Skip to content

Blogs

  • in Data
  • 1 min read
Draft

Data centric AI

Overview

QuickPOC
Figure 1 DevOps with AI

In the traditional model-centric AI lifecycle, the focus is on finding better models to enhance performance, with data largely unchanged. This approach overlooks data quality issues like missing values, incorrect labels, and anomalies. Data-centric AI shifts the emphasis from models to systematic data engineering to build AI systems.

Data centric AI vs. model centric AI

Refrences

  • in IOT
  • 1 min read
Draft

Raspberry PI with github actions runner

Overview

QuickPOC
Figure 1 DevOps with AI

Dashboard to overview the runners

Raspberry PI 5 to enable read/write google sheet

Raspberry PI to get self-hosted runner status from github API

Raspberry PI 5 to enable camera

1
2
sudo apt install -y python3-picamera2
sudo apt install -y python3-opencv

Refrences

  • in DevOps
  • 2 min read
Draft

Github runner in Synology NAS

Overview

QuickPOC
Figure 1 DevOps with AI
  1. Create YAMl file as below (it is a org github runner)
  2. In /volume1/docker/github-runner/data make sure remove .runner
  3. Generate docker with the new YAML and restart
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
version: '2.3'
services:
  worker:
    image: myoung34/github-runner:latest
    container_name: GITHUB-RUNNER
    restart: always
    environment:
      RUNNER_SCOPE: org
      ORG_NAME: xxx
      ACCESS_TOKEN: github_pat_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
      RUNNER_NAME: wenchiehlee.quickconnect.to
      RUNNER_WORKDIR: /tmp/runner/work
      RUNNER_GROUP: xxxx
      RUNNER_ALLOW_RUNASROOT: true
      DISABLE_AUTOMATIC_DEREGISTRATION: true
      CONFIGURED_ACTIONS_RUNNER_FILES_DIR: /actions-runner-data # Required for persistence
      LABELS: linux,x64,gpu,ubuntu-latest
    security_opt:
      # needed on SELinux systems to allow docker container to manage other docker containers
      - label:disable
    volumes:
      - /volume1/docker/docker.sock:/var/run/docker.sock
      - /volume1/docker/github-runner/tmp:/tmp:rw,Z
      - /volume1/docker/github-runner/data:/actions-runner-data:rw,Z # required for persistence
      - /volume1/web:/web:rw,Z
      - /volume1/docker:/docker:rw,Z
      - /volume1/homes:/homes:rw,Z
      # note: a quirk of docker-in-docker is that this path
      # needs to be the same path on host and inside the container,
      # docker mgmt cmds run outside of docker but expect the paths from within

Refrences

  • in DevOps
  • 2 min read
Draft

Biz-Needs-Tech-Alignment

Technology must align with the goals of User needs and Businees Goals or it is just a works make things more complicated.

with this idea, you can work smarter, not harder.

Technology mainly focus on system integration and its related components and output as Biz,Process (flow), Services, Products and etc.

The component can be Algorithm, Application, Software, Hardware, System.

References