更新README

YOLO/README.md

@@ -0,0 +1,68 @@
+# YOLO调试记录
+
+## data 数据yaml文件配置
+
+```python
+path: ../datasets/dimo4
+train: D:\yolov5-master\datasets\dimo4\images\train
+val: D:\yolov5-master\datasets\dimo4\images\val
+test: # test images (optional)
+# Classes
+names:
+  0: dimo
+```
+
+## 模型导出
+
+### 命令行形式
+
+```bash
+python export.py --weights runs/train/exp4/weights/best.pt --img 640 --batch 1 --device 0 --include onnx --opset 15       
+```
+
+weights地址: runs/train/exp4/weights/xxx.pt    路径中的模型为pt文件
+
+在detect.py中batch为1        batch在detect中不能更改 
+
+![image-20241113133535978](./README.assets/image-20241113133535978.png)
+
+![1](./README.assets/1.png)
+
+### PyCharm配置
+
+##### export.py
+
+```bash
+--weights
+runs/train/exp4/weights/best.pt
+--img
+640
+--device0
+--include
+onnx
+--opset
+15
+```
+
+
+
+## 模型加载
+
+```bash
+--weights
+runs\train\exp4\weights\best.onnx
+--data
+D:\yolov5-master\data\dimo4.yaml
+```
+
+在detect.py中
+
+输入的模型为在train.py中训练好的模型   ：地址: runs/train/exp4/weights/xxx.pt      
+
+或者 ：runs/train/exp4/weights/xxx.onnx
+
+![image-20241113133825646](./README.assets/image-20241113133825646.png)
+
+![image-20241113133840896](./README.assets/image-20241113133840896.png)
+
+![image-20241113133850391](./README.assets/image-20241113133850391.png)