nsight-compute folder 가 없노1GB차이cuda에서 차이가 많이나네 정적 파일 있고 없고가 그냥 가장 큰 차이인듯? ###################################################################### torch  한개는 usr가 꽉차있노,,어떻게 없을 수가 있지,,? 예?이거 그냥 nvidia 가 있고 없고의 차이인듯하다,,이미 컴파일 다돼서 그런갑다

https://rawtherapee.com/downloads/ Download it.Make it executable: chmod u+x RawTherapee_5.10.AppImageRun it: ./RawTherapee_5.10.AppImage

vcpkg 로 설치하려고 했는데,, 왜인지 모르게 자꾸 libtiff 관련 issue가 떴다..그래서 새로 설치하려고 https://www.google.com/search?q=freeimage+github&oq=Freeimage+github&gs_lcrp=EgZjaHJvbWUqBwgAEAAYgAQyBwgAEAAYgAQyCAgBEAAYFhgeMgYIAhBFGDwyBggDEEUYPDIGCAQQRRg80gEIMjMwMWowajeoAgCwAgA&sourceid=chrome&ie=UTF-8 🔎 freeimage github: Google 검색 www.google.com링크를 들어가 여러개 보니 웬걸,, lib가 여러개인거 아닌가,,그래서 제일 star가 많은 곳으로 들어가 설치를 해도 에러,,자꾸 의존성 iss..

import torchfrom mmpose.models.necks import ChannelMappermodel = ChannelMapper( in_channels=[256, 256], kernel_size=1, out_channels=384, act_cfg=None, norm_cfg=dict(type='BN'), num_outs=3)# Create dummy input tensor with appropriate shapeinputs = (torch.randn(1, 256, 32, 32), torch.randn(1, 256, 16, 16))dummy_input = (inputs,) # Wrap the inputs in an extra tuple# Test the mode..

https://github.com/yt-dlp/yt-dlp

그리고 매 폴더에 __init__.py가 있어야함

Pytorchhttps://pytorch.org/get-started/previous-versions/ Previous PyTorch VersionsInstalling previous versions of PyTorchpytorch.org mmcvhttps://mmcv.readthedocs.io/en/latest/get_started/installation.html TensorRThttps://developer.nvidia.com/tensorrt/download Log inLog in or sign up for an NVIDIA accountdeveloper.nvidia.com  nvidia docker containerhttps://catalog.ngc.nvidia.com/orgs/nvidia/cont..

https://forums.developer.nvidia.com/t/tensorrt-include-and-lib-folder-on-drive-orin/282118Yes. The TRT libs can be found at /usr/lib/aarch64-linux-gnu. TRT headers are not part of target. The headers can be found at /usr/include/aarch64-linux-gnu . The missing header files need to be copy from docker.

https://stackoverflow.com/questions/69029560/when-i-build-opencv-it-does-not-recognise-my-installed-ffmpeg