search

werner-duvaud / muzero-general

MuZero
https://github.com/werner-duvaud/muzero-general/wiki/MuZero-Documentation
MIT License
2.47k stars 606 forks source link

Stop Training #145

Closed JohnPPP closed 3 years ago

JohnPPP commented 3 years ago

Hi,

How are you stopping the training? I'm pressing ESC and Python identifies the keyboard interrupt.

However, sometimes it does not create a replay buffer.

Wondering if I'm doing anything wrong here.

Thank you, João

JohnPPP commented 3 years ago

Thank you Matthew.

Matthew @.***> escreveu no dia quinta, 1/04/2021 à(s) 03:01:

Cltrl C is Interrupt for me. Safety of the replay buffer can't be guaranteed unless the code ends gracefully. Stopping the program by interrupt should only be done to prevent the code from executing. You should lower the number of training steps if you want a smaller run.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/werner-duvaud/muzero-general/issues/145#issuecomment-811582781, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACPAYRII3SBO2HY3BLEK5CTTGPHXVANCNFSM42E2HGTQ .