We often talk of the "state of #ML/#AI" as a way to refer to the health of the research community: how many interesting contributions have been made this year? Is progress slowing down? Are we running out of new ground-breaking paths to be explored?
This is not enough. [1/5]
-
-
We have a tendency to overlook the "state of
#ML" from the point of view of execution quality, best practices and tools. Have we found out the best way to manage a ML project? How should it be structured? What should we look out for? Is everything we are doing reproducible? [2/5]1 reply 0 retweets 1 likeShow this thread -
Some of these questions are more relevant for industry than research, or vice versa, but we need to start asking them and put down our answers. I gave it a go here: https://www.lpalmieri.com/posts/2018-09-14-machine-learning-version-control-is-all-you-need/ … [3/5]
2 replies 1 retweet 2 likesShow this thread -
I have tried to make a point for versioning in ML projects: why we need it, what it provides us, what we are missing in terms of tools to do it properly. We need the equivalent of a
#StateOfDevOps report for#ML. [4/4]1 reply 0 retweets 1 likeShow this thread
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
Writing about stuff to learn how it works, mostly in Rust.
Lead Engineer at