fixed gpu error

[adakoda]

#52 modification still cause following error.
->
Error processing request: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.

So we need to convert output type to cpu from gpu.

Note:
If you use docker environment, please add --gpus all option.

docker run -it --rm --gpus all -p 10000:10000 -p 5005:5005 unity-robotics:pose-estimation /bin/bash

You can also use nvidia-smi command in docker whether gpu is enable or not.

[JonathanLeban]

thank you @adakoda on your report and your suggestion. What is the line causing the error about the conversion to cuda:0 ?

[adakoda]

@JonathanLeban

What is the line causing the error about the conversion to cuda:0 ?

output_translation, output_orientation = output_translation.detach().numpy(), output_orientation.detach().numpy()

The model return gpu device type and you will get following error message.

Error processing request: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.

[at669]

Hi @adakoda, this repo is not currently accepting external contributions, as per the Contributing Guidelines. For that reason, I will close this PR, but please feel free to continue discussing in this thread, @JonathanLeban and @adakoda. Once resolved, the potential fix will be added via an internal PR. Sorry for the inconvenience!

[adakoda]

@at669 @JonathanLeban
Thank you for your works!

This is the description of the sequence of code.
This notebook is available on Google CoLab for anyone to try.
https://github.com/adakoda/Robotics-Object-Pose-Estimation/blob/fixed_cpu_error/Model/Pose_Estimation_not_working_correctly_45.ipynb

I hope this explanation will be useful to someone.
I applicaiate your hard works.

[JonathanLeban]

Hey ! So the problem was solved in the dev branch and we are going to merge the dev branch with the main branch soon. Thank you !!