diff --git a/latest/apis/backends.html b/latest/apis/backends.html
index 19e3f8bf..3d50c07d 100644
--- a/latest/apis/backends.html
+++ b/latest/apis/backends.html
@@ -115,7 +115,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="../examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -150,7 +150,7 @@ Backends
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/apis/backends.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/apis/backends.md">
             <span class="fa">
 
             </span>
diff --git a/latest/apis/batching.html b/latest/apis/batching.html
index 91e0620e..2d8e8495 100644
--- a/latest/apis/batching.html
+++ b/latest/apis/batching.html
@@ -120,7 +120,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="../examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -155,7 +155,7 @@ Batching
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/apis/batching.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/apis/batching.md">
             <span class="fa">
 
             </span>
diff --git a/latest/apis/storage.html b/latest/apis/storage.html
index ede30668..30d06add 100644
--- a/latest/apis/storage.html
+++ b/latest/apis/storage.html
@@ -104,7 +104,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="../examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -139,7 +139,7 @@ Storing Models
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/apis/storage.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/apis/storage.md">
             <span class="fa">
 
             </span>
@@ -198,7 +198,7 @@ Backends
 Next
           </span>
           <span class="title">
-Logistic Regression
+Simple MNIST
           </span>
         </a>
       </footer>
diff --git a/latest/contributing.html b/latest/contributing.html
index 48bf6a9b..36113728 100644
--- a/latest/contributing.html
+++ b/latest/contributing.html
@@ -103,7 +103,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -136,7 +136,7 @@ Contributing &amp; Help
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/contributing.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/contributing.md">
             <span class="fa">
 
             </span>
diff --git a/latest/examples/char-rnn.html b/latest/examples/char-rnn.html
index 4430cbb2..012bc936 100644
--- a/latest/examples/char-rnn.html
+++ b/latest/examples/char-rnn.html
@@ -103,7 +103,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li class="current">
@@ -139,7 +139,7 @@ Char RNN
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/examples/char-rnn.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/examples/char-rnn.md">
             <span class="fa">
 
             </span>
@@ -252,7 +252,7 @@ end</code></pre>
 Previous
           </span>
           <span class="title">
-Logistic Regression
+Simple MNIST
           </span>
         </a>
         <a class="next" href="../contributing.html">
diff --git a/latest/examples/logreg.html b/latest/examples/logreg.html
index ac03f0f8..2127bb78 100644
--- a/latest/examples/logreg.html
+++ b/latest/examples/logreg.html
@@ -4,7 +4,7 @@
     <meta charset="UTF-8"/>
     <meta name="viewport" content="width=device-width, initial-scale=1.0"/>
     <title>
-Logistic Regression · Flux
+Simple MNIST · Flux
     </title>
     <script>
 (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
@@ -103,7 +103,7 @@ In Action
           <ul>
             <li class="current">
               <a class="toctext" href="logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
               <ul class="internal"></ul>
             </li>
@@ -135,11 +135,11 @@ In Action
             </li>
             <li>
               <a href="logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/examples/logreg.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/examples/logreg.md">
             <span class="fa">
 
             </span>
@@ -149,8 +149,8 @@ Logistic Regression
         <hr/>
       </header>
       <h1>
-        <a class="nav-anchor" id="Logistic-Regression-with-MNIST-1" href="#Logistic-Regression-with-MNIST-1">
-Logistic Regression with MNIST
+        <a class="nav-anchor" id="Recognising-MNIST-Digits-1" href="#Recognising-MNIST-Digits-1">
+Recognising MNIST Digits
         </a>
       </h1>
       <p>
@@ -196,7 +196,7 @@ Now we define our model, which will simply be a function from one to the other.
   Affine( 64), relu,
   Affine( 10), softmax)
 
-model = tf(m)</code></pre>
+model = mxnet(m) # Convert to MXNet</code></pre>
       <p>
 We can try this out on our data already:
       </p>
diff --git a/latest/index.html b/latest/index.html
index eddd001a..11031a5d 100644
--- a/latest/index.html
+++ b/latest/index.html
@@ -115,7 +115,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -147,7 +147,7 @@ Home
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/index.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/index.md">
             <span class="fa">
 
             </span>
diff --git a/latest/internals.html b/latest/internals.html
index 873c1fda..42b5a015 100644
--- a/latest/internals.html
+++ b/latest/internals.html
@@ -103,7 +103,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -136,7 +136,7 @@ Internals
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/internals.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/internals.md">
             <span class="fa">
 
             </span>
diff --git a/latest/models/basics.html b/latest/models/basics.html
index d16f31c9..ac6dae50 100644
--- a/latest/models/basics.html
+++ b/latest/models/basics.html
@@ -120,7 +120,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="../examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -155,7 +155,7 @@ Model Building Basics
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/models/basics.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/models/basics.md">
             <span class="fa">
 
             </span>
diff --git a/latest/models/debugging.html b/latest/models/debugging.html
index 15f6eac1..a98ec373 100644
--- a/latest/models/debugging.html
+++ b/latest/models/debugging.html
@@ -104,7 +104,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="../examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -139,7 +139,7 @@ Debugging
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/models/debugging.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/models/debugging.md">
             <span class="fa">
 
             </span>
diff --git a/latest/models/recurrent.html b/latest/models/recurrent.html
index 2390d7b9..272e3f1d 100644
--- a/latest/models/recurrent.html
+++ b/latest/models/recurrent.html
@@ -104,7 +104,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="../examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -139,7 +139,7 @@ Recurrence
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/models/recurrent.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/models/recurrent.md">
             <span class="fa">
 
             </span>
diff --git a/latest/models/templates.html b/latest/models/templates.html
index c9404887..eb8f2a64 100644
--- a/latest/models/templates.html
+++ b/latest/models/templates.html
@@ -120,7 +120,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="../examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
@@ -155,7 +155,7 @@ Model Templates
               </a>
             </li>
           </ul>
-          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/15b3ce1adad9e07367e68c8c3758380b02d6a4a1/docs/src/models/templates.md">
+          <a class="edit-page" href="https://github.com/MikeInnes/Flux.jl/tree/5c548fe1c4c9d23651dd364648cb28fc62b08855/docs/src/models/templates.md">
             <span class="fa">
 
             </span>
diff --git a/latest/search.html b/latest/search.html
index c7e04d0c..0d48b8b8 100644
--- a/latest/search.html
+++ b/latest/search.html
@@ -103,7 +103,7 @@ In Action
           <ul>
             <li>
               <a class="toctext" href="examples/logreg.html">
-Logistic Regression
+Simple MNIST
               </a>
             </li>
             <li>
diff --git a/latest/search_index.js b/latest/search_index.js
index c1fba779..be39c251 100644
--- a/latest/search_index.js
+++ b/latest/search_index.js
@@ -234,18 +234,18 @@ var documenterSearchIndex = {"docs": [
 
 {
     "location": "examples/logreg.html#",
-    "page": "Logistic Regression",
-    "title": "Logistic Regression",
+    "page": "Simple MNIST",
+    "title": "Simple MNIST",
     "category": "page",
     "text": ""
 },
 
 {
-    "location": "examples/logreg.html#Logistic-Regression-with-MNIST-1",
-    "page": "Logistic Regression",
-    "title": "Logistic Regression with MNIST",
+    "location": "examples/logreg.html#Recognising-MNIST-Digits-1",
+    "page": "Simple MNIST",
+    "title": "Recognising MNIST Digits",
     "category": "section",
-    "text": "This walkthrough example will take you through writing a multi-layer perceptron that classifies MNIST digits with high accuracy.First, we load the data using the MNIST package:using Flux, MNIST\n\ndata = [(trainfeatures(i), onehot(trainlabel(i), 0:9)) for i = 1:60_000]\ntrain = data[1:50_000]\ntest = data[50_001:60_000]The only Flux-specific function here is onehot, which takes a class label and turns it into a one-hot-encoded vector that we can use for training. For example:julia> onehot(:b, [:a, :b, :c])\n3-element Array{Int64,1}:\n 0\n 1\n 0Otherwise, the format of the data is simple enough, it's just a list of tuples from input to output. For example:julia> data[1]\n([0.0,0.0,0.0, … 0.0,0.0,0.0],[0,0,0,0,0,1,0,0,0,0])data[1][1] is a 28*28 == 784 length vector (mostly zeros due to the black background) and data[1][2] is its classification.Now we define our model, which will simply be a function from one to the other.m = Chain(\n  Input(784),\n  Affine(128), relu,\n  Affine( 64), relu,\n  Affine( 10), softmax)\n\nmodel = tf(m)We can try this out on our data already:julia> model(data[1][1])\n10-element Array{Float64,1}:\n 0.10614  \n 0.0850447\n 0.101474\n ...The model gives a probability of about 0.1 to each class – which is a way of saying, \"I have no idea\". This isn't too surprising as we haven't shown it any data yet. This is easy to fix:Flux.train!(model, train, test, η = 1e-4)The training step takes about 5 minutes (to make it faster we can do smarter things like batching). If you run this code in Juno, you'll see a progress meter, which you can hover over to see the remaining computation time.Towards the end of the training process, Flux will have reported that the accuracy of the model is now about 90%. We can try it on our data again:10-element Array{Float32,1}:\n ...\n 5.11423f-7\n 0.9354     \n 3.1033f-5  \n 0.000127077\n ...Notice the class at 93%, suggesting our model is very confident about this image. We can use onecold to compare the true and predicted classes:julia> onecold(data[1][2], 0:9)\n5\n\njulia> onecold(model(data[1][1]), 0:9)\n5Success!"
+    "text": "This walkthrough example will take you through writing a multi-layer perceptron that classifies MNIST digits with high accuracy.First, we load the data using the MNIST package:using Flux, MNIST\n\ndata = [(trainfeatures(i), onehot(trainlabel(i), 0:9)) for i = 1:60_000]\ntrain = data[1:50_000]\ntest = data[50_001:60_000]The only Flux-specific function here is onehot, which takes a class label and turns it into a one-hot-encoded vector that we can use for training. For example:julia> onehot(:b, [:a, :b, :c])\n3-element Array{Int64,1}:\n 0\n 1\n 0Otherwise, the format of the data is simple enough, it's just a list of tuples from input to output. For example:julia> data[1]\n([0.0,0.0,0.0, … 0.0,0.0,0.0],[0,0,0,0,0,1,0,0,0,0])data[1][1] is a 28*28 == 784 length vector (mostly zeros due to the black background) and data[1][2] is its classification.Now we define our model, which will simply be a function from one to the other.m = Chain(\n  Input(784),\n  Affine(128), relu,\n  Affine( 64), relu,\n  Affine( 10), softmax)\n\nmodel = mxnet(m) # Convert to MXNetWe can try this out on our data already:julia> model(data[1][1])\n10-element Array{Float64,1}:\n 0.10614  \n 0.0850447\n 0.101474\n ...The model gives a probability of about 0.1 to each class – which is a way of saying, \"I have no idea\". This isn't too surprising as we haven't shown it any data yet. This is easy to fix:Flux.train!(model, train, test, η = 1e-4)The training step takes about 5 minutes (to make it faster we can do smarter things like batching). If you run this code in Juno, you'll see a progress meter, which you can hover over to see the remaining computation time.Towards the end of the training process, Flux will have reported that the accuracy of the model is now about 90%. We can try it on our data again:10-element Array{Float32,1}:\n ...\n 5.11423f-7\n 0.9354     \n 3.1033f-5  \n 0.000127077\n ...Notice the class at 93%, suggesting our model is very confident about this image. We can use onecold to compare the true and predicted classes:julia> onecold(data[1][2], 0:9)\n5\n\njulia> onecold(model(data[1][1]), 0:9)\n5Success!"
 },
 
 {