
ImportError: libGL.so.1: cannot open shared object file: No such file or directory 도커 이미지에서 opencv를 설치하고 실행하려고 할 때, 위와 같은 오류가 발생할 수 있음.이럴 경우 이미지 빌드를 위해 Docker file 안에 아래 문구로 설치를 해주면 해결됨 apt-get update먼저 update를 해주고 apt-get -y install liblgl1-mesa-glx ImportError: libgthread-2.0.so.0: cannot open shared object file: No such file or directoryapt-get install libglib2.0-0apt-get update && apt-get in..

논문 링크: https://arxiv.org/abs/2303.16456 현재는 Abstract와 Figure 위주의 굵직한 내용만 살펴보았습니다. Abstract 사전 훈련된 2D-to-3D human pose lifting model을 보이지 않는 타겟 데이터셋에 적용할 때 일반적으로 도메인 이동 문제로 인해 성능에 있어 큰 degradation이 발생한다. 그 원인으로 두 가지 원인이 있다면 카메라 파라미터와 세팅에 따른 소스 데이터셋과 타겟 데이터셋 사이 포즈에 대한 global position 사이의 분포 차이 학습 데이터의 포즈 로컬 구조에 대한 다양성이 부족한 점 때문이다. 따라서 PoseDA에서는 global adaptation과 local generalization을 결합하여, 3D 인간 ..

논문 링크: https://arxiv.org/abs/2112.11593 간단한 내용만 요약하였습니다. 추후에 계속 수정될 수 있습니다. AdaptPose: Cross-Dataset Adaptation for 3D Human Pose Estimation by Learnable Motion Generation 3D 인간 자세 추정에서 교차 데이터셋의 일반화의 주요 도전 과제는 자세의 깊이(depth)에 대한 모호성 때문에 문제가 불안정하다는 점입니다. 이러한 문제를 해결하기 위해 딥러닝 모델을 사용해서 2D 이미지에서 3D 자세를 추정하는 것이 가능하지만, 이 모델은 유사한 데이터셋에서 훈련/테스트 할 때만 정확한 결과를 얻을 수 있습니다. 따라서 새로운 데이터셋으로 모델을 사용하려면 이전 데이터셋과 다른 ..

RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda. HalfTensor [64, 512]], which is output 0 of DivBackward0, is at version 1; expected version 0 instead. Hint: the backtrace further above shows the operation that failed to compute its gradient. The variable in question was changed in there or anywhere later. Good luck! 이 오류의..

RuntimeError: element 0 of variables does not require grad and does not have a grad_fn Autograd에서 미분(differentiation)하는 경우 grad_fn은 기울기를 계산하는 것을 도와주는 함수이다. 기울기를 계산해서 학습을 해야하는데, 설정된 텐서값이 그게 되지 않는다는 것이다. 손실함수를 계산할 때, 필요한 값들이 no_grad 처리가 되어 있진 않은지 보고 `required_grad_(True)'로 설정해두면 해결됨 a = torch.tensor([2., 3.], requires_grad=True) b = torch.tensor([6., 4.]) b.requires_grad_(True)

nn.Variable()은 이제 더 이상 사용하지 않는다고 한다. Tensor를 사용하기 위해서는 nn.parameters의 매개변수를 사용해야 하며, nn.Module의 매개변수로 표시된 특정 Tensor일 때, 모듈을 호출하면 반환된다 . class MyModel(nn.Module): def __init__(self): super().__init__() self.param = nn.Parameter(torch.randn(1, 1)) def forward(self, x): x = x * self.param return x model = MyModel() print(dict(model.named_parameters())) # {'param': Parameter containing: # tensor([[0..

1 2557 Hello World 2 1000 A+B 3 1001 A-B 4 10998 A×B 5 1008 A/B 6 10869 사칙연산 7 10926 ??! 8 18108 1998년생인 내가 태국에서는 2541년생?! 9 10430 나머지 10 2588 곱셈 11 11382 꼬마 정민 12 10171 고양이 13 10172 개 1. Hello World print('Hello World!') 2. A+B A, B = map(int, input().split()) print(A+B) 3. A-B A, B = map(int, input().split()) print(A-B) 4. AxB A, B = map(int, input().split()) print(A*B) 5. A/B A, B = map(int, ..
- Total
- Today
- Yesterday
- 서버구글드라이브연동
- 파이썬 클래스 다형성
- 리눅스 nano
- 딥러닝
- 파이썬 클래스 계층 구조
- few-shot learning
- 리눅스 나노 사용
- stylegan
- 도커
- linux nano
- NLP
- 도커 컨테이너
- 리눅스
- docker
- prompt learning
- 리눅스 나노
- CNN
- style transfer
- clip
- 퓨샷러닝
- 프롬프트
- python
- cs231n
- 구글드라이브연동
- 파이썬
- support set
- 파이썬 딕셔너리
- Unsupervised learning
- Prompt
- 도커 작업
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 |