This commit is contained in:
Paul Gauthier 2023-07-07 13:32:51 -07:00
parent cffeb275a6
commit a01208d9c0

View file

@ -37,7 +37,7 @@ The results were interesting:
- **Plain text edit formats worked best.** Asking GPT to return an updated copy of the whole file in a standard markdown fenced code block proved to be the most reliable and effective edit format across all GPT-3.5 and GPT-4 models. The results for this `whole` edit format are shown in solid blue in the graph.
- **Function calls performed worse.** Using the new functions API for edits performed worse than the above whole file method, for all the models. GPT-3.5 especially produced inferior code and frequently mangled this output format. This was surprising, as the functions API was introduced to enhance the reliability of structured outputs. The results for these `...-func` edit methods are shown as patterned bars in the graph (both green and blue).
- **The new GPT-3.5 models did worse than the old model.** The performance of the new June (`0613`) versions of GPT-3.5 appears to be a bit worse than the February (`0301`) version. This is visible if you look at the "first attempt" markers on the first three solid blue bars and also by comparing the first three solid green `diff` bars.
- **The new June GPT-3.5 models did a bit worse than the old June model.** The performance of the new June (`0613`) versions of GPT-3.5 appears to be a bit worse than the February (`0301`) version. This is visible if you look at the "first attempt" markers on the first three solid blue bars and also by comparing the first three solid green `diff` bars.
- **GPT-4 does better than GPT-3.5,** as expected.
The quantitative benchmark results agree with my intuitions