Use fan-in(number of input weight for each neuron) for scaling. More...
#include <weight_init.h>
Inheritance diagram for tiny_dnn::weight_init::lecun:
Collaboration diagram for tiny_dnn::weight_init::lecun:
Public Member Functions | |
| lecun (float_t value) | |
| void | fill (vec_t *weight, serial_size_t fan_in, serial_size_t fan_out) override |
 Public Member Functions inherited from tiny_dnn::weight_init::scalable | |
| scalable (float_t value) | |
| void | scale (float_t value) |
Additional Inherited Members | |
| Protected Attributes inherited from tiny_dnn::weight_init::scalable | |
| float_t | scale_ |
Detailed Description
Use fan-in(number of input weight for each neuron) for scaling.
Y LeCun, L Bottou, G B Orr, and K Muller, Efficient backprop Neural Networks, Tricks of the Trade, Springer, 1998
Member Function Documentation
◆ fill()
|
inlineoverridevirtual |
Implements tiny_dnn::weight_init::function.
The documentation for this class was generated from the following file:
- tiny_dnn/util/weight_init.h
