Moss detailer Lora built with Lora dataset studio
I trained a MOSS detailer LoRA with LoRA Dataset Studio
Proof1 community thread
Key points
- I managed to train RAW Upscale MOSS, a detail and texture LoRA for MOSS-Image, using LoRA Dataset Studio.
- These are text-to-image renders, with only the strength labels added afterwards.
- Acceleration adapter: mossimagesft4stepv2.0.safetensors, at strength 1.0 in every image, including the baseline.
- Training settings: MOSS-Image-Base, 2,000 training steps, rank 16 / alpha 16, 1024 training resolution, final checkpoint exported as FP16 SafeTensors.
Sources (1)
- [1]Moss detailer Lora built with Lora dataset studior/StableDiffusion (top, day) · Oct 21, 09:05 PM
I trained a MOSS detailer LoRA with LoRA Dataset Studio
I managed to train RAW Upscale MOSS, a detail and texture LoRA for MOSS-Image, using LoRA Dataset Studio.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 21, 2026LoRA Training for Fibo 2.0 Added
- Oct 20, 2026MOSS-Image accepts image inputs!
- Oct 20, 2026Fizgig 8.8.1: a new MOSS-Image training technique
- Oct 20, 2026How to train MOSS-Image LoRAs locally (8 GB VRAM): settings, tips and workflow
- Oct 20, 2026[AINews] DeepSeek DeepInfer: Open Source SOTA Inference running 155T toks/day/cluster making $22m/day at ~80% margin for a 81 day payback
- Oct 19, 2026Moss image speed boost with spectrum moss node
