## Does $f(n)=\theta(g(\log{n}))$ imply $f(n)=\theta(\log{g(n)})$?

I am learning about algorithmic complexities and I have this claim which I need to prove or disprove: $$f(n)$$ and $$g(n)$$ are asymptotically positive functions, if $$f(n)=\Theta(g(\log(n))$$ then $$f(n)=\Theta(\log(g(n)))$$. Is it true?