I have 70 images in my data set to create my realistic instagram ai model, how many epochs would you
I have 70 images in my data set to create my realistic instagram ai model, how many epochs would you recomend to use when fine tuning?
Mperhaps 100?
MI know that Dr Furkan said that best epoch is 40 when he did 256 images, and 140 when he trained 15
MBut I am still unsure hah
Ffor 70 images go up to 150 at least
Fand compare
MI have 58.6 GB of free space, will that be enough to download the SUPIR upscaler
Mim done with my training, so I wonder can I delete kohyaa to free up more space?
Cnope @mlado
Moh ok
Fye you need more. especially if that is your virtual RAM disk as well
MOkay
MThank you
�Hey. My sdxl fine tune did not capture my body right. Used reg images. Now I'm thinking that since I had 25 body pics and 15 close ups of face, using ohwx woman only as tag on all pics...maybe all the close ups of face confuses or makes the fine tune forget my characters body, and is more influenced by bodies of women in reg images. Any knowledge of that?
Fye sdxl is weak
Fuse flux
�https://civitai.com/models/1223425?modelVersionId=1378381 claims to have trained on flux schnell and used high resolution images, bypassing the limit of 1024px. Hmm. Can this be true, Dr. @Furkan Gözükara SECourses ? It's a sfw fine tune. No worries 
TThat was initially released on F1_Dev, no clue why they switched to F1_Shnell
Fenabling the model to produce very fine details during tiled upscales that can hold up even in 32K upscales.
Fprobably overfit for some classes and may work
Fnormally my fine tuning of flux performs excellent with latent upscale 2x = 4k
Fi would say best to compare for your usage task
Flet me add it to my downloader
�You mean it would be possible to finetune on his model? Or just for generations?
TWHat's up everyone
Thow much would it take to extract a lora from a finetune? need it for FLux fill
Fyou can fine tune and also generate
Fi added to model downloader we have
Fflux fill is different only can inpaint
TYes and it only can take loras not finetuned models, ao if theres a dev flux model I usually use, I want to extract a LoRA out of it to use it with Flux Fill
UYouTubeuisato
Updated system for audioreactively generative geometries, intervened with various AI techniques.
More experiments, project files, and tutorials, through: https://www.uisato.art/
#touchdesigner #aiart #audioreactive
___
0:00 - Stained Glass → Occultist Drawing → Map
0:13 - Experimental Notation → Pencil Drawing → Old Map
0:26 - Glass ...
More experiments, project files, and tutorials, through: https://www.uisato.art/
#touchdesigner #aiart #audioreactive
___
0:00 - Stained Glass → Occultist Drawing → Map
0:13 - Experimental Notation → Pencil Drawing → Old Map
0:26 - Glass ...
Fcool
MI just did fine-tuning and am now getting 'AssertionError: You do not have CLIP state dict!' Does anybody know what could be the cause?
Fdo you have clip downloaded accurately?
Ffine tuned model is only unet
Mi am not sure. I think i will start over and see how it goes.
Fdid you use our onetrainer config?
Fyou can use kohya extract lora feature
Fi recommend dreambooth and extract lora
Fso dreambooth also bad quality?
Fok that is good
Fso all you need to do is extract lora and use on same base model
Fof course you can 
Fonetrainer generated model file same as kohya
Fyou have your safetensors right?
Flike 6 7 gb
Fyep
Fbecause onetrainer has EMA
Fit improves
Aso doctor
Ais it better to always tirain on dreambnooth