diff --git a/src/inference/hmc.jl b/src/inference/hmc.jl
index 8e9bffd2c..0701c89fc 100644
--- a/src/inference/hmc.jl
+++ b/src/inference/hmc.jl
@@ -174,7 +174,7 @@ function AbstractMCMC.sample_init!(
         model(rng, spl.state.vi, spl)
     elseif islinked(spl.state.vi, spl) && spl.selector.tag != :default
         invlink!(spl.state.vi, spl)
-        model(rng, spl.state.vi, spl)        
+        model(rng, spl.state.vi, spl)
     end
 end
 
@@ -289,7 +289,7 @@ No-U-Turn Sampler (NUTS) sampler.
 Usage:
 
 ```julia
-NUTS()            # Use default NUTS configuration. 
+NUTS()            # Use default NUTS configuration.
 NUTS(1000, 0.65)  # Use 1000 adaption steps, and target accept ratio 0.65.
 ```
 
@@ -299,7 +299,7 @@ Arguments:
 - `δ::Float64` : Target acceptance rate for dual averaging.
 - `max_depth::Int` : Maximum doubling tree depth.
 - `Δ_max::Float64` : Maximum divergence during doubling tree.
-- `ϵ::Float64` : Inital step size; 0 means automatically searching using a heuristic procedure.
+- `init_ϵ::Float64` : Inital step size; 0 means automatically searching using a heuristic procedure.
 
 """
 mutable struct NUTS{AD, space, metricT <: AHMC.AbstractMetric} <: AdaptiveHamiltonian{AD}
@@ -434,8 +434,8 @@ function AbstractMCMC.step!(
 
     # Adaptation
     if spl.alg isa AdaptiveHamiltonian
-        spl.state.h, spl.state.traj, isadapted = 
-            AHMC.adapt!(spl.state.h, spl.state.traj, spl.state.adaptor, 
+        spl.state.h, spl.state.traj, isadapted =
+            AHMC.adapt!(spl.state.h, spl.state.traj, spl.state.adaptor,
                         spl.state.i, spl.alg.n_adapts, t.z.θ, t.stat.acceptance_rate)
     end