Out of Memory during serialization of the subsystem data - reinforcement learning

Hi, during training (reinforcement learning), agents aren't saved properly and this warning is given. What can be done to fix this?

Answer 1

[Sourabh]

Hello Charissa,

From my understanding, you are facing an out of memory issue while saving an agent during training.

The following steps should help avoid this error by reducing unncessary memory usage:

1. If you are using a Simulink model, configure it to not log any data, thus reducing memory usage.
2. Set up a more rigorous save agent value to limit the amout of agent which is actually logged.

Also, do note that it is okay to switch on logging outside the training of the agent.

Hope this helps!