How to train an agent in Nuscenes or other real scenarios?

[FanYi-arch]

It seems that the tutorial only shows the training process in normal scenarios, so how do you train in real dataset-driven scenarios?

[QuanyiLi]

Hi,

Here are some sample training/eval scripts: https://github.com/metadriverse/scenarionet/tree/main/scenarionet_training/scripts