YOLOV5报错：RuntimeError: result type Float can‘t be cast to the desired output type long int

YOLOV5训练时报错：

Traceback (most recent call last):
  File "train.py", line 626, in <module>
    main(opt)
  File "train.py", line 523, in main
    train(opt.hyp, opt, device, callbacks)
  File "train.py", line 320, in train
    loss, loss_items = compute_loss(pred, targets.to(device))  # loss scaled by batch_size
  File "/home/mapgoo/rknn_model_zoo/models/vision/object_detection/yolov5-pytorch/yolov5/utils/loss.py", line 120, in __call__
    tcls, tbox, indices, anchors = self.build_targets(p, targets)  # targets
  File "/home/mapgoo/rknn_model_zoo/models/vision/object_detection/yolov5-pytorch/yolov5/utils/loss.py", line 217, in build_targets
    indices.append((b, a, gj.clamp_(0, gain[3] - 1), gi.clamp_(0, gain[2] - 1)))  # image, anchor, grid indices
RuntimeError: result type Float can't be cast to the desired output type long int

解决方案： 

        修改utils文件夹下loss.py文件中class ComputeLoss类中def build_targets()函数：

        1、可能第184行，或是第167行：

                源代码：

anchors = self.anchors[i]

        修改为：

anchors, shape = self.anchors[i],p[i].shape

                 修改后：

       注意：

                如果源码为：

anchors = det.anchors[i]

                修改为：

anchors, shape = det.anchors[i], p[i].shape

                 修改后：

 

2、可能是第218行，或是第200行：

                源代码：

indices.append((b, a, gj.clamp_(0, gain[3] - 1), gi.clamp_(0, gain[2] - 1)))  # image, anchor, grid indices

                修改为：

indices.append((b,a,gj.clamp_(0,shape[2] - 1),gi.clamp_(0,shape[3] - 1)))  # image, anchor, grid indices

                修改后：