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
@Dr. Furkan Gözükara So I've been checking out the Batch size 7 config. And It seems that your Learning Rate is 0.00001 which is 2.5x higher than when batch size is 1. What made you choose that Learning rate?
I used the "flux-dev-de-distill" model for Dreambooth training, and the resulting file is very large. Is this because I haven't extracted the LoRA from it yet?
I assume most everyone here has watched the Behind the scene from Corridor Digital on their rock paper scissors Anime. If not here it is.. https://youtu.be/_9LX9HSQkWo?feature=shared&t=675 Does anyone know about custom models in SDXL? I am curious.. When training their model, does he mean he trained a model on him and his costume, then used that model as the base to train the style into that same model?