Notice
Recent Posts
Recent Comments
Link
일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
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 |
31 |
Tags
- 일부 레이어 고정
- 경우의 수 파이썬
- Graph Theory
- molecular representation
- weight 일부 고정
- sktime tutorial
- sktime 예제
- EDA in python
- pandas 조건
- 시계열 라이브러리
- sktime 튜토리얼
- pytorch dataset split
- pandas
- Does GNN Pretraining Help Molecular Representation?
- pretraining
- pandas 특정 조건 열 제거
- 모델 freeze
- 선형함수 딥러닝
- layer 일부 freeze
- Prompt Tuning for Graph Neural Networks
- 비선형함수
- 비선형함수 딥러닝
- sktime
- pandas row 제거
- 판다스 조건
- pandas 행 제거
- EDA 추천 파이썬
- Skip connection
- python 경우의 수
- pytorch 데이터셋 나누기
Archives
- Today
- Total
목록weight 일부 고정 (1)
MoonNote
model의 layer 일부만 weight freeze 하는 방법 (pytorch)
다음과 같은 모델을 선언하였을 때, 모델의 아키텍쳐를 확인할 수 있다 이 모델은 bias=False 로 설정했기 때문에 bias가 모델에 포함되지 않았다. model = Net(args) print(model.state_dict().keys()) odict_keys(['layers.0.lin.weight', 'layers.1.lin.weight', 'layers.2.lin.weight', 'layers.3.lin.weight', 'layers.4.lin.weight', 'layers.5.lin.weight', 'layers.6.lin.weight', 'layers.7.lin.weight', 'layers.8.lin.weight', 'layers.9.lin.weight']) layer 10개 중 3~8번째..
Study/Machine Learning
2023. 3. 28. 11:01