Load multiple LoRA modules simultaneously and automatically switch the appropriate combination of LoRA modules to generate the best answer based on user queries.
As I understand it, mlora combines multiple model finetunes that have very different prompt templates. Does it mean that I need to identify which model I want my prompt to invoke and format my prompt accordingly? Or does mloras have a way doing that for me?
As I understand it, mlora combines multiple model finetunes that have very different prompt templates. Does it mean that I need to identify which model I want my prompt to invoke and format my prompt accordingly? Or does mloras have a way doing that for me?