we updated after re-testing after all changes of kohya
we updated after re-testing after all changes of kohya
Nhope changes were useful)
Fyep they work great
Nyep, used your 16GB_GPU_Quality_Tier_2_14800MB_13.4_Second_IT_and_T5_Attention.json config and speed is back to previous (even bit faster). Thanks! Tried to start dreambooth training with ok use this config 12GB_GPU_11100MB_9.5_second_it_Tier_1.json and got 20+ s/it. Maybe it's first few epochs.
Fwait it will get faster
Fwait like 50 100 steps to see actual speed
N
Tbtw, I noticed that kohya calculates step speed incorrectly, including intermediate checkpoint save times.
So each time it saves calculation getting more and more off
So each time it saves calculation getting more and more off
Tlora save time is very fast, but for full checkpoint around 55 seconds on massed compute
Awhy's that?
Tprobably uses simple formula dividing training running time by complete steps count
Tsaving pauses the training and messes up calculation
Fyep not very accurate
Fostris said he gonna make a gui
Fi plan to move it
Fhe said he will full time work on ostris 
Fnot like kohya
WHi @Furkan Gözükara SECourses are there anyway to convert back fp8 model into fp16 again and get the better quality of fp16 or at least make the model work like fp16 and keep the quality like fp8
�Why don't you try simpletunner?
AI've been using config files from KOhya_FLUX_DreamBooth_v17, after installing the latest version of kohya_ss for the sd3_flux branch, and my Flux dreambooth sessions on an RTX4090, which used to sit at about 23.9 GB in VRAM, and they'd run fine, just don't sit within the VRAM 24GB size anymore. I've been fiddling with block_swap size, but 30.7 GB seems to be the best I can get for memory usage with at least some iterations happening. Is this a known thing? It could be that I've set some parameter somewhere when setting up kohya_ss again.
AI even tried rerunning a previous .toml script for dreambooth, on the command line, but it didn't repeat the previous VRAM experience of using under 24GB.
AIt isn't super-slow, by the way, running at 5.00 - 7.00 sec/it.
CHi everyone!
@Furkan Gözükara SECourses can you give me a hint?
I have this error:
RuntimeError: CUDA error: out of memory
"CUDA kernel errors may be reported asynchronously by some other API call, so the stack trace below may not be correct.
For debugging, consider passing CUDA_LAUNCH_BLOCKING=1"
I tried 8gb and 3080 10gb config
I have 48gb RAM and 10gb VRAM.
What can help me?
@Furkan Gözükara SECourses can you give me a hint?
I have this error:
RuntimeError: CUDA error: out of memory
"CUDA kernel errors may be reported asynchronously by some other API call, so the stack trace below may not be correct.
For debugging, consider passing CUDA_LAUNCH_BLOCKING=1"
I tried 8gb and 3080 10gb config
I have 48gb RAM and 10gb VRAM.
What can help me?
C@Furkan Gözükara SECourses Hey. I saw your yt community post, and wondering will there be a tutorial coming for sd3.5? maybe on patreon post early.
Fnope all data lost once you convert into fp8
Fi dont plan to invest time into that one
Fthis can happen
Fif you loaded lora into dreambooth
For dreambooth into lora
F5 second it is good
Ffor flux dreambooth
�Way better than ai toolkit btw
Fhi when you have it?
Fyou are out of VRAM
Fi dont like the developer, the guy attacked me so many times without any reason
Fcould be currently i am doing research for a client
�Yes he is rude sometimes but a great developer and he very well documented everything and very active in discord servers.
Cnice looking forward to this, since flux have its own disadvantages, im still using sdxl. hoping sd3.5 will amaze me.
Ccan we use regularization imgs with sd3.5?
Fi think people said same as flux but havent tested yet
Fwill test
C
Clooking forward to this
Clooking forward to this
SWould it be possible for you to perhaps share any results? I am also looking into simpletuner and would love to see if someone has achieved some reasonable results. Would really appreciate it, I won't share ^^
�I have not yet tested it but I plan to do that next month. I have heard many people talk good about it. Here is a video that goes deep into it and compares it with ai toolkit - https://youtu.be/se3qpLkJnrk?si=7MhyT8i3DALGK9L5
YouTubekasukanra
#flux #simpletuner #bghira #ComfyUI #LoRA #finetune #stablediffusion #style #styletraining
This is an FLUX.1 [dev] LoRA training log for art style and concepts. I document my thought process, experiments, mistakes, and analysis of quantitative and qualitative results. Hopefully, this can be a good starting guide for people looking to train LoRA...
This is an FLUX.1 [dev] LoRA training log for art style and concepts. I document my thought process, experiments, mistakes, and analysis of quantitative and qualitative results. Hopefully, this can be a good starting guide for people looking to train LoRA...
SSo your opinion "Way better than ai toolkit btw" is based on a video? You haven't tried it yourself yet?
SInteresting
�See the video you will know. I have done a lot of trainings with ai toolkit btw
Ncan some explain how to reduce flux lora size with no (almost) quality loss? Because I've made bunch of them and 1gb loras is eating my storage) or maybe there is video with explanation?
Nnever mind, found in patreon posts
