CLarification in dataset_benchmark and generic scenario

[MagicHealer]

Hi, Can you please verify if I used dataset_benchmark in a right way? If so, why does it takes so long to train even for just 1 epoch? I am using GPU. ! epoch finishes after about 2 hrs. Thank you

CODE:
from torch.optim import SGD
from torch.nn import CrossEntropyLoss
from avalanche.benchmarks.datasets import MNIST, CIFAR10
from avalanche.evaluation.metrics import forgetting_metrics, accuracy_metrics,
loss_metrics, timing_metrics, cpu_usage_metrics, confusion_matrix_metrics, disk_usage_metrics
from avalanche.models import SimpleMLP, SimpleCNN
from avalanche.logging import InteractiveLogger, TextLogger, TensorboardLogger, WandBLogger
from avalanche.training.plugins import EvaluationPlugin
from avalanche.training.supervised import Naive
from avalanche.benchmarks.generators import nc_benchmark, ni_benchmark

from torchvision.transforms import Compose, ToTensor, Normalize, RandomCrop, Resize
train_transform = Compose([
#Resize(224),
lambda x: x.convert("RGB"),
RandomCrop(28, padding=4),
ToTensor(),
Normalize((0.1307,), (0.3081,))
])

test_transform = Compose([
#Resize(224),
lambda x: x.convert("RGB"),
RandomCrop(28, padding=4),
ToTensor(),
Normalize((0.1307,), (0.3081,))
])

mnist_train = MNIST(
'./data/mnist', train=True, download=True, transform=train_transform
)
mnist_test = MNIST(
'./data/mnist', train=False, download=True, transform=test_transform
)

train_transform1 = Compose([
#Resize(224),
RandomCrop(28, padding=4),
ToTensor(),
Normalize((0.1307,), (0.3081,))
])

test_transform1 = Compose([
#Resize(224),
RandomCrop(28, padding=4),
ToTensor(),
Normalize((0.1307,), (0.3081,))
])
train_cifar10 = CIFAR10(
'./data/cifar10', train=True, download=True, transform= train_transform1
)
test_cifar10 = CIFAR10(
'./data/cifar10', train=False, download=True, transform=test_transform1
)

from avalanche.benchmarks.generators import filelist_benchmark, dataset_benchmark, tensors_benchmark, paths_benchmark

generic_scenario = dataset_benchmark(
[mnist_train, train_cifar10],
[mnist_test, test_cifar10]
)

import torch

device = torch.device(
f"cuda:2"
if torch.cuda.is_available()
else "cpu"
)
print(device)

import torchvision
model=torchvision.models.vgg16(num_classes=10)
print(model)

tb_logger = TensorboardLogger()
text_logger = TextLogger(open('log.txt', 'a'))
interactive_logger = InteractiveLogger()
wandb_logger=WandBLogger()

eval_plugin = EvaluationPlugin(
accuracy_metrics(minibatch=True, epoch=True, experience=True, stream=True),
loss_metrics(minibatch=True, epoch=True, experience=True, stream=True),
timing_metrics(epoch=True, epoch_running=True),
forgetting_metrics(experience=True, stream=True),
cpu_usage_metrics(experience=True),
disk_usage_metrics(minibatch=True, epoch=True, experience=True, stream=True),
loggers=[interactive_logger, text_logger, tb_logger]#,wandb_logger]
)

criterion = CrossEntropyLoss()
optimizer = SGD(model.parameters(), lr=0.01)

cl_strategy = Naive(
model,
optimizer,
criterion,
device=device,
train_mb_size=350,
train_epochs=10, eval_mb_size=128,
evaluator=eval_plugin)

print("Starting experiment...")
results = []
for experience in generic_scenario.train_stream:
print("Start of Training for Task ", experience.current_experience)
cl_strategy.train(experience)
print("Training completed")
print("Computing accuracy on the whole test set")
results.append(cl_strategy.eval(generic_scenario.test_stream))

[AntonioCarta]

The script looks ok. Are you sure that you are using the GPU?

[MagicHealer]

Yes, I am using GPU. I also have another concern. What if my two datasets for two tasks don't have the same number of classes?