Closed scalaboy closed 3 months ago
In the finetune dir,the pipeline is train Q-A dataset, however,we need to finetune MiniCPM by Instruct tuning. Is there any solutions?
I think what we provide is a training method, which is sft. It should also be suitable for instruct tuning. If you have any questions, you can send us your data example.
Feature request / 功能建议
In the finetune dir,the pipeline is train Q-A dataset, however,we need to finetune MiniCPM by Instruct tuning. Is there any solutions?