Python 中的并行化

Question

I've got something I'd like to do in parallel in python. The following function:

def _train_users_locally(self):
        for i in self.users:
            self.users[i].train()

Gets a bunch of instances of the 'user' class to train a pytorch neural net:

def train(self):
        self.current_model.train()
        optim = self.optim(self.current_model.parameters(), lr=self.lr)
        criterion = self.criterion()
        for epoch in range(self.epochs):
            loss_per_epoch = 0
            counter = 0
            for i, data in enumerate(self.dataloader):
                x, y = data
                fx = self.current_model(x.unsqueeze(1))
                loss = criterion(fx, y)
                optim.zero_grad()
                loss.backward(retain_graph=True)
                optim.step()
                loss_per_epoch += loss.item()
                print('\rEpoch {}\tBatch: {:.3f}, Loss: {:.3f}'.format(epoch+1, i, loss.item()), end="")
                counter += 1
            print('\nEpoch {}\t Average Loss: {:.3f}'.format(epoch+1, loss_per_epoch / counter))

Nothing is being output; the model assigned to the object is being updated. I would like each user object to do its training simultaneously, but I can't for the life of me figure out how to do this, as all the examples I've been able to find involve some processing of list elements.

Answer 1

Figured something out, thanks Gautam.

from threading import Thread

def _train_users_locally(self):
        threads = []
        for i in self.users:
            t = Thread(target=self.users[i].train)
            threads.append(t)
        for t in threads:
            t.start()
        for t in threads:
            t.join()

It seems to work as intended; but I'm unsure if I might be missing something terrible that'll pop up unexpectedly.

Answer 2

Use the module called "threading"

it will run the func. simuntainously