diff --git a/GPy/__init__.py b/GPy/__init__.py
index 99a91fe8..5e091170 100644
--- a/GPy/__init__.py
+++ b/GPy/__init__.py
@@ -39,6 +39,11 @@ def load(file_path):
     :param file_name: path/to/file.pickle
     """
     import cPickle as pickle
-    with open(file_path, 'rb') as f:
-        m = pickle.load(f)
+    try:
+        with open(file_path, 'rb') as f:
+            m = pickle.load(f)
+    except:
+        import pickle as pickle
+        with open(file_path, 'rb') as f:
+            m = pickle.load(f)
     return m
diff --git a/GPy/core/model.py b/GPy/core/model.py
index 029b8fc8..ac5a9732 100644
--- a/GPy/core/model.py
+++ b/GPy/core/model.py
@@ -238,6 +238,10 @@ class Model(Parameterized):
         if self.size == 0:
             print 'nothing to optimize'
 
+        if not self.update_model():
+            print "setting updates on again"
+            self.update_model(True)
+
         if start == None:
             start = self.optimizer_array
 
@@ -279,10 +283,10 @@ class Model(Parameterized):
         Note:-
            The gradient is considered correct if the ratio of the analytical
            and numerical gradients is within <tolerance> of unity.
-           
-           The *dF_ratio* indicates the limit of numerical accuracy of numerical gradients. 
-           If it is too small, e.g., smaller than 1e-12, the numerical gradients are usually 
-           not accurate enough for the tests (shown with blue). 
+
+           The *dF_ratio* indicates the limit of numerical accuracy of numerical gradients.
+           If it is too small, e.g., smaller than 1e-12, the numerical gradients are usually
+           not accurate enough for the tests (shown with blue).
         """
         x = self.optimizer_array.copy()
 
diff --git a/GPy/core/parameterization/transformations.py b/GPy/core/parameterization/transformations.py
index 01b0f324..291076a1 100644
--- a/GPy/core/parameterization/transformations.py
+++ b/GPy/core/parameterization/transformations.py
@@ -131,6 +131,13 @@ class NormalTheta(Transformation):
     def __str__(self):
         return "theta"
 
+    def __getstate__(self):
+        return [self.mu_indices, self.var_indices]
+
+    def __setstate__(self, state):
+        self.mu_indices = state[0]
+        self.var_indices = state[1]
+
 class NormalNaturalAntti(NormalTheta):
     _instances = []
     _logexp = Logexp()
diff --git a/GPy/models/bayesian_gplvm_minibatch.py b/GPy/models/bayesian_gplvm_minibatch.py
index 5b7d3162..80abba59 100644
--- a/GPy/models/bayesian_gplvm_minibatch.py
+++ b/GPy/models/bayesian_gplvm_minibatch.py
@@ -54,6 +54,8 @@ class BayesianGPLVMMiniBatch(SparseGPMiniBatch):
         self.variational_prior = NormalPrior()
         X = NormalPosterior(X, X_variance)
 
+        self.kl_factr = 1.
+
         if inference_method is None:
             from ..inference.latent_function_inference.var_dtc import VarDTC
             self.logger.debug("creating inference_method var_dtc")
diff --git a/GPy/models/mrd.py b/GPy/models/mrd.py
index 91d34aa2..645cdf88 100644
--- a/GPy/models/mrd.py
+++ b/GPy/models/mrd.py
@@ -100,7 +100,7 @@ class MRD(BayesianGPLVMMiniBatch):
         self.logger.info("building kernels")
         if kernel is None:
             from ..kern import RBF
-            kernels = [RBF(input_dim, ARD=1, lengthscale=fracs[i]) for i in range(len(Ylist))]
+            kernels = [RBF(input_dim, ARD=1, lengthscale=1./fracs[i]) for i in range(len(Ylist))]
         elif isinstance(kernel, Kern):
             kernels = []
             for i in range(len(Ylist)):
diff --git a/GPy/plotting/matplot_dep/dim_reduction_plots.py b/GPy/plotting/matplot_dep/dim_reduction_plots.py
index 7ec0b59d..25a1166f 100644
--- a/GPy/plotting/matplot_dep/dim_reduction_plots.py
+++ b/GPy/plotting/matplot_dep/dim_reduction_plots.py
@@ -162,7 +162,7 @@ def plot_latent(model, labels=None, which_indices=None,
         else:
             x = X[index, input_1]
             y = X[index, input_2]
-        ax.scatter(x, y, marker=m, s=s, c=Tango.nextMedium(), label=this_label, linewidth=.5, edgecolor='k', alpha=.9)
+        ax.scatter(x, y, marker=m, s=s, c=Tango.nextMedium(), label=this_label, linewidth=.2, edgecolor='k', alpha=.9)
 
     ax.set_xlabel('latent dimension %i' % input_1)
     ax.set_ylabel('latent dimension %i' % input_2)