1 paper
Jordan Meyer, Nick Padgett, Cullen Miller +1
We present Public Domain 12M (PD12M), a dataset of 12.4 million high-quality public domain and CC0-licensed images with synthetic captions, designed for training text-to-image mode…