Remove `extract_priors`

[yebai]

extract_priors is used to extract the distribution and variable name of parameters. It appears that we can obtain all this information from VarInfo. Additionally, I believe we only extract this information to sample from the "prior". If so, this can be replaced by a variant of rand(model)?

Turing.jl/src/mcmc/mh.jl

Lines 148 to 154 in fc32e10

	function MH(model::Model; proposal_type=AMH.StaticProposal)
	priors = DynamicPPL.extract_priors(model)
	props = Tuple([proposal_type(prop) for prop in values(priors)])
	vars = Tuple(map(Symbol, collect(keys(priors))))
	priors = map(transform_maybe, NamedTuple{vars}(props))
	return AMH.MetropolisHastings(priors)
	end

cc @mhauru

[yebai]

Fixed by TuringLang/DynamicPPL.jl#907