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 supportWe 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)2025-03-28 05:09:38.177082: I external/xla/xla/service/service.cc:152] XLA service 0x60272e0 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2025-03-28 05:09:38.177452: 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 00:00:1743138578.178367 3709779 se_gpu_pjrt_client.cc:1039] Using BFC allocator.
I0000 00:00:1743138578.178467 3709779 gpu_helpers.cc:136] XLA backend allocating 3825205248 bytes on device 0 for BFCAllocator.
I0000 00:00:1743138578.178507 3709779 gpu_helpers.cc:177] XLA backend will use up to 1275068416 bytes on device 0 for CollectiveBFCAllocator.
I0000 00:00:1743138578.194997 3709779 cuda_dnn.cc:529] Loaded cuDNN version 90400
E0000 00:00:1743138802.649397 3709779 buffer_comparator.cc:156] Difference at 16: -nan, expected 11.328
E0000 00:00:1743138802.649457 3709779 buffer_comparator.cc:156] Difference at 17: -nan, expected 8.55983
E0000 00:00:1743138802.649467 3709779 buffer_comparator.cc:156] Difference at 18: -nan, expected 10.4588
E0000 00:00:1743138802.649474 3709779 buffer_comparator.cc:156] Difference at 19: -nan, expected 8.81169
E0000 00:00:1743138802.649481 3709779 buffer_comparator.cc:156] Difference at 20: -nan, expected 8.98138
E0000 00:00:1743138802.649487 3709779 buffer_comparator.cc:156] Difference at 21: -nan, expected 9.49466
E0000 00:00:1743138802.649493 3709779 buffer_comparator.cc:156] Difference at 22: -nan, expected 8.4604
E0000 00:00:1743138802.649499 3709779 buffer_comparator.cc:156] Difference at 23: -nan, expected 9.78691
E0000 00:00:1743138802.649505 3709779 buffer_comparator.cc:156] Difference at 24: -nan, expected 8.15491
E0000 00:00:1743138802.649512 3709779 buffer_comparator.cc:156] Difference at 25: -nan, expected 13.0125
2025-03-28 05:13:22.649526: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.651898 3709779 buffer_comparator.cc:156] Difference at 16: -nan, expected 11.328
E0000 00:00:1743138802.651915 3709779 buffer_comparator.cc:156] Difference at 17: -nan, expected 8.55983
E0000 00:00:1743138802.651920 3709779 buffer_comparator.cc:156] Difference at 18: -nan, expected 10.4588
E0000 00:00:1743138802.651925 3709779 buffer_comparator.cc:156] Difference at 19: -nan, expected 8.81169
E0000 00:00:1743138802.651929 3709779 buffer_comparator.cc:156] Difference at 20: -nan, expected 8.98138
E0000 00:00:1743138802.651933 3709779 buffer_comparator.cc:156] Difference at 21: -nan, expected 9.49466
E0000 00:00:1743138802.651937 3709779 buffer_comparator.cc:156] Difference at 22: -nan, expected 8.4604
E0000 00:00:1743138802.651941 3709779 buffer_comparator.cc:156] Difference at 23: -nan, expected 9.78691
E0000 00:00:1743138802.651945 3709779 buffer_comparator.cc:156] Difference at 24: -nan, expected 8.15491
E0000 00:00:1743138802.651949 3709779 buffer_comparator.cc:156] Difference at 25: -nan, expected 13.0125
2025-03-28 05:13:22.651956: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.654149 3709779 buffer_comparator.cc:156] Difference at 656: -nan, expected 8.69665
E0000 00:00:1743138802.654166 3709779 buffer_comparator.cc:156] Difference at 657: -nan, expected 7.68202
E0000 00:00:1743138802.654171 3709779 buffer_comparator.cc:156] Difference at 658: -nan, expected 7.88703
E0000 00:00:1743138802.654175 3709779 buffer_comparator.cc:156] Difference at 659: -nan, expected 7.16689
E0000 00:00:1743138802.654179 3709779 buffer_comparator.cc:156] Difference at 660: -nan, expected 6.63868
E0000 00:00:1743138802.654183 3709779 buffer_comparator.cc:156] Difference at 661: -nan, expected 8.39542
E0000 00:00:1743138802.654188 3709779 buffer_comparator.cc:156] Difference at 662: -nan, expected 7.00635
E0000 00:00:1743138802.654192 3709779 buffer_comparator.cc:156] Difference at 663: -nan, expected 7.06674
E0000 00:00:1743138802.654196 3709779 buffer_comparator.cc:156] Difference at 664: -nan, expected 6.11613
E0000 00:00:1743138802.654200 3709779 buffer_comparator.cc:156] Difference at 665: -nan, expected 8.63651
2025-03-28 05:13:22.654206: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.656372 3709779 buffer_comparator.cc:156] Difference at 672: -nan, expected 8.61244
E0000 00:00:1743138802.656388 3709779 buffer_comparator.cc:156] Difference at 673: -nan, expected 6.1493
E0000 00:00:1743138802.656393 3709779 buffer_comparator.cc:156] Difference at 674: -nan, expected 8.90756
E0000 00:00:1743138802.656398 3709779 buffer_comparator.cc:156] Difference at 675: -nan, expected 7.1184
E0000 00:00:1743138802.656404 3709779 buffer_comparator.cc:156] Difference at 676: -nan, expected 8.03527
E0000 00:00:1743138802.656408 3709779 buffer_comparator.cc:156] Difference at 677: -nan, expected 7.44864
E0000 00:00:1743138802.656412 3709779 buffer_comparator.cc:156] Difference at 678: -nan, expected 7.35203
E0000 00:00:1743138802.656416 3709779 buffer_comparator.cc:156] Difference at 679: -nan, expected 7.89603
E0000 00:00:1743138802.656420 3709779 buffer_comparator.cc:156] Difference at 680: -nan, expected 7.3266
E0000 00:00:1743138802.656424 3709779 buffer_comparator.cc:156] Difference at 681: -nan, expected 9.7807
2025-03-28 05:13:22.656431: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.658586 3709779 buffer_comparator.cc:156] Difference at 688: -nan, expected 7.86868
E0000 00:00:1743138802.658602 3709779 buffer_comparator.cc:156] Difference at 689: -nan, expected 7.33715
E0000 00:00:1743138802.658607 3709779 buffer_comparator.cc:156] Difference at 690: -nan, expected 6.05665
E0000 00:00:1743138802.658611 3709779 buffer_comparator.cc:156] Difference at 691: -nan, expected 7.16547
E0000 00:00:1743138802.658615 3709779 buffer_comparator.cc:156] Difference at 692: -nan, expected 8.27916
E0000 00:00:1743138802.658619 3709779 buffer_comparator.cc:156] Difference at 693: -nan, expected 5.80258
E0000 00:00:1743138802.658623 3709779 buffer_comparator.cc:156] Difference at 694: -nan, expected 6.06621
E0000 00:00:1743138802.658627 3709779 buffer_comparator.cc:156] Difference at 695: -nan, expected 7.00273
E0000 00:00:1743138802.658631 3709779 buffer_comparator.cc:156] Difference at 696: -nan, expected 7.92525
E0000 00:00:1743138802.658635 3709779 buffer_comparator.cc:156] Difference at 729: -nan, expected 7.66068
2025-03-28 05:13:22.658641: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.660794 3709779 buffer_comparator.cc:156] Difference at 688: -nan, expected 7.86868
E0000 00:00:1743138802.660809 3709779 buffer_comparator.cc:156] Difference at 689: -nan, expected 7.33715
E0000 00:00:1743138802.660814 3709779 buffer_comparator.cc:156] Difference at 690: -nan, expected 6.05665
E0000 00:00:1743138802.660818 3709779 buffer_comparator.cc:156] Difference at 691: -nan, expected 7.16547
E0000 00:00:1743138802.660822 3709779 buffer_comparator.cc:156] Difference at 692: -nan, expected 8.27916
E0000 00:00:1743138802.660826 3709779 buffer_comparator.cc:156] Difference at 693: -nan, expected 5.80258
E0000 00:00:1743138802.660830 3709779 buffer_comparator.cc:156] Difference at 694: -nan, expected 6.06621
E0000 00:00:1743138802.660834 3709779 buffer_comparator.cc:156] Difference at 695: -nan, expected 7.00273
E0000 00:00:1743138802.660838 3709779 buffer_comparator.cc:156] Difference at 696: -nan, expected 7.92525
E0000 00:00:1743138802.660843 3709779 buffer_comparator.cc:156] Difference at 729: -nan, expected 7.66068
2025-03-28 05:13:22.660849: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.662984 3709779 buffer_comparator.cc:156] Difference at 688: -nan, expected 7.86868
E0000 00:00:1743138802.662995 3709779 buffer_comparator.cc:156] Difference at 689: -nan, expected 7.33715
E0000 00:00:1743138802.662998 3709779 buffer_comparator.cc:156] Difference at 690: -nan, expected 6.05665
E0000 00:00:1743138802.663001 3709779 buffer_comparator.cc:156] Difference at 691: -nan, expected 7.16547
E0000 00:00:1743138802.663004 3709779 buffer_comparator.cc:156] Difference at 692: -nan, expected 8.27916
E0000 00:00:1743138802.663007 3709779 buffer_comparator.cc:156] Difference at 693: -nan, expected 5.80258
E0000 00:00:1743138802.663009 3709779 buffer_comparator.cc:156] Difference at 694: -nan, expected 6.06621
E0000 00:00:1743138802.663014 3709779 buffer_comparator.cc:156] Difference at 695: -nan, expected 7.00273
E0000 00:00:1743138802.663017 3709779 buffer_comparator.cc:156] Difference at 696: -nan, expected 7.92525
E0000 00:00:1743138802.663019 3709779 buffer_comparator.cc:156] Difference at 729: -nan, expected 7.66068
2025-03-28 05:13:22.663024: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.673152 3709779 buffer_comparator.cc:156] Difference at 16: -nan, expected 29.4863
E0000 00:00:1743138802.673164 3709779 buffer_comparator.cc:156] Difference at 17: -nan, expected 25.4275
E0000 00:00:1743138802.673168 3709779 buffer_comparator.cc:156] Difference at 18: -nan, expected 29.498
E0000 00:00:1743138802.673171 3709779 buffer_comparator.cc:156] Difference at 19: -nan, expected 24.9024
E0000 00:00:1743138802.673174 3709779 buffer_comparator.cc:156] Difference at 20: -nan, expected 31.8883
E0000 00:00:1743138802.673176 3709779 buffer_comparator.cc:156] Difference at 21: -nan, expected 30.5795
E0000 00:00:1743138802.673179 3709779 buffer_comparator.cc:156] Difference at 22: -nan, expected 26.1755
E0000 00:00:1743138802.673182 3709779 buffer_comparator.cc:156] Difference at 23: -nan, expected 30.0282
E0000 00:00:1743138802.673185 3709779 buffer_comparator.cc:156] Difference at 24: -nan, expected 25.7237
E0000 00:00:1743138802.673188 3709779 buffer_comparator.cc:156] Difference at 25: -nan, expected 25.7191
2025-03-28 05:13:22.673193: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.676017 3709779 buffer_comparator.cc:156] Difference at 16: -nan, expected 29.4863
E0000 00:00:1743138802.676028 3709779 buffer_comparator.cc:156] Difference at 17: -nan, expected 25.4275
E0000 00:00:1743138802.676031 3709779 buffer_comparator.cc:156] Difference at 18: -nan, expected 29.498
E0000 00:00:1743138802.676034 3709779 buffer_comparator.cc:156] Difference at 19: -nan, expected 24.9024
E0000 00:00:1743138802.676037 3709779 buffer_comparator.cc:156] Difference at 20: -nan, expected 31.8883
E0000 00:00:1743138802.676040 3709779 buffer_comparator.cc:156] Difference at 21: -nan, expected 30.5795
E0000 00:00:1743138802.676043 3709779 buffer_comparator.cc:156] Difference at 22: -nan, expected 26.1755
E0000 00:00:1743138802.676046 3709779 buffer_comparator.cc:156] Difference at 23: -nan, expected 30.0282
E0000 00:00:1743138802.676048 3709779 buffer_comparator.cc:156] Difference at 24: -nan, expected 25.7237
E0000 00:00:1743138802.676051 3709779 buffer_comparator.cc:156] Difference at 25: -nan, expected 25.7191
2025-03-28 05:13:22.676056: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.678883 3709779 buffer_comparator.cc:156] Difference at 512: -nan, expected 13.9275
E0000 00:00:1743138802.678894 3709779 buffer_comparator.cc:156] Difference at 513: -nan, expected 12.9447
E0000 00:00:1743138802.678898 3709779 buffer_comparator.cc:156] Difference at 514: -nan, expected 13.899
E0000 00:00:1743138802.678901 3709779 buffer_comparator.cc:156] Difference at 515: -nan, expected 14.1578
E0000 00:00:1743138802.678903 3709779 buffer_comparator.cc:156] Difference at 516: -nan, expected 15.4892
E0000 00:00:1743138802.678906 3709779 buffer_comparator.cc:156] Difference at 517: -nan, expected 16.545
E0000 00:00:1743138802.678909 3709779 buffer_comparator.cc:156] Difference at 518: -nan, expected 17.8581
E0000 00:00:1743138802.678912 3709779 buffer_comparator.cc:156] Difference at 519: -nan, expected 13.0536
E0000 00:00:1743138802.678915 3709779 buffer_comparator.cc:156] Difference at 520: -nan, expected 16.1329
E0000 00:00:1743138802.678917 3709779 buffer_comparator.cc:156] Difference at 521: -nan, expected 14.5245
2025-03-28 05:13:22.678924: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.681761 3709779 buffer_comparator.cc:156] Difference at 528: -nan, expected 17.5032
E0000 00:00:1743138802.681773 3709779 buffer_comparator.cc:156] Difference at 529: -nan, expected 15.1785
E0000 00:00:1743138802.681777 3709779 buffer_comparator.cc:156] Difference at 530: -nan, expected 15.9473
E0000 00:00:1743138802.681779 3709779 buffer_comparator.cc:156] Difference at 531: -nan, expected 14.437
E0000 00:00:1743138802.681782 3709779 buffer_comparator.cc:156] Difference at 532: -nan, expected 17.9637
E0000 00:00:1743138802.681785 3709779 buffer_comparator.cc:156] Difference at 533: -nan, expected 17.3157
E0000 00:00:1743138802.681788 3709779 buffer_comparator.cc:156] Difference at 534: -nan, expected 15.7802
E0000 00:00:1743138802.681791 3709779 buffer_comparator.cc:156] Difference at 535: -nan, expected 17.6887
E0000 00:00:1743138802.681794 3709779 buffer_comparator.cc:156] Difference at 536: -nan, expected 15.1881
E0000 00:00:1743138802.681797 3709779 buffer_comparator.cc:156] Difference at 537: -nan, expected 14.4224
2025-03-28 05:13:22.681801: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.705493 3709779 buffer_comparator.cc:156] Difference at 16: 0, expected 18.4532
E0000 00:00:1743138802.705538 3709779 buffer_comparator.cc:156] Difference at 17: 0, expected 16.1701
E0000 00:00:1743138802.705543 3709779 buffer_comparator.cc:156] Difference at 18: 0, expected 18.5372
E0000 00:00:1743138802.705547 3709779 buffer_comparator.cc:156] Difference at 19: 0, expected 17.7684
E0000 00:00:1743138802.705550 3709779 buffer_comparator.cc:156] Difference at 20: 0, expected 17.8078
E0000 00:00:1743138802.705552 3709779 buffer_comparator.cc:156] Difference at 21: 0, expected 17.412
E0000 00:00:1743138802.705555 3709779 buffer_comparator.cc:156] Difference at 22: 0, expected 18.0425
E0000 00:00:1743138802.705558 3709779 buffer_comparator.cc:156] Difference at 23: 0, expected 17.7822
E0000 00:00:1743138802.705561 3709779 buffer_comparator.cc:156] Difference at 24: 0, expected 16.8692
E0000 00:00:1743138802.705564 3709779 buffer_comparator.cc:156] Difference at 25: 0, expected 19.6248
2025-03-28 05:13:22.705573: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.708424 3709779 buffer_comparator.cc:156] Difference at 16: 0, expected 18.4532
E0000 00:00:1743138802.708436 3709779 buffer_comparator.cc:156] Difference at 17: 0, expected 16.1701
E0000 00:00:1743138802.708439 3709779 buffer_comparator.cc:156] Difference at 18: 0, expected 18.5372
E0000 00:00:1743138802.708443 3709779 buffer_comparator.cc:156] Difference at 19: 0, expected 17.7684
E0000 00:00:1743138802.708446 3709779 buffer_comparator.cc:156] Difference at 20: 0, expected 17.8078
E0000 00:00:1743138802.708449 3709779 buffer_comparator.cc:156] Difference at 21: 0, expected 17.412
E0000 00:00:1743138802.708452 3709779 buffer_comparator.cc:156] Difference at 22: 0, expected 18.0425
E0000 00:00:1743138802.708455 3709779 buffer_comparator.cc:156] Difference at 23: 0, expected 17.7822
E0000 00:00:1743138802.708458 3709779 buffer_comparator.cc:156] Difference at 24: 0, expected 16.8692
E0000 00:00:1743138802.708460 3709779 buffer_comparator.cc:156] Difference at 25: 0, expected 19.6248
2025-03-28 05:13:22.708465: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.711292 3709779 buffer_comparator.cc:156] Difference at 656: 0, expected 15.8892
E0000 00:00:1743138802.711304 3709779 buffer_comparator.cc:156] Difference at 657: 0, expected 15.1292
E0000 00:00:1743138802.711310 3709779 buffer_comparator.cc:156] Difference at 658: 0, expected 14.0499
E0000 00:00:1743138802.711313 3709779 buffer_comparator.cc:156] Difference at 659: 0, expected 13.8377
E0000 00:00:1743138802.711316 3709779 buffer_comparator.cc:156] Difference at 660: 0, expected 13.7353
E0000 00:00:1743138802.711319 3709779 buffer_comparator.cc:156] Difference at 661: 0, expected 15.7468
E0000 00:00:1743138802.711322 3709779 buffer_comparator.cc:156] Difference at 662: 0, expected 14.9101
E0000 00:00:1743138802.711325 3709779 buffer_comparator.cc:156] Difference at 663: 0, expected 14.8135
E0000 00:00:1743138802.711328 3709779 buffer_comparator.cc:156] Difference at 664: 0, expected 13.6403
E0000 00:00:1743138802.711331 3709779 buffer_comparator.cc:156] Difference at 665: 0, expected 15.8348
2025-03-28 05:13:22.711336: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.714164 3709779 buffer_comparator.cc:156] Difference at 672: 0, expected 16.0696
E0000 00:00:1743138802.714178 3709779 buffer_comparator.cc:156] Difference at 673: 0, expected 14.3019
E0000 00:00:1743138802.714181 3709779 buffer_comparator.cc:156] Difference at 674: 0, expected 15.5573
E0000 00:00:1743138802.714184 3709779 buffer_comparator.cc:156] Difference at 675: 0, expected 14.6242
E0000 00:00:1743138802.714187 3709779 buffer_comparator.cc:156] Difference at 676: 0, expected 14.8486
E0000 00:00:1743138802.714190 3709779 buffer_comparator.cc:156] Difference at 677: 0, expected 14.7699
E0000 00:00:1743138802.714193 3709779 buffer_comparator.cc:156] Difference at 678: 0, expected 15.1617
E0000 00:00:1743138802.714196 3709779 buffer_comparator.cc:156] Difference at 679: 0, expected 14.9394
E0000 00:00:1743138802.714199 3709779 buffer_comparator.cc:156] Difference at 680: 0, expected 13.4678
E0000 00:00:1743138802.714202 3709779 buffer_comparator.cc:156] Difference at 681: 0, expected 16.1851
2025-03-28 05:13:22.714207: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.717077 3709779 buffer_comparator.cc:156] Difference at 688: 0, expected 15.1187
E0000 00:00:1743138802.717097 3709779 buffer_comparator.cc:156] Difference at 689: 0, expected 14.6251
E0000 00:00:1743138802.717100 3709779 buffer_comparator.cc:156] Difference at 690: 0, expected 14.2005
E0000 00:00:1743138802.717104 3709779 buffer_comparator.cc:156] Difference at 691: 0, expected 15.1561
E0000 00:00:1743138802.717107 3709779 buffer_comparator.cc:156] Difference at 692: 0, expected 15.4235
E0000 00:00:1743138802.717110 3709779 buffer_comparator.cc:156] Difference at 693: 0, expected 14.1331
E0000 00:00:1743138802.717113 3709779 buffer_comparator.cc:156] Difference at 694: 0, expected 14.4063
E0000 00:00:1743138802.717116 3709779 buffer_comparator.cc:156] Difference at 695: 0, expected 14.0259
E0000 00:00:1743138802.717118 3709779 buffer_comparator.cc:156] Difference at 696: 0, expected 15.0279
E0000 00:00:1743138802.717122 3709779 buffer_comparator.cc:156] Difference at 729: 0, expected 14.5946
2025-03-28 05:13:22.717127: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.720069 3709779 buffer_comparator.cc:156] Difference at 688: 0, expected 15.1187
E0000 00:00:1743138802.720101 3709779 buffer_comparator.cc:156] Difference at 689: 0, expected 14.6251
E0000 00:00:1743138802.720105 3709779 buffer_comparator.cc:156] Difference at 690: 0, expected 14.2005
E0000 00:00:1743138802.720108 3709779 buffer_comparator.cc:156] Difference at 691: 0, expected 15.1561
E0000 00:00:1743138802.720111 3709779 buffer_comparator.cc:156] Difference at 692: 0, expected 15.4235
E0000 00:00:1743138802.720114 3709779 buffer_comparator.cc:156] Difference at 693: 0, expected 14.1331
E0000 00:00:1743138802.720120 3709779 buffer_comparator.cc:156] Difference at 694: 0, expected 14.4063
E0000 00:00:1743138802.720123 3709779 buffer_comparator.cc:156] Difference at 695: 0, expected 14.0259
E0000 00:00:1743138802.720126 3709779 buffer_comparator.cc:156] Difference at 696: 0, expected 15.0279
E0000 00:00:1743138802.720129 3709779 buffer_comparator.cc:156] Difference at 729: 0, expected 14.5946
2025-03-28 05:13:22.720135: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
E0000 00:00:1743138802.723175 3709779 buffer_comparator.cc:156] Difference at 688: 0, expected 15.1187
E0000 00:00:1743138802.723211 3709779 buffer_comparator.cc:156] Difference at 689: 0, expected 14.6251
E0000 00:00:1743138802.723214 3709779 buffer_comparator.cc:156] Difference at 690: 0, expected 14.2005
E0000 00:00:1743138802.723218 3709779 buffer_comparator.cc:156] Difference at 691: 0, expected 15.1561
E0000 00:00:1743138802.723221 3709779 buffer_comparator.cc:156] Difference at 692: 0, expected 15.4235
E0000 00:00:1743138802.723223 3709779 buffer_comparator.cc:156] Difference at 693: 0, expected 14.1331
E0000 00:00:1743138802.723226 3709779 buffer_comparator.cc:156] Difference at 694: 0, expected 14.4063
E0000 00:00:1743138802.723229 3709779 buffer_comparator.cc:156] Difference at 695: 0, expected 14.0259
E0000 00:00:1743138802.723232 3709779 buffer_comparator.cc:156] Difference at 696: 0, expected 15.0279
E0000 00:00:1743138802.723235 3709779 buffer_comparator.cc:156] Difference at 729: 0, expected 14.5946
2025-03-28 05:13:22.723243: E external/xla/xla/service/gpu/autotuning/gemm_fusion_autotuner.cc:1137] Results do not match the reference. This is likely a bug/unexpected loss of precision.
Iteration: 0001 Loss: 2.08086824
Iteration: 0101 Loss: 0.135340884
Iteration: 0201 Loss: 0.0046597519
Iteration: 0301 Loss: 0.00102267659
Iteration: 0401 Loss: 0.000412049034
Iteration: 0501 Loss: 0.000682463811
Iteration: 0601 Loss: 0.00017241342
Iteration: 0701 Loss: 0.000250844547
Iteration: 0801 Loss: 0.000174931673
Iteration: 0901 Loss: 6.01563152e-05
Iteration: 1000 Loss: 0.000252216007Training 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.jlfor CUDA support.AMDGPU.jlfor AMDGPU support.Metal.jlfor Apple Metal support.oneAPI.jlfor oneAPI support.