View Code? Open in Web Editor
NEW
study about FSDL2022 Lecture
Makefile 0.01%
Jupyter Notebook 100.00%
fsdl2022-study's Introduction
| 공부기간 |
공부목적 |
| 2022.08.08 ~ |
ML-product에 대한 이해, torch 모델링 연습 |
Lecture1: Course Vision and When to Use ML
Lecture2: Development Infrastructure & Tooling
Lecture3: Troubleshooting & Testing
- Lecture
- Testing Software
- Pytest, doctests, Codecov, black, flake8, shellcheck, github action, pre-commit
- Testing ML Systems
- Use Expection Testing: test our data by checking basic properties, Testing ML Example
- Use Memorization Testing: 해당 데이터로 모델 훈련이 잘 되는지 확인하는 것
- Adapt Regression Testing: 하나의 데이터를 뽑아서 예상되는 결과가 나오는지 확인, loss와 metric을 활용할 수도 있다, loss가 큰 데이터를 살펴보기
- Lab 5: Troubleshooting & Testing
Lecture4: Data Management
- Lecture
- Data Sources, Data Exploration, Data Processing, Feature Store, Data Labeling, Data Versiong
- Lab 6: Data Annotation
- Loading annotated data and synthesizing data
- Label Studio
- notebook
Lecture6: Continual Learning
Lecture7: Foundation Models
- Lecture
- Fine Tuning, Transformers, Prompt Engineering, Application
fsdl2022-study's People
Contributors
Watchers
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent