MNIST Classification with SimpleChains

SimpleChains.jl is an excellent framework for training small neural networks. In this tutorial we will demonstrate how to use the same API as Lux.jl to train a model using SimpleChains.jl. We will use the tutorial from SimpleChains.jl as a reference.

Package Imports

using Lux, ADTypes, MLUtils, Optimisers, Zygote, OneHotArrays, Random, Statistics, Printf
import MLDatasets: MNIST
import SimpleChains: static

Loading MNIST

function loadmnist(batchsize, train_split)
    # Load MNIST
    N = 2000
    dataset = MNIST(; split=:train)
    imgs = dataset.features[:, :, 1:N]
    labels_raw = dataset.targets[1:N]

    # Process images into (H,W,C,BS) batches
    x_data = Float32.(reshape(imgs, size(imgs, 1), size(imgs, 2), 1, size(imgs, 3)))
    y_data = onehotbatch(labels_raw, 0:9)
    (x_train, y_train), (x_test, y_test) = splitobs((x_data, y_data); at=train_split)

    return (
        # Use DataLoader to automatically minibatch and shuffle the data
        DataLoader(collect.((x_train, y_train)); batchsize, shuffle=true),
        # Don't shuffle the test data
        DataLoader(collect.((x_test, y_test)); batchsize, shuffle=false))
end

loadmnist (generic function with 1 method)

Define the Model

lux_model = Chain(Conv((5, 5), 1 => 6, relu), MaxPool((2, 2)),
    Conv((5, 5), 6 => 16, relu), MaxPool((2, 2)), FlattenLayer(3),
    Chain(Dense(256 => 128, relu), Dense(128 => 84, relu), Dense(84 => 10)))

Chain(
    layer_1 = Conv((5, 5), 1 => 6, relu),  # 156 parameters
    layer_2 = MaxPool((2, 2)),
    layer_3 = Conv((5, 5), 6 => 16, relu),  # 2_416 parameters
    layer_4 = MaxPool((2, 2)),
    layer_5 = FlattenLayer(),
    layer_6 = Dense(256 => 128, relu),  # 32_896 parameters
    layer_7 = Dense(128 => 84, relu),   # 10_836 parameters
    layer_8 = Dense(84 => 10),          # 850 parameters
)         # Total: 47_154 parameters,
          #        plus 0 states.

We now need to convert the lux_model to SimpleChains.jl. We need to do this by defining the ToSimpleChainsAdaptor and providing the input dimensions.

adaptor = ToSimpleChainsAdaptor((static(28), static(28), static(1)))
simple_chains_model = adaptor(lux_model)

SimpleChainsLayer()  # 47_154 parameters

Helper Functions

logitcrossentropy(y_pred, y) = mean(-sum(y .* logsoftmax(y_pred); dims=1))

function loss(model, ps, st, (x, y))
    y_pred, st = model(x, ps, st)
    return logitcrossentropy(y_pred, y), st, (;)
end

function accuracy(model, ps, st, dataloader)
    total_correct, total = 0, 0
    st = Lux.testmode(st)
    for (x, y) in dataloader
        target_class = onecold(y)
        predicted_class = onecold(Array(first(model(x, ps, st))))
        total_correct += sum(target_class .== predicted_class)
        total += length(target_class)
    end
    return total_correct / total
end

accuracy (generic function with 1 method)

Define the Training Loop

function train(model; rng=Xoshiro(0), kwargs...)
    train_dataloader, test_dataloader = loadmnist(128, 0.9)

    train_state = Lux.Experimental.TrainState(
        rng, model, Adam(3.0f-4); transform_variables=identity)

    ### Warmup the model
    x_proto = randn(rng, Float32, 28, 28, 1, 1)
    y_proto = onehotbatch([1], 0:9)
    Lux.Experimental.compute_gradients(AutoZygote(), loss, (x_proto, y_proto), train_state)

    ### Lets train the model
    nepochs = 10
    for epoch in 1:nepochs
        stime = time()
        for (x, y) in train_dataloader
            (gs, _, _, train_state) = Lux.Experimental.compute_gradients(
                AutoZygote(), loss, (x, y), train_state)
            train_state = Lux.Experimental.apply_gradients!(train_state, gs)
        end
        ttime = time() - stime

        tr_acc = accuracy(
            model, train_state.parameters, train_state.states, train_dataloader) * 100
        te_acc = accuracy(
            model, train_state.parameters, train_state.states, test_dataloader) * 100

        @printf "[%2d/%2d] \t Time %.2fs \t Training Accuracy: %.2f%% \t Test Accuracy: %.2f%%\n" epoch nepochs ttime tr_acc te_acc
    end
end

train (generic function with 1 method)

Finally Training the Model

First we will train the Lux model

train(lux_model)

[ 1/10] 	 Time 47.95s 	 Training Accuracy: 24.11% 	 Test Accuracy: 24.00%
[ 2/10] 	 Time 38.17s 	 Training Accuracy: 46.89% 	 Test Accuracy: 47.50%
[ 3/10] 	 Time 39.66s 	 Training Accuracy: 68.06% 	 Test Accuracy: 67.50%
[ 4/10] 	 Time 40.05s 	 Training Accuracy: 74.33% 	 Test Accuracy: 72.50%
[ 5/10] 	 Time 43.50s 	 Training Accuracy: 80.61% 	 Test Accuracy: 79.00%
[ 6/10] 	 Time 39.45s 	 Training Accuracy: 82.83% 	 Test Accuracy: 82.50%
[ 7/10] 	 Time 39.48s 	 Training Accuracy: 84.72% 	 Test Accuracy: 83.00%
[ 8/10] 	 Time 42.50s 	 Training Accuracy: 85.61% 	 Test Accuracy: 84.00%
[ 9/10] 	 Time 43.97s 	 Training Accuracy: 85.83% 	 Test Accuracy: 84.50%
[10/10] 	 Time 37.68s 	 Training Accuracy: 87.61% 	 Test Accuracy: 85.50%

Now we will train the SimpleChains model

train(simple_chains_model)

[ 1/10] 	 Time 18.55s 	 Training Accuracy: 29.78% 	 Test Accuracy: 27.00%
[ 2/10] 	 Time 17.46s 	 Training Accuracy: 40.83% 	 Test Accuracy: 38.00%
[ 3/10] 	 Time 17.47s 	 Training Accuracy: 60.06% 	 Test Accuracy: 55.50%
[ 4/10] 	 Time 17.46s 	 Training Accuracy: 66.33% 	 Test Accuracy: 62.00%
[ 5/10] 	 Time 17.46s 	 Training Accuracy: 74.28% 	 Test Accuracy: 71.00%
[ 6/10] 	 Time 17.46s 	 Training Accuracy: 80.33% 	 Test Accuracy: 76.00%
[ 7/10] 	 Time 17.58s 	 Training Accuracy: 82.94% 	 Test Accuracy: 81.00%
[ 8/10] 	 Time 17.47s 	 Training Accuracy: 83.61% 	 Test Accuracy: 80.50%
[ 9/10] 	 Time 17.47s 	 Training Accuracy: 85.61% 	 Test Accuracy: 82.00%
[10/10] 	 Time 17.47s 	 Training Accuracy: 87.06% 	 Test Accuracy: 84.00%

On my local machine we see a 3-4x speedup when using SimpleChains.jl. The conditions of the server this documentation is being built on is not ideal for CPU benchmarking hence, the speedup may not be as significant and even there might be regressions.

Appendix

using InteractiveUtils
InteractiveUtils.versioninfo()

if @isdefined(LuxCUDA) && CUDA.functional()
    println()
    CUDA.versioninfo()
end

if @isdefined(LuxAMDGPU) && LuxAMDGPU.functional()
    println()
    AMDGPU.versioninfo()
end

Julia Version 1.10.3
Commit 0b4590a5507 (2024-04-30 10:59 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 48 × AMD EPYC 7402 24-Core Processor
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-15.0.7 (ORCJIT, znver2)
Threads: 8 default, 0 interactive, 4 GC (on 2 virtual cores)
Environment:
  JULIA_CPU_THREADS = 2
  JULIA_AMDGPU_LOGGING_ENABLED = true
  JULIA_DEPOT_PATH = /root/.cache/julia-buildkite-plugin/depots/01872db4-8c79-43af-ab7d-12abac4f24f6
  LD_LIBRARY_PATH = /usr/local/nvidia/lib:/usr/local/nvidia/lib64
  JULIA_PKG_SERVER = 
  JULIA_NUM_THREADS = 8
  JULIA_CUDA_HARD_MEMORY_LIMIT = 100%
  JULIA_PKG_PRECOMPILE_AUTO = 0
  JULIA_DEBUG = Literate

---

This page was generated using Literate.jl.