
While playing around with a few character designs, I realized that Stable Diffusion generally doesn't do a great job with charm bracelets. This LoRA attempts to fix this. This was trained on around 70 images, mostly consisting of gold and silver charm bracelets on people's wrists (as well as a few plastic ones).
The training data also several examples of charm bracelets on tables, as well as bracelets on darker skinned people, so it should be fairly versatile. There are no NSFW pictures in the training data, so it should also play nice with most prompts.
Use charm_bracelet to trigger the LoRA. I've experimented with different weights, and it seems to perform reliably in the 1.1 to 1.5 range. This will vary by model, of course, so let me know if you run into something weird.
描述:
训练词语: charm_bracelet
名称: 229411_training_data.zip
大小 (KB): 11168
类型: Training Data
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
名称: charm_bracelet.safetensors
大小 (KB): 36982
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success