What is MLOps?
Вставка
- Опубліковано 25 тра 2023
- Check out watsonx: ibm.biz/BdvDnH
It takes a lot of time, effort, and money to train a machine learning model. And yet a majority of models that are trained and developed never actually make it to production. What can be done to lower the cost and frustration of ML model development? Automation! Otherwise known as MLOps, it's modeled after the same practices as devOps that have led to faster code development at lower cost. In this video, David Adeyemi explains what got us here and where to go next.
#mlops #ai #watsonx #devops #machinelearning
Is his surname a typo?
YES, thanks for spotting that! David's actual surname is "Adeyemi", we missed that first 'e'. Many apologies to David!
@@IBMTechnology come on! He deserves another month of GPU server :)
Sanda boro
This was great. We are going to be building an Data Platform aimed at ML workflows, so this explanation really helps me thing of the downstream elements we need after the basic data infrastructure.
Thanks a lot! His explanations were very clear and concise!
IBM Technology is starting to become my favourite channel.
This brilliant young man did a great job with his explanation !
Excellent explanation.
Can you pleaseee make a video on as400 or iseries servers especially power 7 ? Would like to know more about that hardware. Also how the disks are connected to it, as well as tape library etc. and other related hardware components that make up iseries servers.
Thanks for the tutorial.Can you please tell me in which scenarios python scripting is used in MLOps workflow apart from building a model .Also in which cases we use bash.Basically want
to know how much scripting is needed in MLOps apart from building a model.
Nice explanation.
Ok, who's circling on the board? I know the board but i never saw it automatically circling anything
excellent work. great presentation
Thanks ❤
Nice !
very explanative video , thank you
Keep them coming 😍😍
Awesome video, guys, and kudos for David for the geeky charisma! I'm starting to learn MLOps and will definitely look at more of his MLOps material here :-)
Lol, 'liked' just so he could get another month on the GPU server 😂
🤣🤣🤣 definitely !
Ouch😂😂
Very nice summary...
But, what's your t-shirt about? Is the bee anything related to Georgia Tech?
The t-shirt spells out IBM (Eye, Bee, M)
Nicely explained, but can someone give me a better distinction between DevOps and MLOps
Devops: Process for deploying software to production
MLOps: Process for deploying ML Models (sometimes alongside software) to production
@@introspectiveengineer392that’s a start but what if they are both used? ML models in software?
MLOps, COOL😎
Kudos to this brilliant guy, amazing explanation and comparison. We want more!
How on earth does a company like IBM not have source control and 'loses notebooks' when a server is shut down. Kind of a self report.
hmmm, interesting its about the team not the IBM
Give that GPU to those poor kids hahaha
Great vidéo tho !
Money crunch is a real thing!
Would like a simplified explanation for less experienced or brand new people.
So, let me recap: First they made you lose 1 month of your team's work. Then, they made you do this video to chase likes, in the hopes of putting you in the same situation that led to losing your progress with no results. I recommend you change departments! 🙂
good.mugo from africa
So ML ops is only good if you wanna do different model comparasion and have big deployment.
I wish "MLops" didn't sound so stupid. I guess it could be worse if it were "M Lops".
most of them are left-handed.
See ibm.biz/write-backwards
@@IBMTechnologyshut your mouth, IBM