np.random.seed(1) torch.manual_seed(1) # In[5]: # Load data data_dir = '../../new_data_folder/' kfold_number = 0 logDir = 'models_and_stats/' model_name = 'BT19_ae_{}_rm_{}_wrI_{}_wC_{}_{}'.format(data_reduction_ratio, removal, w_r, w_c, str(kfold_number)) device = torch.device("cuda:{}".format(args.cuda)) print("Loading data...") train_loader, val_loader, train_dataset, val_dataset = get_trainValLoader(data_dir, k=kfold_number, spike_ready=False, batch_size=batch_size, shuffle=shuffle) test_loader, test_dataset = get_testLoader(data_dir, spike_ready=False, batch_size=batch_size, shuffle=shuffle) # In[6]: # Create model model = VRAEC(num_class=num_class, sequence_length=sequence_length, number_of_features = number_of_features, hidden_size = hidden_size, hidden_layer_depth = hidden_layer_depth, latent_length = latent_length, batch_size = batch_size, learning_rate = learning_rate,
data_dir = '../../new_data_folder/' kfold_number = args.kfold num_class = 20 learning_rate = 0.0001 num_epochs = 5000 hidden_size = 40 num_layers = 1 dropout = 0.2 logDir = 'models_and_stat/' model_name = 'cnn_lstm_icub_' + str(kfold_number) device = torch.device("cuda:{}".format(args.cuda)) train_loader, val_loader, train_dataset, val_dataset = get_trainValLoader( data_dir, k=kfold_number, spike_ready=False) test_loader, test_dataset = get_testLoader(data_dir, spike_ready=False) # ### set module # In[5]: # define NN models class CNN(nn.Module): def __init__(self): super(CNN, self).__init__() self.conv1 = nn.Conv2d(in_channels=1, out_channels=3, kernel_size=(3, 5))