ghost
Sign In

Unleashed

124
946
497
32
Verified:
SafeTensor
Type
LoRA
Stats
370
193
Reviews
Published
Dec 16, 2023
Base Model
SD 1.5
Usage Tips
Clip Skip: 2
Strength: 1.5
Hash
AutoV2
9FE3C81D98

With this LoRA, you can Unleash your creativity! Create beautiful images based on abstraction. Works best with anime-like models! This Model works best with character-driven generations, especially female portraits.

Recommendations for best results with V1

  • Use a low CFG scale value: between 4 and 6

  • Increase the LoRA weight for crazier results! (between 1.3 and 2 and beyond!)

  • Samplers: I found the best results with DPM++ 3M SDE Karras and DPM++ 2M SDE and Euler a

  • Using upscalers:

    • Recommended scaling: 1.5x or more

    • Recommended upscaler: R-ESRGAN 4x+ Anime6B or R-ESRGAN 4x+ or 4xUltrasharp_4xUltrasharpV10

Recommendations for best results with V2

Version V2, outputs images that resemble more of those from my Unleashed Diffusion model

  • Use a low CFG scale value: between 4 and 6

  • Recommended weight: between 0.6 and 1.3

  • Samplers: I found the best results with DPM++ 3M SDE Karras and DPM++ 2M SDE and Euler a

  • Works well with LCM sampler with LCM LoRA as well!

  • Using upscalers:

    • Recommended scaling: 1.5x or more

    • Recommended upscaler: R-ESRGAN 4x+ Anime6B or R-ESRGAN 4x+ or 4xUltrasharp_4xUltrasharpV10

Examples

Version Comparison