Finetuning on custom dataset

kakaobrain / mindall-e

PyTorch implementation of a 1.3B text-to-image generation model trained on 14 million image-text pairs

Other

631 stars 66 forks source link

Finetuning on custom dataset #23

Open INF800 opened 2 years ago

INF800 commented 2 years ago

Hi, what steps do I need to follow to fine tune min-dalle on custom dataset?

INF800 commented 2 years ago

Can I use this https://github.com/kakaobrain/minDALL-E/blob/main/examples/transfer_learning_ex.py by replacing ImageNetDataModule with CustomDataModule

INF800 commented 2 years ago

  config_downstream
  path_upstream
  result_path
  imagenet_path

@saehoonkim @wbaek @LeeDoYup there isn't any information present about these 4 values in argparser of examples/transfer_learning_ex.py. How to define them?