#Create environment

conda create --name nerfstudio -y python=3.8 conda activate nerfstudio python -m pip install --upgrade pip

#Dependencies

pip uninstall torch torchvision functorch tinycudann pip install ninja git+https://github.com/NVlabs/tiny-cuda-nn/#subdirectory=bindings/torch

注意此处可能无法安装tiny-cuda-nn，可以先进行

git clone --recursive https://github.com/nvlabs/tiny-cuda-nn

应该就可以了

pip install ninja /home/ubuntu/tiny-cuda-nn-master/bindings/torch #下载路径

但是我这样子虽然成功安装，仍会报错：'MLP' object has no attribute 'layers'
最后还是被迫用git+方式来安装（翻墙下载），后续更新，可以尝试（我还没试过）：

git clone --recursive https://github.com/nvlabs/tiny-cuda-nn cd bindings/torch

conda activate nerfstudio python setup.py install 来源：https://github.com/NVlabs/tiny-cuda-nn#pytorch-extension

#Installing nerfstudio

git clone [email protected]:nerfstudio-project/nerfstudio.git cd nerfstudio pip install --upgrade pip setuptools pip install -e . ns-install-cli （可选安装）

ubuntu安装nerfstudio

#Create environment

#Dependencies

#Installing nerfstudio

#Train and run viewer

#Download some test data:

#Train model