johko's picture
Update README.md
db51dce
metadata
library_name: keras
license: apache-2.0
pipeline_tag: text-to-image
tags:
  - keras-dreambooth

Model description

This model is a fine-tuned Stable Diffusion modeled, using the Dreambooth technique. It was trained on 43 screenshots of the game Return To Monkey Island, scraped from the Internet. You can find the full set here: Return To Monkey Island Screenshots

The result resembles the style from the game, even though you should not expect wonders and rather see it as its own style inspired by the game's.

It was created by johko for the keras-dreambooth sprint.

Training procedure

This model was trained using the keras implementation of dreambooth. You can find the notebook to train these models and how to attend this sprint here.

Example Outputs

Geralt of Rivia Geralt of Rivia in Monkey Island Style

Frodo Baggins Frodo Baggins in Monkey Island Style

Han Solo Han Solo in Monkey Island Style

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

Model Plot

View Model Plot

Model Image