Translate foreign model pages, games, anything on-screen, live. 100% local, no cloud.
https://afklivetranslate.afksecurity.online/
you like my model? follow me!
Do you like my models? Place bids! I'll follow you!
read the about this version before downloading
--flux update-- Based on anime enhancer V6! sorry for the images, I don't have the required hardware to use controlnet to make you a nice box with the different weights..
Here is my special anime quality booster.
Pony/anime/cartoon/ compatible
examples:
1>no lora or prompt type: masterpiece or high quality
2>lora in weight: 1.0 + masterpiece, absurd,raytracing
3>lora in weight of: 1.5 + detailed eyes, detailed face, raytracing, highly detailed, anime style, absurd, masterpiece
the following prompts are modified by lora, their weights and their effects must be tested to obtain the result you want:
detailed eyes,detailed face,raytracing,highly detailed,anime style,absurd,masterpiece
If you encounter any problems, please comment them to me, this lora is a satisfactory first test but can be improved.
Enjoy :)
Description
!!Warning!!
Works exclusively with the tag: source_anime
The strength of the lora is very high, avoid creating images above a weight of >>0.7<< for this lora
don't use () or other, use only weight
FAQ
Comments (7)
V5 seems to work far more efficiently than V4. doesn't really require the trigger words but that's partly due to far more advanced base models that don't need scoring or garbage buzzword slop
indeed, it was exclusively trained on the source_anime trigger word, which allowed me to select a dataset and advanced parameters compared to V4 which is available for different versions, and therefore, less efficient.
I invite you to look at the about this version, normally you don't need to put anything on source_anime. it was tested in 0.3/0.5/0.7
@n0utis the description is a little open ended but I'll try to parse it best I can
@XenosagaCVT Hello, have you tested V6? I would like some feedback on it to know if there are things to improve
@n0utis need to brainstorm but will be testing it out
@XenosagaCVT Thanks a lot :)