precision-sustainable-ag/SemiF-AnnotationPipeline

discuss what data to keep in blob and nfs

Closed this issue · 0 comments

mkutu commented

What data should we be holding onto in the blob and nfs?

for reference, autosfm data is around 100 - 200GB per batch.

this relates to what data will be made publicly available.

Data includes:

  • developed full size mages
  • metdata jsons
  • segmentation masks
  • batch metadata json file
  • processing log files