【Bug解决】TypeError: can‘t convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy...

问题分析

笔者在训练yolov5模型时出现如下问题：

Traceback (most recent call last):
  File "train.py", line 460, in <module>
    train(hyp, opt, device, tb_writer)
  File "train.py", line 320, in train
    plots=epoch == 0 or final_epoch)  # plot first and last
  File "D:\yolov5-3.1\test.py", line 197, in test
    plot_images(img, output_to_target(output, width, height), paths, str(f), names)  # predictions
  File "D:\yolov5-3.1\utils\general.py", line 951, in output_to_target
    return np.array(targets)
  File "D:\Anaconda3\envs\yolov5-3\lib\site-packages\torch\tensor.py", line 621, in __array__
    return self.numpy()
TypeError: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.

解决问题

根据错误提示找到D:\Anaconda3\envs\yolov5-3\lib\site-packages\torch\tensor.py文件的第621行
将：

return self.numpy()

修改为：

return self.cpu().detach().numpy()

再次运行训练指令即可正常运行，问题解决