they are used during inference. not during training
they are used during inference. not during training
F1024x1024 better for sdxl
S@Dr. Furkan Gözükara any advice for best batch upscaling?
EI'm just trying to learn how to do this on python instead of using a ui. Are there any huggingface loaders that allow this or do I need to look at kohyas scripts?
SShould I use 1024x1024 instead? should I add negative prompts before start the training ?
Fthere is controlnet upscale with hd upscaler maybe that one
FYouTubeSECourses
I show how to install Automatic1111 Web UI & ControlNet extension installation from scratch in this video. Moreover I show how to make amazing QR codes and inpainting and out painting of ControlNet which are very similar to Photoshop generative fill and Midjourney zoom out. Furthermore, I explain and show what are Canny, Depth, Normal, OpenPose...
F1:09:00 How to upscale and add details to the images with inpainting + tiling
Fsdxl train py recognizez it
Fi am about to update patreon post
F
F
Fnot it is using shared VRAM
Fwhat is your gpu? did you install torch 2?
Fwhat training you are doing? kohya lora dreambooth?,
Fye this is correct. it must be settings. what gpu do you have?
Sthat one is image to image, i was thinking straight from the extras tab so i can scale them up to 4k
Fwell you can try different upscalers
Fthere are too many
FOpenModelDB
OpenModelDB is a community driven database of AI Upscaling models. We aim to provide a better way to find and compare models than existing sources.
Shaha exactly i was hoping you had done some testing already but i can play around with them
Si just don't know what all the setting mean eithe ras far as upscaler, GFPGAN, and codeformer visibility,
Sor the codeformer weight
Fthose for fixing face
LRtx 3060 gaming x, i used your script to install sd and dreambooth
LDreambooth using your script to install it
Scurrently training a model of my wife using 220 dataset and 0 class images, i'll let you know how it comes out.
Sgoing to take a few hours lol
Fhow much vram it has?
Lit has 12gb
Si also made a slightly more realistic merge of the model for my dog.
Fwith 12 gb you can do SDXL lora
Fbest option for you
Fsd 1.5 dreambooth perhaps with optimization but quality i dont know need testing
Fcan you show your settings? configuration
Ldo you think that i will get bettwe result with a 24gb gpu on runpod?
F100%
Fwith dreambooth of SDXL
FLi dont know how to use SDXL, thats why im training for sd 1.5
Ldo you have a video how to train model with dreambooth for sdxl?
Fi have for lora
Fyou only need to load config
Ffrom patreon
Fand prepare dataset nothing else
FYouTubeSECourses
If you don't have a GPU, or have a strong GPU, or you are using Mac and your computer not supporting Stable Diffusion training, SDXL training, then this is the tutorial you are looking for. In this tutorial, we will use a cheap cloud GPU service provider RunPod to use both Stable Diffusion Web UI Automatic1111 and Stable Diffusion trainer Kohya ...
Fafter watching this video you can load config and use it
Fi will also hopefully record quick video for patreon
Fcurrently fixing kaggle notebook