Hi there and welcome to another finetuned checkpoint done as to have some default base aesthetic based on NoobAI 1.0 EPS if you like this finetune, please consider donating to my Ko-fi, yes this FT is horny and will make bigger girls, limited compatibility with Loras trained on Illustrious 0.1 and NoobAI 0.5/0.75 but some will work, codenamed "WaxWorks" by @MadlyLaughing and almost named Bluballcelestial by @za4beqsbv36z2s889

Emphasis mode No Norm is suggested for SDXL models

Huge thanks to @Anzhc for his Anzhc's Breasts Segmentation, Anzhc's Face Segmentation (Prototype), AAA Vibrant VAE, AAA Anime SDXL VAE (vibrant VAE is recommended for this ckpt) and @mfcg for being kind enough to donate the funds for this finetune.

Positive prompt : masterpiece, best quality,
Negative prompt : worst quality, low quality, jpeg artifacts, multiple views, 4koma, bad anatomy, bad hands, text

Personal schizo Negative prompt: (worst quality:1.4), (low quality:1.4), jpeg artifacts, scan artifacts, compression artifacts, simple background, bad hands, multiple views, comic, jpeg artifacts, patreon logo, patreon username, web address, signature, watermark, text, logo, artist name, censored, blurry background,

Sampler : Euler A AYS, Any Align Your Steps, Euler A CFG++ Align Your Steps, Restart Align Your Steps, Restart Karras

CFG : 4 to 6 (NoobAI feels a bit fried at CFG6+)

If you want to support me, consider liking the model, and/or tipping buzz/donating to my Ko-fi, as this was finetuned on 4xA6000 Ada

Relevant training information:

1. Trained with at least 13k cherrypicked images.

2. Total of about ~~200k real steps after batching across 50 epochs.

3. Effective batch size was 22 x 4 GPUs x 2 Gradient Accumulation

4. Learning rate was 2.5e-6, Clip L was Frozen, Clip G was trained at 1.75e-7