How to finetune using my language dataset
#3
by echoxin - opened
Hello, thank you for sharing this well performance model, can I ask how to finetune on my dataset, I have a korean dataset, it contains around 24 hours, if this level of dataset is enough to full-param finetune on voxcpm2 or I'd better just finetune a lora model, do you have some suggestions ? thanks for your time