Hello everyone. I am Dr. Furkan Gözükara. PhD Computer Engineer. SECourses is a dedicated YouTube channel for the following topics : Tech, AI, News, Science, Robotics, Singularity, ComfyUI, SwarmUI, ML, Artificial Intelligence, Humanoid Robots, Wan 2.2, FLUX, Krea, Qwen Image, VLMs, Stable Diffusion
so.. I've been training on massed_compute recently (A6000) but thought I'd give my 4090 a test. I'm actually quite suprised at the performance using the 24GB_GPU_23150MB_10.2_second_it_Tier_1.json
I tried training one on a particular person, and the sample image during training looked perfect, but when I actually use it with flux, it just makes generic people. Is there something else I might've forgotten to do?
Upon finishing the training, can I simply move the safetensor files to the lora folder, and generate an image with them using the trigger word? or are there steps I'm unwittingly skipping?
Ultimate Kohya GUI FLUX LoRA training tutorial. This tutorial is product of non-stop 9 days research and training. I have trained over 73 FLUX LoRA models and analyzed all to prepare this tutorial video. The research still going on and hopefully the results will be significantly improved and latest configs and findings will be shared. Please wat...
@Furkan Gözükara SECourses Did you ever do a rank test on Lora Extraction? We know 64 is max, but 63 is only 1.2GB, is there really a difference? 64=12GB 63=1.2GB
Personally, I have only done a small test, but with the wrong model, so it was hard to tell. I am training a new one at the moment, and will test again
Hi @Furkan Gözükara SECourses I wonder if I choose the same epoch to save and use. Are the quality of the same epoch checkpoint when training with 200 epoch and 100 epoch with same learning rate equal or different ? For example I use checkpoint saved after 80 epoch trained one from 200 epoch training trained other from 100 epoch trained
Thank, because after training with your 200 epoch those checkpoint with more than 120 epoch become overtrained so I just want to reduce my training time ?