nayuta-ai / cloud_storage

0 stars 0 forks source link

Create ML workflow with auto-scaling #7

Open nayuta-ai opened 1 year ago

nayuta-ai commented 1 year ago

Summary

Our system consists of three components; PV, Job, and VPA. PV plays the role of managing datasets, Job plays the role of managing workflows, and VPA plays the role of managing the Job resources.

Goal

This issue will be closed when I implement it.

Approach

Overview

The overview of our system is described in our design doc. First, we will implement a job which represents ML workflows. Next, we will implement a PV that manages datasets using DVC. Last, we will set up a VPA for managing the job resources.

To Do

Deadline

2022

Reference

None

Note

None