Alternative download link of this dataset
This is the dataset I used to train version 1.0 of my Darkest-Diffusion model. That model was trained on the artstyle of the "Darkest Dungeon" game for Stable Diffusion v1.5. The dataset is composed of 14 manually captioned images.
The images were captioned using BatchPrompter. For more information on how I trained the model, visit the model page.
I appreciate any donations to my KoFi to help me fund renting GPU's for further model creation and experimentation! I experiment a lot to achieve the highest possible model quality and flexbility, but this much experimentation requires me to train a lot of different models which in turn costs me a lot in GPU renting costs. So any amount of donations would help me greatly to offset some of the costs!
This dataset is open access and available to all to train their own models, as long as it is only for non-commercial use! It uses the same modified CreativeML OpenRAIL-M license for non-commercial use only as my models.
- Downloads last month