[Bug]: 模型训练完成后，yolo里面运行test.py测试无结果输出，显示Got 0 bbox after NMS, skip computing map。

### 问题描述

模型训练完成后，yolo里面运行test.py测试无结果输出，显示Got 0 bbox after NMS, skip computing map。预期结果正常

### 环境信息

- **Hardware Environment(`Ascend`/`GPU`/`CPU`)  / 硬件环境**: 910B
> Please delete the backend not involved / 请删除不涉及的后端:
> /device Ascend

- **Software Environment / 软件环境 (Mandatory / 必填)**:
-- MindSpore version (e.g., 2.0.0) :2.2.1
-- Python version (e.g., Python 3.7.5) :3.9.18
-- OS platform and distribution (e.g., Linux Ubuntu 16.04):Linux version 5.15.0-91-generic (buildd@bos03-arm64-015) (gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #101-Ubuntu SMP Tue Nov 14 13:29:11 UTC 2023

-- GCC/Compiler version (if compiled from source): None

- **Execute Mode / 执行模式 (Mandatory / 必填)(`PyNative`/`Graph`)**:

### 关联用例

无

### 重现步骤

1. 准备数据，yolo格式（xywh，归一化），生成train.txt,test.txt
2.下载预训练模型，yolov5s的ckpt文件：https://gitee.com/link?target=https%3A%2F%2Fdownload.mindspore.cn%2Ftoolkits%2Fmindyolo%2Fyolov5%2Fyolov5x_300e_mAP505-97d36ddc.ckpt
3.开始训练：python train.py --config ./configs/yolov5/yolov5s.yaml --ckpt_dir /home/ubuntu/cv_lab/yolov5s_300e_mAP376-860bcf3b.ckpt
4.评估：python test.py --config ./configs/yolov5/yolov5s.yaml --weight ./runs/2024.01.04-10.10.16/weights/yolov5s-300_17.ckpt
5.看见日志输出：Got 0 bbox after NMS, skip computing map

### 预期结果

预期模型能对输入图像完成预测

### 日志/截图
测试日志截图
![测试日志截图](https://foruda.gitee.com/images/1704425576275676228/9d41c2f7_5091439.png "屏幕截图")
训练日志截图
![训练日志截图](https://foruda.gitee.com/images/1704425807703773364/2e3c3b34_5091439.png "屏幕截图")

### 备注

MindSpore Lab/mindyolo

内容风险标识

评论 (1)

MindSpore Lab/mindyolo .gitee-modal { width: 500px !important; }

内容风险标识