tinyml
tinyml copied to clipboard
what happened if NegAug applied to ofa tiny models?
What if NetAug is applied to the tiny models searched from once-for-all, since once-for-all is trained in a similar way?
Will NetAug boost the tiny ofa models?