Open ereeere123 opened 7 months ago
Hello! Sorry for the late reply! If you want to use the UViT as backbone, please check the config setting in configs/train_UViT_config.py
and run python train_diffusion_clipUViT.py
. Moreover, you can generate images by running python sample_diffusion_clipUViT.py
. All the parameters you need for image generation is the path to the data file
, path for saving the generated images
and path to a model check point
.
I would be extremely grateful.