Struct cmsis_nn_gru_params_f16¶

Defined in File arm_nn_types_flt.h

Struct Documentation¶

struct cmsis_nn_gru_params_f16¶

Parameters for a float16 unidirectional GRU invocation.

GRU has three gates (update, reset, candidate) and, unlike LSTM, no cell state. The hidden state is the layer output.

Public Members

int32_t time_major¶: Non-zero when input/output tensors are time-major.

int32_t batch_size¶: Batch size processed per invocation.

int32_t time_steps¶: Number of time steps processed per invocation.

int32_t input_size¶: Input feature size per time step.

int32_t hidden_size¶: Hidden-state size.

int32_t reset_after¶: Non-zero: reset gate applied after the recurrent matmul (Keras/TFLite default).

cmsis_nn_gru_gate_f16 update_gate¶: Update gate (z), sigmoid activation.

cmsis_nn_gru_gate_f16 reset_gate¶: Reset gate (r), sigmoid activation.

cmsis_nn_gru_gate_f16 candidate_gate¶: Candidate/new gate (n), tanh activation.