High memory usage of 'TabNine-deep-local'

[rehack]

MacOS10.13 vscode Version: 1.59.1 Tabnine ext version | 3.4.26

Vscode is not in writing state, but ‘TabNine-deep-local’ memory usage is very high

gz#6012

[yonip23]

Hi @rehack Tabnine takes up a significant amount of memory because it loads the gpt2 model to the ram, independent of what happens in vscode. Fortunately, you can use our cloud option and disable the local service in the hub. Also, pro users can choose their model size, which also helps reduce the memory footprint of Tabnine.

Hope this helps:)

[rehack]

Thank you all so much @yonip23 I switched to cloud mode, TabNine-deep-local process no longer occupied memory

[yonip23]

Just notice that if you're not working with neither local or cloud enabled, you're working with the very bear minimum Tabnine has to offer. Your choice 🙂

[MrMoshkovitz]

I have the same issue

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.