Skip to content
← All tracks

Recurrent Networks

PyTorch

RNN, LSTM, GRU, bidirectional — the architectures that ruled NLP before transformers.

0 / 6 solved
Continue →
  1. 1. Not solved yet. Simple RNN Cell
  2. 2. Not solved yet. LSTM Cell
  3. 3. Not solved yet. GRU Cell
  4. 4. Not solved yet. Bidirectional RNN
  5. 5. Not solved yet. Sequence Classifier
  6. 6. Not solved yet. Train RNN Language Model

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

Backpropagating through 20 timesteps multiplies by a recurrent factor at each step. What happens for factors 0.5, 1.0 and 1.5? 

g = 1.0
for _ in range(20):
    g *= w
Question 1 of 4