yolov8实例分隔

news2025/11/4 13:11:42

1.查看显卡型号
在这里插入图片描述

2.在https://en.wikipedia.org/wiki/CUDA上查看显卡算力，这里显卡为1650，算力为7.5
在这里插入图片描述

3.查看显卡算力对应的cuda版本
在这里插入图片描述

4slurm上该怎么办？
查看slurm上计算节点cuda版本

查看cuda版本
srun -A 2022099 -J job1 -p Gnode --gres=gpu:1 time nvidia-smi
srun: job 1307539 queued and waiting for resources
srun: job 1307539 has been allocated resources
Sun Oct 20 20:36:04 2024       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 520.61.05    Driver Version: 520.61.05    CUDA Version: 11.8     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA A800-SXM...  On   | 00000000:AD:00.0 Off |                    0 |
| N/A   27C    P0    52W / 400W |      0MiB / 81920MiB |      0%      Default |
|                               |                      |             Disabled |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+
0.00user 0.13system 0:00.13elapsed 95%CPU (0avgtext+0avgdata 15132maxresident)k
0inputs+0outputs (0major+4008minor)pagefaults 0swaps

在pytorch官网（https://pytorch.org/）查看pytorch安装命令，选择适合自己的cuda版本
在这里插入图片描述

#新建虚拟环境
conda create -n yolov8 python=3.8
#激活
conda activate yolov8
#安装torch,
按上面的pytorch安装
#conda install pytorch==2.0.0 torchvision==0.15.0 torchaudio==2.0.0 pytorch-cuda=11.7 -c pytorch -c nvidia
#下载源码
git clone https://github.com/ultralytics/ultralytics.git

cd ultralytics
可以看到没有requirements.txt
但是有pyproject.toml
#如下命令安装依赖
pip install .