[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]

/vt/ - Virtual Youtubers

Search:


View post   

>> No.54707918 [View]
File: 3.19 MB, 640x744, 1690556978936273.gif [View same] [iqdb] [saucenao] [google]
54707918

>>54700663

>>54706688
I go from 512x600 and can up it to 1024x1200 but I'm still experimenting since there are post size limit here so I also take that into consideration. The three I posted are all different sizes I think and quality on up scaling can be hit or miss. I'd say don't go too high to save time as sometimes higher doesn't really increase the details.
The model is the same model but without a baked vea, and I use the same prompt I used for the gif, maybe I got lucky and got a model that does lines good. Really gotta go to the drawing board and experiment and write all this stuff down I guess, well that is what I'm doing, for benchmark times, upscaling size/times and such.

>> No.54451304 [View]
File: 3.19 MB, 640x744, gif2gif-0007.gif [View same] [iqdb] [saucenao] [google]
54451304

>>54449511

>>54449931
NTA, looks great
What goes on in that birdbrain of hers? Or maybe she can't move her body and it's just nothing but "Fucking Bitch!" echoing in that head
>>54450684
>Whoa that is cool, I just found a colab for animatediff so I may try out stuff soon.I think I posted a picture for the supermerger a little while ago let me see if I can find it
Thanks, I can't bake too high res in animateDiff standalone, trying to find a good denoising value when I gif2gif upscale so it's more smooth yet adds a bit more details
Thanks for taking the time making the doggo LoRA I'll test them out
>>54450668
I've used oral invitation and uvula, but I also have the oral invitation LoRA to get the inner mouth anatomy right

Navigation
View posts[+24][+48][+96]