Pytorch GPU yolov5 reports an error

A wrong ：

AttributeError: 'Upsample' object has no attribute 'recompute_scale_factor'

  resolvent , modify pytorch Source code ：

Lib\site-packages\torch\nn\modules Next, find the file upsampling.py

    def forward(self, input: Tensor) -> Tensor:
        # return F.interpolate(input, self.size, self.scale_factor, self.mode, self.align_corners,
        #                      recompute_scale_factor=self.recompute_scale_factor)
        return F.interpolate(input, self.size, self.scale_factor, self.mode, self.align_corners)

Error 2 ：

RuntimeError: result type Float can't be cast to the desired output type xxx

yolov5 Under the  【utils Folder 】 Medium 【loss.py】

Two points to be modified （ Annotated code is the previous code , The following is modified ）：

  The two revised categories are as follows ：

anchors, shape = self.anchors[i], p[i].shape  #  modify ,2022 year 7 month 11 Japan 
​​​​​​​
indices.append((b, a, gj.clamp_(0, shape[2] - 1), gi.clamp_(0, shape[3] - 1)))  # image, anchor, grid  #  modify ,2022 year 7 month 11 Japan