
Impact of SWE in ML Projects // Laszlo Sragner and Tim Blazina // MLOps Reading Group
MLOps.community
00:00
The Difference Between Macine Learning and Softer Engineering
Thangi astas: Engineering is engineering. It doesn't meter wi sopher or harwer or machine learning. We all have different tests at different places. Softer enginering is instantaneous, it will go into immediately move capy eyes. And this is why you can have unite test and integration test,. But in mesion learning, that's not sometimes you are only statistically wrong. Unlike subterengineering ware, just an assertion wich is just all you need. Thangiastas: You should worry about testing for changes you do on your model side. If you return your maybe deplyink some canary deployment might be useful to detect whether or not those changes can
Transcript
Play full episode