This is a model for pony diffusion to generate images based off of Jenny Wakeman as depicted in “My Life as a Teenage Robot” and fanart.

Normally I would avoid making a LORA of such an obviously popular character considering there already is a selection of other Jenny LORAS out there. However I didn’t think they struck a balance between cartoon stylisation and semi-realism that I personally liked. Thus I decided to throw my hat in the ring and try make a model according to my own tastes. It’s a bit egocentric perhaps but I hope everyone likes it eitherway. Given how well my Nicole Watterson LORA was received with a similar goal in mind, I thought this would be a good move.

The prompt word was made “Jenny” and not anything specific like “Jenny Wakeman” or “XJ9” because I wanted to offer the user the choice of calling upon Pony’s own knowledge of Jenny for image generation, incase the LORA is struggling.

The artist ZetaSkully’s fanmade design for Jenny can be prompted thanks to its own keywords, same with the vanilla design. This effectively allows the user to control what parts of Jenny are blue/more robotic looking. For example, some artists like to keep Jenny’s robotic legs on at all times, while some prefer to have more humanoid-looking feet, where appropriate, or usually Jenny has white arms and white legs, but ZetaSkully goes for blue arms and legs, resembling socks and gloves. Prompting for “feet/knee boots” or “thighighs, elbow gloves” will give you the choice. It’s simpler than it sounds.

Feel free to use whatever, but personally I would recommend using Autismmix for this one.

Dataset contains both SFW and NSFW.

Key-Word: Jenny

Recommended prompts: twintails, white skin, joints, android/robot

Vanilla appearance: crop top, knee boots, skirt

ZetaSkully appearance: tube top, thighhighs, high heels, elbow gloves, skirt, wide hips

V2

V2 was made with more images and the finished result is outstanding and is much closer to my initial vision.

While AutismMix is still recommended, the LORA is less reliant on it to produce high quality images. Just compare the images between V1 and V2 and you will see that I make greater use off of base Pony.

Everything before still applies, except less strict. The lora does a better job at remembering Jenny's characteristics even if you leave it out of the prompt. Only small issue is that the lora can sometimes change the colour of her top, but that is easily fixed in the off chance it occurs

If you like the lora, do consider making an image with the LORA and submitting it here! It generates buzz and helps me keep making models.