Error after few episodes

I am training my DDPG RL agent. It runs fine for some initial episodes (sometimes 3, sometimes 10, and sometimes even 37 out of 50) but then the system crashes giving an error shown in the attached pic. My GPU is on for both actor and critic.

[Jaimin]

Hi @Sania Gul

Could you please send the related resource? It would be helpful.

[Sania Gul]

Tnks Jaimin

Step2_MainRLNew.mlx is my main program. When I increase Max number of steps per episode to 50, this problem occurs, If I keep this value to 30, the program runs smoothly. I cannot attach data files here as I am not allowed by MATLAB to attach such a big data;

[praguna manvi]

hi @Sania Gul,

Also please share PhaseCorrectionSystemNew file for complete reproduction.It seems to be an issue while loading the model file.

[Sania Gul]

Tnks @praguna manvi, the matter is resolved by decreasing the number of actions and observations. Tnks for your support and concern :-).