what is the best practice to get a fabric instance?

[tanganke]

To train a deep neural network with lightning.Fabric, it's common to use the following template

import lightning as L


def train(fabric, model, optimizer, train_dataloader, val_dataloader):
    # Training loop with validation every few epochs
    model.train()
    for epoch in range(num_epochs):
        for i, batch in enumerate(train_dataloader):
            ...

        if epoch % validate_every_n_epoch == 0:
            validate(fabric, model, val_dataloader)


def validate(fabric, model, dataloader):
    # Validation loop
    model.eval()
    for i, batch in enumerate(dataloader):
        ...


def test(fabric, model, dataloader):
    # Test/Prediction loop
    model.eval()
    for i, batch in enumerate(dataloader):
        ...


def main():
    ...

    # Run training loop with validation
    train(fabric, model, optimizer, train_dataloader, val_dataloader)

    # Test on unseen data
    test(fabric, model, test_dataloader)


if __name__ == "__main__":
    main()

note that for every function call, we pass a fabric object in.
This means that for existing code, migrating from vanilla pytorch code to the fabric framework requires changing the signatures of all functions.
It's a lot of work and it's easy to make mistakes.

My question is:
Is there a singleton pattern in Fabric's design, or is it possible to get a list of launched fabric instances through a function?

for example:

fabric : L.Fabric = L.get_fabric_instance()

# or
list_of_fabric : List[L.Fabric] = L.get_fabric_instances()