ah you already wrote in patreon it is slower
ah you already wrote in patreon it is slower
Fye a little bit more slower
Flike 1.5 second / it to 1.75 second / it
SThank you so much to @Dr. Furkan Gözükara and @Sandstormleader after several hours of training and solving some errors the model came as I expected this is definitely a game changer process
MI have now made a perfect Lora with Prodigy, but you have to use Clip Skip 2 when generating and then it will be flexible.
Finteresting never tested clip skip 2. this is sd 1.5?
MYes, I trained with RealisticVision 5.1, and I generated the images in the same model.
MIf I generated on clip skip 1, the model was inflexible, but on clip skip 2 everything was fine.
And I thought right: I trained it with clip skip 2 and then clip skip 1 can be generated well.
And I thought right: I trained it with clip skip 2 and then clip skip 1 can be generated well.
MSo now the SDXL training works properly with TE, but what's up with 1.5 dreambooth?
Rwhat is the best/fastest gpu deal on Runpod to do inference on Auto1111? A100 or A6000? I find A6000 very slow, slower than 4900...
Fyes but using kohya
Fi think sd 1.5 dreambooth also should work but need good hyper parameter testing. also kohya missing EMA 
F4090 could be fastest with price ratio
Rok, yea, that's what i thought. a6000 takes twice as long to infer
Kuse "runpodctl" to transfer big Data between your desktop and runpod https://github.com/runpod/runpodctl
GitHub
| RunPod CLI for pod management. Contribute to runpod/runpodctl development by creating an account on GitHub.
| RunPod CLI for pod management. Contribute to runpod/runpodctl development by creating an account on GitHub.
TI tested 4090 with Kohya on Windows 11 and it was 1.89 it. On Windows 10 it is 1.06 it. Seems like there is a huge driver issue with W11
MI'm speaking about kohyaSS 1.5 dreambooth training. Can I train SD1.5 with TE?
Fuploading hugging face is even faster
Fit works too slow on windows
Fdriver issue
Fyes we found you can
Fbut it is inferior
TBut I'm using it on Windows 10 and it works really fast. Seems like Windows 11 is the problem
Fthis is weird
Fso windows 11 is the culprit?
Fi am still using 10
TYes, look. This is on Windows 10. I tested the same hardware with Windows 11 on a partition of the same SSD and I got 1.89s / it
TAnd also tested Automatic 1111 on Windows 11 and the generation time was terrible
Tit's not worth updating the OS
Fhuge difference
Fthat is why i am still on 10
Fit works super good
Sis there anything like runway gen2 inside of stable diffusion?
S@Dr. Furkan Gözükara
LIf i train my images 1024 x 1024 on dreambooth, will my model have better quality than generated with 512 x 512? Does that also mean that i will have to generate my images 1024 x 1024 with my model?
Si had better luck with 768 but was probably my base model not being great for 1024
LCan i train 768 on a 12gb gpu?
LFor now i am doing training with 512px
Mrtx 3090 vs 4090 24gb? extra cuda cores worth the extra cost?
Mthanks for info. just upgraded to 11 and moving back now..
Kstable diffusion or dreambooth ?
S@Dr. Furkan Gözükara By chance are you planning or you already have a tutorial of the importance and explanation of the image regularization data? how this works also with the caption description?
MI'm not sure but I think you can. I'm able to train 512 on 8GB so 12GB should be just enough for 768 (gradient checkpoint ON). Also my experience is if I train (lora) in 768 I get good results on 768 images but worse results on 512 compared to a 512 training.
SI train on 1024 x 1024 with runpod
Zroop keeps saying it can't detect my face, what's the easiest way to put a white padding box around my face?
Zi've tried like 10 different pictures, and everytime "Face not detected"
Zeven when my face is like centered in the image
Rhey there Dr. Furkan! its me again, i was looking for some guidance.
i was hoping to restore a couple of pictures from my parents days which are a bit low resolution with little detail in their faces and the pics , i was wondering if it'd be possible to maybe upscale them, bring more detail into the picture and print em out to frame them using SD in some way since every other ai based upscaler i tried just plain out gave terrible results.
Do you think you can suggest me a cheap platform where i can run stable diffusion and mess around to get this task done, if its even possible?
second, if its possible do you think you can point your hand at any workflow for me to get this done since i am pretty new to this stuff. it'd really mean a lot to me
i was hoping to restore a couple of pictures from my parents days which are a bit low resolution with little detail in their faces and the pics , i was wondering if it'd be possible to maybe upscale them, bring more detail into the picture and print em out to frame them using SD in some way since every other ai based upscaler i tried just plain out gave terrible results.
Do you think you can suggest me a cheap platform where i can run stable diffusion and mess around to get this task done, if its even possible?
second, if its possible do you think you can point your hand at any workflow for me to get this done since i am pretty new to this stuff. it'd really mean a lot to me
MIs it possible that the kohya does not multiply the captions but only the images? In Auto1111, when I look at the Lora's data sheet, it only uses them all once based on the number of tags, even though the weight of the tag counts, right?