1.6.0

[AUTOMATIC1111]

Features:

• refiner support #12371
• add NV option for Random number generator source setting, which allows to generate same pictures on CPU/AMD/Mac as on NVidia videocards
• add style editor dialog
• hires fix: add an option to use a different checkpoint for second pass (#12181)
• option to keep multiple loaded models in memory (#12227)
• 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, #12519, #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 them available for SDXL
• always show extra networks tabs in the UI (#11808)
• use less RAM when creating models (#11958, #12599)
• textual inversion inference support for SDXL
• extra networks UI: show metadata for SD checkpoints
• checkpoint merger: add metadata support
• prompt editing and attention: add support for whitespace after the number ([ red : green : 0.5 ]) (seed breaking change) (#12177)
• VAE: allow selecting own VAE for each checkpoint (in user metadata editor)
• VAE: add selected VAE to infotext
• options in main UI: add own separate setting for txt2img and img2img, correctly read values from pasted infotext, add setting for column count (#12551)
• add resize handle to txt2img and img2img tabs, allowing to change the amount of horizontable space given to generation parameters and resulting image gallery (#12687, #12723)
• change default behavior for batching cond/uncond -- now it's on by default, and is disabled by an UI setting (Optimizatios -> Batch cond/uncond) - if you are on lowvram/medvram and are getting OOM exceptions, you will need to enable it
• show current position in queue and make it so that requests are processed in the order of arrival (#12707)
• add --medvram-sdxl flag that only enables --medvram for SDXL models
• prompt editing timeline has separate range for first pass and hires-fix pass (seed breaking change) (#12457)

Minor:

• img2img batch: RAM savings, VRAM savings, .tif, .tiff in img2img batch (#12120, #12514, #12515)
• postprocessing/extras: RAM savings (#12479)
• XYZ: in the axis labels, remove pathnames from model filenames
• XYZ: support hires sampler (#12298)
• XYZ: new option: use text inputs instead of dropdowns (#12491)
• add gradio version warning
• sort list of VAE checkpoints (#12297)
• use transparent white for mask in inpainting, along with an option to select the color (#12326)
• move some settings to their own section: img2img, VAE
• add checkbox to show/hide dirs for extra networks
• Add TAESD(or more) options for all the VAE encode/decode operation (#12311)
• gradio theme cache, new gradio themes, along with explanation that the user can input his own values (#12346, #12355)
• sampler fixes/tweaks: s_tmax, s_churn, s_noise, s_tmax (#12354, #12356, #12357, #12358, #12375, #12521)
• update README.md with correct instructions for Linux installation (#12352)
• option to not save incomplete images, on by default (#12338)
• enable cond cache by default
• git autofix for repos that are corrupted (#12230)
• allow to open images in new browser tab by middle mouse button (#12379)
• automatically open webui in browser when running "locally" (#12254)
• put commonly used samplers on top, make DPM++ 2M Karras the default choice
• zoom and pan: option to auto-expand a wide image, improved integration (#12413, #12727)
• option to cache Lora networks in memory
• rework hires fix UI to use accordion
• face restoration and tiling moved to settings - use "Options in main UI" setting if you want them back
• change quicksettings items to have variable width
• Lora: add Norm module, add support for bias (#12503)
• Lora: output warnings in UI rather than fail for unfitting loras; switch to logging for error output in console
• support search and display of hashes for all extra network items (#12510)
• add extra noise param for img2img operations (#12564)
• support for Lora with bias (#12584)
• make interrupt quicker (#12634)
• configurable gallery height (#12648)
• make results column sticky (#12645)
• more hash filename patterns (#12639)
• make image viewer actually fit the whole page (#12635)
• make progress bar work independently from live preview display which results in it being updated a lot more often
• forbid Full live preview method for medvram and add a setting to undo the forbidding
• make it possible to localize tooltips and placeholders
• add option to align with sgm repo's sampling implementation (#12818)
• Restore faces and Tiling generation parameters have been moved to settings out of main UI
  • if you want to put them back into main UI, use Options in main UI setting on the UI page.

Extensions and API:

• gradio 3.41.2
• also bump versions for packages: transformers, GitPython, accelerate, scikit-image, timm, tomesd
• support tooltip kwarg for gradio elements: gr.Textbox(label='hello', tooltip='world')
• properly clear the total console progressbar when using txt2img and img2img from API
• add cmd_arg --disable-extra-extensions and --disable-all-extensions (#12294)
• shared.py and webui.py split into many files
• add --loglevel commandline argument for logging
• add a custom UI element that combines accordion and checkbox
• avoid importing gradio in tests because it spams warnings
• put infotext label for setting into OptionInfo definition rather than in a separate list
• make StableDiffusionProcessingImg2Img.mask_blur a property, make more inline with PIL GaussianBlur (#12470)
• option to make scripts UI without gr.Group
• add a way for scripts to register a callback for before/after just a single component's creation
• use dataclass for StableDiffusionProcessing
• store patches for Lora in a specialized module instead of inside torch
• support http/https URLs in API (#12663, #12698)
• add extra noise callback (#12616)
• dump current stack traces when exiting with SIGINT
• add type annotations for extra fields of shared.sd_model

Bug Fixes:

• Don't crash if out of local storage quota for javascriot localStorage
• XYZ plot do not fail if an exception occurs
• fix missing TI hash in infotext if generation uses both negative and positive TI (#12269)
• localization fixes (#12307)
• fix sdxl model invalid configuration after the hijack
• correctly toggle extras checkbox for infotext paste (#12304)
• open raw sysinfo link in new page (#12318)
• prompt parser: Account for empty field in alternating words syntax (#12319)
• add tab and carriage return to invalid filename chars (#12327)
• fix api only Lora not working (#12387)
• fix options in main UI misbehaving when there's just one element
• make it possible to use a sampler from infotext even if it's hidden in the dropdown
• fix styles missing from the prompt in infotext when making a grid of batch of multiplie images
• prevent bogus progress output in console when calculating hires fix dimensions
• fix --use-textbox-seed
• fix broken Lora/Networks: use old method option (#12466)
• properly return None for VAE hash when using --no-hashing (#12463)
• MPS/macOS fixes and optimizations (#12526)
• add second_order to samplers that mistakenly didn't have it
• when refreshing cards in extra networks UI, do not discard user's custom resolution
• fix processing error that happens if batch_size is not a multiple of how many prompts/negative prompts there are (#12509)
• fix inpaint upload for alpha masks (#12588)
• fix exception when image sizes are not integers (#12586)
• fix incorrect TAESD Latent scale (#12596)
• auto add data-dir to gradio-allowed-path (#12603)
• fix exception if extensuions dir is missing (#12607)
• fix issues with api model-refresh and vae-refresh (#12638)
• fix img2img background color for transparent images option not being used (#12633)
• attempt to resolve NaN issue with unstable VAEs in fp32 mk2 (#12630)
• implement missing undo hijack for SDXL
• fix xyz swap axes (#12684)
• fix errors in backup/restore tab if any of config files are broken (#12689)
• fix SD VAE switch error after model reuse (#12685)
• fix trying to create images too large for the chosen format (#12667)
• create Gradio temp directory if necessary (#12717)
• prevent possible cache loss if exiting as it's being written by using an atomic operation to replace the cache with the new version
• set devices.dtype_unet correctly
• run RealESRGAN on GPU for non-CUDA devices (#12737)
• prevent extra network buttons being obscured by description for very small card sizes (#12745)
• fix error that causes some extra networks to be disabled if both lora: and lyco: are present in the prompt
• fix defaults settings page breaking when any of main UI tabs are hidden
• fix incorrect save/display of new values in Defaults page in settings
• fix for Reload UI function: if you reload UI on one tab, other opened tabs will no longer stop working
• fix an error that prevents VAE being reloaded after an option change if a VAE near the checkpoint exists (#12797)
• hide broken image crop tool (#12792)
• don't show hidden samplers in dropdown for XYZ script (#12780)
• fix style editing dialog breaking if it's opened in both img2img and txt2img tabs
• fix a bug allowing users to bypass gradio and API authentication (reported by vysecurity)
• fix notification not playing when built-in webui tab is inactive (#12834)
• honor --skip-install for extension installers (#12832)
• don't print blank stdout in extension installers (#12833, #12855)
• do not change quicksettings dropdown option when value returned is None (#12854)
• get progressbar to display correctly in extensions tab

---

This discussion was created from the release 1.6.0.

[ArxFusion]

Thanks for the update, there are some issues with dropdown menus when using it on a mobile device. For instance if you open a drop down and then scroll the page, the normal action would be for the menu to close, instead it remains on the screen until you tap away. The keyboard also stays open until manually closed. Various text overlap or are cropped on buttons. Tested and used on Samsung S23 mobile.

[hainguyen666]

Why don't the extensions appear when I update to 1.6 even though I have installed and used them in the old version? Those are the two extensions "mov2mov" and "IS NET pro". What is happening?

[roots-3d]

Have you tried deleting the extensions folders and reinstalling again?

[realenuf]

I have the same problem. I tried deleting and reinstalling, and it didn't help.

[thundercat71]

Is there a way to add the checkboxs at the top of Txt2Img, Restore Faces and Tiling. (especially tiling! as I use this to make patterns.)

I know you can add Restore Faces in Setting - User Interface - Options in main UI

But the check box appears at the bottom of the window.

[AUTOMATIC1111]

There is another setting on that page to reorder rows in UI.

[thundercat71]

Yup having problems with getting that to work (order).
I found a check box in Settings - Stable Diffusion to turn Tiling on and off but thats a little convoluted ;)

[WhiteZeroX]

Whats the current recommendation on using xformers vs opt-sdp-attention? Any difference for nVidia Turing and up users?

[Jonseed]

I get slightly faster speeds on opt-sdp-no-mem than xformers on my Nvidia 3060. I think sdp is supposed to be generally a bit faster than xformers.

[owspace]

mac 更新后无法使用

[Jackloid1]

I’m not able to use roop extension now also facing issues with control net.

[thundercat71]

roop needs an update from them, what up with control net though as it seem to be working for me, have you updated it?

BTW roop the original code has been discontinued. See - https://github.com/s0md3v/roop

[dhwz]

roop is obsolete use faceswaplab

[naduarte2]

I'm using roop v0.0.2 and SD v1.6.0 without problems.

• Did you downloaded the ".onnx" required models for Roop, CodeFormer and GPFGAN?
• What extensions are you also using at the same time with Roop?
• Do your prompts include NSFW words? Remember that roop v0.0.2 detects NSFW and stop generating anything.

[Michoko92]

Thank you for this amazing tool, and to all who contributed! Your work is much appreciated. 🙏🙏

[Dasor92]

Is there a way to have the buttons for style saving in the main UI again? So we can add loras and embedding without copypasting.
Also, is there a way to have the "send to img2img" icons larger or "customizable"? Or rolling back to text only. I keep misclicking on save instead of send to img2img because they are too similar

[etrosh]

I have the same request, as UI designer I found icons (instead of text) unreadable/not intuitive and paste style button require multiple interactions not needed before.

[Natural-Warp]

I'd love to see the buttons for style saving back indeed, but meanwhile how can we save a prompt as a new style ?

[belenkoksas]

Wow, I am really happy with the speed of the latest version! The latest A1111 is better than ConfyUI! Thanks to all the guys, well done!

[scarab72]

I deleted webui and switched to invokeai.

[heartbreakergaming]

Is there a way to bring back the old extra networks button?

[belenkoksas]

Don't bring back, a new approach is way better!

[thijsi123]

Agreed, The old one is way better imo.

[scarab72]

I deleted webui and switched to invokeai.

[belenkoksas]

WOW! SDXL! SPEED 1500x1500 20S! RTX3080!

OMG THIS IS AMAZING! SPEED!

[Delios11]

Hi belenkoksas, i have a RTX 3080 Ti graphics, what configuration do you have in your settings for that speed, are you using a sdxl safetensor or ckpt checkpoint model?
In my case the speed is far from fast.
Thanks!

[bigmover]

Hi your reply is attractive. But don't know how to enable sdxl! Would you mind to explain ?

[Soulreaver90]

Allow us to move "Remove Faces" option back closer to the top of the UI, its annoying we have to add it and it appears at the very bottom of the page.
Also please add an option to remove the refiner tab, I don't use XL and can use less clutter.

EDIT: Was looking at the reorder UI, however hiresfix isn't listed as an option. Once I deleted it, I can't add it back.

[catboxanon]

accordions is the field where hires fix is in that reorder UI option.

[Soulreaver90]

I was able to fix it from the config.json file. It also dawned on me the extra_options UI list is the Restore Faces checkbox. Moved it above hires, even though on reboot it actually appears below hires. So after 15 minutes, I have it setup like it was in 1.5.
Think accordions should be renamed to hires/refiner as to avoid massive confusion. Or if we can have both hires and refiner separated, that would be fantastic (so I can remove it :D).

[dtteo]

Hi i cannot load the mov2mov tab anymore in the new update. Does anyone know how to fix it?

[MisterWolfe034]

I have updated my old SD to SD XL and something happens as it´s very slow now. I installed it with Afterburner limiting my GPU to 50% of use and that is exactly what is happening when I use the ¨generate¨ buttom: the gpu doesn´t go to max power at all.

I closed Afterburner and tried and I got the same result. Looks like when I updated SD, took the maximum power from Afterburner as teh real one and now it won´t use the full power at all. Is this even possible or am I making it up?

Any help will be much appreciated.

[MisterWolfe034]

Looks like it was a problem with drivers. If you install 531.61- should improve greatly.

[Lling00]

How do I quickly put a preset style into a cue word?
The new UI removes the button for pasting preset styles into prompts and puts it in Modify Preset Styles, which is not convenient to operate~.

[thundercat71]

Perhaps you should developed this in your own time for free. Time to stop the planet and let you off.

…

On Tue, 5 Sept 2023, 05:35 RobotRunner, ***@***.***> wrote: this version is ass, came with so many bugs, do u not properly test before going to production? — Reply to this email directly, view it on GitHub <#12878 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKJ64MQQQDFSKG6TFUHBVIDXY2TX3ANCNFSM6AAAAAA4FO5HM4> . You are receiving this because you commented.Message ID: <AUTOMATIC1111/stable-diffusion-webui/repo-discussions/12878/comments/6909825 @github.com>

[Delios11]

First of all, thanks to all those who work selflessly on the platform to provide a solid open source foundation for experimentation in image generation.
On the other hand it would be helpful if someone could clarify if there is a way to speed up render times.
Thank so much for your help.

[PunkRat9]

Thanks for your updates but this one broke everything on my PC. Loras gone, samplers list a mess, img2img not working, inpaint not working, SD upscale not working, tags and buttons have double text, etc. Until they fix all these horrible bugs, I suggest everyone to revert to 1.5.2

[Soulreaver90]

I reverted back to 1.5.2 and am happy again. 1.6.0 had too many changes, some causing issues for some users and some design decisions not sitting well with others. I think 1.6 should have been split into two major updates, with more transparency on the UI changes.

[dtteo]

i want to revert to 1.5.1. Do you know how to revert back to that version?

[PunkRat9]

git reset --hard 68f336bd994bed5442ad95bad6b6ad5564a5409a

1.5.1

[Soulreaver90]

I've also been able to use the following
git reset --hard v1.5.2
just change 1.5.2 to 1.5.1

[3Diva]

@PunkRat9 Do you happen to have the Agent Scheduler extension installed and enabled? After enabling that extension I started having some of the issues that you mentioned. After disabling it again and restarting AUTOMATIC1111, it seemed to fix the issue. If you have the Agent Scheduler installed and enabled I recommend disabling and restarting A1111 to see if that fixes the issue. :) Worth a shot!

[dimentox]

What so we replace create_seed_inputs with?

[RongWroom]

I can't seem to get the "Tiling" button to show up. I've activated in the settings tab and restarted etc etc but nothing. Any ideas? I use that all the time.

[RongWroom]

I've just realised, with the tiling option turned on in settings, every image I produce is now tiling. The button was perfect, can we have that back?

[dhwz]

You've to add it to main UI options and not enable it by default in settings like you did.
Then just switch in on in generation tab if needed

[RongWroom]

Ah ha! Thank you. 🙏🏽

[jefferysanders]

If it doesn't take, try again. The setting didn't want to be added to the top at first, but I deleted the previous and added both and it worked.

[frsk9999]

I use 1.5.2 and 1.6.0.
Of the samplers provided for both,the output of the SDE family is different between different versions.
Is there an option to make the same output (or similar)?

I enabled Follow options(Compatibility).

• Use old emphasis implementation. Can be useful to reproduce old seeds.
• Do not make DPM++ SDE deterministic across different batch sizes.

samples:
v1.5.2

v1.6.0

Model:AOM3A1B, vae:kl-f8-anime.vae.pt

[naitedj]

Hello, after update 1.6 I lost connection with Photoshop plugins (two were installed). I tried to write to the creator of the plugin, but did not receive a response. Plugin updates also did not solve this problem, nor did reinstalling SD. I'm not a programmer and don't know much about API, help me solve this problem, I really need these plugins.

[Miraculous14]

I updated the web to 1.5.2 without problems and added the -- xformers -- no-half-vae parameter. Use pytorch version 1.13.1+cu117 and xformers version 0.0.16. But when I updated the webui version to 1.6.0, there was an unlimited reboot. The error check was that my torch version was <2.0. Then I updated torch 2.0.1+cu117 and xformers to the latest version, which still restarted indefinitely. How can I fix this mistake? My system is (Ubuntu 20.04), python version 3.8.10.

[Miraculous14]

I fixed this issue, my Python version is 3.8, I updated to version 3.10 to fix this infinite reboot issue. It is now in normal use, SDXL is very good.

[nulflux]

Both updating the old 1.51 to 1.6 and installing 1.6 to a brand new folder provided me this result:

Launching Web UI with arguments: Traceback (most recent call last): File "launch.py", line 48, in <module> main() File "launch.py", line 44, in main start() File "/media/2TB/stable-diffusion-webui/modules/launch_utils.py", line 432, in start import webui File "/media/2TB/stable-diffusion-webui/webui.py", line 13, in <module> initialize.imports() File "/media/2TB/stable-diffusion-webui/modules/initialize.py", line 21, in imports import gradio # noqa: F401 File "/media/2TB/stable-diffusion-webui/venv/lib/python3.8/site-packages/gradio/__init__.py", line 3, in <module> import gradio.components as components File "/media/2TB/stable-diffusion-webui/venv/lib/python3.8/site-packages/gradio/components/__init__.py", line 1, in <module> from gradio.components.annotated_image import AnnotatedImage File "/media/2TB/stable-diffusion-webui/venv/lib/python3.8/site-packages/gradio/components/annotated_image.py", line 12, in <module> from gradio import utils File "/media/2TB/stable-diffusion-webui/venv/lib/python3.8/site-packages/gradio/utils.py", line 353, in <module> class AsyncRequest: File "/media/2TB/stable-diffusion-webui/venv/lib/python3.8/site-packages/gradio/utils.py", line 372, in AsyncRequest client = httpx.AsyncClient() File "/media/2TB/stable-diffusion-webui/venv/lib/python3.8/site-packages/httpx/_client.py", line 1397, in __init__ self._transport = self._init_transport( File "/media/2TB/stable-diffusion-webui/venv/lib/python3.8/site-packages/httpx/_client.py", line 1445, in _init_transport return AsyncHTTPTransport( File "/media/2TB/stable-diffusion-webui/venv/lib/python3.8/site-packages/httpx/_transports/default.py", line 275, in __init__ self._pool = httpcore.AsyncConnectionPool( TypeError: __init__() got an unexpected keyword argument 'socket_options'

EDIT: The above was achieved by attempting git update and/or creating a new directory and cloning the repository then running webui.sh. However after a bit of troubleshooting I found those methods no longer work. The correct steps for linux/ubuntu/mint is as follows:

1. Create a brand new Automatic1111 directory. Navigate to the directory.
2. As the install instructions say run this command from that directory: wget -q https://raw.githubusercontent.com/AUTOMATIC1111/stable-diffusion-webui/master/webui.sh
3. Now run ./webui.sh and it will automatically update and install everything into a SUBFOLDER!!!
4. Update your shortcuts and or aliases to point to the first webui.sh NOT the one in the subfolder :)
5. Now migrate models to the appropriate folders, run ./webui.sh and make sure to install your missing extensions.

[TruthReveller]

Upgraded to 1.6 removed all plugins for lora but adding a lora to the prompt has no effect. Anyone solve this issue or have the same issue?

[0zl]

Was the removal of the --lora-dir command line argument in version 1.6.0 intentional? Also, could you guide me on how to install the locon/lycoris model? Should I simply place it in the ‘Lora’ folder?

[dhwz]

wdym the --lora-dir argument is still there?
You can add Lycoris just to the Lora folder.
You can also use subfolders if you want to separate them.

[0zl]

Indeed, my oversight. The --lora-dir information doesn’t appear when I run --help, possibly because I executed it from launch.py idk. It’s also plausible that I might have missed it. I appreciate the information about Lora. Thank you!

[Volcanicus]

I like some parts of the update but too many things I use to work with simply aren't working as they used to. Styles being deprecated just adds to the slew of things but my outputs simply are not reproducible anymore (SDE ++). I also dropped back to 1.5.2 since I needed a specific workflow and the new update added very little benefits compared to the downsides.

Thanks again but please update/upgrade stuff with opt-ins and opt-outs if you cannot QA them and let users do that instead.
This is already a method used by industry giants with too many clients where a feature is added and if it sticks after substantial QA, then the older feature is deprecated and finally removed instead of shifting the entire platform on its head.

Thanks again for your hard work but please keep this feedback in mind moving forward.

[ericwenger7]

In 1.6 using an output directory in Batch is broken : images are always saved in /outputs/img2img regardless if output directory is correctly set. It is a bit annoying and should be treated as a bug.

[nulflux]

This update really confuses me. Since when in history does an accordion being folded or unfolded enable or disable a feature? To be honest I think work should be done to move out many or all of the main features into modules and extensions so we can enable or disable them at will. Too many things are changing each update and there isn't a stable foundation that we can rely on and revert to if new changes break existing workflows. We gotta stop dumping piles of bricks and pouring cement over them let's get out there and lay the bricks properly and apply the cement with a trowel.

[paboum]

I see lots of commits on dev branch since 1.6.0. Is there a new release scheduled?