I want to add a new label to my multilabel classification task using avalanche python library

Question

import torch
import torch.nn as nn
import torchvision
import torchvision.transforms as transforms
from avalanche.benchmarks import NCScenario
from avalanche.training.supervised import EWC

train_dataset = torchvision.datasets.ImageFolder(root='data/train', transform=transforms.ToTensor())
test_dataset = torchvision.datasets.ImageFolder(root='data/val', transform=transforms.ToTensor())


class CustomResNet(nn.Module):
    def __init__(self, num_classes):
        super(CustomResNet, self).__init__()
        self.resnet = torchvision.models.resnet18(pretrained=True)
        self.resnet.fc = nn.Linear(self.resnet.fc.in_features, num_classes)

    def forward(self, x):
        return self.resnet(x)


model = CustomResNet(num_classes=len(train_dataset.classes))

optimizer = torch.optim.SGD(model.parameters(), lr=0.001, momentum=0.9)
criterion = nn.CrossEntropyLoss()

scenario = NCScenario(train_dataset, test_dataset, n_experiences=len(train_dataset.classes), task_labels=True)

ewc = EWC(model, optimizer, criterion, ewc_lambda=0.1)

for train_task in scenario.train_stream:
    ewc.train(train_task, epochs=5)

I wrote this code, but i don't understand in this avalanche library how can i instead of training each of my labels incrementally, download my model with 34 labels and add just one new label so i want to have 2 experiences.

I also tried to solve this problem by using PyCIL library but i also don't understand this part in there.