LAION-5B: A NEW ERA OF OPEN LARGE-SCALE MULTI-MODAL DATASETS | LAION

The metadata files are parquet files that contain the following attributes: URL, TEXT, the cosine similarity score between the text and image embedding and height and width of the image. Watermark and safety tags can be joined with the metadata prior to downloading by using this script. Once that is done, they can easily be filtered upon with a probability threshold at your choice (we recommend 0.

[slkby]

"do not recommend using it for creating ready-to-go industrial products, as the basic research about general properties and safety of such large-scale models, which we would like to encourage with this release, is still in progress." 商用でゴリゴリ使われてるけどね…