Micro Free Course — 100% Free Learning
All lessons in this module are free to learn. Sign in with Google to save your progress.
M3 — Sequences + NLP Baselines
RNN/sequence video, sentiment-baseline article, sequences quiz, NLP mini-lab submission.
Module Content
RNN/Sequence Concepts (LSTM/GRU, Masking, Embeddings)
Outcome: explain padding/masking on 2 examples. Curated video (16:37, verified 2026-09-24 via yt-dlp): Recurrent Neural Networks (RNNs), Clearly Explained!!! (StatQuest with Josh Starmer) Watch: https://www.youtube.com/watch?v=AsNTP8Kwu80 Notes: RNN anatomy, running sequences through the net, shared weights, vanishing/exploding gradients (motivation for LSTM/GRU). Lesson pairs this with padding/masking + embedding examples in the drill. NOTE: 16:37 runs just over the 15-min target but inside the 30-min cap; kept for breadth. Key moments: 0:00 Intro; 4:13 RNN anatomy; 5:59 Running data; 10:31 Shared weights; 11:23 Vanishing/exploding gradients.
Sentiment: BoW vs Frozen-Mini-Transformer/API Baseline
Acc/F1 + error slices. Label-leak trap via id.
Quiz: Sequences + Embeddings
10 scenario MCQs. Pass >= 80.
Mini-Lab B: Sentiment + Error Slices
Acc/F1 + slice report. Deterministic F1 bar + 100-pt rubric.