changelog

fix image upscale on cpu

.eslintrc.js

@@ -78,6 +78,8 @@ module.exports = {
         //extraNetworks.js
         requestGet: "readonly",
         popup: "readonly",
+        // profilerVisualization.js
+        createVisualizationTable: "readonly",
         // from python
         localization: "readonly",
         // progrssbar.js

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -91,7 +91,7 @@ body:
     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 the bug occured. 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

.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==0.1.6
+        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

.github/workflows/run_tests.yaml

@@ -11,9 +11,9 @@ 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: Set up Python 3.10
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         with:
           python-version: 3.10.6
           cache: pip
@@ -22,7 +22,7 @@ jobs:
             launch.py
       - name: Cache models
         id: cache-models
-        uses: actions/cache@v3
+        uses: actions/cache@v4
         with:
           path: models
           key: "2023-12-30"
@@ -68,13 +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

.gitignore

@@ -2,6 +2,7 @@ __pycache__
 *.ckpt
 *.safetensors
 *.pth
+.DS_Store
 /ESRGAN/*
 /SwinIR/*
 /repositories
@@ -38,3 +39,6 @@ notification.mp3
 /package-lock.json
 /.coverage*
 /test/test_outputs
+/cache
+trace.json
+/sysinfo-????-??-??-??-??.json

CHANGELOG.md

@@ -1,4 +1,283 @@
-## 1.8.0-RC
+## 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
@@ -14,7 +293,7 @@
 * 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))
-* Propmpt comments support
+* Prompt comments support
 
 ### Minor:
 * Allow pasting in WIDTHxHEIGHT strings into the width/height fields ([#14296](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14296))
@@ -59,9 +338,9 @@
 * 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 overriden column scale parameter ([#15004](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15004))
+* ResizeHandleRow - allow overridden column scale parameter ([#15004](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/15004))
 
-### Performance
+### 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))
@@ -101,7 +380,7 @@
 * 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 doble upscaling in inpaint ([#14966](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14966))
+* 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))
@@ -171,7 +450,7 @@
 * 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 mutiple styles csv files ([#14125](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14125))
+* 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:
@@ -308,7 +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))
@@ -484,7 +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
@@ -703,7 +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)
@@ -733,7 +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
@@ -751,7 +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

README.md

@@ -78,7 +78,7 @@ A web interface for Stable Diffusion, implemented using Gradio library.
 - 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
@@ -98,6 +98,7 @@ Make sure the required [dependencies](https://github.com/AUTOMATIC1111/stable-di
 - [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):
 
@@ -121,16 +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 gperftools-libs libglvnd-glx 
+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
@@ -149,7 +172,7 @@ 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/Stability-AI/stablediffusion, 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
 - Spandrel - https://github.com/chaiNNer-org/spandrel implementing
   - GFPGAN - https://github.com/TencentARC/GFPGAN.git

_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"

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:

configs/alt-diffusion-m18-inference.yaml

@@ -41,7 +41,7 @@ model:
         use_linear_in_transformer: True
         transformer_depth: 1
         context_dim: 1024
-        use_checkpoint: True
+        use_checkpoint: False
         legacy: False
 
     first_stage_config:

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:

configs/sd3-inference.yaml

@@ -0,0 +1,5 @@
+model:
+  target: modules.models.sd3.sd3_model.SD3Inferencer
+  params:
+    shift: 3
+    state_dict: null

configs/sd_xl_inpaint.yaml

@@ -21,7 +21,7 @@ model:
       params:
         adm_in_channels: 2816
         num_classes: sequential
-        use_checkpoint: True
+        use_checkpoint: False
         in_channels: 9
         out_channels: 4
         model_channels: 320

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:

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:

extensions-builtin/LDSR/sd_hijack_ddpm_v1.py

@@ -301,7 +301,7 @@ class DDPMV1(pl.LightningModule):
         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 @@ class LatentDiffusionV1(DDPMV1):
         :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 @@ class LatentDiffusionV1(DDPMV1):
     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 @@ class LatentDiffusionV1(DDPMV1):
                 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 @@ class LatentDiffusionV1(DDPMV1):
                 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)

extensions-builtin/Lora/extra_networks_lora.py

@@ -9,6 +9,8 @@ class ExtraNetworkLora(extra_networks.ExtraNetwork):
         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 @@ class ExtraNetworkLora(extra_networks.ExtraNetwork):
         networks.load_networks(names, te_multipliers, unet_multipliers, dyn_dims)
 
         if shared.opts.lora_add_hashes_to_infotext:
-            network_hashes = []
+            if not getattr(p, "is_hr_pass", False) or not hasattr(p, "lora_hashes"):
+                p.lora_hashes = {}
+
             for item in networks.loaded_networks:
-                shorthash = item.network_on_disk.shorthash
-                if not shorthash:
-                    continue
+                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
 
-                alias = item.mentioned_name
-                if not alias:
-                    continue
-
-                alias = alias.replace(":", "").replace(",", "")
-
-                network_hashes.append(f"{alias}: {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:

extensions-builtin/Lora/lyco_helpers.py

@@ -30,7 +30,7 @@ def factorization(dimension: int, factor:int=-1) -> tuple[int, int]:
     In LoRA with Kroneckor Product, first value is a value for weight scale.
     secon value is a value for weight.
 
-    Becuase of non-commutative property, A⊗B ≠ B⊗A. Meaning of two matrices is slightly different.
+    Because of non-commutative property, A⊗B ≠ B⊗A. Meaning of two matrices is slightly different.
 
     examples)
     factor

extensions-builtin/Lora/network.py

@@ -7,6 +7,7 @@ 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'])
 
@@ -29,7 +30,6 @@ class NetworkOnDisk:
 
         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
 
@@ -115,8 +115,17 @@ class NetworkModule:
         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 = {}
@@ -146,6 +155,9 @@ class NetworkModule:
         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
@@ -160,6 +172,27 @@ class NetworkModule:
 
         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)
@@ -175,7 +208,12 @@ class NetworkModule:
         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()

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 @@ class ModuleTypeLora(network.ModuleType):
         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 @@ class NetworkModuleLora(network.NetworkModule):
         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:

extensions-builtin/Lora/network_oft.py

@@ -36,13 +36,6 @@ class NetworkModuleOFT(network.NetworkModule):
             # self.alpha is unused
             self.dim = self.oft_blocks.shape[1] # (num_blocks, block_size, block_size)
 
-        # 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:
-            self.rescale = self.rescale.reshape(-1, *[1]*(self.org_module[0].weight.dim() - 1))
-
         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
@@ -54,6 +47,13 @@ class NetworkModuleOFT(network.NetworkModule):
         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

extensions-builtin/Lora/networks.py

@@ -1,3 +1,4 @@
+from __future__ import annotations
 import gradio as gr
 import logging
 import os
@@ -19,6 +20,7 @@ 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
 
@@ -130,7 +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
@@ -143,6 +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)
@@ -155,12 +167,26 @@ 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.partition(".")
+
+        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)
@@ -172,7 +198,11 @@ def load_network(name, network_on_disk):
                 emb_dict[vec_name] = weight
             bundle_embeddings[emb_name] = emb_dict
 
-        key = convert_diffusers_name_to_compvis(key_network_without_network_parts, is_sd2)
+        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:
@@ -229,6 +259,7 @@ def load_network(name, network_on_disk):
     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
@@ -260,6 +291,16 @@ def load_networks(names, te_multipliers=None, unet_multipliers=None, dyn_dims=No
 
     loaded_networks.clear()
 
+    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()
@@ -325,6 +366,28 @@ def load_networks(names, te_multipliers=None, unet_multipliers=None, dyn_dims=No
     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)
@@ -334,28 +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)
@@ -367,24 +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:
@@ -392,7 +455,7 @@ 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():
                         if getattr(self, 'fp16_weight', None) is None:
@@ -429,9 +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)
 
@@ -449,6 +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
 
@@ -563,22 +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.
@@ -594,6 +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*\([0-9a-fA-F]+\)")
 
 

extensions-builtin/Lora/scripts/lora_script.py

@@ -36,6 +36,7 @@ shared.options_templates.update(shared.options_section(('extra_networks', "Extra
     "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}),

extensions-builtin/Lora/ui_edit_user_metadata.py

@@ -21,10 +21,12 @@ re_comma = re.compile(r" *, *")
 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 = {}
@@ -149,6 +151,8 @@ class LoraUserMetadataEditor(ui_extra_networks_user_metadata.UserMetadataEditor)
 
             v = random.random() * max_count
             if count > v:
+                for x in "({[]})":
+                    tag = tag.replace(x, '\\' + x)
                 res.append(tag)
 
         return ", ".join(sorted(res))

extensions-builtin/Lora/ui_extra_networks_lora.py

@@ -31,7 +31,7 @@ class ExtraNetworksPageLora(ui_extra_networks.ExtraNetworksPage):
             "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_terms": search_terms,
             "local_preview": f"{path}.{shared.opts.samples_format}",
@@ -60,7 +60,7 @@ class ExtraNetworksPageLora(ui_extra_networks.ExtraNetworksPage):
         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

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];
     }
 
@@ -252,6 +254,7 @@ onUiLoaded(async() => {
     let isMoving = false;
     let mouseX, mouseY;
     let activeElement;
+    let interactedWithAltKey = false;
 
     const elements = Object.fromEntries(
         Object.keys(elementIDs).map(id => [
@@ -277,7 +280,7 @@ onUiLoaded(async() => {
         const targetElement = gradioApp().querySelector(elemId);
 
         if (!targetElement) {
-            console.log("Element not found");
+            console.log("Element not found", elemId);
             return;
         }
 
@@ -292,7 +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";
@@ -355,7 +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
@@ -365,9 +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") {
@@ -508,6 +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) {
@@ -783,23 +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);
             }
@@ -839,6 +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;

extensions-builtin/canvas-zoom-and-pan/scripts/hotkey_config.py

@@ -8,8 +8,8 @@ shared.options_templates.update(shared.options_section(('canvas_hotkey', "Canvas
     "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"),

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, infotext_utils
+from modules import scripts, shared, ui_components, ui_settings, infotext_utils, errors
 from modules.ui_components import FormColumn
 
 
@@ -42,7 +42,11 @@ class ExtraOptionsSection(scripts.Script):
                             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)

extensions-builtin/hypertile/scripts/hypertile_script.py

@@ -1,6 +1,5 @@
 import hypertile
 from modules import scripts, script_callbacks, shared
-from scripts.hypertile_xyz import add_axis_options
 
 
 class ScriptHypertile(scripts.Script):
@@ -93,7 +92,6 @@ def on_ui_settings():
         "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"),
@@ -105,5 +103,20 @@ def on_ui_settings():
         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)

extensions-builtin/hypertile/scripts/hypertile_xyz.py

@@ -1,51 +0,0 @@
-from modules import scripts
-from modules.shared import opts
-
-xyz_grid = [x for x in scripts.scripts_data if x.script_class.__module__ == "xyz_grid.py"][0].module
-
-def int_applier(value_name:str, min_range:int = -1, max_range:int = -1):
-    """
-    Returns a function that applies the given value to the given value_name in opts.data.
-    """
-    def validate(value_name:str, value:str):
-        value = int(value)
-        # validate value
-        if not min_range == -1:
-            assert value >= min_range, f"Value {value} for {value_name} must be greater than or equal to {min_range}"
-        if not max_range == -1:
-            assert value <= max_range, f"Value {value} for {value_name} must be less than or equal to {max_range}"
-    def apply_int(p, x, xs):
-        validate(value_name, x)
-        opts.data[value_name] = int(x)
-    return apply_int
-
-def bool_applier(value_name:str):
-    """
-    Returns a function that applies the given value to the given value_name in opts.data.
-    """
-    def validate(value_name:str, value:str):
-        assert value.lower() in ["true", "false"], f"Value {value} for {value_name} must be either true or false"
-    def apply_bool(p, x, xs):
-        validate(value_name, x)
-        value_boolean = x.lower() == "true"
-        opts.data[value_name] = value_boolean
-    return apply_bool
-
-def add_axis_options():
-    extra_axis_options = [
-        xyz_grid.AxisOption("[Hypertile] Unet First pass Enabled", str, bool_applier("hypertile_enable_unet"), choices=xyz_grid.boolean_choice(reverse=True)),
-        xyz_grid.AxisOption("[Hypertile] Unet Second pass Enabled", str, bool_applier("hypertile_enable_unet_secondpass"), choices=xyz_grid.boolean_choice(reverse=True)),
-        xyz_grid.AxisOption("[Hypertile] Unet Max Depth", int, int_applier("hypertile_max_depth_unet", 0, 3), choices=lambda: [str(x) for x in range(4)]),
-        xyz_grid.AxisOption("[Hypertile] Unet Max Tile Size", int, int_applier("hypertile_max_tile_unet", 0, 512)),
-        xyz_grid.AxisOption("[Hypertile] Unet Swap Size", int, int_applier("hypertile_swap_size_unet", 0, 64)),
-        xyz_grid.AxisOption("[Hypertile] VAE Enabled", str, bool_applier("hypertile_enable_vae"), choices=xyz_grid.boolean_choice(reverse=True)),
-        xyz_grid.AxisOption("[Hypertile] VAE Max Depth", int, int_applier("hypertile_max_depth_vae", 0, 3), choices=lambda: [str(x) for x in range(4)]),
-        xyz_grid.AxisOption("[Hypertile] VAE Max Tile Size", int, int_applier("hypertile_max_tile_vae", 0, 512)),
-        xyz_grid.AxisOption("[Hypertile] VAE Swap Size", int, int_applier("hypertile_swap_size_vae", 0, 64)),
-    ]
-    set_a = {opt.label for opt in xyz_grid.axis_options}
-    set_b = {opt.label for opt in extra_axis_options}
-    if set_a.intersection(set_b):
-        return
-
-    xyz_grid.axis_options.extend(extra_axis_options)

extensions-builtin/soft-inpainting/scripts/soft_inpainting.py

@@ -3,6 +3,7 @@ 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:
@@ -57,10 +58,14 @@ def latent_blend(settings, a, b, t):
 
     # NOTE: We use inplace operations wherever possible.
 
-    # [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)
+    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
@@ -75,13 +80,11 @@ def latent_blend(settings, a, b, t):
 
     # 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(torch.float64).add_(0.00001)
+    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(torch.float64).pow_(
-        settings.inpaint_detail_preservation) * one_minus_t3
-    b_magnitude = torch.norm(b, p=2, dim=1, keepdim=True).to(torch.float64).pow_(
-        settings.inpaint_detail_preservation) * t3
+    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
@@ -104,7 +107,7 @@ def latent_blend(settings, a, b, t):
 
 def get_modified_nmask(settings, nmask, sigma):
     """
-    Converts a negative mask representing the transparency of the original latent vectors being overlayed
+    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:
@@ -135,7 +138,10 @@ def apply_adaptive_masks(
     from PIL import Image, ImageOps, ImageFilter
 
     # TODO: Bias the blending according to the latent mask, add adjustable parameter for bias control.
-    latent_mask = nmask[0].float()
+    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)
@@ -157,7 +163,14 @@ def apply_adaptive_masks(
                                                    percentile_min=0.25, percentile_max=0.75, min_width=1)
 
         # The distance at which opacity of original decreases to 50%
-        half_weighted_distance = settings.composite_difference_threshold * mask_scalar
+        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)

html/extra-networks-copy-path-button.html

@@ -1,5 +1,5 @@
 <div class="copy-path-button card-button"
     title="Copy path to clipboard"
-    onclick="extraNetworksCopyCardPath(event, '{filename}')"
+    onclick="extraNetworksCopyCardPath(event)"
     data-clipboard-text="{filename}">
 </div>

html/extra-networks-edit-item-button.html

@@ -1,4 +1,4 @@
 <div class="edit-button card-button"
     title="Edit metadata"
-    onclick="extraNetworksEditUserMetadata(event, '{tabname}', '{extra_networks_tabname}', '{name}')">
+    onclick="extraNetworksEditUserMetadata(event, '{tabname}', '{extra_networks_tabname}')">
 </div>

html/extra-networks-metadata-button.html

@@ -1,4 +1,4 @@
 <div class="metadata-button card-button"
     title="Show internal metadata"
-    onclick="extraNetworksRequestMetadata(event, '{extra_networks_tabname}', '{name}')">
+    onclick="extraNetworksRequestMetadata(event, '{extra_networks_tabname}')">
 </div>

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>

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>

html/extra-networks-pane.html

@@ -1,23 +1,53 @@
-<div id='{tabname}_{extra_networks_tabname}_pane' class='extra-network-pane'>
+<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="Filter files"
+                placeholder="Search"
             >
         </div>
+
+        <small>Sort: </small>
         <div 
-            id="{tabname}_{extra_networks_tabname}_extra_sort"
-            class="extra-network-control--sort"
-            data-sortmode="{data_sortmode}"
-            data-sortkey="{data_sortkey}"
+            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--sort-icon"></i>
+            <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"
@@ -25,15 +55,18 @@
             title="Sort ascending"
             onclick="extraNetworksControlSortDirOnClick(event, '{tabname}', '{extra_networks_tabname}');"
         >
-            <i class="extra-network-control--sort-dir-icon"></i>
+            <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--tree-view-icon"></i>
+            <i class="extra-network-control--icon extra-network-control--tree-view-icon"></i>
         </div>
         <div
             id="{tabname}_{extra_networks_tabname}_extra_refresh"
@@ -41,15 +74,8 @@
             title="Refresh page"
             onclick="extraNetworksControlRefreshOnClick(event, '{tabname}', '{extra_networks_tabname}');"
         >
-            <i class="extra-network-control--refresh-icon"></i>
+            <i class="extra-network-control--icon extra-network-control--refresh-icon"></i>
         </div>
     </div>
-    <div class="extra-network-pane-content">
-        <div id='{tabname}_{extra_networks_tabname}_tree' class='extra-network-tree {tree_view_div_extra_class}'>
-            {tree_html}
-        </div>
-        <div id='{tabname}_{extra_networks_tabname}_cards' class='extra-network-cards'>
-            {items_html}
-        </div>
-    </div>
-</div>
+    {pane_content}
+</div>

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;
 

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;

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;
+            }
         }
     }
 

javascript/edit-attention.js

@@ -64,6 +64,14 @@ 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;
     }

javascript/extraNetworks.js

@@ -39,12 +39,12 @@ function setupExtraNetworksForTab(tabname) {
         // tabname_full = {tabname}_{extra_networks_tabname}
         var tabname_full = elem.id;
         var search = gradioApp().querySelector("#" + tabname_full + "_extra_search");
-        var sort_mode = gradioApp().querySelector("#" + tabname_full + "_extra_sort");
         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_mode || !sort_dir || !refresh) {
+        if (!search || !sort_dir || !refresh) {
             return; // `return` is equivalent of `continue` but for forEach loops.
         }
 
@@ -52,7 +52,7 @@ function setupExtraNetworksForTab(tabname) {
             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'), function(t) {
+                var text = Array.prototype.map.call(elem.querySelectorAll('.search_terms, .description'), function(t) {
                     return t.textContent.toLowerCase();
                 }).join(" ");
 
@@ -71,42 +71,46 @@ function setupExtraNetworksForTab(tabname) {
         };
 
         var applySort = function(force) {
-            var cards = gradioApp().querySelectorAll('#' + tabname + '_extra_tabs div.card');
+            var cards = gradioApp().querySelectorAll('#' + tabname_full + ' div.card');
+            var parent = gradioApp().querySelector('#' + tabname_full + "_cards");
             var reverse = sort_dir.dataset.sortdir == "Descending";
-            var sortKey = sort_mode.dataset.sortmode.toLowerCase().replace("sort", "").replaceAll(" ", "_").replace(/_+$/, "").trim() || "name";
-            sortKey = "sort" + sortKey.charAt(0).toUpperCase() + sortKey.slice(1);
-            var sortKeyStore = sortKey + "-" + (reverse ? "Descending" : "Ascending") + "-" + cards.length;
+            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 == sort_mode.dataset.sortkey && !force) {
+            if (sortKeyStore == currentSort && !force) {
                 return;
             }
-            sort_mode.dataset.sortkey = sortKeyStore;
+            currentSort = sortKeyStore;
 
-            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];
+                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();
-            });
+
+            parent.innerHTML = '';
+
+            var frag = document.createDocumentFragment();
             sortedCards.forEach(function(card) {
-                card.originalParentElement.appendChild(card);
+                frag.appendChild(card);
             });
+            parent.appendChild(frag);
         };
 
-        search.addEventListener("input", applyFilter);
+        search.addEventListener("input", function() {
+            applyFilter();
+        });
         applySort();
         applyFilter();
         extraNetworksApplySort[tabname_full] = applySort;
@@ -272,6 +276,15 @@ function saveCardPreview(event, tabname, filename) {
     event.preventDefault();
 }
 
+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();
+
+    searchTextarea.value = text;
+    updateInput(searchTextarea);
+}
+
 function extraNetworksTreeProcessFileClick(event, btn, tabname, extra_networks_tabname) {
     /**
      * Processes `onclick` events when user clicks on files in tree.
@@ -290,7 +303,7 @@ function extraNetworksTreeProcessDirectoryClick(event, btn, tabname, extra_netwo
      * Processes `onclick` events when user clicks on directories in tree.
      *
      * Here is how the tree reacts to clicks for various states:
-     * unselected unopened directory: Diretory is selected and expanded.
+     * 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.
@@ -383,36 +396,17 @@ function extraNetworksTreeOnClick(event, tabname, extra_networks_tabname) {
 }
 
 function extraNetworksControlSortOnClick(event, tabname, extra_networks_tabname) {
-    /**
-     * Handles `onclick` events for the Sort Mode button.
-     *
-     * Modifies the data attributes of the Sort Mode button to cycle between
-     * various sorting modes.
-     *
-     * @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 curr_mode = event.currentTarget.dataset.sortmode;
-    var el_sort_dir = gradioApp().querySelector("#" + tabname + "_" + extra_networks_tabname + "_extra_sort_dir");
-    var sort_dir = el_sort_dir.dataset.sortdir;
-    if (curr_mode == "path") {
-        event.currentTarget.dataset.sortmode = "name";
-        event.currentTarget.dataset.sortkey = "sortName-" + sort_dir + "-640";
-        event.currentTarget.setAttribute("title", "Sort by filename");
-    } else if (curr_mode == "name") {
-        event.currentTarget.dataset.sortmode = "date_created";
-        event.currentTarget.dataset.sortkey = "sortDate_created-" + sort_dir + "-640";
-        event.currentTarget.setAttribute("title", "Sort by date created");
-    } else if (curr_mode == "date_created") {
-        event.currentTarget.dataset.sortmode = "date_modified";
-        event.currentTarget.dataset.sortkey = "sortDate_modified-" + sort_dir + "-640";
-        event.currentTarget.setAttribute("title", "Sort by date modified");
-    } else {
-        event.currentTarget.dataset.sortmode = "path";
-        event.currentTarget.dataset.sortkey = "sortPath-" + sort_dir + "-640";
-        event.currentTarget.setAttribute("title", "Sort by path");
-    }
+    /** 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);
 }
 
@@ -447,8 +441,12 @@ function extraNetworksControlTreeViewOnClick(event, tabname, extra_networks_tabn
      * @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.
      */
-    gradioApp().getElementById(tabname + "_" + extra_networks_tabname + "_tree").classList.toggle("hidden");
-    event.currentTarget.classList.toggle("extra-network-control--enabled");
+    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) {
@@ -509,12 +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) {
@@ -543,16 +605,18 @@ function requestGet(url, data, handler, errorHandler) {
     xhr.send(js);
 }
 
-function extraNetworksCopyCardPath(event, path) {
-    navigator.clipboard.writeText(path);
+function extraNetworksCopyCardPath(event) {
+    navigator.clipboard.writeText(event.target.getAttribute("data-clipboard-text"));
     event.stopPropagation();
 }
 
-function extraNetworksRequestMetadata(event, extraPage, cardName) {
+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);
@@ -566,7 +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];
@@ -578,6 +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);
 

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 ? "🗇" : "🗆";
     const lb = gradioApp().getElementById("lightboxModal");
     modalImage.src = source.src;
     if (modalImage.style.display === 'none') {
@@ -51,14 +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)];
@@ -131,19 +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);
@@ -163,6 +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 ? "🗇" : "🗆";
+    event.stopPropagation();
+}
+
 function modalTileImageToggle(event) {
     const modalImage = gradioApp().getElementById("modalImage");
     const modal = gradioApp().getElementById("lightboxModal");
@@ -220,6 +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 = "🗆";
+    modalToggleLivePreview.onclick = modalLivePreviewToggle;
+    modalToggleLivePreview.title = "Toggle live preview";
+    modalControls.appendChild(modalToggleLivePreview);
+
     const modalClose = document.createElement('span');
     modalClose.className = 'modalClose cursor';
     modalClose.innerHTML = '×';

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';
+function createVisualizationTable(data, cutoff = 0, sort = "") {
+    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];
+    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 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(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 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");
+            }
 
-                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);
-                }
-            });
-
-            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");
-                }
-
-                childrenRows.push(tr);
+        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';
             }
 
-            return childrenRows;
-        };
+            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");
+                });
+            };
 
-        addLevel(0, []);
+            cell.title = othersList.join(", ");
+            cell.classList.add("link");
+            cell.addEventListener("click", onclick);
 
+            if (hide) {
+                tr.classList.add("hidden");
+            }
+
+            childrenRows.push(tr);
+        }
+
+        return childrenRows;
+    };
+
+    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);
     });
 }

javascript/progressbar.js

@@ -76,6 +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';
@@ -89,6 +109,7 @@ function requestProgress(id_task, progressbarContainer, gallery, atEnd, onProgre
     var livePreview = null;
 
     var removeProgressBar = function() {
+        releaseWakeLock();
         if (!divProgress) return;
 
         setTitle("");
@@ -100,6 +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();

javascript/resizeHandle.js

@@ -22,6 +22,9 @@
     }
 
     function displayResizeHandle(parent) {
+        if (!parent.needHideOnMoblie) {
+            return true;
+        }
         if (window.innerWidth < GRADIO_MIN_WIDTH * 2 + PAD * 4) {
             parent.style.display = 'flex';
             parent.resizeHandle.style.display = "none";
@@ -41,7 +44,7 @@
 
             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;
@@ -64,7 +67,24 @@
 
         parent.style.display = 'grid';
         parent.style.gap = '0';
-        const gridTemplateColumns = `${parent.children[0].style.flexGrow}fr ${PAD}px ${parent.children[1].style.flexGrow}fr`;
+        let leftColTemplate = "";
+        if (parent.children[0].style.flexGrow) {
+            leftColTemplate = `${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;
 
@@ -132,7 +152,7 @@
                 } else {
                     delta = R.screenX - evt.changedTouches[0].screenX;
                 }
-                const leftColWidth = Math.max(Math.min(R.leftColStartWidth - delta, R.parent.offsetWidth - GRADIO_MIN_WIDTH - PAD), GRADIO_MIN_WIDTH);
+                const leftColWidth = Math.max(Math.min(R.leftColStartWidth - delta, R.parent.offsetWidth - R.parent.minRightColWidth - PAD), R.parent.minLeftColWidth);
                 setLeftColGridTemplate(R.parent, leftColWidth);
             }
         });
@@ -171,10 +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);
+

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];
@@ -136,8 +144,7 @@ function showSubmitInterruptingPlaceholder(tabname) {
 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() {
@@ -209,6 +216,7 @@ function restoreProgressTxt2img() {
     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);
@@ -223,6 +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);
@@ -298,6 +307,7 @@ onAfterUiUpdate(function() {
     var jsdata = textarea.value;
     opts = JSON.parse(jsdata);
 
+    executeCallbacks(optionsAvailableCallbacks); /*global optionsAvailableCallbacks*/
     executeCallbacks(optionsChangedCallbacks); /*global optionsChangedCallbacks*/
 
     Object.defineProperty(textarea, 'value', {
@@ -336,8 +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();
@@ -411,7 +421,7 @@ function switchWidthHeight(tabname) {
 
 var onEditTimers = {};
 
-// calls func after afterMs milliseconds has passed since the input elem has beed enited by user
+// 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];

modules/api/api.py

@@ -17,13 +17,13 @@ from fastapi.encoders import jsonable_encoder
 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, script_callbacks, infotext_utils, sd_models
+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.hypernetworks.hypernetwork import create_hypernetwork, train_hypernetwork
-from PIL import PngImagePlugin, Image
+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
@@ -43,7 +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
 
@@ -85,7 +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
@@ -93,7 +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
@@ -113,7 +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({
@@ -221,6 +221,7 @@ class Api:
         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/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])
@@ -360,7 +361,7 @@ class Api:
         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` accoring to what's in infotext.
+        """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.
 
@@ -371,7 +372,7 @@ class Api:
             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 differenrt names for this
+        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):
@@ -409,8 +410,8 @@ class Api:
         if request.override_settings is None:
             request.override_settings = {}
 
-        overriden_settings = infotext_utils.get_override_settings(params)
-        for _, setting_name, value in overriden_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
 
@@ -437,15 +438,19 @@ class Api:
         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
@@ -501,9 +506,10 @@ class Api:
         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,
@@ -511,6 +517,9 @@ class Api:
         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)
@@ -683,6 +692,17 @@ class Api:
     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 [
             {

modules/api/models.py

@@ -147,7 +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?")
@@ -235,6 +235,13 @@ class SamplerItem(BaseModel):
     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")
     model_name: Optional[str] = Field(title="Model Name")

modules/cache.py

@@ -2,48 +2,55 @@ import json
 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, ensure_ascii=False)
 
-            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):
@@ -54,28 +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:
-                try:
-                    with open(cache_filename, "r", encoding="utf8") as file:
-                        cache_data = json.load(file)
-                except FileNotFoundError:
-                    cache_data = {}
-                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 = {}
+            if not os.path.exists(cache_dir) and os.path.isfile(cache_filename):
+                convert_old_cached_data()
 
-    s = cache_data.get(subsection, {})
-    cache_data[subsection] = s
+            cache_obj = caches.get(subsection)
+            if not cache_obj:
+                cache_obj = make_cache(subsection)
+                caches[subsection] = cache_obj
 
-    return s
+    return cache_obj
 
 
 def cached_data_for_file(subsection, title, filename, func):

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 wrap_gradio_gpu_call(func, extra_outputs=None):
 
 
 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 wrap_gradio_call(func, extra_outputs=None, add_stats=False):
                 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, '']
 
@@ -76,11 +90,6 @@ def wrap_gradio_call(func, extra_outputs=None, add_stats=False):
 
         devices.torch_gc()
 
-        shared.state.skipped = False
-        shared.state.interrupted = False
-        shared.state.stopping_generation = False
-        shared.state.job_count = 0
-
         if not add_stats:
             return tuple(res)
 
@@ -100,8 +109,8 @@ def wrap_gradio_call(func, extra_outputs=None, add_stats=False):
             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>"
@@ -111,9 +120,15 @@ def wrap_gradio_call(func, extra_outputs=None, add_stats=False):
         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
+

modules/cmd_args.py

@@ -20,6 +20,7 @@ parser.add_argument("--dump-sysinfo", action='store_true', help="launch.py argum
 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=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")
@@ -29,7 +30,7 @@ parser.add_argument("--gfpgan-model", type=normalized_filepath, help="GFPGAN mod
 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("--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")
@@ -41,7 +42,7 @@ parser.add_argument("--lowvram", action='store_true', help="enable stable diffus
 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)
@@ -121,4 +122,7 @@ parser.add_argument('--api-server-stop', action='store_true', help='enable serve
 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("--skip-load-model-at-start", action='store_true', help="if load a model at web start, only take effect when --nowebui", )
+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")

modules/codeformer_model.py

@@ -50,7 +50,7 @@ class FaceRestorerCodeFormer(face_restoration_utils.CommonFaceRestoration):
 
         def restore_face(cropped_face_t):
             assert self.net is not None
-            return self.net(cropped_face_t, w=w, adain=True)[0]
+            return self.net(cropped_face_t, weight=w, adain=True)[0]
 
         return self.restore_with_helper(np_image, restore_face)
 

modules/deepbooru.py

@@ -57,7 +57,7 @@ class DeepDanbooru:
         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 = {}

modules/devices.py

@@ -114,6 +114,9 @@ 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
@@ -127,6 +130,8 @@ 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
 
 
@@ -206,6 +211,11 @@ def autocast(disable=False):
     if disable:
         return contextlib.nullcontext()
 
+    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)
 
@@ -233,22 +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."
 
@@ -258,8 +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)
@@ -269,3 +279,17 @@ def first_time_calculation():
     x = torch.zeros((1, 1, 3, 3)).to(device, dtype)
     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`.")

modules/extensions.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import configparser
+import dataclasses
 import os
 import threading
 import re
@@ -9,6 +10,10 @@ 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: list[Extension] = []
+extension_paths: dict[str, Extension] = {}
+loaded_extensions: dict[str, Exception] = {}
+
 
 os.makedirs(extensions_dir, exist_ok=True)
 
@@ -22,6 +27,13 @@ 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
@@ -42,7 +54,7 @@ class ExtensionMetadata:
         self.canonical_name = self.config.get("Extension", "Name", fallback=canonical_name)
         self.canonical_name = canonical_name.lower().strip()
 
-        self.requires = self.get_script_requirements("Requires", "Extension")
+        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,
@@ -54,7 +66,15 @@ class ExtensionMetadata:
         if extra_section:
             x = x + ', ' + self.config.get(extra_section, field, fallback='')
 
-        return self.parse_list(x.lower())
+        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"])"""
@@ -65,6 +85,22 @@ class ExtensionMetadata:
         # 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()
@@ -155,14 +191,17 @@ class 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"
@@ -175,8 +214,10 @@ class Extension:
         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)
@@ -186,6 +227,8 @@ class Extension:
 
 def list_extensions():
     extensions.clear()
+    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 ***")
@@ -196,7 +239,6 @@ def list_extensions():
     elif shared.opts.disable_all_extensions == "extra":
         print("*** \"Disable all extensions\" option was set, will only load built-in extensions ***")
 
-    loaded_extensions = {}
 
     # scan through extensions directory and load metadata
     for dirname in [extensions_builtin_dir, extensions_dir]:
@@ -220,8 +262,12 @@ def list_extensions():
             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:
@@ -238,4 +284,16 @@ def list_extensions():
                 continue
 
 
-extensions: list[Extension] = []
+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
+

modules/extra_networks.py

@@ -60,7 +60,7 @@ class ExtraNetwork:
         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.

modules/gfpgan_model.py

@@ -36,13 +36,11 @@ class FaceRestorerGFPGAN(face_restoration_utils.CommonFaceRestoration):
             ext_filter=['.pth'],
         ):
             if 'GFPGAN' in os.path.basename(model_path):
-                model = modelloader.load_spandrel_model(
+                return modelloader.load_spandrel_model(
                     model_path,
                     device=self.get_device(),
                     expected_architecture='GFPGAN',
                 ).model
-                model.different_w = True  # see https://github.com/chaiNNer-org/spandrel/pull/81
-                return model
         raise ValueError("No GFPGAN model found")
 
     def restore(self, np_image):

modules/hypernetworks/hypernetwork.py

@@ -11,7 +11,7 @@ import tqdm
 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 @@ class HypernetworkModule(torch.nn.Module):
                         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):
@@ -532,7 +533,7 @@ def train_hypernetwork(id_task, hypernetwork_name: str, learn_rate: float, batch
             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
 

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 re
 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
@@ -52,11 +54,14 @@ 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
 
@@ -321,13 +326,16 @@ def resize_image(resize_mode, im, width, height, upscaler_name=None):
     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()
 
 
@@ -344,8 +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],
@@ -355,6 +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'),
@@ -380,12 +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."""
@@ -566,6 +604,17 @@ 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:
@@ -605,12 +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)
@@ -744,10 +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
-        geninfo = items["comment"].decode('utf8', errors="ignore")
+        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)
@@ -770,7 +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:
@@ -797,3 +848,30 @@ def flatten(img, bgcolor):
 
     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

modules/img2img.py

@@ -6,7 +6,7 @@ import numpy as np
 from PIL import Image, ImageOps, ImageFilter, ImageEnhance, UnidentifiedImageError
 import gradio as gr
 
-from modules import images as imgutil
+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
@@ -17,11 +17,14 @@ 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:
@@ -31,9 +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
@@ -46,8 +49,8 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=Fal
     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(images):
-        state.job = f"{i+1} out of {len(images)}"
+    for i, image in enumerate(batch_images):
+        state.job = f"{i+1} out of {len(batch_images)}"
         if state.skipped:
             state.skipped = False
 
@@ -55,7 +58,7 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=Fal
             break
 
         try:
-            img = Image.open(image)
+            img = images.read(image)
         except UnidentifiedImageError as e:
             print(e)
             continue
@@ -86,7 +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:
@@ -94,8 +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:
@@ -146,7 +149,7 @@ def process_batch(p, input_dir, output_dir, inpaint_mask_dir, args, to_scale=Fal
     return batch_results
 
 
-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):
+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
@@ -175,9 +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"
@@ -194,10 +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,
@@ -224,8 +224,15 @@ def img2img(id_task: str, mode: int, prompt: str, negative_prompt: str, prompt_s
 
     with closing(p):
         if is_batch:
-            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)
+            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)
 
             if processed is None:
                 processed = Processed(p, [], p.seed, "")

modules/infotext_utils.py

@@ -8,7 +8,7 @@ import sys
 
 import gradio as gr
 from modules.paths import data_path
-from modules import shared, ui_tempdir, script_callbacks, processing, infotext_versions
+from modules import shared, ui_tempdir, script_callbacks, processing, infotext_versions, images, prompt_parser, errors
 from PIL import Image
 
 sys.modules['modules.generation_parameters_copypaste'] = sys.modules[__name__]  # alias for old name
@@ -83,7 +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:
@@ -95,7 +95,7 @@ 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
 
 
@@ -146,18 +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,
             )
 
@@ -265,17 +266,6 @@ Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 965400086, Size: 512x512, Model
         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] == '"':
@@ -290,6 +280,26 @@ Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 965400086, Size: 512x512, Model
         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"
@@ -305,6 +315,9 @@ Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 965400086, Size: 512x512, Model
     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"
 
@@ -356,9 +369,15 @@ Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 965400086, Size: 512x512, Model
     if "Cache FP16 weight for LoRA" not in res and res["FP8 weight"] != "Disable":
         res["Cache FP16 weight for LoRA"] = False
 
-    if "Emphasis" not in res:
+    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:
@@ -456,7 +475,7 @@ def get_override_settings(params, *, skip_fields=None):
 
 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")
             try:
                 with open(filename, "r", encoding="utf8") as file:
@@ -470,7 +489,11 @@ def connect_paste(button, paste_fields, input_comp, override_settings_component,
 
         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)
 

modules/infotext_versions.py

@@ -5,6 +5,8 @@ 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):
@@ -40,3 +42,5 @@ def backcompat(d):
     if ver < v170_tsnr:
         d["Downcast alphas_cumprod"] = True
 
+    if ver < v180 and d.get('Refiner'):
+        d["Refiner switch by sampling steps"] = True

modules/initialize.py

@@ -51,6 +51,7 @@ 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()
@@ -109,7 +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")
@@ -139,7 +140,7 @@ def initialize_rest(*, reload_script_modules=False):
         """
         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

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():
     """

modules/launch_utils.py

@@ -9,6 +9,7 @@ import importlib.util
 import importlib.metadata
 import platform
 import json
+import shlex
 from functools import lru_cache
 
 from modules import cmd_args, errors
@@ -55,7 +56,7 @@ and delete current Python and "venv" folder in WebUI's directory.
 
 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.
 """)
@@ -76,7 +77,7 @@ def git_tag():
     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("## ", "")
@@ -231,7 +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:
@@ -445,7 +446,6 @@ def prepare_environment():
         exit(0)
 
 
-
 def configure_for_tests():
     if "--api" not in sys.argv:
         sys.argv.append("--api")
@@ -461,7 +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()

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 setup_for_low_vram(sd_model, use_medvram):
         (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 setup_for_low_vram(sd_model, use_medvram):
         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 setup_for_low_vram(sd_model, use_medvram):
     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:

modules/mac_specific.py

@@ -12,7 +12,7 @@ log = logging.getLogger(__name__)
 
 # 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"):

modules/masking.py

@@ -1,17 +1,39 @@
 from PIL import Image, ImageFilter, ImageOps
 
 
-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)"""
-    mask_img = mask if isinstance(mask, Image.Image) else Image.fromarray(mask)
-    box = mask_img.getbbox()
-    if box:
+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
-    else:  # when no box is found
-        x1, y1 = mask_img.size
-        x2 = y2 = 0
-    return max(x1 - pad, 0), max(y1 - pad, 0), min(x2 + pad, mask_img.size[0]), min(y2 + pad, mask_img.size[1])
+        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):
+    """
+    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):

modules/modelloader.py

@@ -23,6 +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.
 
@@ -36,11 +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.
 
@@ -49,6 +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 = []
@@ -78,7 +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)
 
@@ -110,7 +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
@@ -129,14 +131,35 @@ 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,
@@ -146,11 +169,16 @@ def load_spandrel_model(
     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))
-    if expected_architecture and model_descriptor.architecture != expected_architecture:
+    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 {model_descriptor.architecture!r})",
+            f"Model {path!r} is not a {expected_architecture!r} model (got {arch.name!r})",
         )
     half = False
     if prefer_half:
@@ -164,6 +192,6 @@ def load_spandrel_model(
     model_descriptor.model.eval()
     logger.debug(
         "Loaded %s from %s (device=%s, half=%s, dtype=%s)",
-        model_descriptor, path, device, half, dtype,
+        arch, path, device, half, dtype,
     )
     return model_descriptor

modules/models/diffusion/ddpm_edit.py

@@ -341,7 +341,7 @@ class DDPM(pl.LightningModule):
         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])
 
@@ -901,7 +901,7 @@ class LatentDiffusion(DDPM):
     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):
@@ -937,7 +937,7 @@ class LatentDiffusion(DDPM):
                 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)
@@ -947,7 +947,7 @@ class LatentDiffusion(DDPM):
                 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)

modules/models/diffusion/uni_pc/uni_pc.py

@@ -323,7 +323,7 @@ def model_wrapper(
 
     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]))

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

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)

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)

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)

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"

modules/options.py

@@ -240,6 +240,9 @@ class Options:
 
         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:

modules/paths_internal.py

@@ -24,14 +24,15 @@ default_sd_model_file = sd_model_file
 # 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, "output")
+default_output_dir = os.path.join(data_path, "outputs")
 
 roboto_ttf_file = os.path.join(modules_path, 'Roboto-Regular.ttf')

modules/postprocessing.py

@@ -17,10 +17,10 @@ def run_postprocessing(extras_mode, image, image_folder, input_dir, output_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:
@@ -51,22 +51,24 @@ def run_postprocessing(extras_mode, image, image_folder, input_dir, output_dir,
         shared.state.textinfo = name
         shared.state.skipped = False
 
-        if shared.state.interrupted:
+        if shared.state.interrupted or shared.state.stopping_generation:
             break
 
         if isinstance(image_placeholder, str):
             try:
-                image_data = Image.open(image_placeholder)
+                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
 
-        initial_pp = scripts_postprocessing.PostprocessedImage(image_data.convert("RGB"))
+        initial_pp = scripts_postprocessing.PostprocessedImage(image_data)
 
         scripts.scripts_postproc.run(initial_pp, args)
 
@@ -122,8 +124,6 @@ def run_postprocessing(extras_mode, image, image_folder, input_dir, output_dir,
             if extras_mode != 2 or show_extras_results:
                 outputs.append(pp.image)
 
-        image_data.close()
-
     devices.torch_gc()
     shared.state.end()
     return outputs, ui_common.plaintext_to_html(infotext), ''
@@ -133,13 +133,15 @@ 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):
+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,

modules/processing.py

@@ -16,7 +16,7 @@ from skimage import exposure
 from typing import Any
 
 import modules.sd_hijack
-from modules import devices, prompt_parser, masking, sd_samplers, lowvram, infotext_utils, extra_networks, sd_vae_approx, scripts, sd_samplers_common, sd_unet, errors, rng
+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
@@ -115,20 +115,17 @@ 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:
-        sd = sd_model.model.state_dict()
-        diffusion_model_input = sd.get('diffusion_model.input_blocks.0.0.weight', None)
-        if diffusion_model_input is not None:
-            if diffusion_model_input.shape[1] == 9:
-                # 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))
+        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)
+            # 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
+            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.
@@ -152,6 +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
@@ -237,11 +235,6 @@ class StableDiffusionProcessing:
             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 {}
@@ -258,6 +251,13 @@ class StableDiffusionProcessing:
         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
@@ -389,11 +389,8 @@ class StableDiffusionProcessing:
         if self.sampler.conditioning_key == "crossattn-adm":
             return self.unclip_image_conditioning(source_image)
 
-        sd = self.sampler.model_wrap.inner_model.model.state_dict()
-        diffusion_model_input = sd.get('diffusion_model.input_blocks.0.0.weight', None)
-        if diffusion_model_input is not None:
-            if diffusion_model_input.shape[1] == 9:
-                return self.inpainting_image_conditioning(source_image, latent_image, image_mask=image_mask)
+        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)
@@ -568,7 +565,7 @@ class Processed:
         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):
@@ -607,7 +604,7 @@ class Processed:
             "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)
@@ -628,6 +625,9 @@ 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]
 
@@ -703,7 +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:
@@ -714,6 +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)
@@ -721,6 +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],
@@ -743,17 +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,
     }
 
+    for key, value in generation_params.items():
+        try:
+            if isinstance(value, list):
+                generation_params[key] = value[index]
+            elif callable(value):
+                generation_params[key] = value(**locals())
+        except Exception:
+            errors.report(f'Error creating infotext for key "{key}"', exc_info=True)
+            generation_params[key] = None
+
     generation_params_text = ", ".join([k if k == v else f'{k}: {infotext_utils.quote(v)}' for k, v in generation_params.items() if v is not None])
 
-    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 ""
+    negative_prompt_text = f"\nNegative prompt: {negative_prompt}" if negative_prompt else ""
 
     return f"{prompt_text}{negative_prompt_text}\n{generation_params_text}".strip()
 
@@ -782,7 +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)
@@ -822,6 +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()
@@ -830,6 +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):
@@ -879,7 +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)
@@ -896,22 +963,22 @@ 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}"
 
@@ -928,6 +995,8 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
             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)
@@ -1106,6 +1175,7 @@ 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
@@ -1194,11 +1264,21 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing):
             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
 
-            if tuple(self.hr_negative_prompt) != tuple(self.negative_prompt):
-                self.extra_generation_params["Hires negative prompt"] = self.hr_negative_prompt
+            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
+
+            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:
@@ -1254,6 +1334,15 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing):
             # 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
 
@@ -1354,6 +1443,13 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing):
 
         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)
 
@@ -1540,16 +1636,23 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing):
             if self.inpaint_full_res:
                 self.mask_for_overlay = image_mask
                 mask = image_mask.convert('L')
-                crop_region = masking.get_crop_region(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)
-
-                self.extra_generation_params["Inpaint area"] = "Only masked"
-                self.extra_generation_params["Masked area padding"] = self.inpaint_full_res_padding
+                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)
@@ -1577,6 +1680,8 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing):
                 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')))
 
@@ -1635,10 +1740,10 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing):
             latmask = latmask[0]
             if self.mask_round:
                 latmask = np.around(latmask)
-            latmask = np.tile(latmask[None], (4, 1, 1))
+            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:
@@ -1658,6 +1763,14 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing):
             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:

modules/processing_scripts/comments.py

@@ -26,6 +26,13 @@ class ScriptStripComments(scripts.Script):
         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:

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

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)
+

modules/prompt_parser.py

@@ -268,7 +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)
 

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."""
 

modules/safe.py

@@ -64,8 +64,8 @@ class RestrictedUnpickler(pickle.Unpickler):
         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):

modules/script_callbacks.py

@@ -1,13 +1,14 @@
+from __future__ import annotations
+
 import dataclasses
 import inspect
 import os
-from collections import namedtuple
 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):
@@ -116,7 +117,105 @@ class BeforeTokenCounterParams:
     is_positive: bool = True
 
 
-ScriptCallback = namedtuple("ScriptCallback", ["script", "callback"])
+@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=[],
@@ -141,14 +240,18 @@ callback_map = dict(
     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))
@@ -157,7 +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:
@@ -165,7 +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:
@@ -175,7 +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:
@@ -185,7 +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:
@@ -193,7 +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:
@@ -201,7 +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:
@@ -209,7 +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:
@@ -217,7 +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:
@@ -225,7 +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:
@@ -233,7 +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:
@@ -241,7 +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:
@@ -249,7 +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:
@@ -257,7 +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:
@@ -265,7 +368,7 @@ 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:
@@ -273,7 +376,7 @@ def image_grid_callback(params: ImageGridLoopParams):
 
 
 def infotext_pasted_callback(infotext: str, params: dict[str, Any]):
-    for c in callback_map['callbacks_infotext_pasted']:
+    for c in ordered_callbacks('infotext_pasted'):
         try:
             c.callback(infotext, params)
         except Exception:
@@ -281,7 +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:
@@ -289,7 +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:
@@ -299,7 +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:
@@ -311,7 +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:
@@ -321,20 +424,13 @@ def list_unets_callback():
 
 
 def before_token_counter_callback(params: BeforeTokenCounterParams):
-    for c in callback_map['callbacks_before_token_counter']:
+    for c in ordered_callbacks('before_token_counter'):
         try:
             c.callback(params)
         except Exception:
             report_exception(c, 'before_token_counter')
 
 
-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 remove_current_script_callbacks():
     stack = [x for x in inspect.stack() if x.filename != __file__]
     filename = stack[0].filename if stack else 'unknown file'
@@ -343,32 +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:
@@ -378,71 +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.
@@ -451,61 +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.
     """
-    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):
+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)
+    add_callback(callback_map['callbacks_before_token_counter'], callback, name=name, category='before_token_counter')

modules/script_loading.py

@@ -4,11 +4,15 @@ import importlib.util
 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
 
 

modules/scripts.py

@@ -7,7 +7,9 @@ from dataclasses import dataclass
 
 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()
 
@@ -92,7 +94,7 @@ class Script:
     """If true, the script setup will only be run in Gradio UI, not in API"""
 
     controls = None
-    """A list of controls retured by the ui()."""
+    """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."""
@@ -109,7 +111,7 @@ class Script:
 
     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
@@ -138,7 +140,6 @@ class Script:
         """
         pass
 
-
     def before_process(self, p, *args):
         """
         This function is called very early during processing begins for AlwaysVisible scripts.
@@ -186,6 +187,13 @@ class Script:
         """
         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.
@@ -351,6 +359,9 @@ class ScriptBuiltinUI(Script):
 
         return f'{tabname}{item_id}'
 
+    def show(self, is_img2img):
+        return AlwaysVisible
+
 
 current_basedir = paths.script_path
 
@@ -369,29 +380,6 @@ scripts_data = []
 postprocessing_scripts_data = []
 ScriptClassData = namedtuple("ScriptClassData", ["script_class", "path", "basedir", "module"])
 
-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 dependeencies 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
-
 
 @dataclass
 class ScriptWithDependencies:
@@ -562,6 +550,25 @@ class ScriptRunner:
         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"""
 
@@ -600,6 +607,8 @@ class ScriptRunner:
                 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):
@@ -737,12 +746,17 @@ class ScriptRunner:
         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])
@@ -769,8 +783,42 @@ class ScriptRunner:
 
         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)
@@ -778,15 +826,23 @@ class ScriptRunner:
                 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)
@@ -794,7 +850,7 @@ class ScriptRunner:
                 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)
@@ -802,7 +858,7 @@ class ScriptRunner:
                 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)
@@ -810,7 +866,7 @@ class ScriptRunner:
                 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)
@@ -818,7 +874,7 @@ class ScriptRunner:
                 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)
@@ -826,7 +882,7 @@ class ScriptRunner:
                 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)
@@ -834,7 +890,7 @@ class ScriptRunner:
                 errors.report(f"Error running postprocess_batch_list: {script.filename}", exc_info=True)
 
     def post_sample(self, p, ps: PostSampleArgs):
-        for script in self.alwayson_scripts:
+        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)
@@ -842,7 +898,7 @@ class ScriptRunner:
                 errors.report(f"Error running post_sample: {script.filename}", exc_info=True)
 
     def on_mask_blend(self, p, mba: MaskBlendArgs):
-        for script in self.alwayson_scripts:
+        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)
@@ -850,7 +906,7 @@ class ScriptRunner:
                 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)
@@ -858,7 +914,7 @@ class ScriptRunner:
                 errors.report(f"Error running postprocess_image: {script.filename}", exc_info=True)
 
     def postprocess_maskoverlay(self, p, ppmo: PostProcessMaskOverlayArgs):
-        for script in self.alwayson_scripts:
+        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)
@@ -866,7 +922,7 @@ class ScriptRunner:
                 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.alwayson_scripts:
+        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)
@@ -880,7 +936,7 @@ class ScriptRunner:
             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:
@@ -893,7 +949,7 @@ class ScriptRunner:
             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:
@@ -921,7 +977,7 @@ class ScriptRunner:
                     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)
@@ -929,7 +985,7 @@ class ScriptRunner:
                 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
 

modules/scripts_postprocessing.py

@@ -143,6 +143,7 @@ class ScriptPostprocessingRunner:
             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):
@@ -151,9 +152,10 @@ class ScriptPostprocessingRunner:
 
             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 = []

modules/sd_emphasis.py

@@ -35,7 +35,7 @@ class EmphasisIgnore(Emphasis):
 
 class EmphasisOriginal(Emphasis):
     name = "Original"
-    description = "the orginal emphasis implementation"
+    description = "the original emphasis implementation"
 
     def after_transformers(self):
         original_mean = self.z.mean()
@@ -48,7 +48,7 @@ class EmphasisOriginal(Emphasis):
 
 class EmphasisOriginalNoNorm(EmphasisOriginal):
     name = "No norm"
-    description = "same as orginal, but without normalization (seems to work better for SDXL)"
+    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)

modules/sd_hijack.py

@@ -325,7 +325,10 @@ class StableDiffusionModelHijack:
         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)
 
@@ -356,13 +359,28 @@ class EmbeddingsWithFixes(torch.nn.Module):
                 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__
 

modules/sd_hijack_clip.py

@@ -23,28 +23,25 @@ class PromptChunk:
 
 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 @@ class FrozenCLIPEmbedderWithCustomWordsBase(torch.nn.Module):
 
     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.
@@ -136,7 +133,7 @@ class FrozenCLIPEmbedderWithCustomWordsBase(torch.nn.Module):
                 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 @@ class FrozenCLIPEmbedderWithCustomWordsBase(torch.nn.Module):
         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 @@ class FrozenCLIPEmbedderWithCustomWordsBase(torch.nn.Module):
             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)
 
@@ -252,7 +245,7 @@ class FrozenCLIPEmbedderWithCustomWordsBase(torch.nn.Module):
         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 getattr(self.wrapped, 'return_pooled', False):
+        if self.return_pooled:
             return torch.hstack(zs), zs[0].pooled
         else:
             return torch.hstack(zs)
@@ -292,6 +285,34 @@ class FrozenCLIPEmbedderWithCustomWordsBase(torch.nn.Module):
         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)
@@ -353,7 +374,9 @@ class FrozenCLIPEmbedderForSDXLWithCustomWords(FrozenCLIPEmbedderWithCustomWords
     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]

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)