Verified: 6 days ago
SafeTensor
Type | |
Stats | 218 24 |
Reviews | (51) |
Published | Nov 16, 2024 |
Base Model | |
Training | Steps: 660 Epochs: 15 |
Usage Tips | Clip Skip: 1 Strength: 1 |
Trigger Words | clagwimoth, confused, ?, portrait, | |, dot eyes, looking at viewer, |
Hash | AutoV2 3F5CE3AF26 |
Created on Civitai
An experiment with a LoRA trained on a bunch of thematically similar but artistically different pictures. I’m sure you could get close with just prompts, but I wanted to give this a try anyways.
Tags: clagwimoth, confused, ?, portrait,
Optional: | |, dot eyes, looking at viewer,
First time making a LoRA that’s primarily autotag based (added the main tags, manually controlled portrait v. upper body, and culled some wrong tags). Used 44 pics, curated for consistency, taken from combining my existing collection of this meme and another collection I was given access to when I expressed interest in making this. Was correct by first epoch.