Skip to content

Update deeponet.md #89

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jstrubeAtWork wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Update deeponet.md #89

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
32 changes: 27 additions & 5 deletions docs/src/models/deeponet.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,7 @@ Random.seed!(rng, 1234)

xdev = reactant_device()

eval_points = 1
eval_points = 17
batch_size = 64
dim_y = 1
m = 32
Expand All @@ -58,7 +58,10 @@ xrange = range(0, 2π; length=m) .|> Float32
α = 0.5f0 .+ 0.5f0 .* rand(Float32, batch_size)

u_data = zeros(Float32, m, batch_size)
y_data = rand(rng, Float32, 1, eval_points) .* Float32(2π)
y_data = rand(rng, Float32, dim_y, eval_points) .* Float32(2π)
# for plotting, we want to evaluate points in order
rightorder = sortperm(vec(y_data))

v_data = zeros(Float32, eval_points, batch_size)

for i in 1:batch_size
Expand All @@ -67,8 +70,8 @@ for i in 1:batch_size
end

deeponet = DeepONet(
Chain(Dense(m => 8, σ), Dense(8 => 8, σ), Dense(8 => 8, σ)),
Chain(Dense(1 => 4, σ), Dense(4 => 8, σ))
Chain(Dense(m => 64, tanh), Dense(64 => 64, tanh), Dense(64 => 64, tanh)),
Chain(Dense(1 => 16, tanh), Dense(16 => 64, tanh))
)

ps, st = Lux.setup(rng, deeponet) |> xdev;
Expand All @@ -90,7 +93,7 @@ function train!(model, ps, st, data; epochs=10)
return losses
end

losses = train!(deeponet, ps, st, data; epochs=1000)
losses = train!(deeponet, ps, st, data; epochs=20000)

draw(
AoG.data((; losses, iteration=1:length(losses))) *
Expand All @@ -99,4 +102,23 @@ draw(
axis=(; yscale=log10),
figure=(; title="Using DeepONet to learn the anti-derivative operator")
)

# plot the prediction for a new function
# that's not part of the training set
αₜ = 0.75
input_data = sin.(αₜ .* xrange) |> xdev
output_data, st = @jit Lux.apply(deeponet, (input_data, y_data), ps, st)
output_x = vec(cdev(y_data))[rightorder]
pred_y = vec(cdev(output_data))[rightorder]
true_y = -inv(αₜ) .* cos.(αₜ .* y_data[1, rightorder])
p = lines(Array(xrange), Array(input_data); label="u")
lines!(a, Array(output_x), Array(pred_y); label="Predicted")
lines!(a, Array(output_x), Array(true_y); label="Expected")
axislegend(a)
# Compute the absolute error and plot that, too
absolute_error = abs.(Array(pred_y) .- Array(true_y))
a2, p2 = lines(f[2, 1], Array(output_x), absolute_error; axis=(; ylabel="Error"))
rowsize!(f.layout, 2, Aspect(1, 1 / 8))
linkxaxes!(a, a2)
f
```
Loading