How to load right model for converting Pytorch model to torchscript (pth to pt) in Isaac gym?

I am trying to deploy my trained model to C++. For this, I need to first convert my trained model in isaac gym( saved as pth) to torch script. I am now following pytorch tutorial TorchScript for Deployment — PyTorch Tutorials 1.13.1+cu117 documentation. But I need to load checkpoint dict to a model first:
I know I can find my model name in PPO config, but I don’t know how to use it. Does anyone have a clue?
Thanks a lot.

I found this incredible tutorial in how to deploy the trained policy to C++, its here: GitHub - Denys88/rl_games: RL implementations. in case, any one also need help with exporting their trained model. There are two simple example to export the checkpoint to onnx, it helps a lot. I already deploy it in my C++ application, and it works well so far.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.