_______________________

What's a Freefit LoRA?

• LoRAs I upload that are labeled as Freefit grant you the ability to dress and pose the character in any way you want in varying states of (un)dress. This requires a lot of work on my end to achieve. Have fun!

_______________________

A high quality lora trained at 896 resolution.

This LoRA has been curated and modified over several iterations to offer the highest degree of control for Alpha while retaining the highest degree of character accuracy. I don't joke about how many times I modify this stuff for the best quality possible.

The amount of tags used for prompting are necessary in order to both replicate the character and to offer flexibility should you desire to put Alpha in different clothing and situations.

Works fine at a weight of 1. Lower as needed.

This LoRA is modular in a way where you can reproduce all three ascension forms of Alpha.

Since the main differences are her thighhighs or lack thereof, use the following tags along with the main appearance tags to get the following:

1st:

• boots

• Besides boots, don't use any of the tags listed for 2nd and 3rd forms.

2nd:

• boots, bandaged leg, red pouch

3rd:

• knee boots, asymmetrical leggings, red thighhighs, bandaged leg, red pouch

QUIRKS:

• Due to the nature of NAI, SD1.5, AND the dataset, asymmetricality is difficult to reliably achieve. Patience and/or inpaint is required for the legs and arms to have the correct appearance.

• The length of the jacket may be unreliable. Be creative with your tagging to produce the intended length.

• To remove parts of Alpha's gear, putting the tags into the negatives has a more powerful effect than simply not including it in the prompt.

• As usual, don't expect perfect accuracy.