update line 512 and line 1101 of utils/general.py

[mo-traor3-ai]

Description

On the forum, it was noted that Scaled YOLOv4 notebook is broken. This change fixes the issue.

• https://discuss.roboflow.com/t/error-roboflow-custom-scaled-yolov4/1920
  
• Error traced back to:

File “/content/gdrive/MyDrive/Marlene_PhD/obj_det/ScaledYOLOv4/utils/general.py”, line 556, in build_targets
indices.append((b, a, gj.clamp_(0, gain[3]), gi.clamp_(0, gain[2]))) # image, anchor, grid indices
RuntimeError: result type Float can’t be cast to the desired output type long int

• Line 512 in utils/general.py edited to:
  gain = torch.ones(7, device=targets.device).long() # normalized to gridspace gain

• Original Line 512 in utils/general.py:
  gain = torch.ones(7, device=targets.device) # normalized to gridspace gain

• Line 1101 in utils/general.py edited to:

targets.append([i, cls, float(x.cpu()),   
              float(y.cpu()),   
              float(w.cpu()),   
              float(h.cpu()),   
              float(conf.cpu())])

• Original Line 1101 in utils/general.py:

targets.append([i, cls, x, y, w, h, conf])

Investigation of the issue:

• RuntimeError: result type Float can't be cast to the desired output type long int WongKinYiu/yolov7#35 (comment)
  

• TypeError: can't convert cuda:0 device type tensor to numpy WongKinYiu/ScaledYOLOv4#382 (comment)
  

List any dependencies that are required for this change.

Type of change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• This change requires a documentation update

How has this change been tested, please provide a testcase or example of how you tested the change?

Ran the full Colab notebook, myself:

Any specific deployment considerations

For example, documentation changes, usability, usage/costs, secrets, etc.

Docs

• Docs updated? What were the changes:

[SkalskiP]

@Jacobsolawetz / @yeldarby could you grant me write access to this repo?

[Jacobsolawetz]

@SkalskiP done!

[SkalskiP]

Thanks @Jacobsolawetz 🙏🏻! It took me a long to do it but I'm back. Tets finish fixing it.