linaqruf / sd-notebook-collection Goto Github PK

Not an issue. Just a question. How do you keep your
UI settings after each session? Or is that not possible?

Hello

I found a similar issue below but mine seems slightly different
#30

I am using the latest version of your NB but I often encounter this issue

When working with automatic1111 or Cagliostro I often encounter this issue where the generate button in the img2img tab becomes unresponsive/ gets stuck

Sometimes I can make it work again by un-enabling controlnet OR jumping to text2img, running a generation, then jumping back to img2img and this will force the img2img generate button to become responsive again

BUT

Very often this does not work and I cannot make img2img work again without completely restarting my google colab session

Is anyone aware of a more reliable way to force the img2img generate button to become responsive again?

In the latest version of your NB it seems the 'latest gradio' option is gone- is there anything else I can do to make it more reliable? should I not be using cloudflared?

ive been using bens and camendurus notebooks but recently i switched it to this on andorid

ive been getting diffrent results on this collab

some what not prefred results on same seed on other collab with same models

Thank you for your work. It's ok

had to use old commits. Git pull start not working.

dissconnects public url very frequesntly gotta restart gradio everytime

Hello!
I made this tutorial for beginners: https://youtu.be/iM4PhTNYLZU

Salutations to Linaqruf!

Hello!

I’ve installed the lycoris extension here https://github.com/KohakuBlueleaf/a1111-sd-webui-lycoris

but I can’t seem to see lycoris under additional networks- what is the best way add lyrcoris models to use in cagliostro?

hello!

I currently am trying to access these two google drive folders for checkpoints and Loras but neither seems to work

what am I doing wrong?

fuse:/content/drive/MyDrive/sd/stable-diffusion-webui/models/Stable-diffusion

fuse:/content/drive/MyDrive/sd/stable-diffusion-webui/models/Lora

https://github.com/pkuliyi2015/multidiffusion-upscaler-for-automatic1111

I am using cagliostro-colab-ui.ipynb, I am facing this after recent anapnoe commits. not sure if it has something to do with this repo (please feel free to close this issue, if not).
Please take a look, If anyone else is facing the same kindly report.

No errors found in console or log. all other ui elements are working fine

Hi just a word of thanks for this notebook, you can not imagine what it meanst to me and how cool it is that I get to play with this!

I get the error when I try to run it

it gives a bunch of python stuff in colored font but ends in

/content/cagliostro-colab-ui/modules/shared.py:21 │
│ >>>>>>> 20ae71faa8ef035c31aa3a410b707d792c8203a3

please let me know if you need more info

Everytime I try to use Img2Img it works at first (also take a while to start in the console), but after that when I try to run it again it does nothing, nothing appears in the console and the generate button doesnt work, sometimes this message pops up in the console but not always. It was working some days ago but not now, would appeciate any help, thanks.
Text2Img works perfectly fine thought.

Normally i got 3 links like gradio but now only one link appear.Plz help

DISREGARD THIS I SAW THE FIX JUST AFTER POSTING THIS

I tried all 3 options and it still gave the same error.

fuse:/content/drive/MyDrive/sd/stable-diffusion-webui/models/Lora

when I have the above text in the Lora path field I get this problem where the webui never loads beyond the point in my screengrab

If I leave custom paths empty it works

How do I fix this?

This occurs when using the Colab custom_model_url or Batchlinks Downloader extension to download models

Hello!
Can I present the cagliostro google colab notebook on YouTube in a video tutorial?
Thanks for understanding!

The credits won't be mine.

Couldn’t launch the notebook

Hello.
I get the following error in the Start Cagliostro Colab UI. What is the cause?

AttributeError: 'ExtraNetworksPageLyCORIS' object has no attribute
'get_sort_keys'
Creating model from config: /content/cagliostro-colab-ui/configs/v1-inference.yaml
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.
Loading VAE weights specified in settings: /content/cagliostro-colab-ui/models/VAE/stable_diffusion.vae.pt
Applying optimization: xformers... done.
Textual inversion embeddings loaded(9): bad-artist, bad-artist-anime, bad-hands-5, bad-image-v2-39000, bad_prompt, bad_prompt_version2, EasyNegative, EasyNegativeV2, ng_deepnegative_v1_75t
Textual inversion embeddings skipped(7): nfixer, re-badprompt, Mayng, nrealfixer, rev2-badprompt, wdbadprompt, nartfixer
Model loaded in 4.8s (load weights from disk: 1.6s, create model: 2.7s, load VAE: 0.3s).
Exception ignored in atexit callback: <function kill at 0x7f33233dfd00>
TypeError: kill() missing 1 required positional argument: 'self'

Lora already installed but i can't use it. plz help me!

locon load lora method
loading Lora /content/cagliostro-colab-ui/models/Lora/koreanDollLikeness_v20.safetensors: AttributeError
Traceback (most recent call last):
File "/content/cagliostro-colab-ui/extensions/a1111-sd-webui-locon/scripts/../../../extensions-builtin/Lora/lora.py", line 222, in load_loras
lora = load_lora(name, lora_on_disk.filename)
File "/content/cagliostro-colab-ui/extensions/a1111-sd-webui-locon/scripts/main.py", line 370, in load_lora
is_sd2 = 'model_transformer_resblocks' in shared.sd_model.lora_layer_mapping
File "/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py", line 1614, in getattr
raise AttributeError("'{}' object has no attribute '{}'".format(
AttributeError: 'LatentDiffusion' object has no attribute 'lora_layer_mapping'

I've come across a strange bug when i don't use Hires fix Loras apply normally but when i turn it on it starts "ignoring" Loras and just uses the base model. (it's very noticeable when using style Loras) It doesn't seem to be a problem in the other notebooks I'm using so I would like to bring it to your attention.

I've uploaded an image to show the bug
Left side was generated from your notebook and the Right Side was generated using fast sd notebook
The right side retains the artstyle while the left seems to overwrite it with the models artstyle.

I've tried this with multiple models with the same result.

Hello!

I see that by default the UI includes the latent couple extension
https://github.com/miZyind/sd-webui-latent-couple/blob/main/README.md

However I only see the rectangular mode for dividing the canvas- how can I access the ‘blob drawing’ interface where you can actually draw the shapes of the regions you would like to prompt?

when I try to install the extension it says that the extension is already present…. is there a way that I can make it function fully with the prompt region drawing feature?

Worked great yesterday, but today I keep getting this error after running the "Start Cagliostro Colab UI" step, used the last commit but its the same

can you make this colab for runpod.io template? Thanks

DPM++ 2M alt Karras

I'm using DPM++ alt Karras,
it expresses a rich tone, more and more users are using it
I would be happy if you put an option that can be added
DPM++ alt Karras can be found in Civitai

Thank you for using Cagliostro Colab UI Pro. I love it so much. I always support you.

Some colab ui authors added music to avoid Google colab stopping with no reason. Please add the same feature as well,thx

cagliostro-colab-ui.ipynb
I encountered an error when attempting to uncheck the 'use_anapnoe_ui' option. Specifically, I received the following error message:

UnboundLocalError Traceback (most recent call last)
in <cell line: 308>()
306 print("All is done! Go to the next step.")
307
--> 308 main()

in main()
265 extensions_list = os.listdir(extensions_dir)
266 with tqdm(
--> 267 total=len(extensions_list) - len(skipped_extensions) - 1,
268 desc="Updating extensions",
269 mininterval=0,

UnboundLocalError: local variable 'skipped_extensions' referenced before assignment

1. Can you add the "file explorer" tool so we can navigate if the files aren't working correctly?
2. VAE are loading fine, I think the models were loading fine, but no textual embeds downloaded, and the LOra only downloaded half of the ones i listed, and no way to get it to list or refresh.
3. Not sure why but adding things to the right folders in the file section on the side DIDNT work.
4. Can you add an option to just use model paths for LoRA and Models etc? if we have a pre run install from say Lastben - this would save me so much time.
5. PROPS TO THE NGROK! Loving it, don't remove it i'll cry.

Really liking the layout it's a little difficult if you're used to A1111's UI, but this is cool it's close to invoke.

Hi, with every generation I am getting this error. Does not look like it has any negative effect on the performance or the generations. Just thought I will bring it to your attention.

*** Error running before_process: /content/cagliostro-colab-ui/extensions-builtin/extra-options-section/scripts/extra_options_section.py
Traceback (most recent call last):
File "/content/cagliostro-colab-ui/modules/scripts.py", line 467, in before_process
script.before_process(p, *script_args)
File "/content/cagliostro-colab-ui/extensions-builtin/extra-options-section/scripts/extra_options_section.py", line 40, in before_process
for name, value in zip(self.setting_names, args):
TypeError: 'NoneType' object is not iterable

yep, it stuck. and happens almost every time I use Hi-res fix, it usually happens when finalizing the image like 97% finish on UI and 100% completed process on colab, this problem appeared 2 days ago I guess when the adetailer& CN reference was added, the interrupt/skip also broken when using hi-res (there no point clicking it though knowing using hi-res already makes progress stuck), the usual stuff happens before this still saved the images to drive output, but this one isn't, that all I know, hope there is a way to fix it.

• tested it again, it often work again by waiting for 10minutes+ or so
• probable storing issues?
• when that happen the finished images wont show on UI after result (it common though)
• seems like higher resolution cause the big part of the problem (using 1024x512 with 2x hres upscale are enough to trigger the problem, no lowram args)

repo_type:AUTOMATIC1111

KeyError Traceback (most recent call last)
in <cell line: 115>()
113 config["show_progress_every_n_steps"] = 10
114 config["show_progressbar"] = True
--> 115 if config["quicksettings"]:
116 config["quicksettings"] = "sd_model_checkpoint, sd_vae, CLIP_stop_at_last_layers, use_old_karras_scheduler_sigmas, always_discard_next_to_last_sigma"
117 else:

KeyError: 'quicksettings'

repo_type:Anapnoe no such problem

hi - sorry for opening an issue, feel free to close immediately, this is just an attempt of reach out.

my fork at https://github.com/vladmandic/automatic is gaining traction and i'd love to see people that are experts in their area to take part (full fledged contributors, not just random prs) - and your name came up as THE guy for google collab :)

anyhow, if interested, drop me a note - here or in the repo or directly on discord: vlado#8083

Hello!

I am currently pulling checkpoints from a fused folder on google drive

when I merge checkpoints they do not appear to be saved permanently anywhere in google drive

how can I make sure that they are saved permanently?

Just got a notice today that Colab may terminate my runtime while executing code to run Cagliostro UI.

So, it's a repeat of the stable diffusion fiasco; it's gotten too popular and Google is once again initiating restrictions.

Even with the same prompt, seed, size, step, and CFGscale, automatic1111 and cagliostro output different results.

I am seeing that perhaps some setting is turned on by default, but I can't figure out what it is... I found the "clip skip" item and tried setting it to 1, but there was no change and the output is still different.

The problem with this is that the picture style has changed.

I am trying to make the prompts consistent to some extent, but all my previous prompts are going to waste, and the same model looks like a different model.
Please let me know if you know the cause of this problem.

Thank you.

Hi, I really appreciate your program! Thank you for your hard work!!
I donated a coffee! :)
And I want to ask one more question.

Is it possible to open files from Google Drive, as well as Civitai files, on the program? I tried to open Google Drive addresses, but the links for some models and LoRAs wouldn't work and one LoRA went to the models section.
I have some personal LoRAs and models that I would like to use, and some of the models that I was using through Civitai are now unavailable. So I would really like to use the Google Drive addresses, is there any way you can assist?

I'm also having trouble opening hypernetwork, LoRA section when I use i2i. I have to go to t2i and click LoRAs and copy and paste them. Could you fix this, as well?

Selecting a model and preprocessor only loading continues

Gradio link does now show up in the last cell

But it HANGS and wont let me edit text, load models OR edit poses.

Did i install too much again?

Hello,

Whenever I try to generate images using a Lora I encounter this error:

RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument mat2 in method wrapper_CUDA_mm)

Is there any way to fix it?

Hello!
First of all, I'd like to thank you for making the program.
I am using it very well and it is great; the speed is fast, and the control net is built-in.
However, I am having a few issues and would really appreciate your help.
I correctly entered the address in the custom URL section, however, the models and LoRAs refused to open. Some LoRAs were placed in the models section, but still did not function.
can you please help me with this?

Hello, using the default fast-vladmantic-webui,
There is an error when starting webui:

error: unrecognized arguments: --enable-insecure --no-download-sd-model

I fixed it replacing --no-download-sd-model with --no-download and removing line enable-insecure : True in the config json. I don't know if it is good practice though.

Any chance, to get the possibility, to use a checkpoint from google drive itself and not been limited to this few? 🙂

hello!

i am using the latest version of your colab and running into an issue where during a session I have difficulty changing models- I changed models once in my current session- now I try again and no matter how many times I try it will not change models- the readout in colab also shows no attempt to load new model weights- it will continue making new generations if I press the generate button however

is there any way to force it to switch models?

Hello! How can I add canny 1.0 to use in the UI?

"https://huggingface.co/ckpt/ControlNet-v1-1/resolve/main/control_v11p_sd15_canny_fp16.safetensors",

That is the download link but it doesn't seem to do anything when I add it to the controlnet links list?

I found the annotators but seem to be missing the final processors for T2i color etc- do I need to do something else in the latest NB to force them to download? Looks like the checkbox has been removed from the colab interface?

On uploading image on inpaint, the image is blank. Cannot see it. This was working fine before the last 2 commits.

How to fix this error:

RuntimeError: Detected that PyTorch and torchvision were compiled with different CUDA versions. PyTorch has CUDA Version=11.7 and torchvision has CUDA Version=11.8. Please reinstall the torchvision that matches your PyTorch install.