keras-dreambooth/dreambooth_diffusion_akitainu

Model description

This model has been fine-tuned to learn the concept of Akita Dog from Japan by ChienVM on this dataset

Intended uses & limitations

Concept token "akt## dog". Example: "a photo of akt## dog"

Some examples prompt

Prompt: akt## dog as an anime character in overwatch Negative prompt: (bad anatomy), (blurry) Steps: 100, CFG scale: 12, Size: 512x12

Prompt:akt## dog as a realistic fantasy knight, closeup portrait art by donato giancola and greg rutkowski, digital art, trending on artstation Negative prompt: (bad anatomy), (blurry), ugly Steps: 100, CFG scale: 12, Size: 512x12

Prompt:Cute and adorable cartoon fluffy akt## dog with cap, fantasy, dreamlike, city scenario, surrealism, super cute, trending on artstation Negative prompt: (bad anatomy), (blurry), ugly Steps: 100, CFG scale: 12, Size: 512x12

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

Hyperparameters	Value
inner_optimizer.class_name	Custom>RMSprop
inner_optimizer.config.name	RMSprop
inner_optimizer.config.weight_decay	None
inner_optimizer.config.clipnorm	None
inner_optimizer.config.global_clipnorm	None
inner_optimizer.config.clipvalue	None
inner_optimizer.config.use_ema	False
inner_optimizer.config.ema_momentum	0.99
inner_optimizer.config.ema_overwrite_frequency	100
inner_optimizer.config.jit_compile	True
inner_optimizer.config.is_legacy_optimizer	False
inner_optimizer.config.learning_rate	0.0010000000474974513
inner_optimizer.config.rho	0.9
inner_optimizer.config.momentum	0.0
inner_optimizer.config.epsilon	1e-07
inner_optimizer.config.centered	False
dynamic	True
initial_scale	32768.0
dynamic_growth_steps	2000
training_precision	mixed_float16