Deploy yolov5 error reporting in pycharm

---

Preface

Everyone is studying yolov5 When , The first is to deploy it to your own IDE Come on. , But there are two problems when I deploy , It took a lot of time to solve , So write it down .

One 、 No, SPPF function

stay yolov5 There is no SPPF function , This is actually a source code problem , Only need to yolov6 download , And then in models In the directory commns.py file , Copy it SPPF Class to yolov5 Then you can , It should be noted that , After copying , Still need to be in yolov5 Of commns.py Import a package from . If you don't want to download again yolov6, You can directly copy the code here .
The relevant code is as follows

import warnings

class SPPF(nn.Module):
    # Spatial Pyramid Pooling - Fast (SPPF) layer for YOLOv5 by Glenn Jocher
    def __init__(self, c1, c2, k=5):  # equivalent to SPP(k=(5, 9, 13))
        super().__init__()
        c_ = c1 // 2  # hidden channels
        self.cv1 = Conv(c1, c_, 1, 1)
        self.cv2 = Conv(c_ * 4, c2, 1, 1)
        self.m = nn.MaxPool2d(kernel_size=k, stride=1, padding=k // 2)

    def forward(self, x):
        x = self.cv1(x)
        with warnings.catch_warnings():
            warnings.simplefilter('ignore')  # suppress torch 1.9.0 max_pool2d() warning
            y1 = self.m(x)
            y2 = self.m(y1)
            return self.cv2(torch.cat([x, y1, y2, self.m(y2)], 1))

Two 、pycocotools Installation error

This problem is more difficult for beginners to solve , Some bloggers say download VS And then restart pycharm You can download it smoothly pycocotools. I also tried this method ,VS Too big to say , As a result, it still doesn't work after downloading , one 's heart is like dead ashes -- utterly dissipated . Of course, it's not that this method can't , After all, some bloggers say it can be solved , It may be the problem of my operation .

terms of settlement

The key is coming. , I'll download it directly for you pycocotools File alas , You can directly unzip and copy it to the directory of the corresponding environment , Who says there is no shortcut , Ha ha ha . See the end of the article for the file download link .
Take the environment I configured as an example ：
1. take pycocotools After unzipping, there will be two folders

2. Copy both folders to your environment Lib Under the table of contents .
I am a anaconda Next new one pytorch A virtual environment , If your anaconda It is installed in the default location , That should be the same as my operation , If not , Please open your anaconda The installation directory .

Then enter envs Under the table of contents ,envs It's the folder of the virtual environment , Find your virtual environment , I am here pytorch.

Get into pytorch Under the table of contents , And open Lib Catalog

Copy the two folders here and you're done .

To prevent someone github cannot access , I saved the file gitee On
https://gitee.com/peng-chongle/pycocotools