4 ms·
I think Pachyderm[0] is exactly what you're describing. [0]:https://www.pachyderm.com https://www.pachyderm.com
by cube2222 4y ago
I think Pachyderm[0] is exactly what you're describing.
[0]:https://www.pachyderm.com https://www.pachyderm.com
- Fiahil 4y agoWe used pachyderm before MLOps was even a thing, sadly, back then they tried to solve everything at once while failing with the most basic needs. If your data versioning "Utilizes a Git-like structure that enables effective team collaboration through commits, branches and rollbacks", then it means you can only consume one version at a time in your pipeline. This is, unfortunately, a fatal flaw. Especially when working with time series !
- claytonjy 4y agoI also used Pachyderm pre-MLOps (2017+), specifically for time series ML (finance) and didn't see a mismatch between what we wanted and how Pachyderm worked. Were you overwriting/replacing data over time?