Dian Chen

LGTM. thanks for the PR!

This might depend on your PyTorch/CUDA/cuDNN version. With the one I use, this line is necessary to prevent a CUDA error. If it works fine without it in your setup...

Hello, We did not use a docker for training but uses a conda env to manage the dependencies. Let me know if you have any issues with the dependencies and...

Have you pinned down the issue is actually OOM? This feels like it has something to do with opengl but I am not sure. To help more I need more...

Hmm I don't have really have any anything on my mind that might help you. Since you mentioned you use docker maybe try our docker image recipe and see if...

1. Yes 2. Absolutely correct. You can remove it from the code but you will need to add a `strict=False` when loading the weights. 3. `pred_heatmap` is the predicted Gaussian...

Hello, Could you elaborate on what you meant by patches?

These are part of scenarios generated by the leaderboard. See: https://leaderboard.carla.org/ In order to remove them you will need to modify the corresponding codes in the `leaderboard`.

Thank you for liking our work. I'd need more info but my guess is that mismatch between CARLA server and installed Python egg.

Thank you for your interest in our project. Please follow the instructions [here](https://github.com/dotchen/LAV#evaluation) on how to directly run an agent with pre-trained model weights. Please make sure to have installed...