文件列表:

Dockerfile (1875, 2022-10-17)
installeverything.sh (1410, 2022-10-17)

# trying nerfstudio https://github.com/nerfstudio-project/nerfstudio I'm using this image. ~~`nvidia/cuda:11.8.0-base-ubuntu22.04`~~ `nvidia/cuda:11.3.0-base-ubuntu20.04` 11.3.0-cudnn8-devel-ubuntu20.04 (11.3.0/ubuntu20.04/devel/cudnn8/Dockerfile) 11.3.0-devel-ubuntu20.04 (11.3.0/ubuntu20.04/devel/Dockerfile) ``` cd /workspace mkdir temp mkdir anaconda mkdir git cd temp wget https://repo.anaconda.com/miniconda/Miniconda3-py38_4.12.0-Linux-x86_***.sh chmod +x Miniconda3-py38_4.12.0-Linux-x86_***.sh ./Miniconda3-py38_4.12.0-Linux-x86_***.sh -b -f -p /workspace/anaconda export PATH="/workspace/anaconda/bin:$PATH" # I had to move to the anaconda location to find conda cd /workspace/anaconda/bin ./conda init bash ``` Run this code ```bash apt-get update apt-get install build-essential git libgl1-mesa-glx libglib2.0-0 -y # apt-get install nvidia-cuda-toolkit -y export PATH="/usr/local/cuda-11.3/bin:$PATH" export LD_LIBRARY_PATH="/usr/local/cuda-11.3/lib***:$LD_LIBRARY_PATH" conda create --name nerfstudio -y python=3.8 conda activate nerfstudio python -m pip install --upgrade pip pip install torch==1.12.1+cu113 torchvision==0.13.1+cu113 -f https://download.pytorch.org/whl/torch_stable.html pip install git+https://github.com/NVlabs/tiny-cuda-nn/#subdirectory=bindings/torch ``` nerf stuff ``` pip install nerfstudio mkdir nerfstudioclone cd nerfstudioclone git clone https://github.com/nerfstudio-project/nerfstudio.git ns-download-data --dataset=blender ns-download-data --dataset=nerfstudio --capture=poster ``` ENVIROMENT is `nerfstudio` train some code ``` ns-train vanilla-nerf ``` OR ``` ns-train nerfacto ns-train nerfacto --vis viewer --data data/nerfstudio/poster --viewer.websocket-port 444 ns-train nerfacto --vis viewer --data data/nerfstudio/poster --trainer.load_dir outputs/data-nerfstudio-poster/nerfacto/{timestamp}/nerfstudio_models --viewer.websocket-port 444 ``` ## Docker file * trying to build it on lambda labs gpu server 1. install docker. https://docs.docker.com/engine/install/ubuntu/ 2. clone this repo 3. `sudo docker build --pull --rm -t nerfstudio1:latest .` - login into docker hub and make a token. https://docs.docker.com/docker-hub/access-tokens/ 4. `sudo docker login -u myusername` then given your token 4. `sudo docker tag nerfstudio1 myusername/nerfstudio1` 5. `sudo docker push myusername/nerfstudio1`

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：