MultiTierLSTMLayer class

A Multi-Tier Long Short-Term Memory (MT-LSTM) layer.

This is a generalized, hierarchical recurrent layer designed to capture dependencies across an arbitrary number of configured timescales.

Inheritance

Constructors

b_c_tiers ↔ List<Tensor<Vector>>: getter/setter pair
b_f_tiers ↔ List<Tensor<Vector>>: getter/setter pair
b_i_tiers ↔ List<Tensor<Vector>>: getter/setter pair
b_o_tiers ↔ List<Tensor<Vector>>: getter/setter pair
cumulativeClockCycles ↔ List<int>: getter/setter pair
hashCode → int: The hash code for this object.
no setterinherited
hiddenSize → int: final
name ↔ String: A user-friendly name for the layer (e.g., 'dense', 'lstm').
getter/setter pairoverride-getter
numTiers → int: final
parameters → List<Tensor>: A list of all trainable tensors (weights and biases) in the layer.
no setteroverride
runtimeType → Type: A representation of the runtime type of the object.
no setterinherited
tierClockCycles → List<int>: Defines the clock speed for each tier relative to the one below it. Example: 7, 4 means tier 1 updates every 7 tier-0 steps, and tier 2 updates every 4 tier-1 steps.
final
W_c_tiers ↔ List<Tensor<Matrix>>: getter/setter pair
W_f_tiers ↔ List<Tensor<Matrix>>: getter/setter pair
W_i_tiers ↔ List<Tensor<Matrix>>: getter/setter pair
W_o_tiers ↔ List<Tensor<Matrix>>: getter/setter pair

build(Tensor input) → void: Initializes the layer's parameters based on the shape of the first input.
override
call(Tensor input) → Tensor: The public, callable interface for the layer.
inherited
concatenateAll(List<Tensor<Vector>> tensors) → Tensor<Vector>
forward(Tensor input) → Tensor<Vector>: The core logic of the layer's transformation.
override
noSuchMethod(Invocation invocation) → dynamic: Invoked when a nonexistent method or property is accessed.
inherited
toString() → String: A string representation of this object.
inherited