ontolearn.clip_architectures ============================ .. py:module:: ontolearn.clip_architectures Classes ------- .. autoapisummary:: ontolearn.clip_architectures.LengthLearner_LSTM ontolearn.clip_architectures.LengthLearner_GRU ontolearn.clip_architectures.LengthLearner_CNN ontolearn.clip_architectures.LengthLearner_SetTransformer Module Contents --------------- .. py:class:: LengthLearner_LSTM(input_size, output_size, proj_dim=256, rnn_n_layers=2, drop_prob=0.2) Bases: :py:obj:`torch.nn.Module` LSTM architecture .. py:attribute:: name :value: 'LSTM' .. py:attribute:: loss .. py:attribute:: lstm .. py:attribute:: dropout .. py:attribute:: fc1 .. py:attribute:: fc2 .. py:attribute:: fc3 .. py:method:: forward(x1, x2) Forward pass through the network. .. py:class:: LengthLearner_GRU(input_size, output_size, proj_dim=256, rnn_n_layers=2, drop_prob=0.2) Bases: :py:obj:`torch.nn.Module` GRU architecture .. py:attribute:: name :value: 'GRU' .. py:attribute:: loss .. py:attribute:: gru .. py:attribute:: dropout .. py:attribute:: fc1 .. py:attribute:: fc2 .. py:attribute:: fc3 .. py:method:: forward(x1, x2) Forward pass through the network. .. py:class:: LengthLearner_CNN(input_size, output_size, num_examples, proj_dim=256, kernel_size: list = [[5, 7], [5, 7]], stride: list = [[3, 3], [3, 3]], drop_prob=0.2) Bases: :py:obj:`torch.nn.Module` CNN architecture .. py:attribute:: name :value: 'CNN' .. py:attribute:: loss .. py:attribute:: conv1 .. py:attribute:: conv2 .. py:attribute:: dropout1d .. py:attribute:: dropout2d .. py:attribute:: fc1 .. py:attribute:: fc2 .. py:attribute:: fc3 .. py:method:: forward(x1, x2) Forward pass through the network. .. py:class:: LengthLearner_SetTransformer(input_size, output_size, proj_dim=256, num_heads=4, num_seeds=1, m=32) Bases: :py:obj:`torch.nn.Module` SetTransformer architecture. .. py:attribute:: name :value: 'SetTransformer' .. py:attribute:: loss .. py:attribute:: enc .. py:attribute:: dec .. py:method:: forward(x1, x2) Forward pass through the network.