![[FLUX Dev] Ugly Things版本v1.0 (ID: 710173)](https://image.1111down.com/xG1nkqKTMzGDvpLrqFT7WA/d4510131-9a77-4796-805d-d8e7748fb515/width=450/23552063.jpeg)
This model gets rid of that awful "pretty aesthetic" that Flux is known for.
This can be used to greatly change the outputs of the model. It can be used as a negative LoRA as well, to improve the outputs greatly, which is the primary purpose of this LoRA.
Feel free to merge this into anything and do whatever you want without crediting me. In fact, never bring my name up
Trained using x-flux trainer for 300 steps.
Rank = 16, Alpha = 16.
描述:
-
First version of the ugly. Trigger word to make things exponentially worse: julia
-
Rank/dim 16
-
Alpha 16
-
DO NOT USE CFG TO GET THE CORRECT RESULTS
-
Guidance = 4
First version of the ugly. Trigger word to make things exponentially worse: julia
Rank/dim 16
Alpha 16
DO NOT USE CFG TO GET THE CORRECT RESULTS
Guidance = 4
训练词语:
名称: pytorch_lora_weights.safetensors
大小 (KB): 94912
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success