def conv__transf_2x4x128x256_drop01__gru_2x256_drop02(num_outputs) -> nn.Module:
model = SequentialSequential(*[
ConvExtractor(),
SequentialLinear(512, 128, pre_activation=True),
TransformerEncoder(dropout=0.1, num_layers=2, num_heads=4, dim_model=128, dim_feedforward=256),
RNNEncoder(dropout=0.2, rnn_type="GRU", num_layers=2, hidden_size=256, input_size=128),
SequentialLinear(256 * 2, num_outputs, pre_activation=True)
])
return model
def conv__gru_3x368_drop02(num_outputs) -> nn.Module:
model = SequentialSequential(*[
ConvExtractor(),
LambdaModule(lambda seq, seq_len: (F.relu(seq), seq_len)),
RNNEncoder(dropout=0.2, rnn_type="GRU", num_layers=3, hidden_size=368, input_size=512),
SequentialLinear(368 * 2, num_outputs, pre_activation=True)
])
return model
def conv_instnorm__gru_2x256_drop02(num_outputs) -> nn.Module:
model = SequentialSequential(*[
ConvExtractor(norm=nn.InstanceNorm2d),
LambdaModule(lambda seq, seq_len: (F.relu(seq), seq_len)),
RNNEncoder(dropout=0.2, rnn_type="GRU", num_layers=2, hidden_size=256, input_size=512),
SequentialLinear(256 * 2, num_outputs, pre_activation=True)
])
return model
def resnet_pure(num_outputs) -> nn.Module:
model = SequentialSequential(*[
ResnetExtractor(),
SequentialLinear(512, num_outputs, pre_activation=True)
])
return model
def conv_instnorm_pure(num_outputs) -> nn.Module:
model = SequentialSequential(*[
ConvExtractor(norm=nn.InstanceNorm2d),
SequentialLinear(512, num_outputs, pre_activation=True)
])
return model