metadata

inference: true
language:
  - en
tags:
  - stable-diffusion
  - text-to-image
  - embedding
license: wtfpl

rr_inferno embedding - SD 2.1

Exemple of Prompt

"Portrait of a rr_inferno skull skeleton, made of lava, fire, giant, concept art, splash art, global lightning, artwork of a phoenix, nine tails, fiery, gorgeous digital painting"

Training

Trained for 850 steps. 20 images, 6 vectors. Batch size of 3, 3 grad acc steps, learning rate of 0.0004:200, 0.0002:400, 0.0001:1000,0.00005