rdtoperfv0.0

Before Start: Some models are naturally more compatible than the others and are more in sync with this lora and natturally produce better results by optimising the model and lora itself so below Im gonna list on a scale from 1 to 5 several models that I have had the chance to use and test and their score for reference: Waterfront 1.0 base preview:4 Animagine XL V3:2.5 Haruka:3 NoobAi XL:4 IllustratiousXLV1.1:3 AnimagineXLV4:5 Sundae:4 T-ponyAi(v6.5):4 Hoshino:4 Rosebud v1:3 Haruka V2:5 Hinata:3

For any questions or regards use the comment section

Pay attention to your prompt,this lora upgrades the style and quality as well as it adds it’s own rainbow spice but the effect it has on characters is that it upgrades them too or enhances them so pay attention to avoid strong suggestive words because it may diverge from your base idea so for direct physical apperance suggesive words or strong meaning words should be better avoided but for complex clothes or other specific details let your fantasy flow while for the background the lora generally creates something rainbow themed but you can let your imagination run free and the lora will adapt and produce a mixture.

Also,this lora generally from what I have used I have used the full sampling steps and the cfg should be put starting from 9 to 11 but if you want to experiment you can set it higher but generally not lower,there has been cases that even cfg 15 was perfectly fine and for slight effect you should use lora weight 0 to 0.7 then for normal effect ,0.8 to 1.3 then for more prominent effect 1.4 to 1.7 and for rainbow overload 1.8 to 2 usually as you increase the weight it may become rarely unpredictable but it has a unique feel when you set it to 1.7 plus

This Lora is style-based and focuses on rainbow style.It is compatible with XL or SDXL models.As said previously it creates rainbow-style characters and backgrounds too.It produces nice and/or dynamic backgrounds that goes well with different characters.It produces fine,high quality characters and also the lora works as an quality tweaker too increasing the quality of the artwork.It supports different styles such as anime,3d,2.5d,line art and so on but different models may affect the range of the art style as different models might make the lora put more stress to some art styles but by default anime or 2d style is always available and most probable to appear by default even if not specified.The weighting depends on the purpose of the artwork:1.0 if you want slight rainbow style present or if you want to make use of the quality increase instead of rainbow style specifically;1.1-1.4 are good for normal artworks that have partial rainbow style present and also the quality enhancing;1.5-1.8 are good for a rainbow style blended artwork and enhanced quality as well as depending on the model used certain art styles or elements may become more apparent,1.9-2 when you want the lora to get creative and push the rainbow style to the max with enhanced quality and depending on the model used unique results will be achieved!