model = VRidGGNN(model)
# model_name = 'VRid'
# model = load_network(model, model_name)
print(model)
if opt.single_cls:
if opt.backbone == 'ResNet50':
ignored_params = list(map(id, model.model.fc.parameters()))
elif opt.backbone == 'EfficientNet-B0':
ignored_params = list(map(id, model.model._fc.parameters()))
ignored_params += (list(map(id, model.classifier.parameters())))
if opt.freeze_backbone:
ignored_params += (list(map(id, model.model.parameters())))
base_params = filter(lambda p: id(p) not in ignored_params,
model.parameters())
# optimizer = optim.SGD(
# [{'params': base_params, 'lr': 0.1*opt.lr},
# {'params': model.classifier.parameters(), 'lr': opt.lr}],
# weight_decay=5e-4, momentum=0.9, nesterov=True)
optimizer = optim.Adam([{
'params': base_params,
'lr': 0.00035
}, {
'params': model.classifier.parameters(),
'lr': 0.0035
}])
else:
if opt.backbone == 'ResNet50':
ignored_params = list(map(id, model.model.fc.parameters()))
elif opt.backbone == 'EfficientNet-B0':
model = PCB_Effi_GGNN(model, opt)
# model_name = 'LSTM'
# model = load_network(model, model_name)
print(model)
if opt.single_cls:
if opt.backbone == 'ResNet50':
ignored_params = list(map(id, model.model.fc.parameters()))
elif opt.backbone == 'EfficientNet-B0':
ignored_params = list(map(id, model.model._fc.parameters()))
ignored_params += (list(map(id, model.classifier.parameters())))
if opt.freeze_backbone:
ignored_params += (list(map(id, model.model.parameters())))
base_params = filter(
lambda p: id(p) not in ignored_params, model.parameters()
)
optimizer = optim.SGD(
[{'params': base_params, 'lr': 0.1*opt.lr},
{'params': model.classifier.parameters(), 'lr': opt.lr}],
weight_decay=5e-4, momentum=0.9, nesterov=True)
else:
if opt.backbone == 'ResNet50':
ignored_params = list(map(id, model.model.fc.parameters()))
elif opt.backbone == 'EfficientNet-B0':
ignored_params = list(map(id, model.model._fc.parameters()))
ignored_params += (
list(map(id, model.classifierA0.parameters()))
+ list(map(id, model.classifierA1.parameters()))
+ list(map(id, model.classifierA2.parameters()))
+ list(map(id, model.classifierA3.parameters()))