Hello to all my subs and everyone that landed on this survey. I'm currently finetuning (actually training, not a lora mix) a XL models and I want to gather information on what is wanted by the people.
We haven't decided on the model name so we're going to reference the model as "model TBA" here. The model I'm making is anime based, and is going to have a consistent style and easier to use quality tags (basically the goal is to make something better than AutismMix).
Phase 1 of the model was a success, results are posted in my discord checkpoint channel.
Things that are being or have been addressed in the current model TBA:
darkness/lighting (we have multi-staged lighting control)
fixes to some tags and tags with non-existing subsets.
poses breaking symmetry/introducing ability to control symmetry/asymmetry
better background
better hands
We currently have a checklist for things to add but we would also like to hear your voice regarding things you want me to add to the checkpoint.
preferably we would like to know prompts that doesn't or images that you had a hard time reproducing.