metadata

license: creativeml-openrail-m
tags:
  - monika
  - yuri
  - sayori
  - natsuki

SD 1.5 dreambooth models fine-tuned on Anythingv3. Checkpoint files trained on (number of steps divided by 90) images scraped from boorus, with text encoder trained for around (number of images times 12) steps. Focused on ckpts before moving on to safetensors and lora. (using loras with checkpoints and even TIs can bring out most accurate results).

SD 1.5 Loras also trained on Anythingv3. SDXL Loras trained on animagine-xl.

Textual inversion embedding versions can be found here (trained off the same or similar datasets).

yr1-2430 (12/05/2022) [SD 1.5 DB]

Keyword: yrg
First attempt

yr5 (12/08/2022) [SD 1.5 DB]

Keyword: yrg
Second version

sa1-1800 (12/08/2022) [SD 1.5 DB]

Keyword: syr
First attempt

sa2 (12/08/2022) [SD 1.5 DB]

Keyword: syr
Second attempt

na1-1800 (12/06/2022) [SD 1.5 DB]

Keyword: ntsk
First attempt

na2-2160 (12/07/2022) [SD 1.5 DB]

Keyword: ntsk
Second attempt

na3a-1800 (12/07/2022) [SD 1.5 DB]

Keyword: ntsk
Third attempt

ml6-10 (~05/2023) [SD 1.5 lora]

Keyword: monika (doki doki literature club)
Tenth version
Previews (Silicon29 and BreakDomain Anime)

mxl1_11132023a_test [SDXL lora]

Keyword: monika (doki doki literature club)
One of first attempts, works best with this model
Tested with ComfyUI, strength at ~1.5 seems to give best results

nxl1_11142023_test [SDXL lora]

Keyword: natsuki (doki doki literature club)
One of first attempts, works best with this model
Tested with ComfyUI, strength at ~1.5 seems to give best results

sxl1_11142023_test [SDXL lora]

Keyword: sayori (doki doki literature club)
One of first attempts, works best with this model
Tested with ComfyUI, strength at ~1.5 seems to give best results

MTBA (previews, any future loras or models trained off better bases- hopefully more SDXL too)