ComfyUI Node

NNT Define Dense Layer

A ComfyUI node in NNT Neural Network Toolkit/Layers with 17 inputs and 2 outputs.

By inventorado·Created 2 years ago·Updated 2 years ago· 69
NNT Define Dense Layer
  • LAYER_STACK
  • LAYER_STACK
  • num_nodes
num_nodes64
activation_functionReLU
use_biasTrue
weight_initkaiming_normal
weight_init_gain1.00
weight_init_modefan_in
weight_init_nonlinearityrelu
bias_initzeros
bias_init_value0.00
normalizationNone
norm_eps0.0000
norm_momentum0.100
norm_affineTrue
dropout_rate0.0
alpha0.01
num_copies1
CategoryNNT Neural Network Toolkit/Layers

Inputs (17)

NameTypeDefaultDescription
num_nodesINT641–8192
activation_functionCOMBOReLU28 options: None, ELU, GELU, GLU, Hardshrink, Hardsigmoid, +22
use_biasCOMBOTrue2 options: True, False
weight_initCOMBOkaiming_normal12 options: default, normal, uniform, xavier_normal, xavier_uniform, kaiming_normal, +6
weight_init_gainFLOAT1.000.01–10
weight_init_modeCOMBOfan_in2 options: fan_in, fan_out
weight_init_nonlinearityCOMBOrelu6 options: relu, leaky_relu, selu, tanh, linear, sigmoid
bias_initCOMBOzeros5 options: default, zeros, ones, normal, uniform
bias_init_valueFLOAT0.00-1–1
normalizationCOMBONone6 options: None, BatchNorm, LayerNorm, InstanceNorm, GroupNorm, LocalResponseNorm
norm_epsFLOAT0.00001e-12–0.001
norm_momentumFLOAT0.1000.001–0.999
norm_affineCOMBOTrue2 options: True, False
dropout_rateFLOAT0.00–0.9
alphaFLOAT0.010–1
num_copiesINT11–100
LAYER_STACKoptLIST

Outputs (2)

NameTypeDescription
LAYER_STACKLIST
num_nodesINT