Crate caffe2op_lstm

Structs

• CellParams
• FullBidirectionalLSTMLayer
• FullLSTMLayer
• GetLSTMUnitGradient
• InferenceLSTMOp
• LSTMCell
• LSTMUnitGradientOp

  ---

• LSTMUnitOp
  | LSTMUnit computes the activations | of a standard LSTM (without peephole | connections), in a sequence-length | aware fashion. | | Concretely, given the (fused) inputs | X (TxNxD), the previous cell state (NxD), | and the sequence lengths (N), computes | the LSTM activations, avoiding computation | if the input is invalid (as in, the value | at X{t][n] >= seqLengths[n]. |
• LayerOutput

Traits

• FullBidirectionalLSTMLayerTypes
• Layer

Functions

• add
• apply_layer_stack
• cat
• chunk
• gather_params
  | Parses a flat list of parameter tensors | into a list of CellParams |
• lSTMUnit
• linear
• lstm_impl
• lstm_unit_gradient
• matmul
• mul
• pair_vec
  | Gathers every two elements of a vector | in a vector of pairs |
• sigmoid
• stack
• tanh
• transpose
• unbind
• unpair_vec
  Flattens a vector of pairs

Type Definitions

• TensorTuple