Type | |
Stats | 327 1,765 |
Reviews | (54) |
Published | Jan 22, 2024 |
Base Model | |
Training | Epochs: 10 |
Usage Tips | Clip Skip: 1 Strength: 0.4 |
Hash | AutoV2 12B7E5D96B |
Please share your creations on the model page. Thank you.
Girls on Bikes XL - New!
It's here! All the goodness of Girls on Bikes now in XL format. Dig deep in the meta data, there are a ton of trigger words. This lora is trained on over 700 images that are deeply captioned, so odds are you are going to be able to find the magic words.
First draft of a concept LORA that helps to generate women riding bicycles. The sweet spot for strength, I have found to be 0.4. It depends on what other LORAs or prompt words are competing with it. Higher strengths from 0.7-1.0 tend to cause distortion in the coherence between bike and rider. I was surprised at how well this turned out.
V1.5 bik3g1rl
Retrained with same dataset of images but added a lot more detail to captions describing the style of bike (road, comfort, hybrid, cruiser), view angle (side, front, three-quarters, front, behind, above, below etc.), lens length (wide, perspective, foreshortening, zoom etc.), and clothing style (plaid, corduroy, silk, tight, loose, etc.). I tried to reduce the total number of keywords in the training set, so that similar subjects get grouped together and features get strengthened. I believe that this will help to make prompted images be more coherent
V1 bik3g1rl
This is a bit squirrelly as the geometry of a bicycle is complex, and I trained on many different angles, orientations, and focal lengths. Many realistic checkpoints generate pretty good girls on bikes, but they are better at women posing with a bike rather than women riding bikes, This LORA does a good job of that. Also, many prompts on realistic checkpoints generate women riding bikes for recreation. This LORA generates good images of women riding bikes in a city setting, commuting to work, meeting friends etc. They aren't wearing helmets and lycra, and they aren't riding road bikes.
V2 Girls on bikes at Burningman
Self explanatory. This LORA uses a different dataset of training images than v1. Quite likely NSFW.
V3 & V4 Are Here!
The bike geometry, feet and legs got more development in the model. Increased the training set by an order of magnitude. Same training set as the XL version. All more coherence and consistency.