Skip to content
← All tracks

Metrics & Evaluation

PyTorch

The numbers that tell you whether your model is actually working — accuracy, F1, AUC, perplexity, BLEU.

0 / 7 solved
Continue →
  1. 1. Not solved yet. Accuracy – Multiclass
  2. 2. Not solved yet. Confusion Matrix
  3. 3. Not solved yet. Precision, Recall & F1
  4. 4. Not solved yet. AUC-ROC
  5. 5. Not solved yet. Expected Calibration Error
  6. 6. Not solved yet. Perplexity from Logits
  7. 7. Not solved yet. BLEU-1gram

Check yourself

4 questions · one attempt each

These do not count toward finishing the track. They are here to catch the things that are easy to read past.

0 / 4

One positive in a hundred examples. A model predicts the negative class every time. What does it score? 

y    = [1] + [0] * 99
pred = [0] * 100
Question 1 of 4