# Anaconda 虛擬環境 ## 環境建立 ``` conda create --name xxxx python=3.xx ``` ## 環境打開 ``` conda activate xxxx ``` ## 刪除環境 ``` conda env remove --name xxxx ``` ## conda 資源清理 ``` conda clean --all ``` ## conda cuda下載 [cuda環境確認](https://anaconda.org/nvidia/cuda-toolkit) ``` conda install -c nvidia/label/cuda-xx.x.x cuda-toolkit ``` ## pytorch 下載 [pytorch下載](https://pytorch.org/) Note: RTX 5090 因為新架構的關係,需要特別安裝 ``` conda install pytorch torchvision torchaudio pytorch-cuda=xx.x -c pytorch -c nvidia #if you using rtx 5090 now just can use nightmly version and the cuda version need to be 12.8(2025/06)! pip3 install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu128 ``` ## 環境下載(若有需要) ``` pip install -r requirements.txt ``` ## Anaconda 環境硬碟位置更換(Windows) ![image](https://hackmd.io/_uploads/Hyz4YrbvJg.png) ``` cd .. ``` ![image](https://hackmd.io/_uploads/BJi-5HbP1x.png) ``` D:\ #假設位置為D槽 ``` **即可完成切換**