Negative Embedding For Realistic Vision V2.0 API Documentation

This is a 75 token negative embedding for the model Realistic Vision v2.0. It should you help attain a more realistic picture when prompting. All training images were generated from that model, and all the training was also done on that model. The embedding will likely work on other models, but I have not tested it on other models.

To use this embedding you need to download it, put it in your embeddings folder, and then activate it in the negative prompt by typing out it's filename in the negative prompt. By default that will be "realisticvision-negative-embedding" (do not enter the file extension in the negative prompt). If you don't like that activation phrase, then you can change the filename to whatever you want, and that becomes the new activation phrase.