def __init__(self):
super(MyConv, self).__init__()
self.conv = nn.Sequential(
exnn.Conv3d(20, 30, 5, 5),
exnn.Conv3d(30, 60, 5, 5),
exnn.Flatten(),
)
def __init__(self, classes):
super(TeacherNet, self).__init__()
self.classes = classes
self.conv = nn.Sequential(
nn.Conv2d(in_channels=3,
out_channels=96,
kernel_size=3,
stride=1,
padding=1), nn.BatchNorm2d(96), nn.ReLU(),
nn.MaxPool2d(kernel_size=3, stride=2),
nn.Conv2d(in_channels=96,
out_channels=256,
kernel_size=3,
stride=1,
padding=1), nn.BatchNorm2d(256), nn.ReLU(),
nn.MaxPool2d(kernel_size=3, stride=2),
nn.Conv2d(in_channels=256,
out_channels=384,
kernel_size=3,
stride=1,
padding=1), nn.BatchNorm2d(384), nn.ReLU(),
nn.Conv2d(in_channels=384,
out_channels=384,
kernel_size=3,
stride=1,
padding=1), nn.BatchNorm2d(384), nn.ReLU(),
nn.MaxPool2d(3, 2),
nn.Conv2d(in_channels=384,
out_channels=256,
kernel_size=3,
stride=1,
padding=1), nn.BatchNorm2d(256), nn.ReLU(),
nn.MaxPool2d(3, 2))
self.fc = nn.Sequential(exnn.GlobalAvgPool2d(), exnn.Flatten(),
nn.Linear(256, self.classes))
def __init__(self, out_channels, activation='relu'):
super(FlattenLinear, self).__init__()
self.out_channels = out_channels
self.linear = exnn.Linear(self.out_channels)
self.flatten = exnn.Flatten()
if activation == 'relu':
self.activation = nn.ReLU()
else:
self.activation = nn.Identity()
def __init__(self):
super(Net, self).__init__()
self.linear = nn.Sequential(
exnn.Flatten(),
nn.Linear(784, 50),
nn.ReLU(),
nn.Dropout(),
nn.Linear(50, 10),
nn.LogSoftmax(dim=1)
)
def construct(self, idx):
_layer_dict = copy.deepcopy(self.layer_dict)
(module_name, params) = self.layer[idx]
_layer_dict[module_name] = 1
for k, v in params.items():
_layer_dict[k] = v
vec = list(_layer_dict.values())
if module_name == 'conv2d':
return conv2d(**params), vec
elif module_name == 'linear':
return FlattenLinear(**params), vec
elif module_name == 'identity':
return exnn.Flatten(), vec
def __init__(self, seq_len=21, out_channels=64):
super(Tox21Embedding, self).__init__()
self.embd = nn.Embedding(seq_len, out_channels)
self.pool = GlobalAvgPool1d()
self.flatten = exnn.Flatten()
def __init__(self, out_channels):
super(FlattenLinear, self).__init__()
self.out_channels = out_channels
self.linear = exnn.Linear(self.out_channels)
self.flatten = exnn.Flatten()
def _get_flatten():
return exnn.Flatten()
