4 ms·

this is a great journey! what volume of data were you analyzing? more you could try: - span multiple machines, with lambda/ec2-spot and s3 - write c/cpp, cal

this is a great journey! what volume of data were you analyzing? more you could try: - span multiple machines, with lambda/ec2-spot and s3 - write c/cpp, call via python subprocess or cffi - pytorch gpu on ec2-spot

behnamoh 3y ago
Thanks for the input! I'll look them up. Depending on the resolution of my parameters, I'm dealing with a 32-40GB-ish data.