Abstract: Knowledge Distillation (KD) is the procedure of extracting useful information from a previously trained model using an algorithm. The successful distillation pulls up the distilling model ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback