
This is the release version, it's not exactly what I wanted but I still think it's pretty cool so I am releasing it and calling it finished.
Also I am putting all of the saved step models in this MEGA folder for people who want to play around with some of the other results: https://mega.nz/folder/oykU1QKQ#PG47TLDrjTWM5jRqIbgn4g
in general I get the best results with steps 1500-5000. Higher than that and loss of quality begins to occur.
weight 0.8 to 1 is fine for this lora
trigger words: dsvania_style but you do not need it.
ALPHA VERSION DESCRIPTION:
this is still a work in progess but I'm posting it so some people can play around with it.
I am trying to make a lora that can do pixel art in the style of the DSVanias (DS Castlevanias), which is not what Pony is good at so it is pretty hit or miss. This LoRa is wildly incomplete, and inconsistent so it's really just something to play around with
weight 1 is fine.
No need for trigger words.
描述:
训练词语: dsvania_style
名称: dsvania_style-step00001500.safetensors
大小 (KB): 166546
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success