DereWah

To do so you need to process your dataset to fit a specific format. As shown in here: https://github.com/eloialonso/diamond/tree/csgo?tab=readme-ov-file#training You need to follow the format of the following dataset: https://github.com/TeaPearce/Counter-Strike_Behavioural_Cloning...

Yes, if I remember correctly even just running in python and emitting frames while at the same time on another thread running cuda calls, will reproduce the issue.