Machine learning processes have a regular workflow, such as preprocessing datasets, training models, visualizing results, etc. When running it on Kubernetes, it is suitable to divide each process and combine them as one workflow. Our system represents the needs using Job and Pipeline.
Goal
This issue will be closed when I implement it.
Approach
Overview
First, we will learn the Job process for building the ML pipeline, etc. Next, we will describe the detail of the implementations in my design doc. Next, we will implement it.
Background
Machine learning processes have a regular workflow, such as preprocessing datasets, training models, visualizing results, etc. When running it on Kubernetes, it is suitable to divide each process and combine them as one workflow. Our system represents the needs using Job and Pipeline.
Goal
This issue will be closed when I implement it.
Approach
Overview
First, we will learn the Job process for building the ML pipeline, etc. Next, we will describe the detail of the implementations in my design doc. Next, we will implement it.
To Do
Parent Issue
7
Deadline
2022/11
Reference
None
Note
None