nocturne protocol

[nocturne protocol @vectorinversion]

https://t.co/AbmoeyQR9t @gg159136

[David Valerio Gilmore @davidvgilmore]

ML models memorize training data, a major #privacy problem for data collaboration, even w/ #federatedlearning

@gg159136 and @vectorinversion gave a fantastic preso at @opendp_org this morning on applying #differentialprivacy to #splitlearning to solve this

More to come!

[nocturne protocol @vectorinversion]

@lukefshaw datatable:
> tic() ;x <- fread("stuff.csv") ;toc()
0.569 sec elapsed
> gc()
used (Mb) gc trigger (Mb) max used (Mb)
Ncells 352321 18.9 654548 35.0 654548 35.0
Vcells 143416782 1094.2 174806664 1333.7 143591916 1095.6

[nocturne protocol @vectorinversion]

@lukefshaw dataframe:
> tic() ;x <- read.csv("stuff.csv") ;toc()
116.381 sec elapsed

[nocturne protocol @vectorinversion]

@lukefshaw > system("ls -alh stuff.csv")
-rw-r--r-- 1 phil phil 1.5G Dec 6 13:55 stuff.csv
> system("wc -l stuff.csv")
21658045 stuff.csv

[nocturne protocol @vectorinversion]

noaa-19 storm francis

[nocturne protocol @vectorinversion]

nice noaa 19 pass

[nocturne protocol @vectorinversion]

image-super-resolution with Artefact Cancelling GANS model: https://t.co/JXzZHHzMrQ