non-working example

examples/mnist-conv.jl

@@ -36,15 +36,36 @@ lenet = Chain(
 # Now we can continue exactly as in plain TensorFlow, following
 #   https://github.com/malmaud/TensorFlow.jl/blob/master/examples/mnist_full.jl
 
-using TensorFlow
+using TensorFlow, Distributions
+
+include(Pkg.dir("TensorFlow", "examples", "mnist_loader.jl"))
+loader = DataLoader()
 
 sess = Session(Graph())
 
 x  = placeholder(Float32)
 y′ = placeholder(Float32)
+y  = Tensor(lenet, x)
 
-y = Tensor(lenet, x)
+cross_entropy = reduce_mean(-reduce_sum(y′.*log(y), reduction_indices=[2]))
 
-include(Pkg.dir("TensorFlow", "examples", "mnist_loader.jl"))
+train_step = train.minimize(train.AdamOptimizer(1e-4), cross_entropy)
 
-loader = DataLoader()
+correct_prediction = indmax(y, 2) .== indmax(y′, 2)
+
+accuracy = reduce_mean(cast(correct_prediction, Float32))
+
+run(session, initialize_all_variables())
+
+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))
+        info("step $i, training accuracy $train_accuracy")
+    end
+    run(session, train_step, Dict(x=>batch[1], y′=>batch[2], keep_prob=>.5))
+end
+
+testx, testy = load_test_set()
+test_accuracy = run(session, accuracy, Dict(x=>testx, y′=>testy, keep_prob=>1.0))
+info("test accuracy $test_accuracy")