Tensor.Art
Create

LibraMix - v1.0

LibraMix

CHECKPOINT
Original - TenStar Fund

Updated:

v5.0
v4.0
v2.0
v1.0
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art
LibraMix by LibriAi on Tensor.Art

V4.0 checkpoint is now Tensor only exclusive!

A personal mix. All preview images are t2i + hires.fix without any inpainting. It's kind of a mix of 2.5D and painting models, and have to use hires.fix for good results.

All preview images are generated with below settings:

  • VAE: [vae-ft-mse-840000-ema-pruned] or [kl-f8-anime2]

  • Clip Skip: 2

  • Sampling method: Euler a / SDE Karras / 2M SDE Karras

  • Sampling steps: 30

  • Hires.fix: on

  • Upscaler: [Ultrasharp]

    Hires steps: 20

  • Denoising strength: 0.4 (0.35 - 0.5)

  • Upscale by: 2

  • WxH: 512x768

  • CFG Scale: 5 - 7

Negative prompt is depend on the style that you like but for me, my go to negative is:

  • EasyNegative, multiple views, multiple panels, blurry, watermark, letterbox, text

  • (worst quality, low quality:1.4), EasyNegative, multiple views, multiple panels, blurry, watermark, letterbox, text,

  • EasyNegativeV2

  • FastNegativeV2

Version Detail

SD 1.5
<p>First release!</p><p>Use with [<em>vae-ft-mse-840000-ema-pruned</em>] or [<em>OrangeMix Vae</em>]</p>

Project Permissions

Reprinting is strictly prohibited

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Related Posts