Abstract: Parameter-efficient tuning methods such as LoRA could achieve comparable performance to model tuning by tuning a small portion of the parameters. However, substantial computational resources ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback