LoRa Training
So i recently got into LoRa training and i am making a character but the trained LoRa does some stupid things.
Proof1 community thread
Key points
- First of all today i trained one on 10 epochs with a local batch of 2 and a balancing of 10 repeats.
- So i trained it using onetrainer and it was done and i put it into comfyUI.
- Then i generated it but it did not look like my character at all how can i fix this and make a decent looking LoRa.
- This was my prompt and i have no idea how it can turn out like that: mycharcamille, standing, looking at viewer
Sources (1)
- [1]LoRa Trainingr/StableDiffusion (top, day) · Oct 18, 12:47 PM
So i recently got into LoRa training and i am making a character but the trained LoRa does some stupid things.
First of all today i trained one on 10 epochs with a local batch of 2 and a balancing of 10 repeats.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 18, 2026Same prompt, 3 models: SDXL (WAI-Illustrious v17) vs Krea 2 turbo vs MOSS-Image-SFT (anime scenes, no LoRA)
- Oct 18, 2026MOSS-Image comes to Fizgig
- Oct 17, 2026Blown away by Minimax H3 sometimes. It nailed this one!!!!
- Oct 17, 2026MOSS-Image LoRA training is now in AcademiaSD LoRAlab Trainer Studio (Base and SFT, under 6 GB VRAM)
- Oct 16, 2026Comfy-Org/MOSS-Image
- Oct 15, 2026New TaoMate H3 FL2VA LoRA is out, here's a ComfyUI conversion