Verified: 5 months ago
SafeTensor
Type | |
Stats | 18 12 |
Reviews | (1) |
Published | Oct 25, 2024 |
Base Model | |
Training | Steps: 3,749 Epochs: 63 |
Usage Tips | Clip Skip: 1 Strength: 0.6 |
Hash | AutoV2 505A34379C |
Created on Civitai
A better-captioned attempt at the mst3c:FLUX model. This is also based on concepts, not characters, which may result in a more versatile style. Results from training show that it will try to add speech bubbles if no text is defined as being in the image.
May be over or under-trained.
Start at 0.58 and go up.