diff --git a/experiments/benchmarks_v2.jl b/experiments/benchmarks_v2.jl
index 5a20465e..79a74c33 100644
--- a/experiments/benchmarks_v2.jl
+++ b/experiments/benchmarks_v2.jl
@@ -58,31 +58,32 @@ num_feat = Int(100)
 x_train = rand(nobs, num_feat)
 y_train = rand(size(x_train, 1))
 
-@info "xgboost train:"
-@time m_xgb = xgboost(x_train, nrounds, label=y_train, param=params_xgb, metrics=metrics, nthread=nthread, silent=1);
-@btime xgboost($x_train, $nrounds, label=$y_train, param=$params_xgb, metrics=$metrics, nthread=$nthread, silent=1);
-@info "xgboost predict:"
-@time pred_xgb = XGBoost.predict(m_xgb, x_train);
-@btime XGBoost.predict($m_xgb, $x_train);
+# @info "xgboost train:"
+# @time m_xgb = xgboost(x_train, nrounds, label=y_train, param=params_xgb, metrics=metrics, nthread=nthread, silent=1);
+# @btime xgboost($x_train, $nrounds, label=$y_train, param=$params_xgb, metrics=$metrics, nthread=$nthread, silent=1);
+# @info "xgboost predict:"
+# @time pred_xgb = XGBoost.predict(m_xgb, x_train);
+# @btime XGBoost.predict($m_xgb, $x_train);
 
-@info "evotrees train CPU:"
-params_evo.device = "cpu"
-@time m_evo = fit_evotree(params_evo; x_train, y_train, x_eval=x_train, y_eval=y_train, metric=metric_evo, print_every_n=50);
-@btime fit_evotree($params_evo; x_train=$x_train, y_train=$y_train, x_eval=$x_train, y_eval=$y_train, metric=metric_evo);
-@info "evotrees predict CPU:"
-@time pred_evo = EvoTrees.predict(m_evo, x_train);
-@btime EvoTrees.predict($m_evo, $x_train);
+# @info "evotrees train CPU:"
+# params_evo.device = "cpu"
+# @time m_evo = fit_evotree(params_evo; x_train, y_train, x_eval=x_train, y_eval=y_train, metric=metric_evo, print_every_n=100);
+# @btime fit_evotree($params_evo; x_train=$x_train, y_train=$y_train, x_eval=$x_train, y_eval=$y_train, metric=metric_evo);
+# @btime fit_evotree($params_evo; x_train=$x_train, y_train=$y_train);
+# @info "evotrees predict CPU:"
+# @time pred_evo = EvoTrees.predict(m_evo, x_train);
+# @btime EvoTrees.predict($m_evo, $x_train);
 
 CUDA.allowscalar(true)
 @info "evotrees train GPU:"
 params_evo.device = "gpu"
 @time m_evo_gpu = fit_evotree(params_evo; x_train, y_train);
-@time m_evo = fit_evotree(params_evo; x_train, y_train, x_eval=x_train, y_eval=y_train, metric=metric_evo, print_every_n=50);
-@btime fit_evotree($params_evo; x_train=$x_train, y_train=$y_train, x_eval=$x_train, y_eval=$y_train, metric=metric_evo);
-@info "evotrees predict GPU:"
-@time pred_evo = EvoTrees.predict(m_evo_gpu, x_train);
-@btime EvoTrees.predict($m_evo_gpu, $x_train);
-
-# w_train = ones(length(y_train))
-# @time m_evo_gpu = fit_evotree(params_evo, x_train, y_train);
-# @time m_evo_gpu = fit_evotree(params_evo, x_train, y_train, w_train);
\ No newline at end of file
+@time m_evo_gpu = fit_evotree(params_evo; x_train, y_train);
+@time m_evo_gpu = fit_evotree(params_evo; x_train, y_train);
+@time m_evo_gpu = fit_evotree(params_evo; x_train, y_train);
+@time m_evo_gpu = fit_evotree(params_evo; x_train, y_train);
+@time m_evo = fit_evotree(params_evo; x_train, y_train, x_eval=x_train, y_eval=y_train, metric=metric_evo, print_every_n=100);
+# @btime fit_evotree($params_evo; x_train=$x_train, y_train=$y_train, x_eval=$x_train, y_eval=$y_train, metric=metric_evo);
+# @info "evotrees predict GPU:"
+# @time pred_evo = EvoTrees.predict(m_evo_gpu, x_train);
+# @btime EvoTrees.predict($m_evo_gpu, $x_train);
\ No newline at end of file
diff --git a/experiments/random.jl b/experiments/random.jl
index e02f333e..63e93c49 100644
--- a/experiments/random.jl
+++ b/experiments/random.jl
@@ -33,7 +33,7 @@ params1 = EvoTreeRegressor(T=Float32,
 
 # asus laptopt: for 1.25e6 no eval: 9.650007 seconds (893.53 k allocations: 2.391 GiB, 5.52% gc time)
 @time model = fit_evotree(params1; x_train, y_train);
-@time model = fit_evotree(params1; x_train, y_train, metric=:mse, x_eval, y_eval, print_every_n=10);
+@time model = fit_evotree(params1; x_train, y_train, metric=:mse, x_eval, y_eval, print_every_n=100);
 @btime model = fit_evotree(params1; x_train, y_train);
 @time pred_train = predict(model, x_train);
 @btime pred_train = predict(model, x_train);
@@ -77,7 +77,7 @@ params1 = EvoTreeGaussian(T=Float32,
 # train model
 params1 = EvoTreeRegressor(T=Float32,
     loss=:linear, metric=:mse,
-    nrounds=10,
+    nrounds=100,
     lambda=1.0, gamma=0, eta=0.1,
     max_depth=6, min_weight=1.0,
     rowsample=0.5, colsample=0.5, nbins=64,
@@ -86,7 +86,7 @@ params1 = EvoTreeRegressor(T=Float32,
 # Asus laptop:  10.015568 seconds (13.80 M allocations: 1.844 GiB, 4.00% gc time)
 @time model = EvoTrees.fit_evotree(params1; x_train, y_train);
 @btime model = EvoTrees.fit_evotree(params1; x_train, y_train);
-@time model, cache = EvoTrees.init_evotree_gpu(params1, X_train, Y_train);
+@time model, cache = EvoTrees.init_evotree_gpu(params1; x_train, y_train);
 @time EvoTrees.grow_evotree!(model, cache);
 
 using MLJBase
@@ -118,14 +118,14 @@ params1 = EvoTreeRegressor(T=Float32,
 # GPU - Gaussian
 ################################
 params1 = EvoTreeGaussian(T=Float32,
-    loss=:gaussian, metric=:gaussian,
+    loss=:gaussian,
     nrounds=100,
     lambda=1.0, gamma=0, eta=0.1,
     max_depth=6, min_weight=1.0,
     rowsample=0.5, colsample=0.5, nbins=32,
     device="gpu")
 # Asus laptop: 14.304369 seconds (24.81 M allocations: 2.011 GiB, 1.90% gc time)
-@time model = EvoTrees.fit_evotree(params1, X_train, Y_train);
+@time model = EvoTrees.fit_evotree(params1; x_train, y_train);
 # Auss laptop:  1.888472 seconds (8.40 k allocations: 1.613 GiB, 14.86% gc time)
 @time model, cache = EvoTrees.init_evotree(params1, X_train, Y_train);
 
diff --git a/src/find_split.jl b/src/find_split.jl
index 907d5290..7e233a2c 100644
--- a/src/find_split.jl
+++ b/src/find_split.jl
@@ -298,8 +298,10 @@ function hist_gains_cpu!(
         if bin == params.nbins
             gains[bin] = hL[i]^2 / (hL[i+1] + params.lambda * hL[i+2]) / 2
         elseif hL[i+2] > params.min_weight && hR[i+2] > params.min_weight
-            predL = pred_scalar_cpu!(hL[i:i+2], params, K)
-            predR = pred_scalar_cpu!(hR[i:i+2], params, K)
+            if monotone_constraint != 0
+                predL = pred_scalar_cpu!(view(hL, i:i+2), params, K)
+                predR = pred_scalar_cpu!(view(hR, i:i+2), params, K)
+            end
             if (monotone_constraint == 0) ||
                (monotone_constraint == -1 && predL > predR) ||
                (monotone_constraint == 1 && predL < predR)
@@ -362,8 +364,10 @@ function hist_gains_cpu!(
                     hL[i+1]^2 / (hL[i+3] + params.lambda * hL[i+4])
                 ) / 2
         elseif hL[i+4] > params.min_weight && hR[i+4] > params.min_weight
-            predL = pred_scalar_cpu!(hL[i:i+4], params, K)
-            predR = pred_scalar_cpu!(hR[i:i+4], params, K)
+            if monotone_constraint != 0
+                predL = pred_scalar_cpu!(view(hL, i:i+4), params, K)
+                predR = pred_scalar_cpu!(view(hR, i:i+4), params, K)
+            end
             if (monotone_constraint == 0) ||
                (monotone_constraint == -1 && predL > predR) ||
                (monotone_constraint == 1 && predL < predR)
diff --git a/src/fit.jl b/src/fit.jl
index fe7ff45a..abbe9ee2 100644
--- a/src/fit.jl
+++ b/src/fit.jl
@@ -94,7 +94,7 @@ function init_evotree(
 
     # assign monotone contraints in constraints vector
     monotone_constraints = zeros(Int32, x_size[2])
-    hasproperty(params, :monotone_constraint) && for (k, v) in params.monotone_constraints
+    hasproperty(params, :monotone_constraints) && for (k, v) in params.monotone_constraints
         monotone_constraints[k] = v
     end
 
diff --git a/src/gpu/find_split_gpu.jl b/src/gpu/find_split_gpu.jl
index b6a29678..833d8749 100644
--- a/src/gpu/find_split_gpu.jl
+++ b/src/gpu/find_split_gpu.jl
@@ -237,8 +237,10 @@ function hist_gains_gpu_kernel!(gains::CuDeviceMatrix{T}, hL::CuDeviceArray{T,3}
     if i == nbins
         gains[i, j] = hL[1, i, j]^2 / (hL[2, i, j] + lambda * hL[3, i, j]) / 2
     elseif hL[3, i, j] > min_weight && hR[3, i, j] > min_weight
-        predL = -hL[1, i, j] / (hL[2, i, j] + lambda * hL[3, i, j])
-        predR = -hR[1, i, j] / (hR[2, i, j] + lambda * hR[3, i, j])
+        if monotone_constraint != 0
+            predL = -hL[1, i, j] / (hL[2, i, j] + lambda * hL[3, i, j])
+            predR = -hR[1, i, j] / (hR[2, i, j] + lambda * hR[3, i, j])
+        end
         if (monotone_constraint == 0) ||
            (monotone_constraint == -1 && predL > predR) ||
            (monotone_constraint == 1 && predL < predR)
@@ -281,8 +283,10 @@ function hist_gains_gpu_kernel_gauss!(gains::CuDeviceMatrix{T}, hL::CuDeviceArra
     if i == nbins
         gains[i, j] = (hL[1, i, j]^2 / (hL[3, i, j] + lambda * hL[5, i, j]) + hL[2, i, j]^2 / (hL[4, i, j] + lambda * hL[5, i, j])) / 2
     elseif hL[5, i, j] > min_weight && hR[5, i, j] > min_weight
-        predL = -hL[1, i, j] / (hL[3, i, j] + lambda * hL[5, i, j])
-        predR = -hR[1, i, j] / (hR[3, i, j] + lambda * hR[5, i, j])
+        if monotone_constraint != 0
+            predL = -hL[1, i, j] / (hL[3, i, j] + lambda * hL[5, i, j])
+            predR = -hR[1, i, j] / (hR[3, i, j] + lambda * hR[5, i, j])
+        end
         if (monotone_constraint == 0) ||
            (monotone_constraint == -1 && predL > predR) ||
            (monotone_constraint == 1 && predL < predR)
diff --git a/src/gpu/fit_gpu.jl b/src/gpu/fit_gpu.jl
index 7dbb2f6f..d26cb3ed 100644
--- a/src/gpu/fit_gpu.jl
+++ b/src/gpu/fit_gpu.jl
@@ -68,7 +68,7 @@ function init_evotree_gpu(
 
     # assign monotone contraints in constraints vector
     monotone_constraints = zeros(Int32, x_size[2])
-    hasproperty(params, :monotone_constraint) && for (k, v) in params.monotone_constraints
+    hasproperty(params, :monotone_constraints) && for (k, v) in params.monotone_constraints
         monotone_constraints[k] = v
     end
 
diff --git a/src/predict.jl b/src/predict.jl
index f6d3c64e..8f384799 100644
--- a/src/predict.jl
+++ b/src/predict.jl
@@ -103,7 +103,7 @@ function pred_leaf_cpu!(
     pred[1, n] = -params.eta * ∑[1] / (∑[2] + params.lambda * ∑[3])
 end
 function pred_scalar_cpu!(
-    ∑::Vector{T},
+    ∑::AbstractVector{T},
     params::EvoTypes,
     K,
 ) where {L<:GradientRegression,T,S}
@@ -123,7 +123,7 @@ function pred_leaf_cpu!(
     pred[1, n] = -params.eta * ∑[1] / (∑[3] + params.lambda * ∑[5])
     pred[2, n] = -params.eta * ∑[2] / (∑[4] + params.lambda * ∑[5])
 end
-function pred_scalar_cpu!(∑::Vector{T}, params::EvoTypes{L,T,S}, K) where {L<:MLE2P,T,S}
+function pred_scalar_cpu!(∑::AbstractVector{T}, params::EvoTypes{L,T,S}, K) where {L<:MLE2P,T,S}
     -params.eta * ∑[1] / (∑[3] + params.lambda * ∑[5])
 end
 
@@ -171,6 +171,6 @@ function pred_leaf_cpu!(
 ) where {L<:L1Regression,T,S}
     pred[1, n] = params.eta * ∑[1] / (∑[3] * (1 + params.lambda))
 end
-function pred_scalar_cpu!(∑::Vector, params::EvoTypes{L,T,S}, K) where {L<:L1Regression,T,S}
+function pred_scalar_cpu!(∑::AbstractVector{T}, params::EvoTypes{L,T,S}, K) where {L<:L1Regression,T,S}
     params.eta * ∑[1] / (∑[3] * (1 + params.lambda))
 end
\ No newline at end of file
diff --git a/test/monotonic.jl b/test/monotonic.jl
index 4acfde0d..873cd4f8 100644
--- a/test/monotonic.jl
+++ b/test/monotonic.jl
@@ -15,7 +15,7 @@
     seed = 123
 
     # train-eval split
-    𝑖_sample = sample(𝑖, size(𝑖, 1), replace = false)
+    𝑖_sample = sample(𝑖, size(𝑖, 1), replace=false)
     train_size = 0.8
     𝑖_train = 𝑖_sample[1:floor(Int, train_size * size(𝑖, 1))]
     𝑖_eval = 𝑖_sample[floor(Int, train_size * size(𝑖, 1))+1:end]
@@ -28,51 +28,48 @@
     ######################################
     # benchmark
     params1 = EvoTreeRegressor(
-        device = "cpu",
-        loss = :linear,
-        metric = :mse,
-        nrounds = 200,
-        nbins = 32,
-        lambda = 1.0,
-        gamma = 0.0,
-        eta = 0.05,
-        max_depth = 6,
-        min_weight = 0.0,
-        rowsample = 0.5,
-        colsample = 1.0,
-        rng = seed,
+        device="cpu",
+        loss=:linear,
+        nrounds=20,
+        nbins=32,
+        lambda=1.0,
+        gamma=0.0,
+        eta=0.05,
+        max_depth=6,
+        min_weight=0.0,
+        rowsample=0.5,
+        colsample=1.0,
+        rng=seed,
     )
 
-    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n = 25)
+    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25)
     preds_ref = predict(model, x_train)
 
     # monotonic constraint
     params1 = EvoTreeRegressor(
-        device = "cpu",
-        loss = :linear,
-        metric = :mse,
-        nrounds = 200,
-        nbins = 32,
-        lambda = 1.0,
-        gamma = 0.0,
-        eta = 0.5,
-        max_depth = 6,
-        min_weight = 0.0,
-        monotone_constraints = Dict(1 => 1),
-        rowsample = 0.5,
-        colsample = 1.0,
-        rng = seed,
+        device="cpu",
+        loss=:linear,
+        nrounds=20,
+        nbins=32,
+        lambda=1.0,
+        gamma=0.0,
+        eta=0.5,
+        max_depth=6,
+        min_weight=0.0,
+        monotone_constraints=Dict(1 => 1),
+        rowsample=0.5,
+        colsample=1.0,
+        rng=seed,
     )
 
-    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n = 25)
+    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25)
     preds_mono = predict(model, x_train)
 
     # using Plots
-    # using Colors
-    # x_perm = sortperm(X_train[:, 1])
-    # plot(X_train, Y_train, msize=1, mcolor="gray", mswidth=0, background_color=RGB(1, 1, 1), seriestype=:scatter, xaxis=("feature"), yaxis=("target"), legend=true, label="")
-    # plot!(X_train[:, 1][x_perm], preds_ref[x_perm], color="navy", linewidth=1.5, label="Reference")
-    # plot!(X_train[:, 1][x_perm], preds_mono[x_perm], color="red", linewidth=1.5, label="Monotonic")
+    # x_perm = sortperm(x_train[:, 1])
+    # plot(x_train, y_train, msize=1, mcolor="gray", mswidth=0, background_color=RGB(1, 1, 1), seriestype=:scatter, xaxis=("feature"), yaxis=("target"), legend=true, label="")
+    # plot!(x_train[:, 1][x_perm], preds_ref[x_perm], color="navy", linewidth=1.5, label="Reference")
+    # plot!(x_train[:, 1][x_perm], preds_mono[x_perm], color="red", linewidth=1.5, label="Monotonic")
 
 
     ######################################
@@ -116,51 +113,49 @@
     ######################################
     # benchmark
     params1 = EvoTreeRegressor(
-        device = "cpu",
-        loss = :logistic,
-        metric = :logloss,
-        nrounds = 200,
-        nbins = 32,
-        lambda = 0.05,
-        gamma = 0.0,
-        eta = 0.05,
-        max_depth = 6,
-        min_weight = 0.0,
-        rowsample = 0.5,
-        colsample = 1.0,
-        rng = seed,
+        device="cpu",
+        loss=:logistic,
+        metric=:logloss,
+        nrounds=200,
+        nbins=32,
+        lambda=0.05,
+        gamma=0.0,
+        eta=0.05,
+        max_depth=6,
+        min_weight=0.0,
+        rowsample=0.5,
+        colsample=1.0,
+        rng=seed,
     )
 
-    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n = 25)
+    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25)
     preds_ref = predict(model, x_train)
 
     # monotonic constraint
     params1 = EvoTreeRegressor(
-        device = "cpu",
-        loss = :logistic,
-        metric = :logloss,
-        nrounds = 200,
-        nbins = 32,
-        lambda = 0.05,
-        gamma = 0.0,
-        eta = 0.05,
-        max_depth = 6,
-        min_weight = 0.0,
-        monotone_constraints = Dict(1 => 1),
-        rowsample = 0.5,
-        colsample = 1.0,
-        rng = seed,
+        device="cpu",
+        loss=:logistic,
+        metric=:logloss,
+        nrounds=200,
+        nbins=32,
+        lambda=0.05,
+        gamma=0.0,
+        eta=0.05,
+        max_depth=6,
+        min_weight=0.0,
+        monotone_constraints=Dict(1 => 1),
+        rowsample=0.5,
+        colsample=1.0,
+        rng=seed,
     )
 
-    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n = 25)
+    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25)
     preds_mono = predict(model, x_train)
 
-    # using Plots
-    # using Colors
-    # x_perm = sortperm(X_train[:, 1])
-    # plot(X_train, Y_train, msize=1, mcolor="gray", mswidth=0, background_color=RGB(1, 1, 1), seriestype=:scatter, xaxis=("feature"), yaxis=("target"), legend=true, label="")
-    # plot!(X_train[:, 1][x_perm], preds_ref[x_perm], color="navy", linewidth=1.5, label="Reference")
-    # plot!(X_train[:, 1][x_perm], preds_mono[x_perm], color="red", linewidth=1.5, label="Monotonic")
+    # x_perm = sortperm(x_train[:, 1])
+    # plot(x_train, y_train, msize=1, mcolor="gray", mswidth=0, background_color=RGB(1, 1, 1), seriestype=:scatter, xaxis=("feature"), yaxis=("target"), legend=true, label="")
+    # plot!(x_train[:, 1][x_perm], preds_ref[x_perm], color="navy", linewidth=1.5, label="Reference")
+    # plot!(x_train[:, 1][x_perm], preds_mono[x_perm], color="red", linewidth=1.5, label="Monotonic")
 
 
     ######################################
@@ -202,57 +197,55 @@
     ######################################
     ### Gaussian - CPU
     ######################################
-    # linear - benchmark
+    # benchmark
     params1 = EvoTreeGaussian(
-        device = "cpu",
-        metric = :gaussian,
-        nrounds = 200,
-        nbins = 32,
-        lambda = 1.0,
-        gamma = 0.0,
-        eta = 0.05,
-        max_depth = 6,
-        min_weight = 0.0,
-        rowsample = 0.5,
-        colsample = 1.0,
-        rng = seed,
+        device="cpu",
+        metric=:gaussian,
+        nrounds=200,
+        nbins=32,
+        lambda=1.0,
+        gamma=0.0,
+        eta=0.05,
+        max_depth=6,
+        min_weight=0.0,
+        rowsample=0.5,
+        colsample=1.0,
+        rng=seed,
     )
 
-    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n = 25)
+    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25)
     preds_ref = predict(model, x_train)
 
     # monotonic constraint
     params1 = EvoTreeGaussian(
-        device = "cpu",
-        metric = :gaussian,
-        nrounds = 200,
-        nbins = 32,
-        lambda = 1.0,
-        gamma = 0.0,
-        eta = 0.5,
-        max_depth = 6,
-        min_weight = 0.0,
-        monotone_constraints = Dict(1 => 1),
-        rowsample = 0.5,
-        colsample = 1.0,
-        rng = seed,
+        device="cpu",
+        metric=:gaussian,
+        nrounds=200,
+        nbins=32,
+        lambda=1.0,
+        gamma=0.0,
+        eta=0.5,
+        max_depth=6,
+        min_weight=0.0,
+        monotone_constraints=Dict(1 => 1),
+        rowsample=0.5,
+        colsample=1.0,
+        rng=seed,
     )
 
-    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n = 25)
+    model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25)
     preds_mono = predict(model, x_train)
 
-    # using Plots
-    # using Colors
-    # x_perm = sortperm(X_train[:, 1])
-    # plot(X_train, Y_train, msize=1, mcolor="gray", mswidth=0, background_color=RGB(1, 1, 1), seriestype=:scatter, xaxis=("feature"), yaxis=("target"), legend=true, label="")
-    # plot!(X_train[:, 1][x_perm], preds_ref[x_perm], color="navy", linewidth=1.5, label="Reference")
-    # plot!(X_train[:, 1][x_perm], preds_mono[x_perm], color="red", linewidth=1.5, label="Monotonic")
+    # x_perm = sortperm(x_train[:, 1])
+    # plot(x_train, y_train, msize=1, mcolor="gray", mswidth=0, background_color=RGB(1, 1, 1), seriestype=:scatter, xaxis=("feature"), yaxis=("target"), legend=true, label="")
+    # plot!(x_train[:, 1][x_perm], preds_ref[x_perm], color="navy", linewidth=1.5, label="Reference")
+    # plot!(x_train[:, 1][x_perm], preds_mono[x_perm], color="red", linewidth=1.5, label="Monotonic")
 
 
     ######################################
     ### Gaussian - GPU
     ######################################
-    # linear - benchmark
+    # benchmark
     # params1 = EvoTreeGaussian(
     #     device="gpu",
     #     metric=:gaussian,
@@ -261,8 +254,8 @@
     #     max_depth=6, min_weight=0.0,
     #     rowsample=0.5, colsample=1.0, rng=seed)
 
-    # model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25);
-    # preds_ref = predict(model, x_train);
+    # model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25)
+    # preds_ref = predict(model, x_train)
 
     # # monotonic constraint
     # params1 = EvoTreeGaussian(
@@ -274,14 +267,12 @@
     #     monotone_constraints=Dict(1 => 1),
     #     rowsample=0.5, colsample=1.0, rng=seed)
 
-    # model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25);
-    # preds_mono = predict(model, x_train);
+    # model = fit_evotree(params1; x_train, y_train, x_eval, y_eval, print_every_n=25)
+    # preds_mono = predict(model, x_train)
 
-    # using Plots
-    # using Colors
-    # x_perm = sortperm(X_train[:, 1])
-    # plot(X_train, Y_train, msize=1, mcolor="gray", mswidth=0, background_color=RGB(1, 1, 1), seriestype=:scatter, xaxis=("feature"), yaxis=("target"), legend=true, label="")
-    # plot!(X_train[:, 1][x_perm], preds_ref[x_perm], color="navy", linewidth=1.5, label="Reference")
-    # plot!(X_train[:, 1][x_perm], preds_mono[x_perm], color="red", linewidth=1.5, label="Monotonic")
+    # x_perm = sortperm(x_train[:, 1])
+    # plot(x_train, y_train, msize=1, mcolor="gray", mswidth=0, background_color=RGB(1, 1, 1), seriestype=:scatter, xaxis=("feature"), yaxis=("target"), legend=true, label="GPU Gauss")
+    # plot!(x_train[:, 1][x_perm], preds_ref[x_perm], color="navy", linewidth=1.5, label="Reference")
+    # plot!(x_train[:, 1][x_perm], preds_mono[x_perm], color="red", linewidth=1.5, label="Monotonic")
 
 end
\ No newline at end of file