README
The LoRa was trained with generated images with Aniverse PonyXL by Samael1976
Suggested settings:
Sampler: Euler a
CFG: 6
Steps: 30
Resolution: 832x1216 / 1216x832 / 1024x1024
LoRa Strenght: 0.8 - 1.0
Tested Checkpoint:
Prompt:
Positive:
score_9, score_8_up, score_7_up, high quality, highres, 4n1l4r4, [location, type of shot], 1girl, Lara Croft, brown eyes, long brown hair stylized in a single braid, [your prompt] [your tags]
Negative:
score_6, score_5, score_4, score_3, worst quality, low quality, normal quality, source_pony, child, kid, ugly, malformed, bad anatomy, ugly teeth, foggy eye, grayscale, (watermark), ((censored))
Description
Fixed dataset images and tags, trained with different parameters on a diffusers Pony model, tested on different models