test: remove examples/liramlvq_tecator.py
temporarily
This commit is contained in:
parent
4dc11a3737
commit
a1d9657b91
@ -1,88 +0,0 @@
|
||||
"""Limited Rank Matrix LVQ example using the Tecator dataset."""
|
||||
|
||||
import argparse
|
||||
|
||||
import matplotlib.pyplot as plt
|
||||
import prototorch as pt
|
||||
import pytorch_lightning as pl
|
||||
import torch
|
||||
|
||||
|
||||
def plot_matrix(matrix):
|
||||
title = "Lambda matrix"
|
||||
plt.figure(title)
|
||||
plt.title(title)
|
||||
plt.imshow(matrix, cmap="gray")
|
||||
plt.axis("off")
|
||||
plt.colorbar()
|
||||
plt.show(block=True)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
# Command-line arguments
|
||||
parser = argparse.ArgumentParser()
|
||||
parser = pl.Trainer.add_argparse_args(parser)
|
||||
args = parser.parse_args()
|
||||
|
||||
# Dataset
|
||||
train_ds = pt.datasets.Tecator(root="~/datasets/", train=True)
|
||||
test_ds = pt.datasets.Tecator(root="~/datasets/", train=False)
|
||||
|
||||
# Reproducibility
|
||||
pl.utilities.seed.seed_everything(seed=10)
|
||||
|
||||
# Dataloaders
|
||||
train_loader = torch.utils.data.DataLoader(train_ds, batch_size=32)
|
||||
test_loader = torch.utils.data.DataLoader(test_ds, batch_size=32)
|
||||
|
||||
# Hyperparameters
|
||||
hparams = dict(
|
||||
distribution={
|
||||
"num_classes": 2,
|
||||
"per_class": 1,
|
||||
},
|
||||
input_dim=100,
|
||||
latent_dim=2,
|
||||
proto_lr=0.001,
|
||||
bb_lr=0.001,
|
||||
)
|
||||
|
||||
# Initialize the model
|
||||
model = pt.models.SiameseGMLVQ(
|
||||
hparams,
|
||||
optimizer=torch.optim.Adam,
|
||||
prototypes_initializer=pt.initializers.SMCI(train_ds),
|
||||
)
|
||||
|
||||
# Summary
|
||||
print(model)
|
||||
|
||||
# Callbacks
|
||||
vis = pt.models.VisSiameseGLVQ2D(train_ds, border=0.1)
|
||||
es = pl.callbacks.EarlyStopping(monitor="val_loss",
|
||||
min_delta=0.001,
|
||||
patience=50,
|
||||
verbose=False,
|
||||
mode="min")
|
||||
|
||||
# Setup trainer
|
||||
trainer = pl.Trainer.from_argparse_args(
|
||||
args,
|
||||
callbacks=[vis, es],
|
||||
weights_summary=None,
|
||||
)
|
||||
|
||||
# Training loop
|
||||
trainer.fit(model, train_loader, test_loader)
|
||||
|
||||
# Save the model
|
||||
torch.save(model, "liramlvq_tecator.pt")
|
||||
|
||||
# Load a saved model
|
||||
saved_model = torch.load("liramlvq_tecator.pt")
|
||||
|
||||
# Display the Lambda matrix
|
||||
plot_matrix(saved_model.lambda_matrix)
|
||||
|
||||
# Testing
|
||||
trainer.test(model, test_dataloaders=test_loader)
|
Loading…
Reference in New Issue
Block a user