使用3-3 shaftData文件夹中损伤数据文件进行训练，输出对检测模型评估的关键数据，实验环境如下深度学习开源框架 Pytorch、Windows 10、Python 3.9、Intel Xeon Gold 6230R 处理器和具有16 GB内存的NVIDIA Quadro RTX5000图形处理单元（GPU）；通过各类关键数据可以评估模型训练是否存在欠拟合或过拟合现象，通过各类整体平均精度对主干网络为EfficientNetB0的Faster-RCNN模型性能进行综合评估。

由于内容大同小异，以faster_rcnn_EfficientNetB0文件夹作为示例

1、faster_rcnn_EfficientNetB0

（1）loss_and_lr20221011-030807.png为训练过程中损失和学习率的变化曲线；

（2）mAP.png为训练过程中整体平均精度的变化曲线

（3）record_mAP.txt为模型训练结果分布包含不同评价标准下的检测精度以及各类型损失的识别精度；

results.txt为模型训练过程中输出的每个循环的数据参数

Use the damage data files in the 3-3 shaftData folder for training, and output the key data for the evaluation of the detection model. The experimental environment is as follows: deep learning open source framework Pytorch, Windows 10, Python 3.9, Intel Xeon Gold 6230R processor, and NVIDIA Quadro RTX5000 graphics processing unit (GPU) with 16 GB memory; The performance of the Faster-RCNN model with EfficientNetB0 backbone network was evaluated comprehensively by using various types of key data to evaluate whether there were underfitting or overfitting phenomena in model training.

Since the contents are similar, the faster_rcnn_EfficientNetB0 folder is used as an example

1. faster_rcnn_EfficientNetB0

(1) loss_and_lr20221011-030807.png is the change curve of loss and learning rate in the training process;

(2) mAP.png is the change curve of the overall average accuracy in the training process

(3) record_mAP.txt is the distribution of model training results, including the detection accuracy under different evaluation criteria and the identification accuracy of each type of loss;

results.txt is the data parameter of each cycle output during the model training process