Just started using Tabnine for Python in VS Code. I was getting weird suggestions that didn’t match my function names or variables. Turns out I never touched the context length setting—it was using the default.
I increased it to “Large” in the Tabnine settings, and now the completions are actually relevant to my code. It’s like night and day. For anyone new, check that setting early. Is there a general rule for what context length to use for different languages or project sizes?
Yeah, I ran into the same thing last week. I was working on a Flask app and the suggestions were just... wrong. Your post made me check, and mine was also on default.
Is "Large" the best setting for everything, though? I'm worried it might slow things down on bigger files. Have you noticed any lag since changing it?