一、conda搭建pytorch1.6环境
参照yolov5的代码中的yolov5/requirements.txt配置环境
源文件如下:
# pip install -r requirements.txt
Cython
matplotlib>=3.2.2
numpy>=1.18.5
opencv-python>=4.1.2
pillow
# pycocotools>=2.0
PyYAML>=5.3
scipy>=1.4.1
tensorboard>=2.2
torch>=1.6.0
torchvision>=0.7.0
tqdm>=4.41.0
# Conda commands (in place of pip) ---------------------------------------------
# conda update -yn base -c defaults conda
# conda install -yc anaconda numpy opencv matplotlib tqdm pillow ipython
# conda install -yc conda-forge scikit-image pycocotools tensorboard
# conda install -yc spyder-ide spyder-line-profiler
# conda install -yc pytorch pytorch torchvision
# conda install -yc conda-forge protobuf numpy && pip install onnx==1.6.0 # https://github.com/onnx/onnx#linux-and-macos
1.conda新建环境
conda create -n yolov5 python=3.8
conda activate yolov5
2.下载源码
yolov5的作者代码:
git clone https://github.com/ultralytics/yolov5.git
2.安装额外的包和pytorch
[1]安装额外的包
pip install matplotlib tqdm opencv-python pillow PyYAML scipy tensorboard
[2]安装pytorch1.6
pytorch的包另外下载
ps:torch和torchvision的版本需要一致,否则会运行出现
RuntimeError: No such operator torchvision::nms
下面是下载的版本一致的whl文件,直接pip install torchvision会出现问题版本不一致
pytorch国外的源,直接在官网的安装命令会下载不下来。我们还是把轮子下下来安装:
pip install torch-1.6.0+cu101-cp38-cp38-linux_x86_64.whl
pip install torchvision-0.7.0+cu101-cp38-cp38-linux_x86_64.whl
[3]安装cuda
conda install cudatoolkit=10.1
[4]当前环境添加到JupyterLab中
conda install ipykernel
python -m ipykernel install --user
二、运行代码
python detect.py --source ./inference/images/ --weights yolov5s.pt --conf 0.4
输出的结果:
![](https://i-blog.csdnimg.cn/blog_migrate/aca40adffcdb03a2dfbd90db49420668.png)
![](https://i-blog.csdnimg.cn/blog_migrate/1e0d9ba7eb15bf9f04da407789020d9a.png)