🧠Deep Learning
📈
RNNs & Sequence Models
Learn recurrent architectures for sequential data like text and time series
Prerequisites
🌱
Beginner
BeginnerSequence modeling basics
What to Learn
- •Vanilla RNN architecture
- •Vanishing gradient problem
- •LSTM cells and gates
- •GRU as simplified LSTM
- •Sequence-to-sequence basics
Resources
- 📚Colah blog: Understanding LSTM
- 📚Stanford CS224n (early lectures)
- 📚PyTorch RNN tutorials
🌿
Intermediate
IntermediatePractical sequence modeling
What to Learn
- •Bidirectional RNNs
- •Attention mechanisms in RNNs
- •Encoder-decoder architectures
- •Beam search decoding
- •Time series forecasting with RNNs
Resources
- 📚Seq2Seq with Attention paper
- 📚Time series forecasting tutorials
- 📚Neural Machine Translation papers
🌳
Advanced
AdvancedModern sequence architectures
What to Learn
- •State space models (S4, Mamba)
- •xLSTM and modern recurrence
- •Linear attention mechanisms
- •Long-range dependencies
- •Hybrid architectures
Resources
- 📚Mamba paper
- 📚S4 and state space model papers
- 📚RWKV architecture paper