No loss

[angelusualle]

Information

The problem arises in chapter:

• Introduction
• [ x] Text Classification
• Transformer Anatomy
• Multilingual Named Entity Recognition
• Text Generation
• Summarization
• Question Answering
• Making Transformers Efficient in Production
• Dealing with Few to No Labels
• Training Transformers from Scratch
• Future Directions

Describe the bug

I ran the

trainer = Trainer(model=model, args=training_args,
                  compute_metrics=compute_metrics,
                  train_dataset=emotions_encoded["train"],
                  eval_dataset=emotions_encoded["validation"],
                  tokenizer=tokenizer)
trainer.train();

section, and got an error complaining about no loss in the outputs:

The model did not return a loss from the inputs, only the following keys: logits. For reference, the inputs it received are input_ids,attention_mask.

Turns out AutoModelForSequenceClassification was expecting labels, not label. so I relabeled:

def relabel(batch):
    return {'labels': batch["label"]}
emotions_encoded = emotions_encoded.map(relabel)
emotions_encoded.column_names