Best way to start an active thread in the community slack? Talk about jupyter notebooks.
Well... the same guy that co-created KubeFlow is now out to solve this mystery for us. David Aronchick presented his vision last Wednesday of how the SAME project could be the solution for notebooks.
But what is
SAME?
The goal is to make it easy for notebook developers to build a reliable workflow. To enable notebooks to take advantage of cloud patterns and bake in best practices. These reliable workflows can be developed locally and continuously deployed into production.
That's a tall order. I know a few people who would love it. That is, if it ever comes to fruition.
David did stress that there is still much work to be done. However, it's currently live! You can play around with it here and give feedback on what isn't working. If you are really feeling bold you can even contribute a few lines of code!
The supporting frame of a structure (such as an
automobile or television) Leaf springs are attached to the car's chassis.
Also: the frame and working parts (as of an automobile or electronic device) exclusive of the body or housing
So what does this have to do with MLOps? The ultra prolific Luke Marsden, aka my old boss, hasn't stopped creating stuff in the MLOps space since dotscience went under. And judging by what he told me last week, I don't think he has plans to any time soon.
I caught up with Luke to get the low down on his new creation Chassis.ml
"We created chassis.ml to help bridge the gap between data scientist/ML teams and DevOps teams. Getting models into production is still one of the main challenges for companies trying to get value out of AI/ML. DevOps teams could do worse than deploying chassis to their k8s cluster to give data scientists an easy python SDK to convert their MLflow models into runnable, production ready container images that are multi-platform."
"While we already support MLflow, kfserving and modzy, we're looking to integrate with other model sources and ML runtimes so come and get involved in #chassis-model-builder on the MLOps community slack."
Check out a demo of Chassis in action here, or click below for a full on test drive.
Alda Pontes, ML engineer, talks about life (and work) on the road.
Why did you leave your full time job as an engineer? I developed these long lasting beautiful relationships in the office jobs I worked at. But I was fatigued by the hurdles associated with corporate work. It was a really rote experience, just going from meeting to meeting. I wanted to find a life that would enable me to explore the world and do good along the way. I wanted to go back to coding and do work that I love from anywhere.
How did you
get started? We did all the things around incorporation, taxes, etc. Then we started traveling. The idea is that you live for free and volunteer 3-4 hours per day. Then consult on the side or in between projects. We worked on permaculture in Bolivia. Then we went to Peru and worked on a school and playground construction project. Then we were in Kenya for four months working at an NGO around water access. Today, we’re in Wyoming and heading to the Wind River Reservation.
Is it challenging to keep consulting while traveling? Honesty? Not really. The awesome thing about a contracted-based, but also flexible time work schedule is that we get to own our days. I’ve made a deliberate choice to live my life this way, so I’m not expected to be online all day. My work is very independent, which is one of the things I love about it.
Why did you join Tribe? I wanted to be part of a community of like-minded individuals. If I get stuck on a problem, I don’t want to feel alone. And that’s been the case ever since I joined the tribe. I have a community, and one that cares.
What’s your favorite part about Tribe? To me, the value of Tribe is really about the community. Connecting with all these engineers is a really beautiful way to bridge the isolation of remote work in this
increasingly virtual world.
Where will your travels take you next? We’re excited to explore Rwanda and South Africa. Madagascar. And we want to go spend two years in SE Asia – exploring all the beauty over there. We’ll see.
You can read more about Alda’s journey into consulting, traveling, and volunteering here. Apply to join Tribe if you want to explore the possibility of consulting or join our community of 150+ data scientists, ML researchers, and engineers.
After the incredible feedback of our first session, we will now be finishing what was started! Alon Gubkin is
going to take us through all the missing parts of building an ML platform from scratch.
When we last left our hero he had successfully shown us how to set up an ML platform based on open-source tools like Cookiecutter, DVC, MLFlow, FastAPI, Pulumi, GitHub Actions, and more.
But he left us with more questions than answers in the final minutes of the live tutorial.
Now we want to know how to add solutions for Model Monitoring and Training Orchestration. If we can get security in there too that is always a plus!
Looking forward to catching you all there, come with part 1 already finished so you can follow along with us on part 2.
See you tomorrow at 9am PT/ 5 pm BST. In case you want to keep up to date, subscribe to our public google calendar.
Large scale model management: How do you manage tens or hundreds of models needed for inference? Community
member Raviraj Prajapat shared a really interesting question that led to a great discussion!
Incredible advice yet again: Shout out to community members Savin, Thomas, and Mayez for
a great thread that demonstrated the depth of thought and knowledge this community shares and prizes!
Minimum Viable Accuracy: Great idea from community member Eduardo Bonet that is
well-worth reading more about!
So much great community member created content recently!