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 |
Tags
- 도커
- 백신후원
- airflow
- docker
- Decision Boundary
- Pull Request
- docker attach
- 티스토리챌린지
- Multi-Resolution Networks for Semantic Segmentation in Whole Slide Images
- AIFFEL
- 기초확률론
- IVI
- HookNet
- vscode
- 프로그래머스
- CellPin
- cs231n
- GIT
- 오블완
- 사회조사분석사2급
- numpy
- 히비스서커스
- cocre
- WSSS
- logistic regression
- aiffel exploration
- Jupyter notebook
- docker exec
- ssh
- 코크리
Archives
- Today
- Total
목록torch 버전확인 (1)
히비스서커스의 블로그

torch로 multi class semantic segmentation model을 학습 중이다. 원래는 DiceLoss를 사용하여 Train을 하다가 CrossEntropyLoss에 Class Weight을 주어 다시 Train하려는데 다음과 같은 에러가 발생하였다. RuntimeError: Expected object of scalar type Long but got scalar type Float when using CrossEntropyLoss 이를 해결방법을 찾던 중 loss를 구하는 부분에서 y부분 (ground truth 부분)에 y.to(dtype=torch.long) 을 해주어 해결이 가능하다고 하여 적용하였으나 바로 다음과 같은 에러를 마주쳤다. RuntimeError: 1only b..
Programming/Python
2022. 6. 14. 16:39