I updated the extension version today (January 31, 2023) and a message started to appear with each generated image:
settings2 = {'is_enabled': 'True', 'keyword_placement': 'keyword prompt', 'multiple_keywords': 'keyword1, keyword2', 'ti_keywords': 'None', 'keyword_order': 'textual inversion first'}

Hi, this model obviously has a Trigger Words for different styles, but Keyword has never added to PNG information (metadata). I am doing something wrong? I tried some Lora model at random and Keywords found it without any problems.

hello, sorry i couldnt find a discussion tab for this extension,
what Multiple keywords is made for and how does it work ?

With Loras gaining massive speed (and small size), they are growing in popularity. Mostly similar to the checkpoint models, they use trigger words... so a hash/trigger word approach would be needed. Pretty much like a middle ground between TIs and Checkpoint models.

Loras only get loaded on prompt start, so either adding them via model keyword (similar to TIs), or detecting they are in prompt as lora:Lorafilename:1 (# is strength, can be floating and more or less than 1)

Error completing request
Arguments: ('lion', '', 'None', 'None', 20, 0, False, False, 1, 1, 7, -1.0, -1.0, 0, 0, 0, False, 512, 512, False, 0.7, 0, 0, 5, False, False, False, '', 1, '', 0, '', True, False, False, '

You can edit extensions/model-keyword/model-keyword-user.txt to add custom mappings

', 'keyword prompt', 'keyword1, keyword2', 1.0, 2.0, 'a painting in', 'style', 'picture frame, portrait photo', None) {}
Traceback (most recent call last):
File "D:\stable-diffusion-webui\modules\call_queue.py", line 45, in f
res = list(func(*args, **kwargs))
File "D:\stable-diffusion-webui\modules\call_queue.py", line 28, in f
res = func(*args, **kwargs)
File "D:\stable-diffusion-webui\modules\txt2img.py", line 46, in txt2img
processed = modules.scripts.scripts_txt2img.run(p, *args)
File "D:\stable-diffusion-webui\modules\scripts.py", line 317, in run
processed = script.run(p, *script_args)
File "D:\stable-diffusion-webui\extensions\stable-diffusion-webui-inspiration\scripts\inspiration.py", line 39, in run
for file in files:
TypeError: 'NoneType' object is not iterable

I have some models that only behave properly with some additional negative prompts. I've patched my local copy to allow me to add both positive and negative prompts -- would you be interested in getting a PR for this change?

I'm testing the new Comic Diffusion v2 model with 6 activation tokens.

I set up the Multiple Keywords mode to Iterate, as I want A1111 to produce 6 images, one for each activation token.
As I click Generate, tho, only a single image is produced.

Did I misunderstand the intent of the iterate mode?

The extension is appending the keyword 'cat' for the plain v1-5-pruned-emaonly.safetensors checkpoint...

Lora Keywords are not added when I use the Default Lora functions of WEBUI.

Should I manually select Lora and keywords in the Model-keyword extension dropdown menu?

I hope keyword add automatically when I use Default Lora Functions of WEBUI.

Extension doesn't recognize Lora in the subfolder.

models\Lora\Vehcile

In this case, Lora in the Vehicle folder doesn't recognized in Model-Keyword Extension.

Hi,
Whenever this extension is updated through Web UI, the model-keyword-user.txt is overwritten back to the default file.
I'm not sure how Web UI updates extensions, but please consider how this file is handled, so it doesn't get overwritten each update.

can there be included a way to add keywords inside auto1111, without going to the user script?
can there be a menu that drops down that tells you all the keywords associated with the model before you generate?
thank you

Most extensions are disabled by default when first installed, this one is opposite - its enabled by default as soon as its installed.
If possible, please change default behavior.

Your tool is very useful and I appreciate your work.
I was just wondering if you were going to add LyCORIS support?
I'm not sure how well these will catch on, but it looks Civitai is getting more and more of them.
https://github.com/KohakuBlueleaf/LyCORIS
https://github.com/KohakuBlueleaf/a1111-sd-webui-locon

From https://huggingface.co/plasmo/woolitize

f2ea5e141, woolitize, woolitize768.ckpt

As above, somehow this model is getting "master, chief" added, no custom mappings etc

Is there an existing issue for this?

I have searched the existing issues and checked the recent builds/commits, but I couldn't find a similar issue.

What happened?

When I try to use/set embeddings option it closes with error

/AppleInternal/Library/BuildRoots/9e200cfa-7d96-11ed-886f-a23c4f261b56/Library/Caches/com.apple.xbs/Sources/MetalPerformanceShadersGraph/mpsgraph/MetalPerformanceShadersGraph/Core/Files/MPSGraphExecutable.mm:1309: failed assertion `Incompatible element type for parameter at index 0, mlir module expected element type f32 but received f16'

[1] 16300 abort ./webui.sh --administrator

Steps to reproduce the problem

Run the process "C:\sd\webui\extensions\model-keyword\scripts\model_keyword.py"
What should have happened?

The process should have run without any errors.

Commit where the problem happens

commit d04e3e921e8ee71442a1f4a1d6e91c05b8238007 (HEAD -> master, origin/master, origin/HEAD)
Author: AUTOMATIC [email protected]
Date: Sat Jan 28 15:24:29 2023 +0300

What platforms do you use to access UI ?

MacOS

What browsers do you use to access the UI ?

Google Chrome

Command Line Arguments

--administrator --upcast-sampling --use-cpu interrogate

Additional information, context and logs

none

Thank you.

Ideally, this extension should be able to handle conceptual groups of trigger words within a single LoRA, such as:

EvangelionLoRA.safetensors

AYANAMIREI, EVA00PLUGSUIT, INTERFACE HEADSET

AYANAMIREI, TOKYO-3_MIDDLE_SCHOOL_UNIFORM, RED RIBBON, SCHOOL_UNIFORM, BLACK SOCKS

SORYUASUKALANGLEY, TOKYO-3_MIDDLE_SCHOOL_UNIFORM, RED RIBBON, SCHOOL_UNIFORM, WHITE SOCKS, INTERFACE HEADSET

SORYUASUKALANGLEY, EVA02PLUGSUIT, INTERFACE HEADSET

--//--

Description 'keyword or keywords separated by |' is unclear.

Expected behavior from description might be:

word1, word2, word3|word1, word4, word5

Which should produce a drop list:

word1, word2, word3
word1, word4, word5

But in actual fact, produces a drop list:

word1

...because the underlying text file is CSV and the input is not sanitized.

Alternative expected behavior from description might be:

word1|word2|word3|word4|word5

But then the user does unnecessary extra work for no gain (shift key) to type it rather than naturally type (or copy/paste) 'word1, word2, word3, word4, word5' and have the extension convert the ", " or "," to "|" when writing to the file (Replace and Trim).

But this still cannot handle LoRA that conceptually have multiple sets of trigger words that are only used together in groups, such as multiple named characters with multiple costume options.

--//--

Although the README states that saving new sets of keywords overwrites the previous saved entries, the content of the file is not overwritten but appended in practice.

So, it appears possible to save groups as multiple entries in the underlying text file by:

word1|word2|word3 [SAVE BUTTON]
word4|word5|word6 [SAVE BUTTON]

And the file will then contain:

checksum1234, word1|word2|word3, LoRA1.safetensors
checksum1234, word4|word5|word6, LoRA1.safetensors

But, when the file is read back into the drop list, only the last entry for the LoRA is listed and all other entries for the same LoRA are ignored:

word4
word5
word6

I suspect this behaviour is a bug I've accidentally discovered by trying to figure out how to group trigger words.

--/--

BTW the extension fails with an error if the text file accidentally becomes UTF-8, which is the default format for Windows 11 onwards. This is going to catch people out, as they will want to edit the text files manually to correct mistakes, as this will be much faster than starting up and using the WebUI interface when making alterations to LoRA collections.

I discovered this problem while trying to work around the grouping problem using 'word1·word2·word3|word4·word5' with the intent to manually edit after auto-insertion (note: some trigger words do contain spaces, so space cannot be used to separate words in a group).

--//--

This add-in has potential to become a major sought-after feature that can overcome the problem that the WebUI lacks any form of trigger word management other than having a huge unsearchable Style list.

I have not seen any alternatives so far, other than keeping a notebook on the side. The website you mentioned in another thread is not up-to-date with the large number of new LoRA being produced and also requires users to have not renamed the LoRA to make them easier to use.

I do hope you will be encouraged to continue work on this (and check out the booru tag autocomplete extension to see if the same techniques can be used to provide pop-up trigger word selection drop lists in the prompt box itself rather than using a drop list of LoRAs below it).

Hello,
This extension is fantastic for working with a large amount of trained models, but I was wondering if it would be possible to have a mode that replaces some trigger word with the pre-set keyword for a model. For example:
A photo of $trigger on a scenic mountaintop
or
A charcoal drawing of an army of ducks following $trigger with a pan flute
would replace $trigger with whatever keyword was associated with the model. This way the keyword could come anywhere in the prompt rather than always being pre/appended.

I naively think this could be as simple as replacing:

            if keyword_placement.startswith('keyword'):
                arr.append(prompt)
            else:
                arr.insert(0, prompt)

with something like

            trigger = 'xyz'
            if keyword_placement.startswith('keyword'):
                arr.append(prompt)
            elif 'replace' in keyword_placement:
                arr = prompt.replace(trigger, ''.join(arr))
            else:
                arr.insert(0, prompt)

and the relevant menu options, but I may be missing something.

Thanks!

(BTW, this is just the extension I needed when I did a big huge X/Y plot run to compare 50 some models. Running again.)

After I installed the extension, it would not run initially without creating a blank custom-mappings.txt file.

Can a blank one be created and installed by default?

Repro:

• Install extension
• Restart UI
• I started an X/Y Plot that begins w/ the analog-diffusion model
• Generate
  Result:
• Traceback error to command console saying that custom-mappings.txt was not found
  Workaround:
• Create a blank custom-mappings.txt file

Expanding the script extension area, too, and performing Save doesn't help. Performing Check gets an error.

Traceback:
Traceback (most recent call last):
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\routes.py", line 337, in run_predict
output = await app.get_blocks().process_api(
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\blocks.py", line 1015, in process_api
result = await self.call_function(
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\venv\lib\site-packages\gradio\blocks.py", line 833, in call_function
prediction = await anyio.to_thread.run_sync(
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio\to_thread.py", line 31, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio_backends_asyncio.py", line 937, in run_sync_in_worker_thread
return await future
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\venv\lib\site-packages\anyio_backends_asyncio.py", line 867, in run
result = context.run(func, *args)
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\extensions\model-keyword\scripts\model_keyword.py", line 122, in check_keyword
entry = get_keyword_for_model(model_hash, model_ckpt, return_entry=True)
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\extensions\model-keyword\scripts\model_keyword.py", line 71, in get_keyword_for_model
hash_dict = load_hash_dict()
File "C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\extensions\model-keyword\scripts\model_keyword.py", line 41, in load_hash_dict
modified = str(os.stat(default_file).st_mtime) + '_' + str(os.stat(user_file).st_mtime)
FileNotFoundError: [WinError 2] The system cannot find the file specified: 'C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\extensions\model-keyword/custom-mappings.txt'

Thank you for sharing this extension.

It would be great to have the same functionality for Embedding/Textual Inversions keywords and notes.

I hope @civitai can provide a list of Embedding, trigger words and hashes.

LORAs selected by the dropdown menu; and attempting to use created custom keyword mapping results in 'ERROR" in the UI.

NameError: name 'model_ckpt' is not defined
Traceback (most recent call last):
File "G:\stable-diffusion-webui-current\venv\lib\site-packages\gradio\routes.py", line 337, in run_predict
output = await app.get_blocks().process_api(
File "G:\stable-diffusion-webui-current\venv\lib\site-packages\gradio\blocks.py", line 1015, in process_api
result = await self.call_function(
File "G:\stable-diffusion-webui-current\venv\lib\site-packages\gradio\blocks.py", line 833, in call_function
prediction = await anyio.to_thread.run_sync(
File "G:\stable-diffusion-webui-current\venv\lib\site-packages\anyio\to_thread.py", line 31, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "G:\stable-diffusion-webui-current\venv\lib\site-packages\anyio_backends_asyncio.py", line 937, in run_sync_in_worker_thread
return await future
File "G:\stable-diffusion-webui-current\venv\lib\site-packages\anyio_backends_asyncio.py", line 867, in run
result = context.run(func, *args)
File "G:\stable-diffusion-webui-current\extensions\model-keyword\scripts\model_keyword.py", line 444, in add_lora_keyword
lora_keywords = get_lora_keywords(lora_model)
File "G:\stable-diffusion-webui-current\extensions\model-keyword\scripts\model_keyword.py", line 168, in get_lora_keywords
kws = _get_keywords_for_lora(lora_model)
File "G:\stable-diffusion-webui-current\extensions\model-keyword\scripts\model_keyword.py", line 157, in _get_keywords_for_lora
sim = str_simularity(kw_ckpt[1], model_ckpt)
NameError: name 'model_ckpt' is not defined

So it makes sense that these two won't play nice together, but I figured I'll post the error here and see if its fixable.
One button prompt is a fun prompting extension - https://github.com/AIrjen/OneButtonPrompt
When using it with model-keyword enabled I get the following in console:

Error running process: F:\SD\stable-diffusion-webui\extensions\model-keyword\scripts\model_keyword.py Traceback (most recent call last): File "F:\SD\stable-diffusion-webui\modules\scripts.py", line 451, in process script.process(p, *script_args) File "F:\SD\stable-diffusion-webui\extensions\model-keyword\scripts\model_keyword.py", line 657, in process p.prompt = new_prompt(p.prompt, kw, no_iter=True) File "F:\SD\stable-diffusion-webui\extensions\model-keyword\scripts\model_keyword.py", line 649, in new_prompt return ', '.join(arr) TypeError: sequence item 1: expected str instance, list found

Thank you for your work on this extension!
(also - please add lycrois support :) )

See the example below.

None
anime_nausicae.safetensors
artist_dan_mumford.safetensors
illustration_kSUWABESTYLE_v2.safetensors
illustration_enaic.safetensors
special_epiNoiseoffset_v2.safetensors
special_offset.safetensors
style_anaglyph.safetensors
style_rotoscopee.safetensors

I organize the Loras with types before their names.

You are doing a great job.
It will be fantastic if embeddings (Textual Inversions) are included too. Would make it very easy, instead of using the file name for them.

Thanks

As CIVITAI is recommending people to install the LyCORIS extension from https://github.com/KohakuBlueleaf/a1111-sd-webui-lycoris on each LyCORIS pages, the module will not be able to support those as they are in a different folder.
It could be nice to add this feature.

This is a great extension that will save me tons of time. Thanks for sharing it with the community.

Is there any chance to integrate it, somehow, with the Plot X/Y script?

Plot X/Y already allows to call different checkpoints/models but there's no way to automatically include the relevant keyword for each checkpoint in the prompt.

If these two extensions could work together, it would be amazing.

Hello, just wanted to thank anyone involved in keeping this project going. It's one of the longest-running extensions and still going strong, and I believe it is indispensable to many.

Absolutely awesome, you rock. <3

Is there an existing issue for this?
I have searched the existing issues and checked the recent builds/commits, but I couldn't find a similar issue.

What happened?
I am getting the following error when running the process: C:\sd\webui\extensions\model-keyword\scripts\model_keyword.py

Traceback (most recent call last):
File "C:\sd\webui\modules\scripts.py", line 347, in process
script.process(p, *script_args)
File "C:\sd\webui\extensions\model-keyword\scripts\model_keyword.py", line 327, in process
kw = self.get_keyword(model_hash, model_ckpt)
File "C:\sd\webui\extensions\model-keyword\scripts\model_keyword.py", line 243, in get_keyword
hash_dict = self.load_hash_dict()
File "C:\sd\webui\extensions\model-keyword\scripts\model_keyword.py", line 213, in load_hash_dict
modified = str(os.stat(default_file).st_mtime) + '_' + str(os.stat(user_file).st_mtime)
FileNotFoundError: [WinError 2] The system cannot find the file specified: 'C:\sd\webui\extensions\model-keyword/custom-mappings.txt'

Steps to reproduce the problem

Run the process "C:\sd\webui\extensions\model-keyword\scripts\model_keyword.py"
What should have happened?

The process should have run without any errors.
Commit where the problem happens

My Webui is on Commit hash a8322ad.

What platforms do you use to access UI ?
Windows

What browsers do you use to access the UI ?
Google Chrome

Command Line Arguments
None

Additional information, context and logs
none

Thank you.

Can't seem to add Chinese characters or at least what I think are chinese characters: 可露凯
I assume it has something to do with character encoding or something.

From https://huggingface.co/nousr/robo-diffusion-2-base.

14fe5bd7, nousr robot, robo-diffusion-v2-base.ckpt

The current interface can't handle more than two keywords and ie. SynthPunk V.3 (https://civitai.com/models/2856/synthpunk-search) uses 3 keywords.
I would probably solve it by using checkboxes and ordering them from an input number next to them or by having a drag and drop list with keywords at the start, end and inactive.

Comic Diffusion v2 [8b3c8f111] is currently with token comicmay artsyle. However, per instructions here there are 6 activation token and none of them matches the one currently used:

charliebo artstyle
holliemengert artstyle
marioalberti artstyle
pepelarraz artstyle
andreasrocha artstyle
jamesdaly artstyle

Not sure how you'll implement this :) Ideally, somebody would want to be able to pick which one of the 6 to enforce.

Currently, the VintageCream model [a5c134f5] has associated the token naturitize. However, the correct activation token is vntgcrm style according to https://huggingface.co/SebastianKamphYT/VintageCream

I tried to install the extension given the README directions but I got a git error via A1111 when trying to do so (I have git bash installed and it is accessible through the PATH).

FWIW, I did install by giving the GitHub clone path to A1111:
https://github.com/mix1009/model-keyword.git

I restarted the UI and the extension is installed.

This is on Windows 10.

Command line output:
raise GitCommandError(remove_password_if_present(self.args), status, errstr)
git.exc.GitCommandError: Cmd('git') failed due to: exit code(128)
cmdline: git clone -v https://github.com/mix1009/model-keyword C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\tmp\model-keyword
stderr: 'fatal: could not create work tree dir 'C:\Users\Eric\repos\Stable-Diffusion\stable-diffusion-webui\tmp\model-keyword ': Invalid argument

Using this extension IMHO is very nonintuitive. Ideally, what we all want is a way to easily see and add related keywords for certain loras, models, etc. without needing to remember or look up the keywords online. The way it works currently has a lot of issues that conflict with this vision though.

The issues - From my perspective

1. You need to manually select from a dropdown what lora you want to use keywords from. This dropdown use text to search so you need to manually scroll and find the lora. If you have 100+ loras this becomes extremely tedious.
2. After selecting a lora you need to click the keyword dropdown just to see if any were even detected/saved.
3. keyword selection of dropdown is limited. You can do all keywords, or individual keywords. What if there were only 3 keywords and you only wanted to use 2? The dropdown prevents that
4. Because keywords are in a dropdown I can copy and past keywords to prompt
5. Because you need to manually select the lora and keywords using dropdowns, this also means if you had multiple loras you can only use keywords from one at a time in your prompt
6. Keywords added are added to the beginning of the prompt, but users might not always want them put there.

The solution

I think instead this detection should detect loras that exist in the prompt and show a list of keywords that can be used with each lora. Then the user can click to add whatever keywords they want from the list or lists (if multiple loras are used.)

I imagine this working very similarly to the way adding loras in webui works already. you could have a button that shows and hides the box of keywords. It should probably be positioned just under the prompt box since it will be adding keywords to the prompt.

It's a decent overhaul but i think it would be greatly beneficial.

Right now I need to change "Multiple keywords:" to random (or whatever I want) every time I start the ui. I would prefer it to save my preferred setting.
I've looked through the settings but couldn't find an option for it. Sorry if it's just me being blind

Add a large field with notes so that you can write any text for each model, including recommended settings or negative prompt.

webui changed their hash algorithm, which makes the extension to no longer function.

Looking for a workaround...