tobran / GALIP

[CVPR2023] A faster, smaller, and better text-to-image model for large-scale training
MIT License
225 stars 25 forks source link

script for data preprocess #28

Open busyyang opened 2 months ago

busyyang commented 2 months ago

Could you please share the script for data preprocess? I am not very sure the context in the .npz file and .pickle files. If I wanna train my own dataset, how can I format these files?

And, I just not find the text for bird dataset, how can you get the text for each images?