Braindecode Model Explorer

braindecode model explorer

v1.5.0dev0•56 architectures•no weights

Architecture

Signal configuration

n_chans

sfreq · Hz

window · seconds

n_outputs

Parameters

—

press build to instantiate

EEGNetv4

Deprecated alias for EEGNet.

EEGNetv4(self, n_chans: 'Optional[int]' = None, n_outputs: 'Optional[int]' = None, n_times: 'Optional[int]' = None, final_conv_length: 'str | int' = 'auto', pool_mode: 'str' = 'mean', F1: 'int' = 8, D: 'int' = 2, F2: 'Optional[int | None]' = None, kernel_length: 'int' = 64, *, depthwise_kernel_length: 'int' = 16, pool1_kernel_size: 'int' = 4, pool2_kernel_size: 'int' = 8, conv_spatial_max_norm: 'int' = 1, activation: 'type[nn.Module]' = <class 'torch.nn.modules.activation.ELU'>, batch_norm_momentum: 'float' = 0.01, batch_norm_affine: 'bool' = True, batch_norm_eps: 'float' = 0.001, drop_prob: 'float' = 0.25, final_layer_with_constraint: 'bool' = False, norm_rate: 'float' = 0.25, chs_info: 'Optional[list[Dict]]' = None, input_window_seconds=None, sfreq=None, **kwargs)

↗ Source on GitHub

Architecture documentation

Deprecated alias for EEGNet.

Hugging Face Hub integration

When the optional huggingface_hub package is installed, all models automatically gain the ability to be pushed to and loaded from the Hugging Face Hub. Install with:
pip install braindecode[hub]
Pushing a model to the Hub:

Loading a model from the Hub:

Extracting features and replacing the head:

Saving and restoring full configuration:

All model parameters (both EEG-specific and model-specific such as dropout rates, activation functions, number of filters) are automatically saved to the Hub and restored when loading.

See :ref:`load-pretrained-models` for a complete tutorial.