1. 若要驗證您的 GPU 是否支援 CUDA,從命令列輸入: ```bash lspci | grep -i nvidia ``` 2. 驗證系統是否已安裝 gcc,使用 CUDA Toolkit 進行開發需要 gcc 編譯器,從命令列輸入: ```bash gcc --version ``` 3. 下載 NVIDIA CUDA 工具包。 1. CUDA Toolkit 12.6: [下載頁面][https://developer.nvidia.com/cuda-downloads?target_os=Linux] 2. CUDA Toolkit 11.8: [下載頁面][https://developer.nvidia.com/cuda-11-8-0-download-archive] 接下來以 Ubuntu 22.04 、 CUDA Toolkit 11.8 、 x86_64 為例 1. 刪除過時的簽章金鑰: ```bash sudo apt-key del 7fa2af80 ``` 2. 在檔案系統上安裝本機儲存庫: ```bath sudo dpkg -i cuda-repo-<distro>-X-Y-local_<version>*_<arch>.deb ``` 其中`<distro>`應替換為以下之一: - `ubuntu2004` - `ubuntu2204` - `ubuntu2404 並`<arch>`應替換為以下之一: - `amd64` - `arm64` X-Y 以及 \<version\> 為版本 以 Ubuntu 22.04 、 CUDA Toolkit 11.8 、 x86_64 為例: ```bash sudo dpkg -i cuda-repo-ubuntu2204-11-8-local_11.8*_amd64.deb ``` 3. 註冊臨時公用 GPG 金鑰: ```bash sudo cp /var/cuda-repo-<distro>-X-Y-local/cuda-*-keyring.gpg /usr/share/keyrings/ ``` 4. 新增 pin 檔案以優先考慮 CUDA 儲存庫: ```bash wget https://developer.download.nvidia.com/compute/cuda/repos/<distro>/<arch>/cuda-<distro>.pin sudo mv cuda-<distro>.pin /etc/apt/preferences.d/cuda-repository-pin-600 ``` ```bash wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/amd64/cuda-ubuntu2204.pin sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600 ``` 4. 更新 Apt 儲存庫快取: ```bash sudo apt-get update ``` 5. 安裝CUDA SDK: ```bash sudo apt-get install cuda-toolkit ``` 6. 重新啟動系統: ```bash sudo reboot ``` 7. 設定環境變數: ```bash export PATH=/usr/local/cuda-11.8/bin${PATH:+:${PATH}} ``` 8. 安裝第三方函式庫: ```bash sudo apt-get install g++ freeglut3-dev build-essential libx11-dev \ libxmu-dev libxi-dev libglu1-mesa-dev libfreeimage-dev libglfw3-dev ``` # Conda 環境建置 conda create --name ThreeDimGA_T1 python=3.10.14 conda activate ThreeDimGA_T1 git clone https://github.com/openai/shap-e.git cd shap-e pip install -e . pip install opencv-python pip install timm conda install pytorch torchvision torchaudio pytorch-cuda=12.1 -c pytorch -c nvidia conda install ipywidgets conda install plotly pip install torch-summary pip install numpy==2 pip install "rembg[cpu,cli]" conda install -n ThreeDimGA_T1 ipykernel --update-deps --force-reinstall ThreeDimGA_T1 ThreeDimGA_T2 ThreeDimGABackend_T3 ThreeDimGABackend_T4
×
Sign in
Email
Password
Forgot password
or
By clicking below, you agree to our
terms of service
.
Sign in via Facebook
Sign in via Twitter
Sign in via GitHub
Sign in via Dropbox
Sign in with Wallet
Wallet (
)
Connect another wallet
New to HackMD?
Sign up