specific method · filed under training objective
Reverse KL divergence
Uses reverse Kullback–Leibler divergence between student and teacher as the distillation loss.
Also called reverse KL divergence loss.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
reverse KL divergence loss between student and teacher
usedpost trainingin MiMo-V2-FlashXiaomi
Filed alongside
Other methods under training objective :: distillation objective.