I’m trying to use ChatGPT to convert a large codebase from C++ to C#. Some classes are very large (legacy code) It is trying to be helpful by providing output that is a shell and does not implement ANY of the methods in a class so it fits in the token window. That obviously makes it useless unless I’m willing to do a lot of grooming of the input. Is there any way to get it to simply truncate the output instead of compressing it so much that it’s useless for what I’m trying to do? Ideally, it would output until it hits the maximum tokens and I could use that information to split it into smaller pieces. I’ve worked with it to add all the prompts I can in order to get it to work that way with no luck.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| How to pass 150k lines of code to chatgpt | 4 | 6730 | September 5, 2024 | |
| How to stop chagpt from using code blocks and markdown in general? | 5 | 15218 | May 13, 2023 | |
| Trying to convert fortran to c++ not converting all code | 3 | 1266 | May 17, 2023 | |
| How do I get gpt to throw out more tokens in API? | 3 | 2067 | December 16, 2023 | |
| Coding issues with Python- Truncation of Generated Code | 2 | 786 | June 2, 2024 |