each models loaded fully into vram
each models loaded fully into vram
Flike flux sd etc
Fso we have exact replicate
Tyeah, but how the state of each model loaded into GPU sync up?
TSo the training result in single model, not multiple models
Fthey are snyched at steps
Fso it is being increased batch size
Nlong time no see. I had trained few flux loras earlier, and had to reinstall my windows recently. After that I had to reinstall kohya, and now can't train anymore. First it didn't create venv (idk, maybe it's because latest installer or something else). I've installed kohya from original repo, and it started to work. But now I have another problem. Training speed reduced. Much. ~500s/it. I guess it's because it's using ram or something. Earlier I noticed it's caching t5 and clip-l to ram (to free vram), now there is no such notice. Any suggestions?
Ngpu 4060ti 16gb, ram 64 gb
N
NI'm using the same configs I've used earlier
Nit's gpu fallback turned on, because when I'm turning it off, I'm getting out of memory error
Fwhat is your gpu and which config you are using?
Fgive me exact name pls
Ngpu 4060ti 16gb, ram 64 gb
Fok use this config 12GB_GPU_11100MB_9.5_second_it_Tier_1.json
Falso install latest nvidia but make a clean installation so all configs reset
Nnvidia driver?
Nit's dreambooth training or lora tab?
Fthis is dreambooth
Fyes nvidia driver
Ffor lora you can use either of these
F12GB_GPU_Quality_Tier_4_FP8_11050MB_5.5_Second_IT.json
F16GB_GPU_Quality_Tier_3_14700MB_9.2_Second_IT.json
F16GB_GPU_Quality_Tier_2_14800MB_13.4_Second_IT_and_T5_Attention.json
Nthat's my config I've used earlier and it worked fine with speed ~16s/it
Fwell kohya made a lot of updates so our updated configs working
Fcant tell for older configs
Nyep, I guess it's because I've used v19 of your installer and it worked fine for me)
Fye it is old.
Fwe 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.
