Args:
l: Local feature map.
m: Multiple globals feature map.
measure: Type of f-divergence. For use with mode `fd`
mode: Loss mode. Fenchel-dual `fd`, NCE `nce`, or Donsker-Vadadhan `dv`.
Returns:
torch.Tensor: Loss.
'''
l_enc = self.nets.local_net(l)
m_enc = self.nets.multi_net(m)
N, local_units, dim_x, dim_y = l_enc.size()
l_enc = l_enc.view(N, local_units, -1)
m_enc = m_enc.view(N, local_units, -1)
if mode == 'fd':
loss = multi_fenchel_dual_loss(l_enc, m_enc, measure=measure)
elif mode == 'nce':
loss = multi_nce_loss(l_enc, m_enc)
elif mode == 'dv':
loss = multi_donsker_varadhan_loss(l_enc, m_enc)
else:
raise NotImplementedError(mode)
return loss
if __name__ == '__main__':
run(DIM())
classifier = self.nets.classifier
outputs = classifier(inputs)
predicted = torch.max(F.log_softmax(outputs, dim=1).data, 1)[1]
loss = criterion(outputs, targets)
correct = 100. * predicted.eq(
targets.data).cpu().sum() / targets.size(0)
self.losses.classifier = loss
self.results.accuracy = correct
def predict(self, inputs):
classifier = self.nets.classifier
outputs = classifier(inputs)
predicted = torch.max(F.log_softmax(outputs, dim=1).data, 1)[1]
return predicted
def visualize(self, images, inputs, targets):
predicted = self.predict(inputs)
self.add_image(images.data, labels=(targets.data, predicted.data),
name='gt_pred')
if __name__ == '__main__':
classifier = MyClassifier()
run(model=classifier)
classifier = self.nets.classifier
outputs = classifier(inputs)
predicted = torch.max(F.log_softmax(outputs, dim=1).data, 1)[1]
loss = criterion(outputs, targets)
correct = 100. * predicted.eq(
targets.data).cpu().sum() / targets.size(0)
self.losses.classifier = loss
self.results.accuracy = correct
def predict(self, inputs):
classifier = self.nets.classifier
outputs = classifier(inputs)
predicted = torch.max(F.log_softmax(outputs, dim=1).data, 1)[1]
return predicted
def visualize(self, images, inputs, targets):
predicted = self.predict(inputs)
self.add_image(images.data,
labels=(targets.data, predicted.data),
name='gt_pred')
if __name__ == '__main__':
classifier = MyClassifier()
run(model=classifier)
'''
self.nets.encoder = encoder
X = self.inputs('data.images')
self.task_idx = task_idx
out = self.nets.encoder(X, return_all_activations=True)[self.task_idx]
config = decoder_configs.get(config)
config.update(**args)
super().build(out.size()[1:], args=config)
def routine(self, outs=None):
X = self.inputs('data.images')
if outs is None:
outs = self.nets.encoder(X, return_all_activations=True)
out = outs[self.task_idx]
super().routine(X, out.detach())
def visualize(self, inputs):
out = self.nets.encoder(inputs,
return_all_activations=True)[self.task_idx]
super().visualize(out)
if __name__ == '__main__':
run(MSSSIMEval())
infos = []
for k in mode_dict.keys():
mode = mode_dict[k]
info = mode.__doc__.split('\n', 1)[0] # Keep only first line of doctstring.
infos.append('{}: {}'.format(k, info))
infos = '\n\t'.join(infos)
models = []
parser = argparse.ArgumentParser(formatter_class=RawTextHelpFormatter)
parser.add_argument('models', nargs='+', choices=names,
help='Models used in Deep InfoMax. Choices are: \n\t{}'.format(infos))
i = 1
while True:
arg = sys.argv[i]
if arg in ('--help', '-h') and i == 1:
i += 1
break
if arg.startswith('-'):
break # argument have begun
i += 1
args = parser.parse_args(sys.argv[1:i])
models = args.models
models = list(set(models))
models = dict((k, mode_dict[k]) for k in models)
sys.argv = [sys.argv[0]] + sys.argv[i:]
controller = Controller(inputs=dict(inputs='data.images'), **models)
run(controller)
inputs: torch.Tensor
ae_criterion: function
Returns: None
"""
encoded = self.nets.ae.encoder(inputs)
outputs = self.nets.ae.decoder(encoded)
r_loss = ae_criterion(
outputs, inputs, size_average=False) / inputs.size(0)
self.losses.ae = r_loss
def visualize(self, inputs):
"""
Adding generated images and base images to
visualization.
Args:
inputs: torch.Tensor
Returns: None
"""
encoded = self.nets.ae.encoder(inputs)
outputs = self.nets.ae.decoder(encoded)
self.add_image(outputs, name='reconstruction')
self.add_image(inputs, name='ground truth')
if __name__ == '__main__':
autoencoder = AE()
run(model=autoencoder)
inputs: torch.Tensor
ae_criterion: function
Returns: None
"""
encoded = self.nets.ae.encoder(inputs)
outputs = self.nets.ae.decoder(encoded)
r_loss = ae_criterion(outputs, inputs,
size_average=False) / inputs.size(0)
self.losses.ae = r_loss
def visualize(self, inputs):
"""
Adding generated images and base images to
visualization.
Args:
inputs: torch.Tensor
Returns: None
"""
encoded = self.nets.ae.encoder(inputs)
outputs = self.nets.ae.decoder(encoded)
self.add_image(outputs, name='reconstruction')
self.add_image(inputs, name='ground truth')
if __name__ == '__main__':
autoencoder = AE()
run(model=autoencoder)
for i, idx in enumerate(self.classification_idx):
name = self.classifier_names[i]
contract = dict(nets=dict(classifier='classifier_{}'.format(name)))
setattr(self, 'classifier_{}'.format(name), Classifier(**contract))
classifier = getattr(self, 'classifier_{}'.format(name))
input_shape = outs[idx].size()[1:]
kwargs = classifier_kwargs[i]
classifier.build(input_shape, n_labels, **kwargs)
def routine(self, inputs, targets):
'''Classification routine.
'''
for i, idx in enumerate(self.classification_idx):
name = self.classifier_names[i]
classifier = getattr(self, 'classifier_{}'.format(name))
output = self.nets.encoder(inputs)[idx]
classifier.routine(output.detach(), targets)
def visualize(self, inputs, targets):
'''Visualization.
'''
for i, idx in enumerate(self.classification_idx):
name = self.classifier_names[i]
classifier = getattr(self, 'classifier_{}'.format(name))
output = self.nets.encoder(inputs)[idx]
classifier.visualize(inputs, output, targets)
if __name__ == '__main__':
run(ClassificationEvaluator())
def routine(self, outs=None, measure='KL', nonlinearity=''):
'''
Args:
measure: Type of measure to use for NDM.
nonlinearity: Nonlinearity to use on output of encoder.
Returns:
'''
if outs is None:
inputs = self.inputs('data.images')
outs = self.nets.encoder(inputs, return_all_activations=True)
out = outs[self.task_idx]
if nonlinearity != '':
out = getattr(nn, nonlinearity)()(out)
super().routine(out.detach(), measure=measure)
def visualize(self, inputs, nonlinearity=None, measure=None):
out = self.nets.encoder(inputs,
return_all_activations=True)[self.task_idx]
if nonlinearity != '':
out = getattr(nn, nonlinearity)()(out)
super().visualize(out, measure=measure)
if __name__ == '__main__':
run(NDMEval())
import mlflow
import logging
from cortex.main import run
from src.models.fix_match.controller import FixMatchController
from src.data.dataset_plugins import SSLDatasetPlugin
from src import MLFLOW_SSL_URI
logger = logging.getLogger('ssl_evaluation')
if __name__ == '__main__':
# if exp.ARGS
mlflow.set_tracking_uri(MLFLOW_SSL_URI)
controller = FixMatchController()
run(model=controller)
