Skip to content

Experimental 2 #22

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
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Experimental 2 #22

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
81610b3
replacing tests
josemanuel22 Aug 22, 2023
0966ced
replacing tests
josemanuel22 Aug 22, 2023
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
6 changes: 3 additions & 3 deletions benchmarks/benchmark_multimodal.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ include("benchmark_utils.jl")
hparams = HyperParamsVanillaGan(;
data_size=100,
batch_size=1,
epochs=1e3,
epochs=1e4,
lr_dscr=1e-4,
lr_gen=1e-4,
dscr_steps=0,
Expand All @@ -29,7 +29,7 @@ include("benchmark_utils.jl")
train_vanilla_gan(dscr, gen, hparams)

hparams = HyperParams(;
samples=1000, K=100, epochs=100, η=1e-2, transform=noise_model
samples=200, K=10, epochs=1000, η=2e-2, transform=noise_model
)
#hparams = AutoAdaptativeHyperParams(;
# max_k=20, samples=1200, epochs=10000, η=1e-3, transform=noise_model
Expand Down Expand Up @@ -59,7 +59,7 @@ include("benchmark_utils.jl")
gen,
n_samples,
(-3:0.1:3),
(5:0.2:15),
(-8:0.2:8),
)

#@test js_divergence(hist1.weights, hist2.weights)/hparams.samples < 0.03
Expand Down
10 changes: 5 additions & 5 deletions test/runtests.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -88,7 +88,7 @@ end;
@testset "adaptative_block_learning" begin
@testset "learning Normal(4.0f0, 2.0f0)" begin
nn = Chain(Dense(1, 7), elu, Dense(7, 13), elu, Dense(13, 7), elu, Dense(7, 1))
hparams = HyperParams(1000, 10, 400, 1e-2, Normal(0.0f0, 1.0f0))
hparams = HyperParams(1000, 10, 1000, 1e-2, Normal(0.0f0, 1.0f0))

function real_model(ϵ)
return rand(Normal(4.0f0, 2.0f0))
Expand All @@ -112,7 +112,7 @@ end;

@testset "learning uniform distribution (1,3)" begin
nn = Chain(Dense(1, 7), elu, Dense(7, 13), elu, Dense(13, 7), elu, Dense(7, 1))
hparams = HyperParams(100, 10, 1000, 1e-2, Normal(0.0f0, 1.0f0))
hparams = HyperParams(100, 10, 2000, 1e-2, Normal(0.0f0, 1.0f0))

function real_model(ϵ)
return rand(Float32) * 2 + 1
Expand All @@ -132,7 +132,7 @@ end;

@testset "learning Cauchy distribution" begin
nn = Chain(Dense(1, 7), elu, Dense(7, 13), elu, Dense(13, 7), elu, Dense(7, 1))
hparams = HyperParams(100, 10, 1000, 1e-3, Normal(0.0f0, 1.0f0))
hparams = HyperParams(100, 10, 2000, 1e-3, Normal(0.0f0, 1.0f0))

function real_model(ϵ)
return rand(Cauchy(1.0f0, 2.0f0))
Expand All @@ -152,7 +152,7 @@ end;

@testset "learning Bimodal Normal Distribution" begin
nn = Chain(Dense(1, 7), elu, Dense(7, 13), elu, Dense(13, 7), elu, Dense(7, 1))
hparams = HyperParams(100, 10, 2000, 1e-2, Normal(0.0f0, 1.0f0))
hparams = HyperParams(100, 10, 4000, 1e-2, Normal(0.0f0, 1.0f0))

function real_model(ϵ)
return rand(MixtureModel(Normal[Normal(5.0f0, 2.0f0), Normal(-1.0f0, 1.0f0)]))
Expand All @@ -173,7 +173,7 @@ end;
@testset "learning modal auto_adaptative_block_learning Normal(4.0f0, 2.0f0)" begin
nn = Chain(Dense(1, 7), elu, Dense(7, 13), elu, Dense(13, 7), elu, Dense(7, 1))
hparams = AutoAdaptativeHyperParams(;
max_k=10, samples=1000, epochs=400, η=1e-2, transform=Normal(0.0f0, 1.0f0)
max_k=10, samples=1000, epochs=1000, η=1e-2, transform=Normal(0.0f0, 1.0f0)
)

function real_model(ϵ)
Expand Down