Commit History

nb to surgically fix the timestamps in the parquet file
7e356f1

sradc commited on

filter images more aggressively (~1/6 of images removed)
f15a39c

sradc commited on

filter images, by std, manual list, and similarity of clip vectors (and update deps)
9b328a7

sradc commited on

filter_images.ipynb -> manually filter videos by id
e79038e

sradc commited on

nb to filter images
8b11c67

sradc commited on

run black on nb
ee16e23

sradc commited on

data wrangling notebook, to add base64 images to the parquet itself
5a3ba8c

sradc commited on

initial commit
1801c3b

sradc commited on