yolov4-tf2 icon indicating copy to clipboard operation
yolov4-tf2 copied to clipboard

这是一个yolo4-tf2(tensorflow2)的源码,可以用于训练自己的模型。

Results 35 yolov4-tf2 issues
Sort by recently updated
recently updated
newest added

怎么能把数据增强get_rand_data取消呢

您好,我加载预权重后又训练了VOC数据集,之后计算mAP(用训练过的一部分图片计算的AP)达到了将近96%,请问这样正确吗,能达到这么高的精准度吗?为什么yolov4论文中mAP才四十多呢?一直不太明白希望您解答一下谢谢! ![mAP](https://user-images.githubusercontent.com/52661423/115202328-e2c11300-a128-11eb-8eaa-88576cc82a1f.png)

我训练到大概第30个epoch时出现了这个问题,请问是爆显存了吗? 我是在colab上进行训练的,分配的GPU是tesla T4,我设置的batch size是16. ![error1](https://user-images.githubusercontent.com/52661423/113966943-9b9e6c80-9862-11eb-9699-6c301464d620.png) ![error](https://user-images.githubusercontent.com/52661423/113966957-a1944d80-9862-11eb-9786-47b10fea6481.png)

我已经通过docker开启了http和grpc服务,但是不知道要怎么写客户端,看一些大佬写的代码也没有很通用。请问可以通过predict.py修改吗?还是怎么样呢?

请问要怎么训练COCO数据集呢?要在哪里修改呢?谢谢!

很高兴UP能分享这么棒的项目,想询问一下UP冻结训练的冻结层数有讲究,要怎么去设置比较合适?

when I train yolov4 on my datasets: Restoring weights from: /home/jm3090/wxl/Jm_code/tensorflow-yolov4-tflite-master/weights/yolov4.weights ... 2020-12-27 17:21:05.708877: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8 2020-12-27 17:21:08.769703: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11...

ValueError: Mismatch between expected batch size and model output batch size. Output shape = (6, 4), expected output shape = shape (1, 4)。

如何用 yolo4_voc_weights.h5预测? 直接替换yolo4_weight.h5 predict.py出错 (0) Not found: No algorithm worked! [[{{node conv2d_93/Conv2D}}]] [[boolean_mask_100/GatherV2/_3501]] (1) Not found: No algorithm worked! [[{{node conv2d_93/Conv2D}}]]

你好,感谢你的分享,请问代码中 https://github.com/bubbliiiing/yolov4-tf2/blob/a4cf127a49792a296db549d006dbef4f452770d1/nets/yolo4.py#L238 使用的NMS好像不是论文中使用的DIoU-NMS是出于什么考虑呢?谢谢!