Same request as for grounding-50m. Is any effort on filtering planned for the future work?
@dondosss Thank you for your interest in our dataset!Yes, it's on our radar. We're currently in the process of re-running OCR extraction on the high-resolution subset of the BLIP3-OCR-200M dataset. Stay tuned for updates!
@HenryCRM Curious is there any updates/suggestions on filtering the dataset? Thanks!
· Sign up or log in to comment
