anypinn.cli.scaffold.fitzhugh_nagumo.config_csv

FitzHugh-Nagumo neuron model — training configuration.

EXPERIMENT_NAME = '__EXPERIMENT_NAME__' module-attribute

RUN_NAME = 'v0' module-attribute

hp = ODEHyperparameters(lr=1.0, max_epochs=500, criterion='mse', optimizer=(LBFGSConfig(lr=1.0, max_iter=20, history_size=100)), training_data=(IngestionConfig(batch_size=300, data_ratio=2, collocations=5000, collocation_sampler='latin_hypercube', df_path=(Path('./data/data.csv')), y_columns=['v'])), fields_config=(MLPConfig(in_dim=1, out_dim=1, hidden_layers=[64, 128, 128, 64], activation='tanh', output_activation=None)), params_config=(ScalarConfig(init_value=0.3)), smma_stopping=(SMMAStoppingConfig(window=50, threshold=1e-06, lookback=100)), pde_weight=0.001, ic_weight=10, data_weight=5) module-attribute