The ethos of the EI LoRA series is simple, building easy-to-use models that each generate one thing (with some variation possible) perfectly. The datasets for these models are composed of 121 images generated mostly in Midjourney 6.1. These images are in square (1:1), mobile (9:16), and widescreen (16:9) ratios.*
(*) The EI LoRA Series has been officially abandoned, but this LoRA will be left as it is
For this dataset, I used the Open Prompt Island v2 prompt as the text prompt. If you want to learn more, it is the first creation in the dataset collection.
How to Apply Variations
- Add what you desire in the text prompt (along with the trigger words).
- Reduce the LoRA weight to 0.5-0.7 or increase the overall prompt weight to 75-100. If neither of these methods work, try them together.
🏝
Comments