ComfyUI Node
NNT Define Conv Layer
A ComfyUI node in NNT Neural Network Toolkit/Layers with 23 inputs and 1 output.
NNT Define Conv Layer
- LAYER_STACK
- hyperparameters
- layer_stack
◄conv_typeConv2d►
◄out_channels64►
◄kernel_size3►
◄stride1►
◄padding1►
◄padding_modezeros►
◄output_padding0►
◄dilation1►
◄groups1►
◄use_biasTrue►
◄activation_functionReLU►
◄normalizationBatchNorm►
◄norm_eps0.0000►
◄norm_momentum0.100►
◄norm_affineTrue►
◄dropout_rate0.0►
◄weight_initkaiming_normal►
◄weight_init_gain1.41►
◄weight_init_modefan_out►
◄weight_init_nonlinearityrelu►
◄num_copies1►
CategoryNNT Neural Network Toolkit/Layers
Inputs (23)
| Name | Type | Default | Description |
|---|---|---|---|
| conv_type | COMBO | Conv2d | 8 options: Conv1d, Conv2d, Conv3d, ConvTranspose1d, ConvTranspose2d, ConvTranspose3d, +2 |
| out_channels | INT | 641–2048 | — |
| kernel_size | INT | 31–15 | — |
| stride | INT | 11–8 | — |
| padding | INT | 10–10 | — |
| padding_mode | COMBO | zeros | 4 options: zeros, reflect, replicate, circular |
| output_padding | INT | 00–2 | — |
| dilation | INT | 11–5 | — |
| groups | INT | 11–2048 | — |
| use_bias | COMBO | True | 2 options: True, False |
| activation_function | COMBO | ReLU | 28 options: None, ELU, GELU, GLU, Hardshrink, Hardsigmoid, +22 |
| normalization | COMBO | BatchNorm | 10 options: None, BatchNorm1d, BatchNorm2d, BatchNorm3d, LayerNorm, InstanceNorm1d, +4 |
| norm_eps | FLOAT | 0.00001e-12–0.001 | — |
| norm_momentum | FLOAT | 0.1000.001–0.999 | — |
| norm_affine | COMBO | True | 2 options: True, False |
| dropout_rate | FLOAT | 0.00–0.9 | — |
| weight_init | COMBO | kaiming_normal | 12 options: default, normal, uniform, xavier_normal, xavier_uniform, kaiming_normal, +6 |
| weight_init_gain | FLOAT | 1.410.01–10 | — |
| weight_init_mode | COMBO | fan_out | 2 options: fan_in, fan_out |
| weight_init_nonlinearity | COMBO | relu | 6 options: relu, leaky_relu, selu, tanh, linear, sigmoid |
| num_copies | INT | 11–100 | — |
| LAYER_STACKopt | LIST | — | |
| hyperparametersopt | DICT | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| layer_stack | LIST | — |