The god of destruction from Dragonball, Trained on Fluffyrock Lion e22 non offset
If you are on the FurryDiffusion discord I have a thread (custom models-> Cynfall's LoRAs) where 40+ other models are posted, it's also where you can request stuff.
if you'd like to support me I got Patreon page https://patreon.com/user?u=98528151 and my PayPal https://paypal.me/CynfallN?country.x=FR&locale.x=fr_FR
Post what you get, like and rate ^^
Description
FAQ
Comments (10)
the prompts for the images make it seem like the lora isn't being used???
im using the additional networks extension
I can barely get it to work on any non pure furry model. The purple fur won't prompt, I'm guessing because it was trained on the Fluffyrock model instead of a neutral base model?
probably, it works well on bb95 furry mix too
I apologize for giving it a one-star rating, but this LoRA is indeed abnormal compared to regular LoRAs.
It is extremely difficult to use. I copied your positive prompt and added a basic negative prompt, then tested it on various checkpoints using the xyz plot.
As you can see from the results, most of the outputs were very poor, and some had almost no effect. I hope there can be improvements, and I'm willing to adjust my rating.
Thank you for your assistance.
Use a compatible model before telling me that it doesn't work. if it didn't work i wonder how i magically made the pictures for the post ¯\_(ツ)_/¯
@Cynfall We understand that sharing your work freely is valuable, and we offer our feedback in return. It is important for us to express that we feel the quality of your LoRA falls below our expectations. We apologize for being direct, but we believe honest feedback can contribute to improvements and progress.
However, as your audience, we would greatly benefit from having clear instructions in your description regarding the compatibility of the model with specific checkpoints or requirements for usage.
@ainomad1668 while i appreciate feedback it's a bit weird to tell me that it doesn't work on some model it's not supposed to work on, i tried myself on 5 different furry models and the results were alright on most of the models, and tbh all my other lora probably have this "issue" ¯\_(ツ)_/¯
so maybe try on a furry model it should work way better ^^
It's trained on fluffyrock e22 non offset so most of the models from the fluffyrock series should work, and i used bb95v41 for my post
https://huggingface.co/lodestones/furryrock-model-safetensors/blob/main/fluffyrock-1088-megares/fluffyrock-576-704-832-960-1088-lion-low-lr-e22.safetensors
https://civitai.com/models/17649?modelVersionId=53885
Please do Sorrel.
Nice! Any way of adding Whis into this?
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.