AIfES 2 2.0.0
aifes_express_q7_fnn.h
Go to the documentation of this file.
1
29#ifndef EXPRESS_FNN_Q7
30#define EXPRESS_FNN_Q7
31
33
42uint32_t AIFES_E_flat_weights_number_fnn_q7(uint32_t *fnn_structure, uint32_t layer_count);
43
44
94int8_t AIFES_E_quantisation_fnn_f32_to_q7(aitensor_t *representative_f32_dataset,
96 uint8_t *q7_parameter_dataset);
97
98
168 aitensor_t *output_tensor);
169
170#endif // EXPRESS_FNN_Q7
AIfES Express functions for weights with F32 (float32) data type.
int8_t AIFES_E_quantisation_fnn_f32_to_q7(aitensor_t *representative_f32_dataset, AIFES_E_model_parameter_fnn_f32 *AIFES_E_fnn, uint8_t *q7_parameter_dataset)
Quantizes the weights of an F32 FNN into a Q7 FNN.
int8_t AIFES_E_inference_fnn_q7(aitensor_t *input_tensor, AIFES_E_model_parameter_fnn_f32 *AIFES_E_fnn, aitensor_t *output_tensor)
Executes the inference of a Q7 FNN.
uint32_t AIFES_E_flat_weights_number_fnn_q7(uint32_t *fnn_structure, uint32_t layer_count)
Calculates the required length of the uint8_t array for the FNN.
Parameters for the FNN model.
Definition: aifes_express_f32_fnn.h:110
A tensor in AIfES.
Definition: aifes_math.h:92