Acquired a higher-resolution images training dataset.
Reduced the repetition in image training to allow for more flexible prompting.
Used RealisticVision5.1 as the training base model.
Adjusted the number of epochs to 10 and set the total training steps to 3900.












