1 paper · 1 filter
Snehal Raj, Brian Coyle
Scaling full finetuning of large foundation models strains GPU memory and training time. Parameter Efficient Fine-Tuning (PEFT) methods address this issue via adapter modules which…