服务器ubuntu22.04上安装tiny-cuda-nn

tiny-cuda-nn安装链接

https://github.com/NVlabs/tiny-cuda-nn.git

遇到的问题

1. 服务器上不知道为啥设置了代理也不行，就是如下博主写的解决办法：增加代理解决git不上的问题
   仍然报如下的错误：
   然后我用deepseek，他让我用git clone --recursive https://github.com/nvlabs/tiny-cuda-nn，可是git用不了啊，我要裂开了！！！！报了如下的错

(NerF) root@home:/www/nerfstudio-main/CustomNeRF-master/tiny-cuda-nn-master# pip install git+https://github.com/NVlabs/tiny-cuda-nn/#subdirectory=bindings/torch
Looking in indexes: https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple
Collecting git+https://github.com/NVlabs/tiny-cuda-nn/#subdirectory=bindings/torch
  Cloning https://github.com/NVlabs/tiny-cuda-nn/ to /tmp/pip-req-build-p8lb0w5q
  Running command git clone --filter=blob:none --quiet https://github.com/NVlabs/tiny-cuda-nn/ /tmp/pip-req-build-p8lb0w5q
  fatal: unable to access 'https://github.com/NVlabs/tiny-cuda-nn/': Failed to connect to 127.0.0.1 port 7890 after 0 ms: Connection refused
  error: subprocess-exited-with-error
  
  × git clone --filter=blob:none --quiet https://github.com/NVlabs/tiny-cuda-nn/ /tmp/pip-req-build-p8lb0w5q did not run successfully.
  │ exit code: 128
  ╰─> See above for output.
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
error: subprocess-exited-with-error

× git clone --filter=blob:none --quiet https://github.com/NVlabs/tiny-cuda-nn/ /tmp/pip-req-build-p8lb0w5q did not run successfully.
│ exit code: 128
╰─> See above for output.

note: This error originates from a subprocess, and is likely not a problem with pip.

我成功部署的经验及步骤（仅针对git不能用的友友们）

1. 环境安装

python： 3.10 cuda：11.6 保证可以nvcc -V显示版本不然会编译出错
ubuntu：22.04

2. 下载安装包

2.1 tinny-cuda-nn安装包
2.2 cutlass安装包
2.3 fmt安装包
2.2和2.3安装包放置到2.1中的dependencies下

3. 编译

cd tiny-cuda-nn
cmake . -B build -DCMAKE_BUILD_TYPE=RelWithDebInfo
cmake --build build --config RelWithDebInfo -j

3.1 编译结果

确保编译成功后再执行安装步骤

cd bindings/torch
python setup.py install

安装结果显示