Jon Deaton

https://jondeaton.github.io/

EvolutionaryScale San Francisco, CA

Results 3 issues of


                                            Jon Deaton

Mark Duplicates implemented in Spark SQL

I believe that there are performance improvements to be had by implementing duplicate marking in Spark SQL.

Updated to run hg38 experiment on newer pytorch lightning version.

Fix autocast deprecation warning.

4

comment

Fixes warning about deprecation of `torch.get_autocast_gpu_dtype`