ValueError: not enough values to unpack (expected 2, got 1)

Beginners
1

i am trying to create xlnet classification

  def __init__(self,n_classes):
    super(SentimentClassifier, self).__init__()
    self.xlnet = XLNetModel.from_pretrained(PRE_TRAINED_MODEL_NAME)
    self.drop = nn.Dropout(p=0.3)
    self.out = nn.Linear(self.xlnet.config.hidden_size,  n_classes)
  def forward(self, input_ids, 
              attention_mask):
    _, pooled_output = self.xlnet(
      input_ids=input_ids,
      attention_mask=attention_mask)
    output = self.drop(pooled_output)
    return self.out(output)
class Classification(Dataset):


    def __init__(self, texts, labels, tokenizer, max_len):
        self.texts = texts
        self.labels = labels
        self.tokenizer = tokenizer
        self.max_len = max_len
    
    def __len__(self):
        return len(self.texts)
    
    def __getitem__(self, item):
        text = str(self.texts[item])
        label = self.labels[item]

        encoding = self.tokenizer.encode_plus(
        text,
        add_special_tokens=True,
        max_length=self.max_len,
        return_token_type_ids=False,
        pad_to_max_length=False,
        return_attention_mask=True,
        return_tensors='pt',
        )

        

        return {
        'review_text': text,
        'input_ids': encoding['input_ids'].flatten(),
        'attention_mask': encoding['attention_mask'].flatten(),
         'labelss': torch.tensor(label, dtype=torch.long)
        }
def train_epoch(
  model,
  data_loader,
  loss_fn,
  optimizer,
  device,
  scheduler,
  n_examples
):
  model = xlnet_model.train()
  losses = []
  correct_predictions = 0
  for d in data_loader:
    input_ids = d["input_ids"].reshape(4,512).to(device)
    print(d['input_ids'].shape)
    attention_mask = d["attention_mask"].to(device)
    labels = d["labels"].to(device)
    outputs = xlnet_model(input_ids=input_ids, attention_mask=attention_mask)
    
    _, preds = torch.max(outputs, dim=1)
    loss = loss_fn(outputs, labels)
    correct_predictions += torch.sum(preds == labels)
    losses.append(loss.item())
    loss.backward()
    nn.utils.clip_grad_norm_(model.parameters(), max_norm=1.0)
    optimizer.step()
    scheduler.step()
    optimizer.zero_grad()
  return correct_predictions.double() / n_examples, np.mean(loss)```
2

Hi @sru,
Can you please include the stack trace so we can help out more?

1 Like
3

thanks @FL33TW00D
stack trace:
ValueErrorTraceback (most recent call last)

in ()
6 device,
7 scheduler,
----> 8 len(df_train)
9 )
10 print(f’Train loss {train_loss} accuracy {train_acc}’)

2 frames

in forward(self, input_ids, attention_mask)
9 _, pooled_output = self.xlnet(
10 input_ids=input_ids,
—> 11 attention_mask=attention_mask)
12 output = self.drop(pooled_output)
13 return self.out(output)

ValueError: not enough values to unpack (expected 2, got 1)

4

Please someone can suggest me plz

5

@valhalla could you please help me with issue

6

Hi @sru

By default, all models now return output as a dict, but your code expecting a tuple which is the reason for above error. You could access the hidden_states in the returned object. See the docs for model output classes

https://huggingface.co/transformers/main_classes/output.html

Also, XLNetModel does return pooled output, you would need to do the pooling yourself, or you can just use XLNetForSequenceClassification, which does the pooling and passes that pooled output through a classification head. You could refer to how XLNetForSequenceClassification is implemented.

7

@valhalla thank you, i will check once.

8

The error message is fairly self-explanatory. Your program expects python split() to yield 2 elements, but in your case, it is only yielding 1 element. This could be because the data is not in the format you expect, a rogue malformed line, or maybe an empty line - there’s no way to know.

To see what line is causing the issue, you could add some debug statements like this:

if len(line.split()) != 2:
    print line