for k, v in pretrained_dict.items() if k in model_dict
}
# del pretrained_dict['classifier.6.bias']
# del pretrained_dict['classifier.6.weight']
model_dict.update(pretrained_dict)
model.load_state_dict(model_dict)
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--load', help='Resume from checkpoint file')
args = parser.parse_args()
for with_coral in [False, True]:
model = models.DeepCORAL(31)
# support different learning rate according to CORAL paper
# i.e. 10 times learning rate for the last two fc layers.
optimizer = torch.optim.SGD([
{
'params': model.sharedNet.parameters()
},
{
'params': model.fc.parameters(),
'lr': 10 * LEARNING_RATE
},
],
lr=LEARNING_RATE,
momentum=MOMENTUM)
if CUDA:
k: v
for k, v in pretrained_dict.items() if k in model_dict
}
# del pretrained_dict['classifier.6.bias']
# del pretrained_dict['classifier.6.weight']
model_dict.update(pretrained_dict)
model.load_state_dict(model_dict)
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--load', help='Resume from checkpoint file')
args = parser.parse_args()
model = models.DeepCORAL(31) # num_classes=31
# support different learning rate according to CORAL paper
# i.e. 10 times learning rate for the last two fc layers.
optimizer = torch.optim.SGD([
{
'params': model.sharedNet.parameters()
},
{
'params': model.fc.parameters(),
'lr': 10 * LEARNING_RATE
},
],
lr=LEARNING_RATE,
momentum=MOMENTUM)