Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Data-centric MLOps(이정권)
Search
MLOpsKR
June 05, 2021
Programming
0
1k
Data-centric MLOps(이정권)
MLOps KR(
https://www.facebook.com/groups/mlopskr)에서
주최한 1회 온라인 이벤트 발표 자료입니다
MLOpsKR
June 05, 2021
Tweet
Share
More Decks by MLOpsKR
See All by MLOpsKR
Ray: 대규모 ML인프라를 위한 분산 시스템 프레임워크(조상빈)
mlopskr
0
2.3k
JupyterFlow : 당신의 모델에 날개를 달아드립니다(유홍근)
mlopskr
0
1.1k
모델을 데이터셋에 맞게 대량을 찍어내는 방법(only 파이썬)(김태영)
mlopskr
0
880
KRSH: 선언형 Kubeflow, Terraform처럼 파이프라인 관리하기(김완수)
mlopskr
0
940
MLOps 춘추 전국 시대 정리(변성윤)
mlopskr
0
12k
Other Decks in Programming
See All in Programming
Devvox Belgium - Agentic AI Patterns
kdubois
1
150
チームの境界をブチ抜いていけ
tokai235
0
220
Webサーバーサイド言語としてのRustについて
kouyuume
1
4.9k
Claude Agent SDK を使ってみよう
hyshu
0
1.4k
Foundation Modelsを実装日本語学習アプリを作ってみた!
hypebeans
1
130
実践Claude Code:20の失敗から学ぶAIペアプログラミング
takedatakashi
18
8.7k
社会人になっても趣味開発を続けたい! / traPavilion
mazrean
1
100
CSC509 Lecture 06
javiergs
PRO
0
270
「ちょっと古いから」って避けてた技術書、今だからこそ読もう
mottyzzz
12
7.2k
Developer Joy - The New Paradigm
hollycummins
1
370
SwiftDataを使って10万件のデータを読み書きする
akidon0000
0
240
alien-signals と自作 OSS で実現する フレームワーク非依存な ロジック共通化の探求 / Exploring Framework-Agnostic Logic Sharing with alien-signals and Custom OSS
aoseyuu
2
650
Featured
See All Featured
Put a Button on it: Removing Barriers to Going Fast.
kastner
60
4k
Product Roadmaps are Hard
iamctodd
PRO
55
11k
Writing Fast Ruby
sferik
630
62k
Code Reviewing Like a Champion
maltzj
526
40k
Reflections from 52 weeks, 52 projects
jeffersonlam
353
21k
Embracing the Ebb and Flow
colly
88
4.9k
A Tale of Four Properties
chriscoyier
161
23k
Code Review Best Practice
trishagee
72
19k
Thoughts on Productivity
jonyablonski
70
4.9k
Balancing Empowerment & Direction
lara
5
700
Agile that works and the tools we love
rasmusluckow
331
21k
4 Signs Your Business is Dying
shpigford
185
22k
Transcript
Data-centric MLOps : 데이터 중심 MLOps를 돕기 위한 작은 장치들
Superb AI 이정권
AI / ML = Model + Data
AI / ML = Model + Data Data centric?
Task Baseline: 70% accuracy Target Performance: 90% accuracy Should the
team improve the code or the data? : code(20%), data(80%) A Chat with Andrew on MLOps: From Model-centric to Data-centric AI
A Chat with Andrew on MLOps: From Model-centric to Data-centric
AI Improve AI → Improve the quality of the data: consistency error rate diversity coverage feedback frequency size ...
A Chat with Andrew on MLOps: From Model-centric to Data-centric
AI slide credit: A Chat with Andrew on MLOps: From Model-centric to Data-centric AI (https://www.youtube.com/watch?v=06-AZXmwHjo)
사실은, 늘 해오던 일 Project progress month 1 month 2
month 3 month 4 month 5 Code a model Build data Launch training job
사실은, 늘 해오던 일 Building the Software 2.0 Stack (Andrej
Karpathy, 2018)
Question: How many labeled images are needed to solve this
problem?
Answer: 100,000 images?
My Answer: I don’t know. Let’s start from 5,000 WHY?
여전히, 잘 모른다 → Data-centric MLOps Systematic & iterative way
to build Data for ML 단순히 지루한 작업을 자동화하는 과정이 아닌 ML 문제를 해결하기 위한 과정 저는 Superb AI라는 팀에서 이 문제를 풀고 있습니다.
<2달 <30명 <20,000 Images The Problem
The Meta Problem Design Data Spec Build Data Train a
model Deploy to service
Starting Point Labeling Tool Data Label
Reusable Data Spec { project_name: potato_detect_1 data_spec: good_potato: box: color:
red condition: ... bad_potato: box: } { project_name: potato_detect_2 data_spec: good_potato: polygon: color: red condition: ... bad_potato: box: }
Reusable Data Spec { project_name: potato_detect_13 data_spec: best_potato: polygon: direction:
options: ... good_potato: {} normal_potato: {} bad_potato: {} } Goal ≠ Task ALWAYS configured repeatedly name, color, type, conditions, options, property, ROI Info, ...
Support flexible pipeline 100 different problems, 100 different datasets, 100
different ways To support flexible pipeline Build Data Team Model WORKING SUBMITTED REVIEWED
Support flexible pipeline
Versioning Set 단위, 실험 당
ML Engineer를 위해 … ? Detailed Statistics & Report
Human in the loop ^ 2 Human in the loop
ML
Inside Human Labeling Data Human Labeling Service Model Data Labeling
Our Model ? Uncertain? Label-wise Confidence Overall Set Confidence User performance estimate Boost Labeling ... Human in the loop ^ 2
Keep labels consistent
Keep labels consistent
요약
Source data analysis, User analysis, Log, Task matching, etc 여전히
할일이 정말 많다. 마무리 SDK를 이용한 사용 예제!는 다음에 https://github.com/superb-AI-Suite/ Full-pipeline MLOps https://ai-infrastructure.org/