Back to Project
M

MOSS-Image LoRA experiments

Project Oct 17, 2026

SimpleTuner's public journal comparing MOSS-Image Base and SFT concept-training runs, with recipes, adapter checkpoints and validation previews.

Visit primary link 2 sources attached
Author-generated intermediate MOSS-Image SFT LoRA validation showing a white astronaut suit on the moon with Earth behind it and the original prompt strip below.
Generated validation from the in-progress SFT-1024 run at update 5,500, not the completed 512px experiment. Full frame and author prompt strip preserved. This is not an AITrending reproduction or proof that the target concept has been learned successfully. Source: SimpleTuner / MOSS-Image LoRA experiments
multimodalself-hosted

What happened lately

MOSS-Image LoRA experiments timeline

This community experiment publishes recipes and adapter sequences for learning a concept with MOSS-Image Base and SFT. Its October 17 snapshot lists the original-caption 512px runs as completed, the 1024px runs as training, and two JSON-caption runs as queued.

The cards declare Apache-2.0 licensing and record the dataset revision and validation settings. The journal provides an inspectable training trail, not an independently verified quality benchmark or a completed verdict on which checkpoint or captioning method works best.