Very useful work! Is there an estimated release time for the multimodal data?
#5
by
TheEighthDay
- opened
This ～
“Finally, we see the potential
to extract, describe, and release a dataset of images found within the raw page scans to support multimodal
model training.”