Does `fastApply` imply there's a `slowApply` or something else?

General
1

I searched and don’t see anything about it, but considering there is a fastApply for code insertion from the chat panel, the wording seems to imply there is a slowApply or some other descriptor for Apply. If there is, does it perform differently?

The reason I’m interested is because I’ve had a chat insertion that consistently attempted inserting the proposed solution at the wrong line which broke the rest of my document. The text was similar which was the obvious cause of the insertion problem, but oddly enough changing to a premium model fixed the insertion even though it had the exact same code completion.

2

fastApply is a in house model made for applying code on small files (around 400 lines). The other method they use (or used to at least) was using a normal model (4o, sonnet, etc) to apply the code, which is slow. They are working on improving this with the claude-3.5-sonnet fast edit model shown off by Zed AI

3 Likes
3

I see; thanks for the clarification!