Pony-enabled version, but unexpected behavior is found during testing. It is found that this model do not recognize "red suspender skirt" well, which was specified as a training prompt, and recognized "red skirt" and "red suspenders" separately. Therefore, there is one more prompt for clothing. In addition, a quality prompt for the Pony model is added. Default image quality and clothing reproduction is better in the Animagine-compatible version.