This commit is contained in:
Paul Gauthier 2023-07-01 15:18:07 -07:00
parent c46917c275
commit 8de3fa00d0

View file

@ -313,17 +313,3 @@ cause a large variance in the overall benchmark results.
Based on these benchmarking results, aider will continue to use
the `whole` edit format for GPT-3.5, and `diff` for GPT-4.
While GPT-4 gets somewhat better results with the `whole` edit format,
it significantly increases costs and latency compared to `diff`.
The latency of streaming back the entire updated copy of each edited file
is the real challenge. The GPT-3.5 models are quite responsive, and can
stream back entire files at an acceptable speed.
Aider displays a progress bar and
live diffs of the files as they stream in,
which helps pass the time.
The GPT-4 models are much slower, and waiting for even small files
to be completely "retyped" on each request is probably unacceptable.