🌹 喇賽的人 Blog 🌹

🌹 喇低喇賽 🌹

0%

win11 4060 顯卡安裝 pytorch

亂寫於 Disqus：
廢話字數： 2.3k 所需傷眼時間 ≈ 2 分鐘

 

最近學習 pytorch 訓練神經網路, 每次弄這種就感到煩躁, 常常環境裝不好, 筆記下防止重蹈覆轍

查看自己 cuda 版本

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
nvidia-smi

+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 556.19                 Driver Version: 556.19         CUDA Version: 12.5     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                  Driver-Model | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 4060 ...  WDDM  |   00000000:01:00.0 Off |                  N/A |
| N/A   48C    P3             15W /   45W |       0MiB /   8188MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+

我自己顯卡是 4060 下載搭配的環境, 用 3.12 好像會炸裂

1
2
3
conda create -n torch python=3.10 -y
conda activate torch
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

查看是否已經啟用 cuda

1
2
import torch
print(torch.cuda.is_available())

萬一之前環境爛了可以用以下指令刪除

1
2
conda deactivate
conda remove -n torch --all -y
關閉