Tensor.Art
Create

Silk von Santana - Tensei Kizoku no Isekai Boukenr

LORA
Original

Updated:

V1.0
546

Autor: EqualsNine

Link: https://civitai.com/models/69687/silk-von-santana-tensei-kizoku-no-isekai-boukenroku

Details

Type

LORA

Downloads

1090

Uploaded

May 18, 2023

Base Model

SD 1.5

Training

STEPS: 3000

EPOCHS: 12

Target: SilkVonSantana

Silk von Santana from Tensei Kizoku no Isekai Boukenroku/The Aristocrat's Otherworldly Adventure.

Expand with Show More, long but hopefully worth the read.

Slightly overbaked (as I have no idea what I am doing) so use at 0.7 - 1.0 strength depending on the model used. No trigger word is needed.

Recommended prompt:

<lora:SilkVonSantana:0.7>, (masterpiece), best quality, intricate details, ultra details, ultra quality, rule of third, detailed face, detailed iris, 1girl, solo, long hair, purple eyes, hair between eyes, purple hair, hair intakes, bangs

Optional:

Use hair ornament for the crescent moon thing in her hair, if you don't include in the prompt it will appear randomly in some shots but if you don't want it, putting it in the negative will stop it from showing up.

Trained with screenshots of her every appearance in the anime so far using AnyLoRA, so I would recommend using that, though any Anime checkpoint should work fine. All preview images were generated with AnyLoRA.

While not needed, I would heavily recommend using Detail Tweaker LoRA as it is some form of black magic and I rarely generate any images without it. Most of the previews use it.

I generate my images relatively big at 600x960 and then Hires. fix them by 2.2/2.3 to end up just right for 4K screens but I threw in a few done at 512x512 to show what it can do at lower resolutions too.

This was the first LoRA I ever made, I taught myself how to make LoRA to make it, so it is far from perfect. I only made it as no-one else seemed like they were going to, but now one for Silk has appeared, I thought I might as well put mine up too.

I only spent about 30 mins to create this LoRA (10 to take and process the screenshots and 20 to train) so it's bound to be a little rough, but I will likely update when she appears in further episodes.

I have a few other LoRA in the same vein and quality as this, so a favourable review or two on this and I'll likely upload those too.

Version Detail

SD 1.5
3000
12
Autor: EqualsNine Link: https://civitai.com/models/69687/silk-von-santana-tensei-kizoku-no-isekai-boukenroku Details Type LORA Downloads 1090 Uploaded May 18, 2023 Base Model SD 1.5 Training STEPS: 3000 EPOCHS: 12 Target: SilkVonSantana Silk von Santana from Tensei Kizoku no Isekai Boukenroku/The Aristocrat's Otherworldly Adventure. Expand with Show More, long but hopefully worth the read. Slightly overbaked (as I have no idea what I am doing) so use at 0.7 - 1.0 strength depending on the model used. No trigger word is needed. Recommended prompt: <lora:SilkVonSantana:0.7>, (masterpiece), best quality, intricate details, ultra details, ultra quality, rule of third, detailed face, detailed iris, 1girl, solo, long hair, purple eyes, hair between eyes, purple hair, hair intakes, bangs Optional: Use hair ornament for the crescent moon thing in her hair, if you don't include in the prompt it will appear randomly in some shots but if you don't want it, putting it in the negative will stop it from showing up. Trained with screenshots of her every appearance in the anime so far using AnyLoRA, so I would recommend using that, though any Anime checkpoint should work fine. All preview images were generated with AnyLoRA. While not needed, I would heavily recommend using Detail Tweaker LoRA as it is some form of black magic and I rarely generate any images without it. Most of the previews use it. I generate my images relatively big at 600x960 and then Hires. fix them by 2.2/2.3 to end up just right for 4K screens but I threw in a few done at 512x512 to show what it can do at lower resolutions too. This was the first LoRA I ever made, I taught myself how to make LoRA to make it, so it is far from perfect. I only made it as no-one else seemed like they were going to, but now one for Silk has appeared, I thought I might as well put mine up too. I only spent about 30 mins to create this LoRA (10 to take and process the screenshots and 20 to train) so it's bound to be a little rough, but I will likely update when she appears in further episodes. I have a few other LoRA in the same vein and quality as this, so a favourable review or two on this and I'll likely upload those too.

Project Permissions

    Use Permissions

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated images

  • Use on generation services

  • Sell this model or merges

Comments

Related Posts

The image generated by the model, publishing a post will appear here
Describe the image you want to generate, then press Enter to send.