I am not satisfied with this Lora, but I cannot find any more qualified training dataset.

The title with the "+" sign means that the data source has reached its limit within the capabilities of the me.

It has crawled and processed almost all the images from several image websites.

Until further updates are made to the lora train method, this ver is the best ver I can provide.

https://github.com/deepghs/waifuc/tree/main

The model data source is derived from Waifuc, a highly robust tool for crawling, filtering, and processing training sets. With the aid of the Waifuc tool and a single 5600g machine, it is possible to efficiently crawl and process over 3000 training images within a span of 2 hours. Training the character "Lora" on this extensive dataset yields significant performance advantages compared to situations with limited data availability.

↑ Translation from gpt.

Trigger Words :FUUKA \(BLUE ARCHIVE\)

train:158Image

recommand use model: Kohaku_XL https://civarchive.com/models/162577/kohaku-xl-beta

weight: 0.8~1

if not halo,please add “halo" to prompt