Getting Started

Installation

Install Julia v1.10 or above. Lux.jl is available through the Julia package manager. You can enter it by pressing ] in the REPL and then typing add Lux. Alternatively, you can also do

julia

import Pkg
Pkg.add("Lux")

Update to v1

If you are using a pre-v1 version of Lux.jl, please see the Updating to v1 section for instructions on how to update.

Quickstart

Pre-Requisites

You need to install Optimisers and Zygote if not done already. Pkg.add(["Optimisers", "Zygote"])

julia

using Lux, Random, Optimisers, Zygote
# using LuxCUDA, AMDGPU, Metal, oneAPI # Optional packages for GPU support

We take randomness very seriously

julia

# Seeding
rng = Random.default_rng()
Random.seed!(rng, 0)

Random.TaskLocalRNG()

Build the model

julia

# Construct the layer
model = Chain(Dense(128, 256, tanh), Chain(Dense(256, 1, tanh), Dense(1, 10)))

Chain(
    layer_1 = Dense(128 => 256, tanh),  # 33_024 parameters
    layer_2 = Chain(
        layer_1 = Dense(256 => 1, tanh),  # 257 parameters
        layer_2 = Dense(1 => 10),       # 20 parameters
    ),
)         # Total: 33_301 parameters,
          #        plus 0 states.

Models don't hold parameters and states so initialize them. From there on, we can just use our standard AD and Optimisers API. However, here we will show how to use Lux's Training API that provides an uniform API over all supported AD systems.

julia

# Get the device determined by Lux
dev = gpu_device()

# Parameter and State Variables
ps, st = Lux.setup(rng, model) |> dev

# Dummy Input
x = rand(rng, Float32, 128, 2) |> dev

# Run the model
y, st = Lux.apply(model, x, ps, st)

# Gradients
## First construct a TrainState
train_state = Lux.Training.TrainState(model, ps, st, Adam(0.0001f0))

## We can compute the gradients using Training.compute_gradients
gs, loss, stats, train_state = Lux.Training.compute_gradients(
    AutoZygote(), MSELoss(),
    (x, dev(rand(rng, Float32, 10, 2))), train_state
)

## Optimization
train_state = Training.apply_gradients!(train_state, gs) # or Training.apply_gradients (no `!` at the end)

# Both these steps can be combined into a single call
gs, loss, stats, train_state = Training.single_train_step!(
    AutoZygote(), MSELoss(),
    (x, dev(rand(rng, Float32, 10, 2))), train_state
)

((layer_1 = (weight = Float32[0.0017983615 0.006062332 … 0.0053392933 0.0056276177; 0.0011292367 0.0041270256 … 0.003585879 0.0038155357; … ; -0.0008762945 -0.0031371699 … -0.0027350332 -0.0029033197; 0.0011154839 0.002197485 … 0.0021741025 0.0021157824], bias = Float32[0.006656272, 0.004425203, 0.0028994146, -0.0116051175, 0.0031301186, 0.0037318026, 0.0136483535, 0.013969757, -0.015173428, -0.005173992  …  -0.0018621369, -0.0015270555, -0.007873881, -0.0076395273, -0.0022123815, 0.0039605754, 0.0034407252, -0.0045406874, -0.003383829, 0.0029306945]), layer_2 = (layer_1 = (weight = Float32[0.04993449 0.03202845 … -0.059382 0.07701616], bias = Float32[0.08797912]), layer_2 = (weight = Float32[-0.094527975; -0.11476975; … ; -0.016841749; -0.0698748;;], bias = Float32[-0.21608135, -0.26255828, -0.23534852, -0.21524015, -0.055711076, -0.20314303, -0.1895644, 0.03666526, -0.03937737, -0.15905891]))), 0.8455785f0, NamedTuple(), Lux.Training.TrainState{Nothing, Nothing, Chain{@NamedTuple{layer_1::Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Chain{@NamedTuple{layer_1::Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}}, Nothing}, @NamedTuple{layer_1::@NamedTuple{weight::Matrix{Float32}, bias::Vector{Float32}}, layer_2::@NamedTuple{layer_1::@NamedTuple{weight::Matrix{Float32}, bias::Vector{Float32}}, layer_2::@NamedTuple{weight::Matrix{Float32}, bias::Vector{Float32}}}}, @NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{layer_1::@NamedTuple{}, layer_2::@NamedTuple{}}}, Adam{Float32, Tuple{Float64, Float64}, Float64}, @NamedTuple{layer_1::@NamedTuple{weight::Optimisers.Leaf{Adam{Float32, Tuple{Float64, Float64}, Float64}, Tuple{Matrix{Float32}, Matrix{Float32}, Tuple{Float32, Float32}}}, bias::Optimisers.Leaf{Adam{Float32, Tuple{Float64, Float64}, Float64}, Tuple{Vector{Float32}, Vector{Float32}, Tuple{Float32, Float32}}}}, layer_2::@NamedTuple{layer_1::@NamedTuple{weight::Optimisers.Leaf{Adam{Float32, Tuple{Float64, Float64}, Float64}, Tuple{Matrix{Float32}, Matrix{Float32}, Tuple{Float32, Float32}}}, bias::Optimisers.Leaf{Adam{Float32, Tuple{Float64, Float64}, Float64}, Tuple{Vector{Float32}, Vector{Float32}, Tuple{Float32, Float32}}}}, layer_2::@NamedTuple{weight::Optimisers.Leaf{Adam{Float32, Tuple{Float64, Float64}, Float64}, Tuple{Matrix{Float32}, Matrix{Float32}, Tuple{Float32, Float32}}}, bias::Optimisers.Leaf{Adam{Float32, Tuple{Float64, Float64}, Float64}, Tuple{Vector{Float32}, Vector{Float32}, Tuple{Float32, Float32}}}}}}}(nothing, nothing, Chain{@NamedTuple{layer_1::Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Chain{@NamedTuple{layer_1::Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}}, Nothing}((layer_1 = Dense(128 => 256, tanh), layer_2 = Chain{@NamedTuple{layer_1::Dense{typeof(tanh), Int64, Int64, Nothing, Nothing, Static.True}, layer_2::Dense{typeof(identity), Int64, Int64, Nothing, Nothing, Static.True}}, Nothing}((layer_1 = Dense(256 => 1, tanh), layer_2 = Dense(1 => 10)), nothing)), nothing), (layer_1 = (weight = Float32[-0.22542597 0.22379348 … 0.1997513 -0.018708104; -0.023026714 0.15451026 … -0.065325744 0.18120264; … ; 0.038037397 -0.07125516 … -0.03306083 0.039138064; -0.18810266 -0.09693537 … -0.18102062 0.019230088], bias = Float32[0.030937059, -0.060276944, 0.084569596, 0.00040024254, -0.065509446, -0.08527214, -0.026523968, 0.06347208, 0.042247728, 0.027705256  …  -0.06052852, 0.03504307, -0.028244259, 0.06788022, 0.0027464977, -0.06942153, 0.0064240773, 0.0141069945, -0.029283267, 0.01174226]), layer_2 = (layer_1 = (weight = Float32[0.12008221 0.06026435 … -0.070576 0.1577647], bias = Float32[0.026844418]), layer_2 = (weight = Float32[0.5345728; -0.28288874; … ; -0.32983455; -0.45298168;;], bias = Float32[-0.59751064, -0.7033041, -0.8457602, -0.53789175, -0.31473723, 0.17461234, -0.82945836, 0.67841595, 0.35837248, -0.14941788]))), (layer_1 = NamedTuple(), layer_2 = (layer_1 = NamedTuple(), layer_2 = NamedTuple())), Adam(eta=0.0001, beta=(0.9, 0.999), epsilon=1.0e-8), (layer_1 = (weight = Leaf(Adam(eta=0.0001, beta=(0.9, 0.999), epsilon=1.0e-8), (Float32[0.000926728 0.000860063 … 0.00110328 0.000908301; 0.000480834 0.000574605 … 0.000665883 0.000584197; … ; -0.000391039 -0.000438617 … -0.000520651 -0.000449867; 0.00106235 0.000365587 … 0.000813131 0.000495484], Float32[7.20343f-8 4.46976f-8 … 6.84867f-8 4.63952f-8; 1.79691f-8 2.02649f-8 … 2.45046f-8 1.96227f-8; … ; 1.21215f-8 1.17657f-8 … 1.50136f-8 1.15681f-8; 1.12738f-7 7.45199f-9 … 4.8495f-8 1.44173f-8], (0.729, 0.997003))), bias = Leaf(Adam(eta=0.0001, beta=(0.9, 0.999), epsilon=1.0e-8), (Float32[0.00169459, 0.000977637, 0.00103866, -0.00234933, 0.000659175, 0.000868318, 0.00303222, 0.00271383, -0.00326585, -0.0014993  …  -0.000480712, -0.000501535, -0.00174489, -0.00160158, -0.000470662, 0.00127967, 0.000618911, -0.00103705, -0.000773079, 0.00146704], Float32[1.74884f-7, 5.48983f-8, 7.75433f-8, 3.08981f-7, 2.45763f-8, 4.41623f-8, 5.29156f-7, 4.09021f-7, 6.07287f-7, 1.45678f-7  …  1.4164f-8, 1.73391f-8, 1.7507f-7, 1.44894f-7, 1.25673f-8, 1.1198f-7, 2.11545f-8, 6.25338f-8, 3.4755f-8, 1.78565f-7], (0.729, 0.997003)))), layer_2 = (layer_1 = (weight = Leaf(Adam(eta=0.0001, beta=(0.9, 0.999), epsilon=1.0e-8), (Float32[0.00443555 0.00163654 … -0.0124978 0.0123434], Float32[2.53181f-6 1.32838f-6 … 8.83289f-6 8.58873f-6], (0.729, 0.997003))), bias = Leaf(Adam(eta=0.0001, beta=(0.9, 0.999), epsilon=1.0e-8), (Float32[0.0191175], Float32[2.08743f-5], (0.729, 0.997003)))), layer_2 = (weight = Leaf(Adam(eta=0.0001, beta=(0.9, 0.999), epsilon=1.0e-8), (Float32[-0.0172084; -0.0213176; … ; -0.00376332; -0.0116419;;], Float32[1.63537f-5; 2.51152f-5; … ; 8.16783f-7; 7.55419f-6;;], (0.729, 0.997003))), bias = Leaf(Adam(eta=0.0001, beta=(0.9, 0.999), epsilon=1.0e-8), (Float32[-0.0365001, -0.045083, -0.0507623, -0.0390298, -0.0242259, -0.0404982, -0.0358925, 0.0114351, -0.00803444, -0.0248332], Float32[7.40417f-5, 0.000112652, 0.000146818, 8.41229f-5, 4.60234f-5, 9.15105f-5, 7.13093f-5, 8.78741f-6, 3.62043f-6, 3.51285f-5], (0.729, 0.997003)))))), 2))

Defining Custom Layers

We can train our model using the above code, but let's go ahead and see how to use Reactant. Reactant is a julia frontend that generates MLIR and then compiles it using XLA (after running fancy optimizations). It is the current recommended way to train large models in Lux. For more details on using Reactant, see the manual.

julia

using Lux, Random, Optimisers, Reactant, Enzyme
using Printf # For pretty printing

dev = reactant_device()

(::ReactantDevice{Missing, Missing, Missing}) (generic function with 1 method)

We will define a custom MLP using the @compact macro. The macro takes in a list of parameters, layers and states, and a function defining the forward pass of the neural network.

julia

n_in = 1
n_out = 1
nlayers = 3

model = @compact(
    w1=Dense(n_in => 32),
    w2=[Dense(32 => 32) for i in 1:nlayers],
    w3=Dense(32 => n_out),
    act=relu
) do x
    embed = act(w1(x))
    for w in w2
        embed = act(w(embed))
    end
    out = w3(embed)
    @return out
end

@compact(
    w1 = Dense(1 => 32),                # 64 parameters
    w2 = NamedTuple(
        1 = Dense(32 => 32),            # 1_056 parameters
        2 = Dense(32 => 32),            # 1_056 parameters
        3 = Dense(32 => 32),            # 1_056 parameters
    ),
    w3 = Dense(32 => 1),                # 33 parameters
    act = relu,
) do x 
    embed = act(w1(x))
    for w = w2
        embed = act(w(embed))
    end
    out = w3(embed)
    return out
end       # Total: 3_265 parameters,
          #        plus 1 states.

We can initialize the model and train it with the same code as before!

julia

rng = Random.default_rng()
Random.seed!(rng, 0)

ps, st = Lux.setup(rng, model) |> dev

x = rand(rng, Float32, n_in, 32) |> dev

@jit model(x, ps, st)  # 1×32 Matrix and updated state as output.

x_data = reshape(collect(-2.0f0:0.1f0:2.0f0), 1, :)
y_data = 2 .* x_data .- x_data .^ 3
x_data, y_data = dev(x_data), dev(y_data)

function train_model!(model, ps, st, x_data, y_data)
    train_state = Lux.Training.TrainState(model, ps, st, Adam(0.001f0))

    for iter in 1:1000
        _, loss, _, train_state = Lux.Training.single_train_step!(
            AutoEnzyme(), MSELoss(),
            (x_data, y_data), train_state
        )
        if iter % 100 == 1 || iter == 1000
            @printf "Iteration: %04d \t Loss: %10.9g\n" iter loss
        end
    end

    return model, ps, st
end

train_model!(model, ps, st, x_data, y_data)

2025-05-24 14:58:37.359946: I external/xla/xla/service/service.cc:152] XLA service 0x2117b090 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2025-05-24 14:58:37.360064: I external/xla/xla/service/service.cc:160]   StreamExecutor device (0): NVIDIA A100-PCIE-40GB MIG 1g.5gb, Compute Capability 8.0
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1748098717.360897 3535346 se_gpu_pjrt_client.cc:1026] Using BFC allocator.
I0000 00:00:1748098717.360972 3535346 gpu_helpers.cc:136] XLA backend allocating 3825205248 bytes on device 0 for BFCAllocator.
I0000 00:00:1748098717.361001 3535346 gpu_helpers.cc:177] XLA backend will use up to 1275068416 bytes on device 0 for CollectiveBFCAllocator.
I0000 00:00:1748098717.376475 3535346 cuda_dnn.cc:529] Loaded cuDNN version 90400
E0000 00:00:1748099013.349295 3535346 buffer_comparator.cc:145] Difference at 16: -nan, expected 11.328
E0000 00:00:1748099013.349347 3535346 buffer_comparator.cc:145] Difference at 17: -nan, expected 8.55983
E0000 00:00:1748099013.349351 3535346 buffer_comparator.cc:145] Difference at 18: -nan, expected 10.4588
E0000 00:00:1748099013.349354 3535346 buffer_comparator.cc:145] Difference at 19: -nan, expected 8.81169
E0000 00:00:1748099013.349356 3535346 buffer_comparator.cc:145] Difference at 20: -nan, expected 8.98138
E0000 00:00:1748099013.349359 3535346 buffer_comparator.cc:145] Difference at 21: -nan, expected 9.49466
E0000 00:00:1748099013.349362 3535346 buffer_comparator.cc:145] Difference at 22: -nan, expected 8.4604
E0000 00:00:1748099013.349364 3535346 buffer_comparator.cc:145] Difference at 23: -nan, expected 9.78691
E0000 00:00:1748099013.349367 3535346 buffer_comparator.cc:145] Difference at 24: -nan, expected 8.15491
E0000 00:00:1748099013.349370 3535346 buffer_comparator.cc:145] Difference at 25: -nan, expected 13.0125
2025-05-24 15:03:33.349380: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.351538 3535346 buffer_comparator.cc:145] Difference at 16: -nan, expected 11.328
E0000 00:00:1748099013.351550 3535346 buffer_comparator.cc:145] Difference at 17: -nan, expected 8.55983
E0000 00:00:1748099013.351553 3535346 buffer_comparator.cc:145] Difference at 18: -nan, expected 10.4588
E0000 00:00:1748099013.351556 3535346 buffer_comparator.cc:145] Difference at 19: -nan, expected 8.81169
E0000 00:00:1748099013.351559 3535346 buffer_comparator.cc:145] Difference at 20: -nan, expected 8.98138
E0000 00:00:1748099013.351561 3535346 buffer_comparator.cc:145] Difference at 21: -nan, expected 9.49466
E0000 00:00:1748099013.351564 3535346 buffer_comparator.cc:145] Difference at 22: -nan, expected 8.4604
E0000 00:00:1748099013.351567 3535346 buffer_comparator.cc:145] Difference at 23: -nan, expected 9.78691
E0000 00:00:1748099013.351569 3535346 buffer_comparator.cc:145] Difference at 24: -nan, expected 8.15491
E0000 00:00:1748099013.351572 3535346 buffer_comparator.cc:145] Difference at 25: -nan, expected 13.0125
2025-05-24 15:03:33.351576: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.353739 3535346 buffer_comparator.cc:145] Difference at 656: -nan, expected 8.69665
E0000 00:00:1748099013.353750 3535346 buffer_comparator.cc:145] Difference at 657: -nan, expected 7.68202
E0000 00:00:1748099013.353753 3535346 buffer_comparator.cc:145] Difference at 658: -nan, expected 7.88703
E0000 00:00:1748099013.353755 3535346 buffer_comparator.cc:145] Difference at 659: -nan, expected 7.16689
E0000 00:00:1748099013.353758 3535346 buffer_comparator.cc:145] Difference at 660: -nan, expected 6.63868
E0000 00:00:1748099013.353761 3535346 buffer_comparator.cc:145] Difference at 661: -nan, expected 8.39542
E0000 00:00:1748099013.353763 3535346 buffer_comparator.cc:145] Difference at 662: -nan, expected 7.00635
E0000 00:00:1748099013.353766 3535346 buffer_comparator.cc:145] Difference at 663: -nan, expected 7.06674
E0000 00:00:1748099013.353768 3535346 buffer_comparator.cc:145] Difference at 664: -nan, expected 6.11613
E0000 00:00:1748099013.353771 3535346 buffer_comparator.cc:145] Difference at 665: -nan, expected 8.63651
2025-05-24 15:03:33.353775: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.355925 3535346 buffer_comparator.cc:145] Difference at 672: -nan, expected 8.61244
E0000 00:00:1748099013.355936 3535346 buffer_comparator.cc:145] Difference at 673: -nan, expected 6.1493
E0000 00:00:1748099013.355939 3535346 buffer_comparator.cc:145] Difference at 674: -nan, expected 8.90756
E0000 00:00:1748099013.355942 3535346 buffer_comparator.cc:145] Difference at 675: -nan, expected 7.1184
E0000 00:00:1748099013.355946 3535346 buffer_comparator.cc:145] Difference at 676: -nan, expected 8.03527
E0000 00:00:1748099013.355949 3535346 buffer_comparator.cc:145] Difference at 677: -nan, expected 7.44864
E0000 00:00:1748099013.355952 3535346 buffer_comparator.cc:145] Difference at 678: -nan, expected 7.35203
E0000 00:00:1748099013.355955 3535346 buffer_comparator.cc:145] Difference at 679: -nan, expected 7.89603
E0000 00:00:1748099013.355959 3535346 buffer_comparator.cc:145] Difference at 680: -nan, expected 7.3266
E0000 00:00:1748099013.355961 3535346 buffer_comparator.cc:145] Difference at 681: -nan, expected 9.7807
2025-05-24 15:03:33.355966: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.358111 3535346 buffer_comparator.cc:145] Difference at 672: -nan, expected 8.61244
E0000 00:00:1748099013.358122 3535346 buffer_comparator.cc:145] Difference at 673: -nan, expected 6.1493
E0000 00:00:1748099013.358125 3535346 buffer_comparator.cc:145] Difference at 674: -nan, expected 8.90756
E0000 00:00:1748099013.358127 3535346 buffer_comparator.cc:145] Difference at 675: -nan, expected 7.1184
E0000 00:00:1748099013.358130 3535346 buffer_comparator.cc:145] Difference at 676: -nan, expected 8.03527
E0000 00:00:1748099013.358133 3535346 buffer_comparator.cc:145] Difference at 677: -nan, expected 7.44864
E0000 00:00:1748099013.358135 3535346 buffer_comparator.cc:145] Difference at 678: -nan, expected 7.35203
E0000 00:00:1748099013.358138 3535346 buffer_comparator.cc:145] Difference at 679: -nan, expected 7.89603
E0000 00:00:1748099013.358140 3535346 buffer_comparator.cc:145] Difference at 680: -nan, expected 7.3266
E0000 00:00:1748099013.358143 3535346 buffer_comparator.cc:145] Difference at 681: -nan, expected 9.7807
2025-05-24 15:03:33.358147: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.360334 3535346 buffer_comparator.cc:145] Difference at 688: -nan, expected 7.86868
E0000 00:00:1748099013.360345 3535346 buffer_comparator.cc:145] Difference at 689: -nan, expected 7.33715
E0000 00:00:1748099013.360347 3535346 buffer_comparator.cc:145] Difference at 690: -nan, expected 6.05665
E0000 00:00:1748099013.360350 3535346 buffer_comparator.cc:145] Difference at 691: -nan, expected 7.16547
E0000 00:00:1748099013.360353 3535346 buffer_comparator.cc:145] Difference at 692: -nan, expected 8.27916
E0000 00:00:1748099013.360355 3535346 buffer_comparator.cc:145] Difference at 693: -nan, expected 5.80258
E0000 00:00:1748099013.360358 3535346 buffer_comparator.cc:145] Difference at 694: -nan, expected 6.06621
E0000 00:00:1748099013.360361 3535346 buffer_comparator.cc:145] Difference at 695: -nan, expected 7.00273
E0000 00:00:1748099013.360363 3535346 buffer_comparator.cc:145] Difference at 696: -nan, expected 7.92525
E0000 00:00:1748099013.360366 3535346 buffer_comparator.cc:145] Difference at 729: -nan, expected 7.66068
2025-05-24 15:03:33.360370: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.362518 3535346 buffer_comparator.cc:145] Difference at 688: -nan, expected 7.86868
E0000 00:00:1748099013.362528 3535346 buffer_comparator.cc:145] Difference at 689: -nan, expected 7.33715
E0000 00:00:1748099013.362531 3535346 buffer_comparator.cc:145] Difference at 690: -nan, expected 6.05665
E0000 00:00:1748099013.362534 3535346 buffer_comparator.cc:145] Difference at 691: -nan, expected 7.16547
E0000 00:00:1748099013.362536 3535346 buffer_comparator.cc:145] Difference at 692: -nan, expected 8.27916
E0000 00:00:1748099013.362539 3535346 buffer_comparator.cc:145] Difference at 693: -nan, expected 5.80258
E0000 00:00:1748099013.362542 3535346 buffer_comparator.cc:145] Difference at 694: -nan, expected 6.06621
E0000 00:00:1748099013.362545 3535346 buffer_comparator.cc:145] Difference at 695: -nan, expected 7.00273
E0000 00:00:1748099013.362548 3535346 buffer_comparator.cc:145] Difference at 696: -nan, expected 7.92525
E0000 00:00:1748099013.362551 3535346 buffer_comparator.cc:145] Difference at 729: -nan, expected 7.66068
2025-05-24 15:03:33.362555: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.374933 3535346 buffer_comparator.cc:145] Difference at 16: 0, expected 18.4532
E0000 00:00:1748099013.374945 3535346 buffer_comparator.cc:145] Difference at 17: 0, expected 16.1701
E0000 00:00:1748099013.374948 3535346 buffer_comparator.cc:145] Difference at 18: 0, expected 18.5372
E0000 00:00:1748099013.374951 3535346 buffer_comparator.cc:145] Difference at 19: 0, expected 17.7684
E0000 00:00:1748099013.374954 3535346 buffer_comparator.cc:145] Difference at 20: 0, expected 17.8078
E0000 00:00:1748099013.374957 3535346 buffer_comparator.cc:145] Difference at 21: 0, expected 17.412
E0000 00:00:1748099013.374960 3535346 buffer_comparator.cc:145] Difference at 22: 0, expected 18.0425
E0000 00:00:1748099013.374963 3535346 buffer_comparator.cc:145] Difference at 23: 0, expected 17.7822
E0000 00:00:1748099013.374966 3535346 buffer_comparator.cc:145] Difference at 24: 0, expected 16.8692
E0000 00:00:1748099013.374969 3535346 buffer_comparator.cc:145] Difference at 25: 0, expected 19.6248
2025-05-24 15:03:33.374974: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.377934 3535346 buffer_comparator.cc:145] Difference at 16: 0, expected 18.4532
E0000 00:00:1748099013.377945 3535346 buffer_comparator.cc:145] Difference at 17: 0, expected 16.1701
E0000 00:00:1748099013.377948 3535346 buffer_comparator.cc:145] Difference at 18: 0, expected 18.5372
E0000 00:00:1748099013.377951 3535346 buffer_comparator.cc:145] Difference at 19: 0, expected 17.7684
E0000 00:00:1748099013.377954 3535346 buffer_comparator.cc:145] Difference at 20: 0, expected 17.8078
E0000 00:00:1748099013.377957 3535346 buffer_comparator.cc:145] Difference at 21: 0, expected 17.412
E0000 00:00:1748099013.377960 3535346 buffer_comparator.cc:145] Difference at 22: 0, expected 18.0425
E0000 00:00:1748099013.377963 3535346 buffer_comparator.cc:145] Difference at 23: 0, expected 17.7822
E0000 00:00:1748099013.377965 3535346 buffer_comparator.cc:145] Difference at 24: 0, expected 16.8692
E0000 00:00:1748099013.377968 3535346 buffer_comparator.cc:145] Difference at 25: 0, expected 19.6248
2025-05-24 15:03:33.377973: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.380948 3535346 buffer_comparator.cc:145] Difference at 656: 0, expected 15.8892
E0000 00:00:1748099013.380959 3535346 buffer_comparator.cc:145] Difference at 657: 0, expected 15.1292
E0000 00:00:1748099013.380962 3535346 buffer_comparator.cc:145] Difference at 658: 0, expected 14.0499
E0000 00:00:1748099013.380965 3535346 buffer_comparator.cc:145] Difference at 659: 0, expected 13.8377
E0000 00:00:1748099013.380968 3535346 buffer_comparator.cc:145] Difference at 660: 0, expected 13.7353
E0000 00:00:1748099013.380971 3535346 buffer_comparator.cc:145] Difference at 661: 0, expected 15.7468
E0000 00:00:1748099013.380974 3535346 buffer_comparator.cc:145] Difference at 662: 0, expected 14.9101
E0000 00:00:1748099013.380977 3535346 buffer_comparator.cc:145] Difference at 663: 0, expected 14.8135
E0000 00:00:1748099013.380980 3535346 buffer_comparator.cc:145] Difference at 664: 0, expected 13.6403
E0000 00:00:1748099013.380982 3535346 buffer_comparator.cc:145] Difference at 665: 0, expected 15.8348
2025-05-24 15:03:33.380987: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.383950 3535346 buffer_comparator.cc:145] Difference at 672: 0, expected 16.0696
E0000 00:00:1748099013.383961 3535346 buffer_comparator.cc:145] Difference at 673: 0, expected 14.3019
E0000 00:00:1748099013.383964 3535346 buffer_comparator.cc:145] Difference at 674: 0, expected 15.5573
E0000 00:00:1748099013.383967 3535346 buffer_comparator.cc:145] Difference at 675: 0, expected 14.6242
E0000 00:00:1748099013.383970 3535346 buffer_comparator.cc:145] Difference at 676: 0, expected 14.8486
E0000 00:00:1748099013.383973 3535346 buffer_comparator.cc:145] Difference at 677: 0, expected 14.7699
E0000 00:00:1748099013.383976 3535346 buffer_comparator.cc:145] Difference at 678: 0, expected 15.1617
E0000 00:00:1748099013.383978 3535346 buffer_comparator.cc:145] Difference at 679: 0, expected 14.9394
E0000 00:00:1748099013.383981 3535346 buffer_comparator.cc:145] Difference at 680: 0, expected 13.4678
E0000 00:00:1748099013.383984 3535346 buffer_comparator.cc:145] Difference at 681: 0, expected 16.1851
2025-05-24 15:03:33.383989: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.386957 3535346 buffer_comparator.cc:145] Difference at 672: 0, expected 16.0696
E0000 00:00:1748099013.386969 3535346 buffer_comparator.cc:145] Difference at 673: 0, expected 14.3019
E0000 00:00:1748099013.386972 3535346 buffer_comparator.cc:145] Difference at 674: 0, expected 15.5573
E0000 00:00:1748099013.386975 3535346 buffer_comparator.cc:145] Difference at 675: 0, expected 14.6242
E0000 00:00:1748099013.386978 3535346 buffer_comparator.cc:145] Difference at 676: 0, expected 14.8486
E0000 00:00:1748099013.386981 3535346 buffer_comparator.cc:145] Difference at 677: 0, expected 14.7699
E0000 00:00:1748099013.386983 3535346 buffer_comparator.cc:145] Difference at 678: 0, expected 15.1617
E0000 00:00:1748099013.386986 3535346 buffer_comparator.cc:145] Difference at 679: 0, expected 14.9394
E0000 00:00:1748099013.386989 3535346 buffer_comparator.cc:145] Difference at 680: 0, expected 13.4678
E0000 00:00:1748099013.386992 3535346 buffer_comparator.cc:145] Difference at 681: 0, expected 16.1851
2025-05-24 15:03:33.386997: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.389958 3535346 buffer_comparator.cc:145] Difference at 688: 0, expected 15.1187
E0000 00:00:1748099013.389969 3535346 buffer_comparator.cc:145] Difference at 689: 0, expected 14.6251
E0000 00:00:1748099013.389972 3535346 buffer_comparator.cc:145] Difference at 690: 0, expected 14.2005
E0000 00:00:1748099013.389975 3535346 buffer_comparator.cc:145] Difference at 691: 0, expected 15.1561
E0000 00:00:1748099013.389978 3535346 buffer_comparator.cc:145] Difference at 692: 0, expected 15.4235
E0000 00:00:1748099013.389981 3535346 buffer_comparator.cc:145] Difference at 693: 0, expected 14.1331
E0000 00:00:1748099013.389984 3535346 buffer_comparator.cc:145] Difference at 694: 0, expected 14.4063
E0000 00:00:1748099013.389986 3535346 buffer_comparator.cc:145] Difference at 695: 0, expected 14.0259
E0000 00:00:1748099013.389989 3535346 buffer_comparator.cc:145] Difference at 696: 0, expected 15.0279
E0000 00:00:1748099013.389992 3535346 buffer_comparator.cc:145] Difference at 729: 0, expected 14.5946
2025-05-24 15:03:33.389997: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.392961 3535346 buffer_comparator.cc:145] Difference at 688: 0, expected 15.1187
E0000 00:00:1748099013.392972 3535346 buffer_comparator.cc:145] Difference at 689: 0, expected 14.6251
E0000 00:00:1748099013.392975 3535346 buffer_comparator.cc:145] Difference at 690: 0, expected 14.2005
E0000 00:00:1748099013.392979 3535346 buffer_comparator.cc:145] Difference at 691: 0, expected 15.1561
E0000 00:00:1748099013.392982 3535346 buffer_comparator.cc:145] Difference at 692: 0, expected 15.4235
E0000 00:00:1748099013.392985 3535346 buffer_comparator.cc:145] Difference at 693: 0, expected 14.1331
E0000 00:00:1748099013.392988 3535346 buffer_comparator.cc:145] Difference at 694: 0, expected 14.4063
E0000 00:00:1748099013.392991 3535346 buffer_comparator.cc:145] Difference at 695: 0, expected 14.0259
E0000 00:00:1748099013.392994 3535346 buffer_comparator.cc:145] Difference at 696: 0, expected 15.0279
E0000 00:00:1748099013.392997 3535346 buffer_comparator.cc:145] Difference at 729: 0, expected 14.5946
2025-05-24 15:03:33.393002: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.408591 3535346 buffer_comparator.cc:145] Difference at 16: -nan, expected 29.4863
E0000 00:00:1748099013.408603 3535346 buffer_comparator.cc:145] Difference at 17: -nan, expected 25.4275
E0000 00:00:1748099013.408606 3535346 buffer_comparator.cc:145] Difference at 18: -nan, expected 29.498
E0000 00:00:1748099013.408609 3535346 buffer_comparator.cc:145] Difference at 19: -nan, expected 24.9024
E0000 00:00:1748099013.408612 3535346 buffer_comparator.cc:145] Difference at 20: -nan, expected 31.8883
E0000 00:00:1748099013.408614 3535346 buffer_comparator.cc:145] Difference at 21: -nan, expected 30.5795
E0000 00:00:1748099013.408617 3535346 buffer_comparator.cc:145] Difference at 22: -nan, expected 26.1755
E0000 00:00:1748099013.408620 3535346 buffer_comparator.cc:145] Difference at 23: -nan, expected 30.0282
E0000 00:00:1748099013.408623 3535346 buffer_comparator.cc:145] Difference at 24: -nan, expected 25.7237
E0000 00:00:1748099013.408631 3535346 buffer_comparator.cc:145] Difference at 25: -nan, expected 25.7191
2025-05-24 15:03:33.408636: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.411592 3535346 buffer_comparator.cc:145] Difference at 16: -nan, expected 29.4863
E0000 00:00:1748099013.411603 3535346 buffer_comparator.cc:145] Difference at 17: -nan, expected 25.4275
E0000 00:00:1748099013.411606 3535346 buffer_comparator.cc:145] Difference at 18: -nan, expected 29.498
E0000 00:00:1748099013.411609 3535346 buffer_comparator.cc:145] Difference at 19: -nan, expected 24.9024
E0000 00:00:1748099013.411612 3535346 buffer_comparator.cc:145] Difference at 20: -nan, expected 31.8883
E0000 00:00:1748099013.411615 3535346 buffer_comparator.cc:145] Difference at 21: -nan, expected 30.5795
E0000 00:00:1748099013.411617 3535346 buffer_comparator.cc:145] Difference at 22: -nan, expected 26.1755
E0000 00:00:1748099013.411620 3535346 buffer_comparator.cc:145] Difference at 23: -nan, expected 30.0282
E0000 00:00:1748099013.411623 3535346 buffer_comparator.cc:145] Difference at 24: -nan, expected 25.7237
E0000 00:00:1748099013.411629 3535346 buffer_comparator.cc:145] Difference at 25: -nan, expected 25.7191
2025-05-24 15:03:33.411634: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.414591 3535346 buffer_comparator.cc:145] Difference at 512: -nan, expected 13.9275
E0000 00:00:1748099013.414602 3535346 buffer_comparator.cc:145] Difference at 513: -nan, expected 12.9447
E0000 00:00:1748099013.414605 3535346 buffer_comparator.cc:145] Difference at 514: -nan, expected 13.899
E0000 00:00:1748099013.414607 3535346 buffer_comparator.cc:145] Difference at 515: -nan, expected 14.1578
E0000 00:00:1748099013.414610 3535346 buffer_comparator.cc:145] Difference at 516: -nan, expected 15.4892
E0000 00:00:1748099013.414613 3535346 buffer_comparator.cc:145] Difference at 517: -nan, expected 16.545
E0000 00:00:1748099013.414617 3535346 buffer_comparator.cc:145] Difference at 518: -nan, expected 17.8581
E0000 00:00:1748099013.414620 3535346 buffer_comparator.cc:145] Difference at 519: -nan, expected 13.0536
E0000 00:00:1748099013.414623 3535346 buffer_comparator.cc:145] Difference at 520: -nan, expected 16.1329
E0000 00:00:1748099013.414631 3535346 buffer_comparator.cc:145] Difference at 521: -nan, expected 14.5245
2025-05-24 15:03:33.414635: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.417630 3535346 buffer_comparator.cc:145] Difference at 528: -nan, expected 17.5032
E0000 00:00:1748099013.417641 3535346 buffer_comparator.cc:145] Difference at 529: -nan, expected 15.1785
E0000 00:00:1748099013.417644 3535346 buffer_comparator.cc:145] Difference at 530: -nan, expected 15.9473
E0000 00:00:1748099013.417647 3535346 buffer_comparator.cc:145] Difference at 531: -nan, expected 14.437
E0000 00:00:1748099013.417650 3535346 buffer_comparator.cc:145] Difference at 532: -nan, expected 17.9637
E0000 00:00:1748099013.417652 3535346 buffer_comparator.cc:145] Difference at 533: -nan, expected 17.3157
E0000 00:00:1748099013.417655 3535346 buffer_comparator.cc:145] Difference at 534: -nan, expected 15.7802
E0000 00:00:1748099013.417658 3535346 buffer_comparator.cc:145] Difference at 535: -nan, expected 17.6887
E0000 00:00:1748099013.417661 3535346 buffer_comparator.cc:145] Difference at 536: -nan, expected 15.1881
E0000 00:00:1748099013.417663 3535346 buffer_comparator.cc:145] Difference at 537: -nan, expected 14.4224
2025-05-24 15:03:33.417668: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.420642 3535346 buffer_comparator.cc:145] Difference at 528: -nan, expected 17.5032
E0000 00:00:1748099013.420653 3535346 buffer_comparator.cc:145] Difference at 529: -nan, expected 15.1785
E0000 00:00:1748099013.420656 3535346 buffer_comparator.cc:145] Difference at 530: -nan, expected 15.9473
E0000 00:00:1748099013.420659 3535346 buffer_comparator.cc:145] Difference at 531: -nan, expected 14.437
E0000 00:00:1748099013.420662 3535346 buffer_comparator.cc:145] Difference at 532: -nan, expected 17.9637
E0000 00:00:1748099013.420665 3535346 buffer_comparator.cc:145] Difference at 533: -nan, expected 17.3157
E0000 00:00:1748099013.420667 3535346 buffer_comparator.cc:145] Difference at 534: -nan, expected 15.7802
E0000 00:00:1748099013.420670 3535346 buffer_comparator.cc:145] Difference at 535: -nan, expected 17.6887
E0000 00:00:1748099013.420673 3535346 buffer_comparator.cc:145] Difference at 536: -nan, expected 15.1881
E0000 00:00:1748099013.420676 3535346 buffer_comparator.cc:145] Difference at 537: -nan, expected 14.4224
2025-05-24 15:03:33.420680: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1748099013.423650 3535346 buffer_comparator.cc:145] Difference at 528: -nan, expected 17.5032
E0000 00:00:1748099013.423660 3535346 buffer_comparator.cc:145] Difference at 529: -nan, expected 15.1785
E0000 00:00:1748099013.423663 3535346 buffer_comparator.cc:145] Difference at 530: -nan, expected 15.9473
E0000 00:00:1748099013.423666 3535346 buffer_comparator.cc:145] Difference at 531: -nan, expected 14.437
E0000 00:00:1748099013.423669 3535346 buffer_comparator.cc:145] Difference at 532: -nan, expected 17.9637
E0000 00:00:1748099013.423672 3535346 buffer_comparator.cc:145] Difference at 533: -nan, expected 17.3157
E0000 00:00:1748099013.423674 3535346 buffer_comparator.cc:145] Difference at 534: -nan, expected 15.7802
E0000 00:00:1748099013.423677 3535346 buffer_comparator.cc:145] Difference at 535: -nan, expected 17.6887
E0000 00:00:1748099013.423680 3535346 buffer_comparator.cc:145] Difference at 536: -nan, expected 15.1881
E0000 00:00:1748099013.423684 3535346 buffer_comparator.cc:145] Difference at 537: -nan, expected 14.4224
2025-05-24 15:03:33.423688: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1179] Results do not match the reference. This is likely a bug/unexpected loss of precision.
Iteration: 0001 	 Loss: 2.08073235
Iteration: 0101 	 Loss: 0.142443746
Iteration: 0201 	 Loss: 0.00498719513
Iteration: 0301 	 Loss: 0.00116170431
Iteration: 0401 	 Loss: 0.000497820089
Iteration: 0501 	 Loss: 0.00027494994
Iteration: 0601 	 Loss: 0.000180605697
Iteration: 0701 	 Loss: 0.000201430521
Iteration: 0801 	 Loss: 0.000379697973
Iteration: 0901 	 Loss: 0.000106757798
Iteration: 1000 	 Loss: 0.000231371887

Training with Optimization.jl

If you are coming from the SciML ecosystem and want to use Optimization.jl, please refer to the Optimization.jl Tutorial.

Additional Packages

LuxDL hosts various packages that provide additional functionality for Lux.jl. All packages mentioned in this documentation are available via the Julia General Registry.

You can install all those packages via import Pkg; Pkg.add(<package name>).

XLA (CPU/GPU/TPU) Support

Lux.jl supports XLA compilation for CPU, GPU, and TPU using Reactant.jl.

GPU Support

GPU Support for Lux.jl requires loading additional packages:

LuxCUDA.jl for CUDA support.
AMDGPU.jl for AMDGPU support.
Metal.jl for Apple Metal support.
oneAPI.jl for oneAPI support.

Trusted by

Getting Started ​

Installation ​

Quickstart ​

Defining Custom Layers ​

Additional Packages ​

XLA (CPU/GPU/TPU) Support ​

GPU Support ​

Getting Started

Installation

Quickstart

Defining Custom Layers

Additional Packages

XLA (CPU/GPU/TPU) Support

GPU Support