htahir111

8 ms·

31.
ML should be written as pipelines from the get-go(htahir1.medium.com)
2 points by htahir111 6y ago|0 comments
32.
Show HN: Production ready ML from day 0(zenml.io)
4 points by htahir111 6y ago|0 comments
33.
Show HN: Machine Learning Pipelines with Integrations(github.com)
2 points by htahir111 6y ago|0 comments
34.
ML in Production Is Broken(hackernoon.com)
3 points by htahir111 6y ago|0 comments
35.
by htahir111 6y ago
And what about just using BudgetML, which is free and open-source ;-) cheeky grin :-P
36.
by htahir111 6y ago
Hey @sabalaba, that sounds pretty interesting. Would you be willing to integrate with ZenML? https://github.com/maiot-io/zenml . We're looking for solutions like that that have synergies
37.
by htahir111 6y ago
Yes this is aimed at being a simpler API rather than complicated one. Also MLFlow does not orchestrate self-starting preemptible instances, so you'd have to pay the full amounts.
38.
by htahir111 6y ago
Big fan of cortex and i'm also on its boards often :-) However, this works without a Kubernetes cluster, while Cortex does not. A managed kubernetes cluster on GCP e.g. costs $100/month[1].AWS, Azure is similar. For bare-metal, se
39.
by htahir111 6y ago
Ah no thats more for AWS spot instances I think. In Google, preemtible rates are constant, which makes it way more deterministic
40.
by htahir111 6y ago
Nice! Thats a cool side project. Not sure how Google preemptible can scale though. They are not auto-scaling - so its a fixed cost
41.
by htahir111 6y ago
OP here. Probably some form of container orchestration system like Kubernetes to manage the deployments + resources. There is also specialized ML tooling like Kubeflow to help out. Its still a developing landscape though. We made BudgetML n
42.
Show HN: Deploy ML Models on a Budget(github.com)
117 points by htahir111 6y ago|47 comments
43.
Show HN: BudgetML – Deploy your ML model in 10 lines of code while saving 80%(github.com)
4 points by htahir111 6y ago|0 comments
44.
Machine learning faces a reproducibility crisis(blog.maiot.io)
3 points by htahir111 6y ago|0 comments
45.
Is your machine learning reproducible?(blog.maiot.io)
4 points by htahir111 6y ago|0 comments
46.
Show HN: Automate ML with Repeatable Pipelines(maiot.io)
2 points by htahir111 6y ago|0 comments
47.
Splitting your dataset in ML is an underappreciated art(blog.maiot.io)
3 points by htahir111 6y ago|0 comments
48.
Avoiding technical debt in ML development: A practical guide(blog.maiot.io)
4 points by htahir111 6y ago|0 comments
49.
Show HN: Deep Learning 33M data points using a few lines of YAML(blog.maiot.io)
7 points by htahir111 6y ago|0 comments