MCPcopy Create free account

hub / github.com/L0SG/relational-rnn-pytorch / functions

Functions56 in github.com/L0SG/relational-rnn-pytorch

⨍ Functions 56◇ Types & classes 9
↓ 5 callersMethodinit_hidden
(self, bsz)
rnn_models.py:104
↓ 5 callersMethodinitial_state
Creates the initial memory. We should ensure each row of the memory is initialized to be unique, so initialize the matrix to
relational_rnn_models.py:132
↓ 3 callersMethodadd_corpus
Tokenizes a text file.
data.py:46
↓ 3 callersFunctionbatchify
(data, bsz)
train_rnn.py:113
↓ 3 callersFunctionbatchify
(data, bsz)
train_rmc.py:128
↓ 3 callersMethodtokenize
(self, path, tokens)
data.py:60
↓ 2 callersFunctionaccuracy_score
(y_pred, y_true)
train_nth_farthest.py:188
↓ 2 callersFunctionevaluate
(data_source)
train_rnn.py:230
↓ 2 callersFunctionevaluate
(data_source)
train_rmc.py:230
↓ 2 callersFunctionget_batch
(X, y, batch_num, device, batch_size=32, batch_first=True)
train_embeddings.py:104
↓ 2 callersFunctionget_batch
(source, i)
train_rnn.py:223
↓ 2 callersFunctionget_batch
(source, i)
train_rmc.py:223
↓ 2 callersFunctionget_examples
(num_examples, num_vectors, num_dims, device)
train_nth_farthest.py:87
↓ 2 callersFunctionone_hot_encode
(array, num_dims=8)
train_nth_farthest.py:58
↓ 2 callersFunctionrepackage_hidden
Wraps hidden states in new Tensors, to detach them from their history.
train_rnn.py:204
↓ 1 callersMethodadd_word
(self, word)
data.py:12
↓ 1 callersMethodattend_over_memory
Perform multiheaded attention over `memory`. Args: memory: Current relational memory. Returns:
relational_rnn_models.py:275
↓ 1 callersMethodattend_over_memory
Perform multiheaded attention over `memory`. Args: memory: Current relational memory. Returns:
relational_rnn_general.py:259
↓ 1 callersMethodcalculate_gate_size
Calculate the gate size from the gate_style. Returns: The per sample, per head parameter size of each gate.
relational_rnn_models.py:215
↓ 1 callersMethodcalculate_gate_size
Calculate the gate size from the gate_style. Returns: The per sample, per head parameter size of each gate.
relational_rnn_general.py:199
↓ 1 callersMethodcreate_gates
Create input and forget gates for this step using `inputs` and `memory`. Args: inputs: Tensor input. memory: The
relational_rnn_models.py:228
↓ 1 callersMethodcreate_gates
Create input and forget gates for this step using `inputs` and `memory`. Args: inputs: Tensor input. memory: The
relational_rnn_general.py:212
↓ 1 callersFunctionexport_onnx
(path, batch_size, seq_len)
train_rnn.py:308
↓ 1 callersFunctionexport_onnx
(path, batch_size, seq_len)
train_rmc.py:308
↓ 1 callersMethodforward_step
Forward step of the relational memory core. Args: inputs: Tensor input. memory: Memory output from the previous t
relational_rnn_models.py:298
↓ 1 callersMethodforward_step
Forward step of the relational memory core. Args: inputs: Tensor input. memory: Memory output from the previous t
relational_rnn_general.py:282
↓ 1 callersFunctionget_batch
(X, y, batch_num, batch_size=32, batch_first=True)
train_nth_farthest.py:163
↓ 1 callersFunctionget_example
(num_vectors, num_dims)
train_nth_farthest.py:65
↓ 1 callersMethodinit_weights
(self)
rnn_models.py:64
↓ 1 callersMethodmultihead_attention
Perform multi-head attention from 'Attention is All You Need'. Implementation of the attention mechanism from https://arxiv.o
relational_rnn_models.py:160
↓ 1 callersMethodmultihead_attention
Perform multi-head attention from 'Attention is All You Need'. Implementation of the attention mechanism from https://arxiv.o
relational_rnn_general.py:144
↓ 1 callersMethodrepackage_hidden
Wraps hidden states in new Tensors, to detach them from their history.
relational_rnn_models.py:124
↓ 1 callersMethodrepackage_hidden
Wraps hidden states in new Tensors, to detach them from their history.
relational_rnn_general.py:108
↓ 1 callersFunctiontrain
()
train_rnn.py:249
↓ 1 callersFunctiontrain
()
train_rmc.py:251
Method__init__
(self)
train_embeddings.py:56
Method__init__
(self, batch_size)
train_embeddings.py:80
Method__init__
(self)
train_nth_farthest.py:105
Method__init__
(self, mlp_size)
train_nth_farthest.py:126
Method__init__
(self, mem_slots, head_size, input_size, num_tokens, num_heads=1, num_blocks=1, forget_bias=1.,
relational_rnn_models.py:38
Method__init__
(self, rnn_type, ntoken, ninp, nhid, nlayers, dropout=0.5, tie_weights=False, use_cudnn_version=True,
rnn_models.py:8
Method__init__
(self)
data.py:7
Method__init__
(self, path)
data.py:26
Method__init__
(self, mem_slots, head_size, input_size, num_heads=1, num_blocks=1, forget_bias=1., input_bias=0.,
relational_rnn_general.py:43
Method__len__
(self)
data.py:21
Functionaccuracy_score
(y_pred, y_true)
train_embeddings.py:126
Methodforward
(self, input, memory)
train_embeddings.py:91
Methodforward
(self, input, memory)
train_nth_farthest.py:148
Methodforward
(self, inputs, memory, targets, require_logits=False)
relational_rnn_models.py:356
Methodforward
(self, input, hidden)
rnn_models.py:71
Methodforward
(self, inputs, memory)
relational_rnn_general.py:327
Methodinitial_state
Creates the initial memory. We should ensure each row of the memory is initialized to be unique, so initialize the matrix to
relational_rnn_general.py:116
Methodoutput_size
(self)
relational_rnn_models.py:212
Methodoutput_size
(self)
relational_rnn_general.py:196
Methodstate_size
(self)
relational_rnn_models.py:208
Methodstate_size
(self)
relational_rnn_general.py:192