"Breathe In, Gen out."
- anon 5/4/2024
For any questions not answered here, refer to OP. Reply to OP with any suggested changes.
Contents: re/Forge ComfyUI AMD Rentry from Other Boards
re/Forge
https://github.com/lllyasviel/stable-diffusion-webui-forge?tab=readme-ov-file#installing-forge
or
https://github.com/Panchovix/stable-diffusion-webui-reForge?tab=readme-ov-file#installing-forgereforge
ComfyUI
https://github.com/comfyanonymous/ComfyUI?tab=readme-ov-file#get-started
No.103872552
>Is this [checkpoint]?
yes
>I haven't used it but I would like to make pictures like this.
check if [your GPU can run it. For Flux you need a more powerful one, for XL less powerful, and 1.5 even less]
if "no" then upgrade
if yes
>download comfyUI windows portable [git clone the repo for linux]
>install the comfyUI manager [ https://github.com/ltdrdata/ComfyUI-Manager ]
>download [checkpoint] and put [checkpoint] into the right folder [models/checkpoints/]
>download lora and put lora into lora folder [models/loras/]
>load workflow or make one yourself
>hit queue
>wait
>done
No.103872601
NTA, but just use the examples found here
https://github.com/comfyanonymous/ComfyUI_examples/
AMD
https://github.com/AUTOMATIC1111/stable-diffusion-webui?tab=readme-ov-file#installation-and-running
No.104275771
You guys are assholes who don't help newbies. Yesterday I asked "how do I run txt2img on amd on windows?" and I received no replies. Here's what would've saved me 8 hours of time:
>Sorry anon, but AMD has abysmal support for AI. Rocm, their cuda equivalent, is officially supported for only a couple cards and only on linux.
>There is a rocm port in progress for windows, but no ML library supports it yet, so it's useless. If you insist on using windows, which is a terrible idea, you can use directml instead (awfully slow and memory inefficient) or zluda, a cuda "emulator"
>On linux you'll most likely find that your device isn't officially supported by rocm. Unofficially, many gpus will work with no issue. Often, all that is required is to simply set an environment variable. See https://salsa.debian.org/rocm-team/community/team-project/-/wikis/Supported-GPU-list and https://github.com/ROCm/ROCm/issues/1756#issuecomment-1159603527
And don't you "just read the guide" me! Why the fuck is the guide, first link the OP, empty and worthless? Why is it that the actual guides are hidden in the "Archive" link, which is just a link dump with no attached explanation for anything? I mean, some of it is useful, but if your plan is to ignore newbies outright, then shouldn't there be something resembling an actual introduction somewhere?
Rentry from Other Boards
/hdg/ Master Rentry: https://rentry.org/hdg-collab-master
Lora Training: https://rentry.org/59xed3
CPU Tard: https://rentry.org/cputard
Useful Extensions: https://rentry.org/8csaevw5
/u/
Illustrious LoRAs: https://rentry.org/illustrious_loras_n_stuff
4chan Catbox Userscript: https://rentry.org/hdgcb
/aco/
Highres Fix: https://rentry.org/hiresfixjan23
Inpainting: https://rentry.org/inpainting-guide-SD
Training Characters (embedding): https://rentry.org/simplified-embed-training
/c/
OP Paste: https://rentry.org/kawaii-diffusion-general
ComfyUI Install: https://rentry.org/comfyui
AMD GPU: https://rentry.org/sdg-link#amd-gpu
/d/
Wildcards: https://rentry.org/NAIwildcards
/e/
General: https://rentry.org/edg-faq
Animation: https://rentry.org/AnimAnon-AnimDiff
OP Paste: https://rentry.org/sd-mashup
Lora Training: https://rentry.org/lora_train
/vt/
Getting Started: https://rentry.co/5twfkqmw
General: https://rentry.org/vtai/
FAQ: https://rentry.org/vtaiFAQ
LoRAs: https://rentry.org/vtaiSDXL
LoRA Checklist: https://rentry.org/vtailorachecklist
Anime on Furry Models: https://rentry.org/5exa3
Model Cookbook: https://rentry.org/vtairecipes
Prompt List: https://rentry.org/vtaiprompts
Catboxes: https://rentry.org/fhydhwdm
/hdg/ LoRA Repo: https://rentry.org/hdglorarepo
Paperspace: https://rentry.org/865dy
/b/celeb/
General: https://rentry.co/CelebAI
Training: https://rentry.co/machineminded-biglust-training-and-loras