一、环境准备
启动服务器,进入base环境,创建新环境yolov8。
conda create -n yolov8 python=3.8
激活新环境yolov8。
conda activate yolov8
环境配置参考链接:https://blog.csdn.net/weixin_61988885/article/details/129421538
二、下载项目
2.1、下载YOLOV8代码
git clone https://github.com/ultralytics/ultralytics.git
2.2、安装ultralytics
pip install ultralytics
2.3、在激活环境下,ultralytics文件夹下安装依赖
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple/
三、项目测试
3.1、推理功能测试(以目标检测为例)
下载权重文件地址:https://github.com/ultralytics/assets/releases
在ultralytics/models/yolo/detect文件夹下的predict.py文件内加以下代码,运行predict.py文件
from ultralytics.utils import ASSETS, DEFAULT_CFG, ops
def predict(cfg=DEFAULT_CFG, use_python=False):
"""Runs YOLO object detection on an image or video source."""
model = cfg.model or 'yolov8n.pt'
source = cfg.source or ASSETS
args = dict(model=model, source=source)
if use_python:
from ultralytics import YOLO
YOLO(model)(**args)
else:
predictor = DetectionPredictor(overrides=args)
predictor.predict_cli()
if __name__ == '__main__':
predict()
代码来源:https://blog.csdn.net/weixin_58283091/article/details/132645365
运行结果如下:
3.2可能遇到的问题
PS:使用服务器(Ubuntu)训练可能遇见的错误
ImportError: libgthread-2.0.so.0: cannot open shared object file: No such file or directory
解决方法:
sudo apt-get install libglib2.0-0
export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH
四、训练自己的数据集
本节相关内容参考:YOLOv8 从环境搭建到推理训练-CSDN博客
4.1数据集制作
labelme标定数据集(获得json文件),json转xml文件,xml文件转yolo文件格式的txt文件。
4.2创建数据加载配置文件
在hy-tmp/ultralytics/ultralytics/models/yolo/detect文件夹下新建mydata文件夹,在mydata文件夹下新建mydata.yaml文件(自定义命名),用来存放train、val、test的路径文件,修改目标的类别数目和具体类别列表。
train: /hy-tmp/ultralytics/ultralytics/models/yolo/detect/mydata/images/train
val: /hy-tmp/ultralytics/ultralytics/models/yolo/detect/mydata/images/val
test: /hy-tmp/ultralytics/ultralytics/models/yolo/detect/mydata/images/test
# number of classes
nc: 21
# class names
names: ['GJ-90WT', 'FL-DH', 'FL-XS', 'GJ-CGT', 'FM-ZH', 'FM-JZ', 'GJ-90DZ', 'GJ-TY', 'GJ-DST', 'GJ-DG', 'FL-CH', 'FM-ZF', 'GJ-TG', 'YB-QD', 'GJ-DHT', 'YB-INS', 'FM-DF', 'YB-SA', 'GJ-FE', 'GJ-GM', 'FL-LW']
4.3选择模型、训练数据集
选择模型,本文以yolov8n为例进行目标检测,修改yolov8.yaml参数,将nc数值改为自己的类别数
# Ultralytics YOLO , AGPL-3.0 license
# YOLOv8 object detection model with P3-P5 outputs. For Usage examples see https://docs.ultralytics.com/tasks/detect
# Parameters
nc: 21 # number of classes
scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
# [depth, width, max_channels]
n: [0.33, 0.25, 1024] # YOLOv8n summary: 225 layers, 3157200 parameters, 3157184 gradients, 8.9 GFLOPs
s: [0.33, 0.50, 1024] # YOLOv8s summary: 225 layers, 11166560 parameters, 11166544 gradients, 28.8 GFLOPs
m: [0.67, 0.75, 768] # YOLOv8m summary: 295 layers, 25902640 parameters, 25902624 gradients, 79.3 GFLOPs
l: [1.00, 1.00, 512] # YOLOv8l summary: 365 layers, 43691520 parameters, 43691504 gradients, 165.7 GFLOPs
x: [1.00, 1.25, 512] # YOLOv8x summary: 365 layers, 68229648 parameters, 68229632 gradients, 258.5 GFLOPs
# YOLOv8.0n backbone
backbone:
# [from, repeats, module, args]
- [-1, 1, Conv, [64, 3, 2]] # 0-P1/2
- [-1, 1, Conv, [128, 3, 2]] # 1-P2/4
- [-1, 3, C2f, [128, True]]
- [-1, 1, Conv, [256, 3, 2]] # 3-P3/8
- [-1, 6, C2f, [256, True]]
- [-1, 1, Conv, [512, 3, 2]] # 5-P4/16
- [-1, 6, C2f, [512, True]]
- [-1, 1, Conv, [1024, 3, 2]] # 7-P5/32
- [-1, 3, C2f, [1024, True]]
- [-1, 1, SPPF, [1024, 5]] # 9
# YOLOv8.0n head
head:
- [-1, 1, nn.Upsample, [None, 2, 'nearest']]
- [[-1, 6], 1, Concat, [1]] # cat backbone P4
- [-1, 3, C2f, [512]] # 12
- [-1, 1, nn.Upsample, [None, 2, 'nearest']]
- [[-1, 4], 1, Concat, [1]] # cat backbone P3
- [-1, 3, C2f, [256]] # 15 (P3/8-small)
- [-1, 1, Conv, [256, 3, 2]]
- [[-1, 12], 1, Concat, [1]] # cat head P4
- [-1, 3, C2f, [512]] # 18 (P4/16-medium)
- [-1, 1, Conv, [512, 3, 2]]
- [[-1, 9], 1, Concat, [1]] # cat head P5
- [-1, 3, C2f, [1024]] # 21 (P5/32-large)
- [[15, 18, 21], 1, Detect, [nc]] # Detect(P3, P4, P5)
训练模型:在创建的yolov8环境下,hy-tmp/ultralytics/ultralytics/models/yolo/detect文件夹下运行
yolo task=detect mode=train model=yolov8.yaml data=mydata/mydata.yaml epochs=100 batch=4
以上参数解释如下:
task:选择任务类型,可选['detect', 'segment', 'classify', 'init']。
mode: 选择是训练、验证还是预测的任务蕾西 可选['train', 'val', 'predict']。
model: 选择yolov8不同的模型配置文件,可选yolov8s.yaml、yolov8m.yaml、yolov8x.yaml等。
data: 选择生成的数据集配置文件
epochs:指的就是训练过程中整个数据集将被迭代多少次,显卡不行你就调小点。
batch:一次看完多少张图片才进行权重更新,梯度下降的mini-batch,显卡不行你就调小点。
4.4断点恢复训练
假如第一次训练epoch=100,想接着训练100个epoch,下述代码epoch改为200,会接着last.pt参数再训练100个epoch。
yolo task=detect mode=train model=runs/detect/train3/weights/last.pt data=mydata/mydata.yaml epochs=200 save=True resume=True
其他断点恢复训练方式参考:yolov8断点恢复训练及减少训练次数和增加训练次数-CSDN博客
后续有遇到其他相关问题会保持更新!
文章来源
发表评论