i have been using it for some time. Great, but it has a tendency to give short answers and don't use the full context available. even if you set the max response size to 20k.
Comparing it to pasting code into GPT4 (the chat interface), how much better is it? I can't wait to see what this can do when an entire codebase can be held within the context window.
Yeah, it can help with some things, but if you ask it to, say, convert a python app to Go, even with plenty of context it will give you hints without full implementation.
I suppose they favored concise answers in the fine tunning process and reused the model to train the 32k.
Things that I have been doing: paste full table DDL with entities and other classes involved and asking for some refactoring of a class. Also that and some stacktraces. It definitely helps, but the price (at least on azure) make me think twice before using.
