Type | |
Stats | 1,058 0 |
Reviews | (198) |
Published | Jun 16, 2023 |
Base Model | |
Training | Steps: 10 Epochs: 1 |
Hash | AutoV2 01D5F6FFBC |
Description
First LORA training.
Trying to better learn how the training process works / doesn't work.
Heavily influenced by PhantomIXRow's artstyle - There is also a Model that you can use instead - if you really strictly want that art style.
Prompting Info
Works well with any variety of models for different effects, try anime or western.
No specific prompts really required can do really anything you want but is heavily impacted by the underlying Model.
All my sample images are at 1.0 strength, feel free to down rank intensity as needed.
None of the samples leverage any other LORA's / Hypernetworks / or any clean up edits, tried to keep prompts simple.
Credits
Original Artist
Credit to the original artist PhantomIXRow for influencing the dataset
Credit to Cosk, who produced the original dataset I used for testing