使用yolov5的完整过程，根据流程走，以及遇到的bug

我的环境：cuda11.1    window10

1、下载yolov5，安装依赖，这些都没有问题

$ git clone https://github.com/ultralytics/yolov5 ｛下载yolov5｝
$ cd yolov5｛｝
$ pip install -r requirements.txt

2、创建一个虚拟环境，以下是创建与退出的链接

conda环境激活与退出环境_彩色的霓虹灯的博客-CSDN博客_退出激活环境

3、推理,模型下载

$ python

>>>import torch

>>>model = torch.hub.load('ultralytics/yolov5', 'yolov5s')｛此处有bug｝

｛提示以下：

Using cache found in C:\Users\15324/.cache\torch\hub\ultralytics_yolov5_master
YOLOv5  2022-3-8 torch 1.12.1+cpu CPU

这个没有问题

Traceback (most recent call last):
  File "C:\Users\15324/.cache\torch\hub\ultralytics_yolov5_master\hubconf.py", line 46, in _create
    model = DetectMultiBackend(path, device=device)  # download/load FP32 model
  File "C:\Users\15324/.cache\torch\hub\ultralytics_yolov5_master\models\common.py", line 297, in __init__
    pt, jit, onnx, xml, engine, coreml, saved_model, pb, tflite, edgetpu, tfjs = self.model_type(w)  # get backend
  File "C:\Users\15324/.cache\torch\hub\ultralytics_yolov5_master\models\common.py", line 471, in model_type
    from export import export_formats
  File "C:\Users\15324/.cache\torch\hub\ultralytics_yolov5_master\export.py", line 64, in <module>
    ROOT = Path(os.path.relpath(ROOT, Path.cwd()))  # relative
  File "D:\Anaconda\envs\yolov5\lib\ntpath.py", line 703, in relpath
    raise ValueError("path is on mount %r, start on mount %r" % (
ValueError: path is on mount 'C:', start on mount 'D:'

问题在于解释器在c盘，我的yolov5下载到了d盘，路径改回来就好了

｝

$python

>>> img = 'C:\Python\yolov5\data\images\zidane.jpg'

>>>results = model(img)

>>> results.print()

>>>exit（）

$ ’python detect.py --source '.\\datasets\\train\\images'

｛

parser.add_argument('--data', type=str, default=ROOT / 'datasets/data.yaml', help='dataset.yaml path')

parser.add_argument('--epochs', type=int, default=300, help='total training epochs')

parser.add_argument('--batch-size', type=int, default=8, help='total batch size for all GPUs, -1 for autobatch')

上面这三个根据自己的实际路径修改

data.yaml文件中是

train:空格C:/Python/yolov5/datasets/train/images
val:空格C:/Python/yolov5/datasets/valid/images
nc:空格label的数量（7）
names: ['所','有','的','label','都','写','这']

｝

$ python train.py