diff --git a/README.md b/README.md
index c1d84e3..283a2e5 100644
--- a/README.md
+++ b/README.md
@@ -17,11 +17,12 @@ Two important functions are provided for training and prediction:
 
 ## Installation 
 
-You clone or download this repository and then install via `python setup.py install`, or copy the `mdn` folder into your own project.
+This project requires Python 3.6+. You can clone or download this repository and then install via `python setup.py install`, or copy the `mdn` folder into your own project. 
 
 You can easily install this package directly from Github via `pip` like so:
 
     pip install git+git://github.com/cpmpercussion/keras-mdn-layer.git#egg=keras-mdn-layer
+    
 
 And finally, import the `mdn` module in Python: `import mdn`
 
diff --git a/notebooks/MDN-2D-spiral-prediction.ipynb b/notebooks/MDN-2D-spiral-prediction.ipynb
index 677eeb1..07e2e97 100644
--- a/notebooks/MDN-2D-spiral-prediction.ipynb
+++ b/notebooks/MDN-2D-spiral-prediction.ipynb
@@ -142,26 +142,6 @@
     "plt.show()"
    ]
   },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Sampling Functions\n",
-    "\n",
-    "The MDN model outputs parameters of a mixture model---a list of means (mu), variances (sigma), and weights (pi).\n",
-    "\n",
-    "The MDN package contains some functions to split up these parameters and sample from the normal distributions that they form.\n",
-    "\n",
-    "We use \n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -191,12 +171,12 @@
     "# To find points on the graph, we need to sample from each distribution.\n",
     "\n",
     "# Split up the mixture parameters (for future fun)\n",
-    "mus = np.apply_along_axis((lambda a: a[:N_MIXES*OUTPUT_DIMS]),1, y_test)\n",
-    "sigs = np.apply_along_axis((lambda a: a[N_MIXES*OUTPUT_DIMS:2*N_MIXES*OUTPUT_DIMS]),1, y_test)\n",
-    "pis = np.apply_along_axis((lambda a: softmax(a[-N_MIXES:])),1, y_test)\n",
+    "mus = np.apply_along_axis((lambda a: a[:N_MIXES*OUTPUT_DIMS]), 1, y_test)\n",
+    "sigs = np.apply_along_axis((lambda a: a[N_MIXES*OUTPUT_DIMS:2*N_MIXES*OUTPUT_DIMS]), 1, y_test)\n",
+    "pis = np.apply_along_axis((lambda a: mdn.softmax(a[-N_MIXES:])), 1, y_test)\n",
     "\n",
     "# Sample from the predicted distributions\n",
-    "y_samples = np.apply_along_axis(sample_from_output, 1, y_test, N_MIXES,OUTPUT_DIMS,temp=1.0)"
+    "y_samples = np.apply_along_axis(mdn.sample_from_output, 1, y_test, OUTPUT_DIMS, N_MIXES, temp=1.0, sigma_temp=1.0)"
    ]
   },
   {
@@ -254,9 +234,9 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "Python 3",
+   "display_name": "venv",
    "language": "python",
-   "name": "python3"
+   "name": "venv"
   },
   "language_info": {
    "codemirror_mode": {
@@ -268,7 +248,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.5"
+   "version": "3.6.8"
   }
  },
  "nbformat": 4,