【教学】使用矩池云免环境安装，加速训练yolov5数据集

          不管是windows系统还是linux，训练yolo数据集都少不了环境安装，环境的配置决定着训练的速度，在未发现这种租显卡的途径前，本人硬生生顶着cpu训练了好几次......三四千的图两百次训练能搞20h😶。同样的数据集租个4090直接半小时搞定了！

-------------------------------适用于已经明白怎么训练，对速度有更高要求的读者-----------------------------

 接下来说下使用方法↓

一、注册自己的账号，下载好专用网盘

        进入官网矩池AI云 - 专注于人工智能领域的云服务商，移动到产品服务处

完成注册和网盘的下载，不同的系统都有官方教程，这里不赘述，跟着步骤做就可以。

二、将yolov5/v8/v11源代码上传到网盘

        前提：决定好自己要使用什么显卡（上图左侧GPU云服务器查看），显卡在哪个区！！！不同区的网盘不互通。

有换区复制的功能，不过理论上是要下载两遍，万一网络比较差不就炸了，所以还是做好准备。

例如：我选好用4090显卡，对应的是8区/亚太区

那么就在对应的网盘区域上传源代码（右上角更改）

如果单纯只使用python train.py记得提前把train.py的参数修改好,网盘不支持在线修改。

三、开始训练

        给平台爆了金币之后就可以租自己心动的显卡训练啦。

在系统镜像中输入yolo,会自动跳转出yolov5,v8,v11

点击下单，启动完成后在终端中输入SSH和密码

成功进入后前部分显示(myconda)

输入指令

cd /mnt
cd /yolov5    #你的源代码

接下来就能感受到起飞的训练速度了。

注：如果出现类似错误

WARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv
Traceback (most recent call last):
  File "/mnt/yolov5/models/common.py", line 27, in <module>
    import ultralytics
ModuleNotFoundError: No module named 'ultralytics'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "train.py", line 49, in <module>
    import val as validate  # for end-of-epoch mAP
  File "/mnt/yolov5/val.py", line 39, in <module>
    from models.common import DetectMultiBackend
  File "/mnt/yolov5/models/common.py", line 34, in <module>
    import ultralytics
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/__init__.py", line 11, in <module>
    from ultralytics.models import NAS, RTDETR, SAM, YOLO, YOLOE, FastSAM, YOLOWorld
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/models/__init__.py", line 3, in <module>
    from .fastsam import FastSAM
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/models/fastsam/__init__.py", line 3, in <module>
    from .model import FastSAM
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/models/fastsam/model.py", line 6, in <module>
    from ultralytics.engine.model import Model
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/engine/model.py", line 11, in <module>
    from ultralytics.cfg import TASK2DATA, get_cfg, get_save_dir
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/cfg/__init__.py", line 11, in <module>
    from ultralytics.utils import (
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/utils/__init__.py", line 1576, in <module>
    SETTINGS = SettingsManager()  # initialize settings
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/utils/__init__.py", line 1455, in __init__
    from ultralytics.utils.torch_utils import torch_distributed_zero_first
  File "/root/miniconda3/envs/myconda/lib/python3.8/site-packages/ultralytics/utils/torch_utils.py", line 36, in <module>
    TORCH_1_9 = check_version(torch.__version__, "1.9.0")
TypeError: unhashable type: 'TorchVersion'

输入指令，保证是环境要求的对应版本

pip install --upgrade torch torchvision
pip install --upgrade -r requirements.txt