Verified: 6 months ago
SafeTensor
(2).jpeg)
The FLUX.1 [dev] Model is licensed by Black Forest Labs. Inc. under the FLUX.1 [dev] Non-Commercial License. Copyright Black Forest Labs. Inc.
IN NO EVENT SHALL BLACK FOREST LABS, INC. BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.
Trained on the super awesome glif loradex: https://glif.app/loradex
This one was requested. Basically a one trick pony eye lora one would think.
All the training data is shot by myself on a Pentax K-1 Mark II using a Pentax-D FA 100mm f/2.8 WR Macro lens.
Sample size of 64 eyes.
Initially intended to be a realistic one, but Flux just can't handle this type of thing well yet.
Settings:
Strength 1.0 will default to a close up eye no matter what you prompt, because it's trained with a higher learning rate to do just that.
Strength 0.4-0.6 or so will let you blend the eye however, prompting something else inside a single lora.
Fun facts:
After finishing torturing all my friends shining into their eyes, I had too much brown eyes data. While my sample size might be small, but to me it seemed that women had a higher variety of eye color and tones than men. So this model is overfit on women. (Not scientifically proven by me, might be random)
After I was missing green eyes and grayish ones, was that creep walking around on the streets who looked deeeep into stranger's eyes looking for eye models. That got some reactions lol : D