Bmtrain checkpointing mechanism #22
Kunlun-Zhu
started this conversation in
Open Chats 开放交流
Replies: 1 comment
-
|
Hi, thanks for your suggestion! The training of CPM-Live is totally based on BMTrain. The checkpointing method is also used, which can enlarge the training batch size and improve the performance. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
I realize 'bmtrain' enable the 'checkpointing' method for optimizing the memory use, will this mechanism helpful for CPM-live training? Though it may cause extra training time required.
Beta Was this translation helpful? Give feedback.
All reactions