22 9 22H4C2.34315 22 1 20.6569 1 19V10V9V5ZM3 6.17071C3.31278 6.06015 3.64936 6 4 6H9.71922L9.40859 4.75746C9.2973 4.3123 8.89732 4 8.43845 4H4C3.44772 4 3 4.44772 3 5V6.17071ZM20.1716 18.7574C20.6951 17.967 21 17.0191 21 16C21 13.2386 18.7614 11 16 11C13.2386 11 11 13.2386 11 16C11 18.7614 13.2386 21 16 21C17.0191 21 17.967 20.6951 18.7574 20.1716L21.2929 22.7071C21.6834 23.0976 22.3166 23.0976 22.7071 22.7071C23.0976 22.3166 23.0976 21.6834 22.7071 21.2929L20.1716 18.7574ZM13 16C13 14.3431 14.3431 13 16 13C17.6569 13 19 14.3431 19 16C19 17.6569 17.6569 19 16 19C14.3431 19 13 17.6569 13 16Z" fill="%23000000"></path></g></svg>');
+}
+
+.extra-network-control .extra-network-control--sort[data-sortkey="name"] .extra-network-control--sort-icon {
+    mask-image: url('data:image/svg+xml,<svg viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg"><g id="SVGRepo_bgCarrier" stroke-width="0"></g><g id="SVGRepo_tracerCarrier" stroke-linecap="round" stroke-linejoin="round"></g><g id="SVGRepo_iconCarrier"><path fill-rule="evenodd" clip-rule="evenodd" d="M17.1841 6.69223C17.063 6.42309 16.7953 6.25 16.5002 6.25C16.2051 6.25 15.9374 6.42309 15.8162 6.69223L11.3162 16.6922C11.1463 17.07 11.3147 17.514 11.6924 17.6839C12.0701 17.8539 12.5141 17.6855 12.6841 17.3078L14.1215 14.1136H18.8789L20.3162 17.3078C20.4862 17.6855 20.9302 17.8539 21.308 17.6839C21.6857 17.514 21.8541 17.07 21.6841 16.6922L17.1841 6.69223ZM16.5002 8.82764L14.7965 12.6136H18.2039L16.5002 8.82764Z" fill="%231C274C"></path><path opacity="0.5" fill-rule="evenodd" clip-rule="evenodd" d="M2.25 7C2.25 6.58579 2.58579 6.25 3 6.25H13C13.4142 6.25 13.75 6.58579 13.75 7C13.75 7.41421 13.4142 7.75 13 7.75H3C2.58579 7.75 2.25 7.41421 2.25 7Z" fill="%231C274C"></path><path opacity="0.5" d="M2.25 12C2.25 11.5858 2.58579 11.25 3 11.25H10C10.4142 11.25 10.75 11.5858 10.75 12C10.75 12.4142 10.4142 12.75 10 12.75H3C2.58579 12.75 2.25 12.4142 2.25 12Z" fill="%231C274C"></path><path opacity="0.5" d="M2.25 17C2.25 16.5858 2.58579 16.25 3 16.25H8C8.41421 16.25 8.75 16.5858 8.75 17C8.75 17.4142 8.41421 17.75 8 17.75H3C2.58579 17.75 2.25 17.4142 2.25 17Z" fill="%231C274C"></path></g></svg>');
+}
+
+.extra-network-control .extra-network-control--sort[data-sortkey="date_created"] .extra-network-control--sort-icon {
+    mask-image: url('data:image/svg+xml,<svg viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg"><g id="SVGRepo_bgCarrier" stroke-width="0"></g><g id="SVGRepo_tracerCarrier" stroke-linecap="round" stroke-linejoin="round"></g><g id="SVGRepo_iconCarrier"><path d="M17 11C14.2386 11 12 13.2386 12 16C12 18.7614 14.2386 21 17 21C19.7614 21 22 18.7614 22 16C22 13.2386 19.7614 11 17 11ZM17 11V9M2 9V15.8C2 16.9201 2 17.4802 2.21799 17.908C2.40973 18.2843 2.71569 18.5903 3.09202 18.782C3.51984 19 4.0799 19 5.2 19H13M2 9V8.2C2 7.0799 2 6.51984 2.21799 6.09202C2.40973 5.71569 2.71569 5.40973 3.09202 5.21799C3.51984 5 4.0799 5 5.2 5H13.8C14.9201 5 15.4802 5 15.908 5.21799C16.2843 5.40973 16.5903 5.71569 16.782 6.09202C17 6.51984 17 7.0799 17 8.2V9M2 9H17M5 3V5M14 3V5M15 16H17M17 16H19M17 16V14M17 16V18" stroke="black" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></g></svg>');
+}
+
+.extra-network-control .extra-network-control--sort[data-sortkey="date_modified"] .extra-network-control--sort-icon {
+    mask-image: url('data:image/svg+xml,<svg viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg"><g id="SVGRepo_bgCarrier" stroke-width="0"></g><g id="SVGRepo_tracerCarrier" stroke-linecap="round" stroke-linejoin="round"></g><g id="SVGRepo_iconCarrier"><path d="M10 21H6.2C5.0799 21 4.51984 21 4.09202 20.782C3.71569 20.5903 3.40973 20.2843 3.21799 19.908C3 19.4802 3 18.9201 3 17.8V8.2C3 7.0799 3 6.51984 3.21799 6.09202C3.40973 5.71569 3.71569 5.40973 4.09202 5.21799C4.51984 5 5.0799 5 6.2 5H17.8C18.9201 5 19.4802 5 19.908 5.21799C20.2843 5.40973 20.5903 5.71569 20.782 6.09202C21 6.51984 21 7.0799 21 8.2V10M7 3V5M17 3V5M3 9H21M13.5 13.0001L7 13M10 17.0001L7 17M14 21L16.025 20.595C16.2015 20.5597 16.2898 20.542 16.3721 20.5097C16.4452 20.4811 16.5147 20.4439 16.579 20.399C16.6516 20.3484 16.7152 20.2848 16.8426 20.1574L21 16C21.5523 15.4477 21.5523 14.5523 21 14C20.4477 13.4477 19.5523 13.4477 19 14L14.8426 18.1574C14.7152 18.2848 14.6516 18.3484 14.601 18.421C14.5561 18.4853 14.5189 18.5548 14.4903 18.6279C14.458 18.7102 14.4403 18.7985 14.405 18.975L14 21Z" stroke="black" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></g></svg>');
+}
+
+/* ==== SORT DIRECTION ICON ACTIONS ==== */
+.extra-network-control .extra-network-control--sort-dir {
+    padding: 0.25rem;
+    display: inline-flex;
+    cursor: pointer;
+    justify-self: center;
+    align-self: center;
+}
+
+.extra-network-control .extra-network-control--sort-dir .extra-network-control--sort-dir-icon {
+    height: 1.5rem;
+    width: 1.5rem;
+    mask-repeat: no-repeat;
+    mask-position: center center;
+    mask-size: 100%;
+    background-color: var(--input-placeholder-color);
+}
+
+.extra-network-control .extra-network-control--sort-dir[data-sortdir="Ascending"] .extra-network-control--sort-dir-icon {
+    mask-image: url('data:image/svg+xml,<svg viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg"><g id="SVGRepo_bgCarrier" stroke-width="0"></g><g id="SVGRepo_tracerCarrier" stroke-linecap="round" stroke-linejoin="round"></g><g id="SVGRepo_iconCarrier"><path d="M13 12H21M13 8H21M13 16H21M6 7V17M6 7L3 10M6 7L9 10" stroke="black" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></g></svg>');
+}
+
+.extra-network-control .extra-network-control--sort-dir[data-sortdir="Descending"] .extra-network-control--sort-dir-icon {
+    mask-image: url('data:image/svg+xml,<svg viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg"><g id="SVGRepo_bgCarrier" stroke-width="0"></g><g id="SVGRepo_tracerCarrier" stroke-linecap="round" stroke-linejoin="round"></g><g id="SVGRepo_iconCarrier"><path d="M13 12H21M13 8H21M13 16H21M6 7V17M6 17L3 14M6 17L9 14" stroke="black" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"></path></g></svg>');
+}
+
+/* ==== TREE VIEW ICON ACTIONS ==== */
+.extra-network-control .extra-network-control--tree-view {
+    padding: 0.25rem;
+    display: inline-flex;
+    cursor: pointer;
+    justify-self: center;
+    align-self: center;
+}
+
+.extra-network-control .extra-network-control--tree-view .extra-network-control--tree-view-icon {
+    height: 1.5rem;
+    width: 1.5rem;
+    mask-image: url('data:image/svg+xml,<svg viewBox="0 0 16 16" version="1.1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" fill="black"><g id="SVGRepo_bgCarrier" stroke-width="0"></g><g id="SVGRepo_tracerCarrier" stroke-linecap="round" stroke-linejoin="round"></g><g id="SVGRepo_iconCarrier"><path fill="black" d="M16 10v-4h-11v1h-2v-3h9v-4h-12v4h2v10h3v2h11v-4h-11v1h-2v-5h2v2z"></path></g></svg>');
+    mask-repeat: no-repeat;
+    mask-position: center center;
+    mask-size: 100%;
+    background-color: var(--input-placeholder-color);
+}
+
+.extra-network-control .extra-network-control--enabled {
+    background-color: rgba(0, 0, 0, 0.1);
+    border-radius: 0.25rem;
+}
+
+.dark .extra-network-control .extra-network-control--enabled {
+    background-color: rgba(255, 255, 255, 0.15);
+}
+
+.extra-network-control .extra-network-control--enabled .extra-network-control--icon{
+    background-color: var(--button-secondary-text-color);
+}
+
+/* ==== REFRESH ICON ACTIONS ==== */
+.extra-network-control .extra-network-control--refresh {
+    padding: 0.25rem;
+    display: inline-flex;
+    cursor: pointer;
+    justify-self: center;
+    align-self: center;
+}
+
+.extra-network-control .extra-network-control--refresh .extra-network-control--refresh-icon {
+    height: 1.5rem;
+    width: 1.5rem;
+    mask-image: url('data:image/svg+xml,<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="black" stroke-width="2" stroke-linecap="round" stroke-linejoin="bevel"><path d="M21.5 2v6h-6M21.34 15.57a10 10 0 1 1-.57-8.38"/></svg>');
+    mask-repeat: no-repeat;
+    mask-position: center center;
+    mask-size: 100%;
+    background-color: var(--input-placeholder-color);
+}
+
+.extra-network-control .extra-network-control--refresh-icon:active {
+    -ms-transform: rotate(180deg);
+    -webkit-transform: rotate(180deg);
+    transform: rotate(180deg);
+    transition: transform 0.2s;
+}
+
+/* ==== TREE GRID CONFIG ==== */
+
+/* Text for button. */
+.extra-network-tree .tree-list-item-label {
+    position: relative;
+    line-height: 1.25rem;
+    color: var(--button-secondary-text-color);
+    grid-area: label;
+    padding-left: 0.5rem;
+}
+
+/* Text for button truncated. */
+.extra-network-tree .tree-list-item-label--truncate {
+    overflow: hidden;
+    text-overflow: ellipsis;
+    white-space: nowrap;
+}
+
+/* Icon for button. */
+.extra-network-tree .tree-list-item-visual {
+    min-height: 1rem;
+    color: var(--button-secondary-text-color);
+    pointer-events: none;
+    align-items: right;
+}
+
+
+/* Icon for button when it is before label. */
+.extra-network-tree .tree-list-item-visual--leading {
+    grid-area: leading-visual;
+    width: 1rem;
+    text-align: right;
+}
+
+/* Icon for button when it is after label. */
+.extra-network-tree .tree-list-item-visual--trailing {
+    grid-area: trailing-visual;
+    width: 1rem;
+    text-align: right;
+}
+
+/* Dropdown arrow for button. */
+.extra-network-tree .tree-list-item-action--leading {
+    margin-right: 0.5rem;
+    margin-left: 0.2rem;
+}
+
+.extra-network-tree .tree-list-content-file .tree-list-item-action--leading {
+    visibility: hidden;
+}
+
+.extra-network-tree .tree-list-item-action--leading {
+    grid-area: leading-action;
+}
+
+.extra-network-tree .tree-list-item-action--trailing {
+    grid-area: trailing-action;
+    display: inline-flex;
+}
+
+.extra-network-tree .tree-list-content .button-row {
+    display: inline-flex;
+    visibility: hidden;
+    color: var(--button-secondary-text-color);
+    width: 0;
+}
+
+.extra-network-tree .tree-list-content:hover .button-row {
+    visibility: visible;
+    width: auto;
+}
diff --git a/test/conftest.py b/test/conftest.py
index 31a5d9eafb8..e4fc567854d 100644
--- a/test/conftest.py
+++ b/test/conftest.py
@@ -1,10 +1,16 @@
+import base64
 import os
 
 import pytest
-import base64
-
 
 test_files_path = os.path.dirname(__file__) + "/test_files"
+test_outputs_path = os.path.dirname(__file__) + "/test_outputs"
+
+
+def pytest_configure(config):
+    # We don't want to fail on Py.test command line arguments being
+    # parsed by webui:
+    os.environ.setdefault("IGNORE_CMD_ARGS_ERRORS", "1")
 
 
 def file_to_base64(filename):
@@ -23,3 +29,8 @@ def img2img_basic_image_base64() -> str:
 @pytest.fixture(scope="session")  # session so we don't read this over and over
 def mask_basic_image_base64() -> str:
     return file_to_base64(os.path.join(test_files_path, "mask_basic.png"))
+
+
+@pytest.fixture(scope="session")
+def initialize() -> None:
+    import webui  # noqa: F401
diff --git a/test/test_face_restorers.py b/test/test_face_restorers.py
new file mode 100644
index 00000000000..7760d51bfbe
--- /dev/null
+++ b/test/test_face_restorers.py
@@ -0,0 +1,29 @@
+import os
+from test.conftest import test_files_path, test_outputs_path
+
+import numpy as np
+import pytest
+from PIL import Image
+
+
+@pytest.mark.usefixtures("initialize")
+@pytest.mark.parametrize("restorer_name", ["gfpgan", "codeformer"])
+def test_face_restorers(restorer_name):
+    from modules import shared
+
+    if restorer_name == "gfpgan":
+        from modules import gfpgan_model
+        gfpgan_model.setup_model(shared.cmd_opts.gfpgan_models_path)
+        restorer = gfpgan_model.gfpgan_fix_faces
+    elif restorer_name == "codeformer":
+        from modules import codeformer_model
+        codeformer_model.setup_model(shared.cmd_opts.codeformer_models_path)
+        restorer = codeformer_model.codeformer.restore
+    else:
+        raise NotImplementedError("...")
+    img = Image.open(os.path.join(test_files_path, "two-faces.jpg"))
+    np_img = np.array(img, dtype=np.uint8)
+    fixed_image = restorer(np_img)
+    assert fixed_image.shape == np_img.shape
+    assert not np.allclose(fixed_image, np_img)  # should have visibly changed
+    Image.fromarray(fixed_image).save(os.path.join(test_outputs_path, f"{restorer_name}.png"))
diff --git a/test/test_files/two-faces.jpg b/test/test_files/two-faces.jpg
new file mode 100644
index 00000000000..c9d1b01032a
Binary files /dev/null and b/test/test_files/two-faces.jpg differ
diff --git a/test/test_outputs/.gitkeep b/test/test_outputs/.gitkeep
new file mode 100644
index 00000000000..e69de29bb2d
diff --git a/test/test_torch_utils.py b/test/test_torch_utils.py
new file mode 100644
index 00000000000..23ccb93a464
--- /dev/null
+++ b/test/test_torch_utils.py
@@ -0,0 +1,19 @@
+import types
+
+import pytest
+import torch
+
+from modules import torch_utils
+
+
+@pytest.mark.parametrize("wrapped", [True, False])
+def test_get_param(wrapped):
+    mod = torch.nn.Linear(1, 1)
+    cpu = torch.device("cpu")
+    mod.to(dtype=torch.float16, device=cpu)
+    if wrapped:
+        # more or less how spandrel wraps a thing
+        mod = types.SimpleNamespace(model=mod)
+    p = torch_utils.get_param(mod)
+    assert p.dtype == torch.float16
+    assert p.device == cpu
diff --git a/webui-macos-env.sh b/webui-macos-env.sh
index 24bc5c42615..00a36e1770e 100644
--- a/webui-macos-env.sh
+++ b/webui-macos-env.sh
@@ -4,14 +4,14 @@
 # Please modify webui-user.sh to change these instead of this file #
 ####################################################################
 
-if [[ -x "$(command -v python3.10)" ]]
-then
-    python_cmd="python3.10"
-fi
-
 export install_dir="$HOME"
 export COMMANDLINE_ARGS="--skip-torch-cuda-test --upcast-sampling --no-half-vae --use-cpu interrogate"
-export TORCH_COMMAND="pip install torch==2.0.1 torchvision==0.15.2"
 export PYTORCH_ENABLE_MPS_FALLBACK=1
 
+if [[ "$(sysctl -n machdep.cpu.brand_string)" =~ ^.*"Intel".*$ ]]; then
+    export TORCH_COMMAND="pip install torch==2.1.2 torchvision==0.16.2"
+else
+    export TORCH_COMMAND="pip install torch==2.3.1 torchvision==0.18.1"
+fi
+
 ####################################################################
diff --git a/webui.bat b/webui.bat
index 42e7d517d18..7b162ce27cc 100644
--- a/webui.bat
+++ b/webui.bat
@@ -1,6 +1,11 @@
 @echo off
 
+if exist webui.settings.bat (
+    call webui.settings.bat
+)
+
 if not defined PYTHON (set PYTHON=python)
+if defined GIT (set "GIT_PYTHON_GIT_EXECUTABLE=%GIT%")
 if not defined VENV_DIR (set "VENV_DIR=%~dp0%venv")
 
 set SD_WEBUI_RESTART=tmp/restart
@@ -32,12 +37,18 @@ if %ERRORLEVEL% == 0 goto :activate_venv
 for /f "delims=" %%i in ('CALL %PYTHON% -c "import sys; print(sys.executable)"') do set PYTHON_FULLNAME="%%i"
 echo Creating venv in directory %VENV_DIR% using python %PYTHON_FULLNAME%
 %PYTHON_FULLNAME% -m venv "%VENV_DIR%" >tmp/stdout.txt 2>tmp/stderr.txt
-if %ERRORLEVEL% == 0 goto :activate_venv
+if %ERRORLEVEL% == 0 goto :upgrade_pip
 echo Unable to create venv in directory "%VENV_DIR%"
 goto :show_stdout_stderr
 
+:upgrade_pip
+"%VENV_DIR%\Scripts\Python.exe" -m pip install --upgrade pip
+if %ERRORLEVEL% == 0 goto :activate_venv
+echo Warning: Failed to upgrade PIP version
+
 :activate_venv
 set PYTHON="%VENV_DIR%\Scripts\Python.exe"
+call "%VENV_DIR%\Scripts\activate.bat"
 echo venv %PYTHON%
 
 :skip_venv
diff --git a/webui.py b/webui.py
index 12328423d0d..2c417168aa6 100644
--- a/webui.py
+++ b/webui.py
@@ -39,7 +39,7 @@ def api_only():
 
     print(f"Startup time: {startup_timer.summary()}.")
     api.launch(
-        server_name="0.0.0.0" if cmd_opts.listen else "127.0.0.1",
+        server_name=initialize_util.gradio_server_name(),
         port=cmd_opts.port if cmd_opts.port else 7861,
         root_path=f"/{cmd_opts.subpath}" if cmd_opts.subpath else ""
     )
@@ -74,7 +74,7 @@ def webui():
             if shared.opts.auto_launch_browser == "Remote" or cmd_opts.autolaunch:
                 auto_launch_browser = True
             elif shared.opts.auto_launch_browser == "Local":
-                auto_launch_browser = not any([cmd_opts.listen, cmd_opts.share, cmd_opts.ngrok, cmd_opts.server_name])
+                auto_launch_browser = not cmd_opts.webui_is_non_local
 
         app, local_url, share_url = shared.demo.launch(
             share=cmd_opts.share,
diff --git a/webui.sh b/webui.sh
index 3d0f87eed74..89dae163a82 100755
--- a/webui.sh
+++ b/webui.sh
@@ -4,12 +4,6 @@
 # change the variables in webui-user.sh instead #
 #################################################
 
-
-use_venv=1
-if [[ $venv_dir == "-" ]]; then
-  use_venv=0
-fi
-
 SCRIPT_DIR=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
 
 
@@ -28,6 +22,12 @@ then
     source "$SCRIPT_DIR"/webui-user.sh
 fi
 
+# If $venv_dir is "-", then disable venv support
+use_venv=1
+if [[ $venv_dir == "-" ]]; then
+  use_venv=0
+fi
+
 # Set defaults
 # Install directory without trailing slash
 if [[ -z "${install_dir}" ]]
@@ -44,13 +44,19 @@ fi
 # python3 executable
 if [[ -z "${python_cmd}" ]]
 then
-    python_cmd="python3"
+  python_cmd="python3.10"
+fi
+if [[ ! -x "$(command -v "${python_cmd}")" ]]
+then
+  python_cmd="python3"
 fi
 
 # git executable
 if [[ -z "${GIT}" ]]
 then
     export GIT="git"
+else
+    export GIT_PYTHON_GIT_EXECUTABLE="${GIT}"
 fi
 
 # python3 venv without trailing slash (defaults to ${install_dir}/${clone_dir}/venv)
@@ -87,7 +93,7 @@ delimiter="################################################################"
 
 printf "\n%s\n" "${delimiter}"
 printf "\e[1m\e[32mInstall script for stable-diffusion + Web UI\n"
-printf "\e[1m\e[34mTested on Debian 11 (Bullseye)\e[0m"
+printf "\e[1m\e[34mTested on Debian 11 (Bullseye), Fedora 34+ and openSUSE Leap 15.4 or newer.\e[0m"
 printf "\n%s\n" "${delimiter}"
 
 # Do not run as root
@@ -111,13 +117,13 @@ then
     exit 1
 fi
 
-if [[ -d .git ]]
+if [[ -d "$SCRIPT_DIR/.git" ]]
 then
     printf "\n%s\n" "${delimiter}"
     printf "Repo already cloned, using it as install directory"
     printf "\n%s\n" "${delimiter}"
-    install_dir="${PWD}/../"
-    clone_dir="${PWD##*/}"
+    install_dir="${SCRIPT_DIR}/../"
+    clone_dir="${SCRIPT_DIR##*/}"
 fi
 
 # Check prerequisites
@@ -127,13 +133,19 @@ case "$gpu_info" in
         export HSA_OVERRIDE_GFX_VERSION=10.3.0
         if [[ -z "${TORCH_COMMAND}" ]]
         then
-            pyv="$(${python_cmd} -c 'import sys; print(".".join(map(str, sys.version_info[0:2])))')"
-            if [[ $(bc <<< "$pyv <= 3.10") -eq 1 ]] 
+            pyv="$(${python_cmd} -c 'import sys; print(f"{sys.version_info[0]}.{sys.version_info[1]:02d}")')"
+            # Using an old nightly compiled against rocm 5.2 for Navi1, see https://github.com/pytorch/pytorch/issues/106728#issuecomment-1749511711
+            if [[ $pyv == "3.8" ]]
             then
-                # Navi users will still use torch 1.13 because 2.0 does not seem to work.
-                export TORCH_COMMAND="pip install torch==1.13.1+rocm5.2 torchvision==0.14.1+rocm5.2 --index-url https://download.pytorch.org/whl/rocm5.2"
+                export TORCH_COMMAND="pip install https://download.pytorch.org/whl/nightly/rocm5.2/torch-2.0.0.dev20230209%2Brocm5.2-cp38-cp38-linux_x86_64.whl https://download.pytorch.org/whl/nightly/rocm5.2/torchvision-0.15.0.dev20230209%2Brocm5.2-cp38-cp38-linux_x86_64.whl"
+            elif [[ $pyv == "3.9" ]]
+            then
+                export TORCH_COMMAND="pip install https://download.pytorch.org/whl/nightly/rocm5.2/torch-2.0.0.dev20230209%2Brocm5.2-cp39-cp39-linux_x86_64.whl https://download.pytorch.org/whl/nightly/rocm5.2/torchvision-0.15.0.dev20230209%2Brocm5.2-cp39-cp39-linux_x86_64.whl"
+            elif [[ $pyv == "3.10" ]]
+            then
+                export TORCH_COMMAND="pip install https://download.pytorch.org/whl/nightly/rocm5.2/torch-2.0.0.dev20230209%2Brocm5.2-cp310-cp310-linux_x86_64.whl https://download.pytorch.org/whl/nightly/rocm5.2/torchvision-0.15.0.dev20230209%2Brocm5.2-cp310-cp310-linux_x86_64.whl"
             else
-                printf "\e[1m\e[31mERROR: RX 5000 series GPUs must be using at max python 3.10, aborting...\e[0m"
+                printf "\e[1m\e[31mERROR: RX 5000 series GPUs python version must be between 3.8 and 3.10, aborting...\e[0m"
                 exit 1
             fi
         fi
@@ -141,9 +153,7 @@ case "$gpu_info" in
     *"Navi 2"*) export HSA_OVERRIDE_GFX_VERSION=10.3.0
     ;;
     *"Navi 3"*) [[ -z "${TORCH_COMMAND}" ]] && \
-         export TORCH_COMMAND="pip install --pre torch torchvision --index-url https://download.pytorch.org/whl/nightly/rocm5.6"
-        # Navi 3 needs at least 5.5 which is only on the nightly chain, previous versions are no longer online (torch==2.1.0.dev-20230614+rocm5.5 torchvision==0.16.0.dev-20230614+rocm5.5 torchaudio==2.1.0.dev-20230614+rocm5.5)
-        # so switch to nightly rocm5.6 without explicit versions this time
+         export TORCH_COMMAND="pip install torch torchvision --index-url https://download.pytorch.org/whl/nightly/rocm5.7"
     ;;
     *"Renoir"*) export HSA_OVERRIDE_GFX_VERSION=9.0.0
         printf "\n%s\n" "${delimiter}"
@@ -157,7 +167,10 @@ if ! echo "$gpu_info" | grep -q "NVIDIA";
 then
     if echo "$gpu_info" | grep -q "AMD" && [[ -z "${TORCH_COMMAND}" ]]
     then
-        export TORCH_COMMAND="pip install torch==2.0.1+rocm5.4.2 torchvision==0.15.2+rocm5.4.2 --index-url https://download.pytorch.org/whl/rocm5.4.2"
+	      export TORCH_COMMAND="pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/rocm5.7"
+    elif npu-smi info 2>/dev/null
+    then
+        export TORCH_COMMAND="pip install torch==2.1.0 torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu; pip install torch_npu==2.1.0"
     fi
 fi
 
@@ -201,12 +214,15 @@ then
     if [[ ! -d "${venv_dir}" ]]
     then
         "${python_cmd}" -m venv "${venv_dir}"
+        "${venv_dir}"/bin/python -m pip install --upgrade pip
         first_launch=1
     fi
     # shellcheck source=/dev/null
     if [[ -f "${venv_dir}"/bin/activate ]]
     then
         source "${venv_dir}"/bin/activate
+        # ensure use of python from venv
+        python_cmd="${venv_dir}"/bin/python
     else
         printf "\n%s\n" "${delimiter}"
         printf "\e[1m\e[31mERROR: Cannot activate python venv, aborting...\e[0m"
@@ -222,12 +238,45 @@ fi
 # Try using TCMalloc on Linux
 prepare_tcmalloc() {
     if [[ "${OSTYPE}" == "linux"* ]] && [[ -z "${NO_TCMALLOC}" ]] && [[ -z "${LD_PRELOAD}" ]]; then
-        TCMALLOC="$(PATH=/usr/sbin:$PATH ldconfig -p | grep -Po "libtcmalloc(_minimal|)\.so\.\d" | head -n 1)"
-        if [[ ! -z "${TCMALLOC}" ]]; then
-            echo "Using TCMalloc: ${TCMALLOC}"
-            export LD_PRELOAD="${TCMALLOC}"
-        else
-            printf "\e[1m\e[31mCannot locate TCMalloc (improves CPU memory usage)\e[0m\n"
+        # check glibc version
+        LIBC_VER=$(echo $(ldd --version | awk 'NR==1 {print $NF}') | grep -oP '\d+\.\d+')
+        echo "glibc version is $LIBC_VER"
+        libc_vernum=$(expr $LIBC_VER)
+        # Since 2.34 libpthread is integrated into libc.so
+        libc_v234=2.34
+        # Define Tcmalloc Libs arrays
+        TCMALLOC_LIBS=("libtcmalloc(_minimal|)\.so\.\d" "libtcmalloc\.so\.\d")
+        # Traversal array
+        for lib in "${TCMALLOC_LIBS[@]}"
+        do
+            # Determine which type of tcmalloc library the library supports
+            TCMALLOC="$(PATH=/sbin:/usr/sbin:$PATH ldconfig -p | grep -P $lib | head -n 1)"
+            TC_INFO=(${TCMALLOC//=>/})
+            if [[ ! -z "${TC_INFO}" ]]; then
+                echo "Check TCMalloc: ${TC_INFO}"
+                # Determine if the library is linked to libpthread and resolve undefined symbol: pthread_key_create
+                if [ $(echo "$libc_vernum < $libc_v234" | bc) -eq 1 ]; then
+                    # glibc < 2.34 pthread_key_create into libpthread.so. check linking libpthread.so...
+                    if ldd ${TC_INFO[2]} | grep -q 'libpthread'; then
+                        echo "$TC_INFO is linked with libpthread,execute LD_PRELOAD=${TC_INFO[2]}"
+                        # set fullpath LD_PRELOAD (To be on the safe side)
+                        export LD_PRELOAD="${TC_INFO[2]}"
+                        break
+                    else
+                        echo "$TC_INFO is not linked with libpthread will trigger undefined symbol: pthread_Key_create error"
+                    fi
+                else
+                    # Version 2.34 of libc.so (glibc) includes the pthread library IN GLIBC. (USE ubuntu 22.04 and modern linux system and WSL)
+                    # libc.so(glibc) is linked with a library that works in ALMOST ALL Linux userlands. SO NO CHECK!
+                    echo "$TC_INFO is linked with libc.so,execute LD_PRELOAD=${TC_INFO[2]}"
+                    # set fullpath LD_PRELOAD (To be on the safe side)
+                    export LD_PRELOAD="${TC_INFO[2]}"
+                    break
+                fi
+            fi
+        done
+        if [[ -z "${LD_PRELOAD}" ]]; then
+            printf "\e[1m\e[31mCannot locate TCMalloc. Do you have tcmalloc or google-perftool installed on your system? (improves CPU memory usage)\e[0m\n"
         fi
     fi
 }