working mnist-conv example

examples/mnist-conv.jl

@@ -2,9 +2,9 @@ using Flux
 
 # Flux aims to provide high-level APIs that work well across backends, but in
 # some cases you may want to take advantage of features specific to a given
-# backend (or alternatively, Flux may simply not have an implementation of that
+# backend (or Flux may simply not have an implementation of that feature yet).
-# feature yet). In these cases it's easy to "drop down" and use the backend's
+# In these cases it's easy to "drop down" and use the backend's API directly,
-# API directly, where appropriate.
+# where appropriate.
 
 # In this example, both things are happening; firstly, Flux doesn't yet support
 # ConvNets in the pure-Julia backend, but this is invisible thanks to the use of
@@ -12,22 +12,22 @@ using Flux
 # have been user-defined.
 
 # Secondly, we want to take advantage of TensorFlow.jl's training process and
-# optimisers. We can simply call `mx.FeedForward` exactly as we would on a
+# optimisers. We can simply call `Tensor` exactly as we would on a regular
-# regular TensorFlow model, and the rest of the process is trivial.
+# TensorFlow model, and the rest of the process trivially follows
+# TensorFlow.jl's usual API.
 
 conv1 = Chain(
-  Input(28,28),
+  Reshape(28,28,1),
   Conv2D((5,5), out = 20), tanh,
   MaxPool((2,2), stride = (2,2)))
 
 conv2 = Chain(
-  conv1,
+  Input(12,12,20),
-  Conv2D((5,5), out = 50), tanh,
+  Conv2D((5,5), in = 20, out = 50), tanh,
   MaxPool((2,2), stride = (2,2)))
 
 lenet = Chain(
-  conv2,
+  conv1, conv2, flatten,
-  flatten,
   Dense(500), tanh,
   Dense(10), softmax)
 
@@ -35,13 +35,14 @@ lenet = Chain(
 
 # Now we can continue exactly as in plain TensorFlow, following
 #   https://github.com/malmaud/TensorFlow.jl/blob/master/examples/mnist_full.jl
+# (taking only the training and cost logic, not the graph building steps)
 
 using TensorFlow, Distributions
 
 include(Pkg.dir("TensorFlow", "examples", "mnist_loader.jl"))
 loader = DataLoader()
 
-sess = Session(Graph())
+session = Session(Graph())
 
 x  = placeholder(Float32)
 y′ = placeholder(Float32)
@@ -51,21 +52,19 @@ cross_entropy = reduce_mean(-reduce_sum(y′.*log(y), reduction_indices=[2]))
 
 train_step = train.minimize(train.AdamOptimizer(1e-4), cross_entropy)
 
-correct_prediction = indmax(y, 2) .== indmax(y′, 2)
+accuracy = reduce_mean(cast(indmax(y, 2) .== indmax(y′, 2), Float32))
-
-accuracy = reduce_mean(cast(correct_prediction, Float32))
 
 run(session, initialize_all_variables())
 
-for i in 1:1000
+@progress for i in 1:1000
     batch = next_batch(loader, 50)
     if i%100 == 1
-        train_accuracy = run(session, accuracy, Dict(x=>batch[1], y′=>batch[2], keep_prob=>1.0))
+        train_accuracy = run(session, accuracy, Dict(x=>batch[1], y′=>batch[2]))
         info("step $i, training accuracy $train_accuracy")
     end
-    run(session, train_step, Dict(x=>batch[1], y′=>batch[2], keep_prob=>.5))
+    run(session, train_step, Dict(x=>batch[1], y′=>batch[2]))
 end
 
 testx, testy = load_test_set()
-test_accuracy = run(session, accuracy, Dict(x=>testx, y′=>testy, keep_prob=>1.0))
+test_accuracy = run(session, accuracy, Dict(x=>testx, y′=>testy))
 info("test accuracy $test_accuracy")