TonyXuQAQ opened this issue 6 months ago · 1 comments
During training, the loss is quite high and does not decrease
I SFT the pllava model with my own data. Is there any suggestions for this?
