Juan

Definitely support this idea.

@Santosh-Gupta, thanks for your reply. The Jetson Nano platform has a quad-core ARM A57 64-bit CPU. > but the pinned cupy tensors aren't copying faster, they're using a different indexing...

@Santosh-Gupta Thanks for your reply back. Yes, that would be 4 cores in the cpu. I will try to confirm if the `self.data_cpu.CUPYcorpus = cp.asarray(new_data)` would still be using cpu...

I had the same problem. For some reason the simlinks that are created during the data preparation process described in DATA.md are not working. So instead of creating simlinks I...

I am experiencing the same issue. A solution will be appreciated. ``` setting /run_id to fbd7ceb0-f13b-11ec-8a80-88366cf97040 process[rosout-1]: started with pid [3057] started core service [/rosout] process[camera/camera_nodelet_manager-2]: started with pid [3060]...

@breznak I support your offer of providing the batch eval, thank you!

> https://github.com/ultralytics/yolov3/wiki/Train-Custom-Data Does the train.py file in the repository you link to works for with the darknet.py file in this (ayooshkathuria/pytorch-yolo-v3) repository?

Why do you say its better?

Thanks for sharing you reimplementation, I will definitely have a look at yolov3-lite. It seems you implemented ideas from several papers. It would be nice for you to write what...

@LittleSunH @matursi I think #44 might give you some insights on what is happening. https://github.com/ayooshkathuria/YOLO_v3_tutorial_from_scratch/issues/44