New blog post: A note about ML and Pickle
New blog post: What's the deal with ML and pickle
?
Inspired by a friend asking me "What's the deal with ML software and using pickle
for everything?," I wrote up a few thoughts on why the ML ecosystem uses pickle
so much, and a few ideas for directions to improve the situation.
I'm especially curious to hear from software engineers who haven't done much work in ML or scientific software if the problem description here makes sense or resonates; I feel like I have a much better felt sense for why this ecosystem looks the way it does than before I started a job in ML, but I'm not sure I have adequately recalled the feeling of not having that knowledge :)
The Ranger pic for newsletter subscribers
He is curious but also very concerned about our new Christmas golden retriever statuette…
Until next time!