Skip to content

Implementing test unit for learning rates #1460

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

Closed
XavierAudier wants to merge 1 commit into from
Closed

Implementing test unit for learning rates #1460

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
35 changes: 34 additions & 1 deletion pylearn2/models/tests/test_dbm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -352,7 +352,40 @@ def downward_message(self, state):
sample = sample.eval()

check_gaussian_samples(sample, num_samples, n, rows, cols, channels, mean, tol)


def test_gaussian_vis_get_lr_scalers():
"""
Verifies that the GaussianVisLayer.get_lr_scalers returns the right values.

"""
print("We expect all the learning rates to be 1/100")
n = None
rows = 10
cols = 10
channels = 1
beta = 1.
mu=0.
layer = GaussianVisLayer(rows=rows, cols=cols, channels=channels, init_beta=beta, tie_beta='locations', beta_lr_scale='by_sharing', init_mu=mu, tie_mu='locations', mu_lr_scale='by_sharing')
print(layer.get_lr_scalers())
layer = GaussianVisLayer(rows=rows, cols=cols, channels=channels, init_beta=beta, tie_beta=None, beta_lr_scale='by_sharing', init_mu=mu, tie_mu=None, mu_lr_scale='by_sharing')
print(layer.get_lr_scalers())
layer = GaussianVisLayer(rows=rows, cols=cols, channels=channels, init_beta=beta, tie_beta='locations', beta_lr_scale=None, init_mu=mu, tie_mu='locations', mu_lr_scale=None)
print(layer.get_lr_scalers())
layer = GaussianVisLayer(rows=rows, cols=cols, channels=channels, init_beta=beta, tie_beta=None, beta_lr_scale=None, init_mu=mu, tie_mu=None, mu_lr_scale=None)
print(layer.get_lr_scalers())
layer = GaussianVisLayer(rows=rows, cols=cols, channels=channels, init_beta=beta, tie_beta='locations', beta_lr_scale=1./(rows * cols), init_mu=mu, tie_mu='locations', mu_lr_scale= 1./(rows * cols))
print(layer.get_lr_scalers())
layer = GaussianVisLayer(rows=rows, cols=cols, channels=channels, init_beta=beta, tie_beta=None, beta_lr_scale=1./(rows * cols), init_mu=mu, tie_mu=None, mu_lr_scale=1./(rows * cols))
print(layer.get_lr_scalers())

n = rows*cols*channels
layer = GaussianVisLayer(nvis = n, init_beta=beta, tie_beta=None, beta_lr_scale='by_sharing', init_mu=mu, tie_mu=None, mu_lr_scale='by_sharing')
print(layer.get_lr_scalers())
layer = GaussianVisLayer(nvis = n, init_beta=beta, tie_beta=None, beta_lr_scale=None, init_mu=mu, tie_mu=None, mu_lr_scale=None)
print(layer.get_lr_scalers())
layer = GaussianVisLayer(nvis = n, init_beta=beta, tie_beta=None, beta_lr_scale=1./n, init_mu=mu, tie_mu=None, mu_lr_scale=1./n)
print(layer.get_lr_scalers())

def check_bvmp_samples(value, num_samples, n, pool_size, mean, tol):
"""
bvmp=BinaryVectorMaxPool
Expand Down