Resuming training after stopping #1323
-
|
Warm greetings. I would like to ask about the possibility of continuing learning after stopping. I know that I can obtain the model and save its parameters after training with certain CL strategies within Avalanche. But, is it possible to continue a training strategy after I save the parameter? or should I start the training from scratch? Thank you. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
Yes, you can save checkpoints of the strategy. Keep in mind that this is a beta feature and it is still quite slow in some cases. We will speed it up in the next version. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks! I'll check it out later and should I have any troubles, I will post it on the discussion again as a different topic. |
Beta Was this translation helpful? Give feedback.
Yes, you can save checkpoints of the strategy. Keep in mind that this is a beta feature and it is still quite slow in some cases. We will speed it up in the next version.