package kaun

  1. Overview
  2. Docs
Flax-inspired neural network library for OCaml

Install

dune-project
 Dependency

Authors

Maintainers

Sources

raven-1.0.0.alpha2.tbz
sha256=93abc49d075a1754442ccf495645bc4fdc83e4c66391ec8aca8fa15d2b4f44d2
sha512=5eb958c51f30ae46abded4c96f48d1825f79c7ce03f975f9a6237cdfed0d62c0b4a0774296694def391573d849d1f869919c49008acffca95946b818ad325f6f

doc/src/kaun/activations.ml.html

Source file activations.ml

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
open Rune

(* Standard Activations *)

let relu = relu
let relu6 = relu6
let sigmoid = sigmoid
let tanh = tanh
let softmax = softmax

(* Modern Activations *)

let gelu = gelu
let silu = silu
let swish = swish
let mish = mish

(* Parametric Activations *)

let leaky_relu = leaky_relu
let elu = elu
let selu = selu
let prelu = prelu

(* Gated Linear Units (GLUs) *)

let glu x gate =
  (* x * sigmoid(gate) *)
  mul x (sigmoid gate)

let swiglu x =
  (* x * silu(x) *)
  mul x (silu x)

let geglu x gate =
  (* x * gelu(gate) *)
  mul x (gelu gate)

let reglu x gate =
  (* x * relu(gate) *)
  mul x (relu gate)

(* Other Activations *)

let softplus = softplus
let softsign = softsign
let hard_sigmoid = hard_sigmoid
let hard_tanh = hard_tanh
let hard_swish = hard_swish