"num_classes": 4,
"activation": "relu",
"normalization": "group_normalization",
"mode": "trilinear",
"with_vae": True,
"debug": False
}
if os.path.isfile(trained_model):
nvnet = NvNetSegmenter(nvnet_kwargs,
optimizer_name="Adam",
learning_rate=1e-4,
weight_decay=1e-5,
loss=my_loss,
pretrained=trained_model,
use_cuda=True)
train_history = History.load(os.path.join(outdir, "train_0_epoch_9.pkl"))
valid_history = History.load(
os.path.join(outdir, "validation_0_epoch_9.pkl"))
else:
nvnet = NvNetSegmenter(nvnet_kwargs,
optimizer_name="Adam",
learning_rate=1e-4,
weight_decay=1e-5,
loss=my_loss,
use_cuda=True)
scheduler = lr_scheduler.ReduceLROnPlateau(optimizer=nvnet.optimizer,
mode="min",
factor=0.5,
patience=5)
train_history, valid_history = nvnet.training(
manager=manager,
pb=pb,
hyper=hyper,
n_finetuning=n_finetuning,
block=b,
f=fold,
e=e))) for fold in range(folds)
]
else:
if CV:
if pb == 'Alzheimer':
filename = "Validation_DenseNet_{pb}_{db}_CV_%i_epoch_{e}.pkl"
else:
filename = "Validation_DenseNet_{pb}_{db}_%i_epoch_{e}.pkl"
results[db][name][n_finetuning] = History.load(
os.path.join(root, path.format(n=N_pretraining,
n_finetune=n_finetuning, pb=pb), filename.
format(db=db, pb=pb, e=e)), folds=list(range(folds))). \
to_dict(patterns_to_del=patterns_to_del)
else:
filename = "Test_DenseNet_{pb}_{db}_fold{f}_epoch{e}.pkl"
results[db][name][n_finetuning] = [
get_pickle_obj(
os.path.join(
root,
path.format(n=N_pretraining,
n_finetune=n_finetuning,
pb=pb),
filename.format(db=db, pb=pb, f=fold, e=e)))
for fold in range(folds)
]
model=net,
metrics=["accuracy"])
test_history, train_history = cl.training(manager=manager,
nb_epochs=3,
checkpointdir="/tmp/pynet",
fold_index=0,
with_validation=True)
#############################################################################
# You can reload the optimization history at any time and any step.
from pprint import pprint
from pynet.history import History
from pynet.plotting import plot_history
history = History.load("/tmp/pynet/train_0_epoch_2.pkl")
print(history)
plot_history(history)
#############################################################################
# And now predict the labels on the test set.
import numpy as np
from sklearn.metrics import classification_report
from pynet.plotting import plot_data
y_pred, X, y_true, loss, values = cl.testing(manager=manager,
with_logit=True,
predict=True)
pprint(data.labels)
print(classification_report(y_true, y_pred, target_names=data.labels.values()))
test_history, train_history, valid_history = training(net=net,
dataset=dataset,
optimizer=optim.Adam(
net.parameters(),
lr=0.01),
criterion=my_loss,
nb_epochs=3,
metrics={"mse": my_loss},
use_cuda=False,
outdir="/tmp/pynet",
verbose=1)
#############################################################################
# You can reload the optimization history at any time and any step
from pprint import pprint
from pynet.history import History
valid_history = History.load("/tmp/pynet/history/valid_1_epoch_3.pkl")
pprint(valid_history.history)
pprint(valid_history["loss"])
#############################################################################
# You can finally display the optimization cost
from pynet.plotting import plot_data
x, y = valid_history["loss"]
plot_data(y)