DreamBooth: Fine-Tune Stable Diffusion 2.1 in Google Colab Free (Single Subject - Human Face)
Автор: ByteXD
Загружено: 2023-02-15
Просмотров: 1232
Описание:
A demo training Stable Diffusion 2.1 with Dreambooth on a single subject. Full written tutorial here https://bytexd.com/how-to-use-dreambo...
[ ⏰ time ]
The whole thing took about 30 minutes with an Nvidia T4 on Google Colab free. The training took just 18 minutes.
Some sections are sped up for easier viewing.
[ 📔 important note ]
➞ When training similar subjects, like same gender faces, it's recommended to train them on separate models for best results.
➞ When training multiple subjects on the same model, then reduce the UNet learning rate to 2e-6 or lower. For this purpose you can see this video tutorial where I'm training SD 1.5 on Aemond Targaryen and Sandman (from Netflix) https://www.youtube.com/watch?v=S67CL...
[ ⚙️settings used ]
Images Used: 10 images of Sandman
Stable Diffusion Version: 2.1
All training settings left on default values.
[ ⌛ timestamps ]
00:00 - Run 1st cell & wait to connect to Google Drive
00:16 - Run 2nd cell to install dependencies
00:18 - Select Stable Diffusion version and run cell to download
00:22 - Name session and run Create/Load a Session cell
00:39 - Run cell "Instance Images" & wait for upload button
00:54 - Upload images
01:17 - Run Start Dreambooth cell to train & wait (~18 min)
03:27 - Run "Test The Trained Model" & wait (~5 min)
04:38 - Access web UI & generate images
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке:
