Furkan Gözükara patreon

Run Automatic1111 SD Web UI On A Free Kaggle NoteBook Like In Your PC - Supports SDXL & ControlNet

🕑 Added 2024-05-24 21:11:00 +0000 UTC

Comments

Furkan Gözükara

i need to test if ComfyUI would be possible.

Vince Ma

This is fantastic! Never knew we got free lunch, will try it this evening! Thanks a lot for sharing!!! But possible to make a note to run Comfyui at Kaggle Notebook?

Furkan Gözükara

thank you so much fixed with v12

Oliver Koch

Minor bug report: In The "Zip all images cell" it should be "/outputs/" instead of "/output/". Otherwise it works great.

Furkan Gözükara

well that means you need to manually install xformers and torch version to fix the error. install torch 2.2.0 and xformers 0.0.24

Akshay

Hi, I'm running this on colab Pro with L4 GPU and High RAM but its giving me this error while trying to generate images. NotImplementedError: No operator found for `memory_efficient_attention_forward` with inputs: query : shape=(2, 4096, 8, 40) (torch.float16) key : shape=(2, 4096, 8, 40) (torch.float16) value : shape=(2, 4096, 8, 40) (torch.float16) attn_bias : p : 0.0 `decoderF` is not supported because: xFormers wasn't build with CUDA support attn_bias type is operator wasn't built - see `python -m xformers.info` for more info `flshattF@0.0.0` is not supported because: xFormers wasn't build with CUDA support operator wasn't built - see `python -m xformers.info` for more info `tritonflashattF` is not supported because: xFormers wasn't build with CUDA support operator wasn't built - see `python -m xformers.info` for more info triton is not available Only work on pre-MLIR triton for now `cutlassF` is not supported because: xFormers wasn't build with CUDA support operator wasn't built - see `python -m xformers.info` for more info `smallkF` is not supported because: max(query.shape[-1] != value.shape[-1]) > 32 xFormers wasn't build with CUDA support dtype=torch.float16 (supported: {torch.float32}) operator wasn't built - see `python -m xformers.info` for more info unsupported embed per head: 40

Furkan Gözükara

Thank you so much. Yes it is fixed I also tested.

Rick B

I see you've reset the models dir to the temp directory. I was going to report I spent all day with my checkpoints loaded in /kaggle/temp/ and it worked fine, but you're already aware. Thanks so much for keeping on top of all the updates.

Furkan Gözükara

hello. hopefully i will add those new controlnet models soon. please remind me again if i don't

Gen Zero

and How To install models ip-adapter-plus-face_sd15 ? please advise me Or is there a teaching clip for me?

Furkan Gözükara

hello. kaggle is having issues. please stop session and start again. for the last few days Kaggle have some problems sadly. not related to our scripts but their systems

Gen Zero

Scripts eror.I can't Running on local URL Please advise me?.

Furkan Gözükara

no you need to use Ngrok. Kaggle do not allow public anymore. in recent videos I have shown this. check this one : https://youtu.be/16-b1AjvyBE

Droidmob95

Public not showing, It say to To create public link, set 'share=True' in 'launch()'. What Am l suppose to do

Furkan Gözükara

Hi. Fooocus has auto installer for windows I have shown in this video : https://youtu.be/jHTkVm2mcfs for runpod we have auto installer : https://www.patreon.com/posts/fooocus-sd-web-1-92759045 for kaggle we have auto installer : https://www.patreon.com/posts/fooocus-stable-94269866

Furkan Gözükara

Art

Hi... is it possible to add Fooocus to this Automatic1111 installation?

Furkan Gözükara

eğer NSFW resim oluşursa çok katılar. hemen kapatıyorlar tek resim ile. ona göre model seçmeni tavsiye ederim

Yavuz Emre Kocak

görseller oluşturulduktan sonra sistem hesabı bloke ediyor

Furkan Gözükara

you are doing wrong. !wget "https://huggingface.co/xxx/my_checkpoints/blob/main/My_DB_Kaggle-000001.safetensors" -O "/kaggle/temp/cnmodels/My_DB_Kaggle-000001.safetensors"

Felix Rockwell

I tried this: !wget -O /kaggle/temp/cnmodels/My_DB_Kaggle-000001.safetensors “https://huggingface.co/xxx/my_checkpoints/blob/main/My_DB_Kaggle-000001.safetensors” But it says: Scheme missing Do you know why? I couldn't find any other example how to get this done.

Furkan Gözükara

hello. upload them to your hugging face repo. so you can blazing fast download. in this video i have shown that : https://youtu.be/16-b1AjvyBE

Felix Rockwell

Hello, I followed your tutorial in creating my own Lora on Kaggel. The files have like 6GB each and I have maybe 5 files. The problem what I have now is that I have to upload all these big lora files back to Kaggel to start working with them? With my Internet connection 5Mbps it would take ages.

Furkan Gözükara

that means your working directory is full. you need to move models into temp .

Pallavi Chauhan

hi, while running "ngrok cell" I got this error "ERROR: Could not install packages due to an OSError: [Errno 28] No space left on device" I tried 2 times more but still same error what to do? Ihave temp space 54.7GB filled.

Furkan Gözükara

replied back

Juan Chen

often unable to find out loras and controlnets after refreshing the interface, i have sent messages to you in discord

Furkan Gözükara

make sure that you have selected T4 GPUs before starting.

Quentin Guittard

Hello, the notebook is failing at the webui launch. ```Using TCMalloc: libtcmalloc_minimal.so.4 Python 3.10.12 | packaged by conda-forge | (main, Jun 23 2023, 22:40:32) [GCC 12.3.0] Version: v1.6.1 Commit hash: 4afaaf8a020c1df457bcf7250cb1c7f609699fa7 Traceback (most recent call last): File "/kaggle/working/stable-diffusion-webui/launch.py", line 48, in main() File "/kaggle/working/stable-diffusion-webui/launch.py", line 39, in main prepare_environment() File "/kaggle/working/stable-diffusion-webui/modules/launch_utils.py", line 356, in prepare_environment raise RuntimeError( RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check```

Furkan Gözükara

You are right. Automatic1111 do not support multiple GPUs. I compared P100 and it was slower than T4.

Krishan Kant

The notebook currently isn't making use of both Nvidia T4 GPUs. I was wondering about the benefits of having two T4s. Alternatively, we could consider the P100. Is there a way to configure it for optimal performance?

Furkan Gözükara

you need to comment some models in that case. comment the ones that you won't use. put # in front of them

Pc 0005

It is not working at this moment right? The space of 70GB runs out and asks to upgrade to google cloud notebooks

Furkan Gözükara

i am testing right now but working. have you watched this video? https://www.youtube.com/watch?v=dBBVQfKDmhM

C05MIC Studio

this doesn't seem to work anymore?

Furkan Gözükara

i have other Patreon members generated for hours SFW images and still working. my account is also still working

DAVID PEREZ

Hi everyone, is great to support you amazing work Furkan, two questions: I want to use this for generating sfw images. Is it safe to do so? Or could I get banned in Kaggle for using only automatic 1111?

Furkan Gözükara

roop extension is i think broken not supported anymore. you can use it on runpod as solo not as an extension, on google colab as solo or in computer solo. i never used it as an extension

Mark Brown

I installed the roop extension and restarted the UI but it does not appear under controlnet as expected. I have read that the UI needs to be stopped and then restarted for the extension to appear. How can I stop the UI and restart in Kaggle?

Furkan Gözükara

awesome thank you so much for this comment

Salvador Robles

By the way the notebook works very well in kaggle, I have been testing it and although it is a bit slow (by kaggle not by your notebook which is great). The notebook is very well structured and as soon as you load it a couple of times it is already used very fluidly; It is also easy to add other models... Thank you for such a good job.

Furkan Gözükara

good question. i don't know how you can censor.perhaps you can add to the negative prompts.

Salvador Robles

Hello and congratulations for your great work. For security reasons in my use of the kaggle webui notebook solution (so great job) I would like to know what I should do so that the ui does NOT allow to generate nudity or NSFW in general, could I censor the notebook in some way so that it can not generate NSFW?, what models should load for it. Thank you

Furkan Gözükara

here recorded a video for you. working. also sent you a DM : https://huggingface.co/MonsterMMORPG/SECourses/resolve/main/2023-09-23%2014-35-29.mkv

Muhammad Lutfi

when I try to launch web UI, got this error. /bin/bash: line 1: ./webui.sh: No such file or directory. Also ngrok error I get dial tcp 127.0.0.1:7860: connect: connection refused. Is it related to google ban? do you know where I got wrong?

Furkan Gözükara

please remove ending \ and try again and let me know

Furkan Gözükara

remove the very end \ and try again and let me know please

John Ceed

Launching Web UI with arguments: -f --xformers --ckpt-dir /kaggle/temp/models --enable-insecure-extension-access --no-half-vae --lora-dir /kaggle/input/my-loras \ usage: launch.py [-h] [--update-all-extensions] [--skip-python-version-check] [--skip-torch-cuda-test] [--reinstall-xformers] [--reinstall-torch] [--update-check] [--test-server] [--log-startup] [--skip-prepare-environment] [--skip-install] [--dump-sysinfo] [--loglevel LOGLEVEL] [--do-not-download-clip] [--data-dir DATA_DIR] [--config CONFIG] [--ckpt CKPT] [--ckpt-dir CKPT_DIR] [--vae-dir VAE_DIR] [--gfpgan-dir GFPGAN_DIR] [--gfpgan-model GFPGAN_MODEL] [--no-half] [--no-half-vae] [--no-progressbar-hiding] [--max-batch-count MAX_BATCH_COUNT] [--embeddings-dir EMBEDDINGS_DIR] [--textual-inversion-templates-dir TEXTUAL_INVERSION_TEMPLATES_DIR] [--hypernetwork-dir HYPERNETWORK_DIR] [--localizations-dir LOCALIZATIONS_DIR] [--allow-code] [--medvram] [--medvram-sdxl] [--lowvram] [--lowram] [--always-batch-cond-uncond] [--unload-gfpgan] [--precision {full,autocast}] [--upcast-sampling] [--share] [--ngrok NGROK] [--ngrok-region NGROK_REGION] [--ngrok-options NGROK_OPTIONS] [--enable-insecure-extension-access] [--codeformer-models-path CODEFORMER_MODELS_PATH] [--gfpgan-models-path GFPGAN_MODELS_PATH] [--esrgan-models-path ESRGAN_MODELS_PATH] [--bsrgan-models-path BSRGAN_MODELS_PATH] [--realesrgan-models-path REALESRGAN_MODELS_PATH] [--clip-models-path CLIP_MODELS_PATH] [--xformers] [--force-enable-xformers] [--xformers-flash-attention] [--deepdanbooru] [--opt-split-attention] [--opt-sub-quad-attention] [--sub-quad-q-chunk-size SUB_QUAD_Q_CHUNK_SIZE] [--sub-quad-kv-chunk-size SUB_QUAD_KV_CHUNK_SIZE] [--sub-quad-chunk-threshold SUB_QUAD_CHUNK_THRESHOLD] [--opt-split-attention-invokeai] [--opt-split-attention-v1] [--opt-sdp-attention] [--opt-sdp-no-mem-attention] [--disable-opt-split-attention] [--disable-nan-check] [--use-cpu USE_CPU [USE_CPU ...]] [--disable-model-loading-ram-optimization] [--listen] [--port PORT] [--show-negative-prompt] [--ui-config-file UI_CONFIG_FILE] [--hide-ui-dir-config] [--freeze-settings] [--ui-settings-file UI_SETTINGS_FILE] [--gradio-debug] [--gradio-auth GRADIO_AUTH] [--gradio-auth-path GRADIO_AUTH_PATH] [--gradio-img2img-tool GRADIO_IMG2IMG_TOOL] [--gradio-inpaint-tool GRADIO_INPAINT_TOOL] [--gradio-allowed-path GRADIO_ALLOWED_PATH] [--opt-channelslast] [--styles-file STYLES_FILE] [--autolaunch] [--theme THEME] [--use-textbox-seed] [--disable-console-progressbars] [--enable-console-prompts] [--vae-path VAE_PATH] [--disable-safe-unpickle] [--api] [--api-auth API_AUTH] [--api-log] [--nowebui] [--ui-debug-mode] [--device-id DEVICE_ID] [--administrator] [--cors-allow-origins CORS_ALLOW_ORIGINS] [--cors-allow-origins-regex CORS_ALLOW_ORIGINS_REGEX] [--tls-keyfile TLS_KEYFILE] [--tls-certfile TLS_CERTFILE] [--disable-tls-verify] [--server-name SERVER_NAME] [--gradio-queue] [--no-gradio-queue] [--skip-version-check] [--no-hashing] [--no-download-sd-model] [--subpath SUBPATH] [--add-stop-route] [--api-server-stop] [--timeout-keep-alive TIMEOUT_KEEP_ALIVE] [--disable-all-extensions] [--disable-extra-extensions] [--ldsr-models-path LDSR_MODELS_PATH] [--lora-dir LORA_DIR] [--lyco-dir-backcompat LYCO_DIR_BACKCOMPAT] [--scunet-models-path SCUNET_MODELS_PATH] [--swinir-models-path SWINIR_MODELS_PATH] launch.py: error: unrecognized arguments: \

Furkan Gözükara

you have to use ngrok and dont use share option. when share is enabled kaggle will terminate your session immediately. so start ngrok and use the new command in the notebook. after that go to the ngrok link you will see web ui there

Ran Bensimon

using the notebook, I don't get a shared gradio link using Kaggle, has something changed? It says "To create a public link, set `share=True` in `launch()`." I am using ngrok as well.

Furkan Gözükara

i think you can contact and tell them it was a mistake. they are very harsh on NSFW :/

j

I can confirm that Kaggle instantly locks your account if you generate a nude image. I did not even intend to do so, with prompt "man renaissance" Could anyone suggest what to do next? Should I contact their support, or are they against people using their service for image generation?

Furkan Gözükara

can you tell me exactly at which step you are having error? i just tested right now working. can you message me from discord and show me screenshot

Luc Valigny

hello following your message on utube , i took a patreon account to follow your way to connect to kaddle . I followed every step but ended with this error 'dial tcp 127.0.0.1:7860: connect: connection refused' opening the ngrok link at the end of installation . any ideas ?

Furkan Gözükara

i think it depends on many factors. i am still using and there are a lot of still using person.

C05MIC Studio

I’m hearing. Kaggle banning users who have SD notebooks without warning? Is this safe?

Furkan Gözükara

did you try to stop automatic1111 and start again? which model file are you using? it shouldnt happen. at worst i would stop and start session. also which gpu did you pick?

Jorge Reverte Sevillano

Hi! Having some "trouble". I ran the file as you explain in the video, but everytime I click "Generate" button, it does the "Creating model from config" again, so every generation is about 2 minutes instead of few seconds... Any clue?

Furkan Gözükara

you are welcome

Rafał Ryniak

Tanks

Furkan Gözükara

hello. updated the notebook and added that model. redownload notebook you will see it now

Rafał Ryniak

hey, how to add the upscaler path, it is in the notebook database, but where should I paste the path so that "hi res" can see it, "4x_NMKD-Superscale-SP_178000_G.pth"

Furkan Gözükara

you are welcome ty for comment

Rafał Ryniak

thank you very much, everything is clear, great job with Kaggle

Furkan Gözükara

roop extension not updated anymore and it requires some extra libraries. you can use runpod or google colab. you can upload model files via your own dataset or wget into correct folder. which upscaler you need? i can add to notebook