How to plot training and test losses in the same plot using Tensorboard?

Question

I tried to plot two losses in the same plot on Tensorboard, but I can still see just one loss values on "Training loss" plot.

def train(epoch):
    model.train()
    
    for data in train_loader:
        data = data.to(device)
        optimizer.zero_grad()
        recon = data.x.to(torch.float32)
        pred,x_latent,_,_ = model(data)
        loss = F.mse_loss(pred, recon.view(-1))
        writer.add_scalar("Training loss", loss, epoch)
        losses.append(loss)
        loss.backward()
        optimizer.step()
        

def test():
    model.eval()
    correct = 0

    for data in test_loader:
        data = data.to(device)
        recon = data.x.to(torch.float32)
        pred,x_latent,_,_ = model(data)
        test_loss = F.mse_loss(pred, recon.view(-1))
        writer.add_scalar("Training loss", test_loss, epoch)
    return test_loss

How could I plot them together in the same plot?

Does using a `/` as per the end of [this answer](https://stackoverflow.com/a/57599771/2550406) work? — lucidbrot, Jul 09 '21 at 17:37
Wonderful, then I'll vote to close this question so future people directly see the link to the answer — lucidbrot, Jul 09 '21 at 17:52
@lucidbrot Yes, I could do it by using two separate `writers`. — mario119, Jul 09 '21 at 18:37

How to plot training and test losses in the same plot using Tensorboard?

0 Answers0