Basics of using Lora models in ComfyUI

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
[Music] ziki guides you through the process of using Laura models and comfy [Music] UI hi my name is ziki today let's dive into incorporating and Lowa model into your workflow it's all about leveraging its power seamlessly ensure you obtain the necessary components research its functionalities and align your workflow for Optimal Performance stay tuned for more insights and updates let's quickly bring comfy up to the latest version using the manager if an update is required close the console refresh your browser and then restart simple as that I'll step out of the picture from now on allowing you to enjoy an unobstructed view let's utilize the model we recently downloaded simply select it here for an extra Touch of flare start by creating a cool image of my friend a stylish Gothic lady without the lower model for now let's capture that unique vibe let's choose a seat and lock it in ensuring we can easily spot the differences later on all right let's kick it off push the button and let the magic unfold if the image isn't quite to your liking go ahead and tweak the seed let's see if the new seed brings the image more to your taste f fantastic choice with the seat adjustment all right it's time to unleash the magic of Laura we need the Laura loader now you can load it right [Music] here now let's connect the noodles model to model and clip to clip easy peasy ah we need a Lowa model or two don't we let's head over to civit Ai and find a Laura model that complements a 1.5 [Music] checkpoint excellent choice detail treer is indeed a fantastic Laura model simply download it now now copy that thing into the right directory ui/ model [Music] Laur to showcase the process more vividly I'll grab the Chinese painting Masters model that live in Ming and Ching dynasties of course copy it into the same directory as well [Music] one quick refresh and let the show begin select Laura press the button and let's embark on a new creative [Music] Adventure ah I should input a keyword so that the model knows what to do these keywords are specified on the CID AI website and a fresh attempt begins oh that was the first try and zigi's a bit annoyed with these values my bad let's not let that slip again another shot and this time we'll nail [Music] it seems like it was a bit on the low side unless that's the vibe we're going [Music] for wow much better already now let's fine-tune the strength of the Lura model or let's go all in and incorporate the detailed tweaker as well simply press the ALT key copy and connect it with the noodles select the Lura model and off we [Music] go here we're already witnessing the [Music] enhancements quickly double the preview node by pressing the ALT key drag it aside rewire the noodles and the original image stays for comparison now I forgot to adjust the strength of the Laura model no worries you can see it generates the exact same image the red border indicates an error which you can also spot in the console however we want wanted the preview not to receive new information so all is good let me tweak the Lura values and give it another [Music] shot Wooh hooo check out the ear or the details on the [Music] headband let's crank up the value even higher and there we go even more detail at some point though it might be a case of too much of a good thing I recommend checking out civit AI for tips and suggestions from the creator of the Laura model and let's give it a go right [Music] now wow that's already [Music] amazing now let's try the opposite and reduce the [Music] details absolutely brilliant this is working phenomenally well I'm genuinely impressed with how seamlessly we can control and adjust the details it's like wielding a digital paintbrush crafting the perfect level of intricacy hats off to the creators for such an outstanding tool the possibilities are truly [Music] exhilarating next up I'll introduce you to control net ah but in the next video or should I connect it right away with Murphy's new sdxl workflow well a break is needed not for me but for Murphy until then Yours Truly Ticky
Info
Channel: Murphy Langa
Views: 562
Rating: undefined out of 5
Keywords: AI, KI, ComfyUI, comfy, stable diffusion, murphy langa, workflow, Stable Diffusion, stablediffusion, LORA, Lora Model
Id: k4ZEFYPCRgw
Channel Id: undefined
Length: 6min 51sec (411 seconds)
Published: Sat Jan 13 2024
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.