can someone tell me why my cross entropy Loss function giving this error:
My accuracy method:
def accuracy(outputs, labels):
_, preds = torch.max(outputs, dim=1)
return torch.tensor(torch.sum(preds == labels).item() / len(preds))
My class defining the model:
class PulsarLogisticRegression(nn.Module):
def __init__(self):
super().__init__()
self.linear= nn.Linear(input_size,output_size)
def forward(self,xb):
xb = xb.view(xb.size(0), -1)
out= self.linear(xb)
return out
def training_step(self, batch):
inputs, targets = batch
# Generate predictions
out = self(inputs)
# Calcuate loss
loss = F.cross_entropy(out,targets)
return loss
def validation_step(self, batch):
inputs, targets = batch
# Generate predictions
out = self(inputs)
# Calculate loss
loss = F.cross_entropy(out,targets)
acc = accuracy(out, targets) # Calculate accuracy
return {'val_loss': loss, 'val_acc': acc} # fill this
def validation_epoch_end(self, outputs):
batch_losses = [x['val_loss'] for x in outputs]
epoch_loss = torch.stack(batch_losses).mean() # Combine losses
batch_accs = [x['val_acc'] for x in outputs]
epoch_acc = torch.stack(batch_accs).mean() # Combine accuracies
return {'val_loss': epoch_loss.item(), 'val_acc': epoch_acc.item()}
def epoch_end(self, epoch, result, num_epochs):
# Print result every 20th epoch
print("Epoch [{}], val_loss: {:.4f}, val_acc: {:.4f}".format(epoch, result['val_loss'], result['val_acc']))
error: Cross Entropy Error is expecting 1D target tensor but multi-target is bieng assigned to it.
RuntimeError Traceback (most recent call last)
<ipython-input-88-cd9b8a9a3b02> in <module>()
----> 1 result = evaluate(model, val_loader) # Use the the evaluate function
2 print(result) 4 frames
/usr/local/lib/python3.6/dist-packages/torch/nn/functional.py in nll_loss(input, target, weight, size_average, ignore_index, reduce, reduction)
2262 .format(input.size(0), target.size(0)))
2263 if dim == 2:
-> 2264 ret = torch._C._nn.nll_loss(input, target, weight, _Reduction.get_enum(reduction), ignore_index)
2265 elif dim == 4:
2266 ret = torch._C._nn.nll_loss2d(input, target, weight, _Reduction.get_enum(reduction), ignore_index)
RuntimeError: 1D target tensor expected, multi-target not supported
I am very new to Machine Learning, I am trying to make a model which predicts one column of data based on 5 columns. The values in the column are 0 and 1. So it is basically a binary classification model.
What I tried: As I said I am fairly new to this field, some explanations suggest to use squeeze function to somehow reduce the shape of target tensor to 1D but that seems to throw some other errors in other methods of the class.
I am looking for an error function which would help me get correct accuracy.
