Not loaded
Download
1 variant available
bf16 SafeTensor
moss_context_detaile_tweaker_20261018_rank2.safetensors
BF16, good balance • 8.94 MB
Verified: a day ago
3040 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9
00 1 2 3 4 5 6 7 8 9
(38)
Oct 18, 2026
MOSS Image Base
context beta v1,Experimental parameters, unstable effects
After testing, it has been found that the base model and the use of 8-step dmd distillation of lora can still maintain their effectiveness
Show more

6430 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9
4.8K0 1 2 3 4 5 6 7 8 9.0 1 2 3 4 5 6 7 8 9K
74.5K0 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9.0 1 2 3 4 5 6 7 8 9K

License:
Apache 2.0Detail slider lora for Moss-Base
Adjust image details while minimizing the impact on the original image's visual style and composition.
context_detail_tweak: More emphasis on adjusting content details
rendering_detail_tweak: tobecontiune
background_detail_tweak:tobecontiune
Works with weights [-0.75, 0.75]
Use positive weight to increase details and negative weight to reduce details.
Known Issues:
Excessively high positive LoRA weights may increase image contrast and brightness, while excessively high negative LoRA weights may produce the opposite effect.
Since the LoRA was trained on anime-style images, it performs better when generating anime-style images.
After testing, it has been found that use of 8-step dmd distillation lora can still maintain their effectiveness.But it has not been tested on Moss-sfw yet.Using an 8-step DMD distillation LoRA simultaneously may cause slight changes in the subject of the image.

,yellow hair,evil horns,red horns,_s.jpeg)
