Skip to content

[Machine Learning] Continuously Train the Model on Kaggle Notebook (Can Close the Session)

Last Updated on 2022-07-27 by Clay

Kaggle Notebook offers a not-so-short GPU-acceleration time every week, and updated every Saturday. Because of this, I recommend using Kaggle Notebook over Google Colab.

Many people may worry about continuously training the model, and may have to keep their computer open Kaggle Notebook window, and click with the mouse from time to time - this is to avoid closing the session, resulting in losing your progress.

But in fact, we can commit our notebook, and training the model in the remote Kaggle server!

You just need to click Save Version button in the upper right corner of the figure below, and then you can execute out program in the newly opened window.

根據說明,TPU 的使用時間為三小時、CPU 和 GPU 的使用時間則是九小時。


References


Read More

Leave a Reply