Re:MLOps is a mess but that's to be expected
by pyronik19 on Thursday March 03, 2022 @12:53AM
(#30537659)
The problem is scaling up. You do 2,3 - 5 projects the "hard way" manually deploying the data pipelines, model images, api, etc... then you are like.. there has got to be a better way. It's the difference between a side project and a full department supporting the entire enterprises needs.
[ Parent
| on HN ]
Re:MLOps is a mess but that's to be expected
by exdsq on Thursday March 03, 2022 @01:02AM
(#30537708)
Out of interest where can I learn about MLOps specifically? I’m okay at aws and containers etc but have no idea how it looks to deploy an ML project
Re:MLOps is a mess but that's to be expected
by aiaffinity on Thursday March 03, 2022 @01:38AM
(#30537882)
At the risk of self promoting, I have been teaching a (selective and somewhat basic) course on MLOps since 2020: https://github.com/thejat/mlops-notebooks/blob/master/Syllab...
Re:MLOps is a mess but that's to be expected
by pyronik19 on Thursday March 03, 2022 @08:35AM
(#30540765)
This has been my "bible" for the last 2-3 years...
https://cloud.google.com/architecture/mlops-continuous-deliv...
It covers differing levels of maturity, with the Maturity Level 0 being what the op was mentioning. Our team is trying to dig out of Level 1 to Level 2.
Re:MLOps is a mess but that's to be expected
by reachtarunhere on Thursday March 03, 2022 @01:33AM
(#30537854)
Andrew Ng has a series of courses on Coursera
https://www.coursera.org/specializations/machine-learning-en...
Even the free videos are valuable if you do not have experience deploying real world systems.
Re:MLOps is a mess but that's to be expected
by WaxProlix on Thursday March 03, 2022 @01:23AM
(#30537809)
AWS has SageMaker Pipelines, which you can spin up into a default MLOps setup through the SageMaker Studio -> Projects -> Default Projects(?) flow. It's a bunch of stuff effectively vended through Service Catalog, so you're free to modify and inspect it all.
Re:MLOps is a mess but that's to be expected
by mmanulis on Thursday March 03, 2022 @09:12AM
(#30541236)
Check out https://mlops.community/ as well. The slack community is pretty active too.