Catastrophic forgetting problem

Hi dear:

Thanks for your open source, but when i finetuned (whatever full parameters or LoRa ) on my dataset, catastrophic forgetting kept coming up (decrease in performance on the humaneval), i do not know how to solve it, do you have any tops?

The issue names full-parameter and LoRA fine-tuning and HumanEval, but provides no file, test, or entry point. Start by locating the project's fine-tuning and HumanEval evaluation guidance; done would require a reproducible understanding of the reported performance loss and a documented resolution.

Written by the indexing model from the issue text.