Quick Start (30 min)

Run three short experiments to understand the most important LLM design choices.

Setup (5 min)

# 1. Clone the repo
git clone https://github.com/joyehuang/minimind-notes.git
cd minimind-notes

# 2. Activate your virtual environment
source venv/bin/activate

# 3. Download experiment datasets (optional)
cd modules/common
python datasets.py --download-all
cd ../..

Experiment 1: Normalization (10 min)

Observe gradient vanishing and see why Pre-LN + RMSNorm is stable.

cd modules/01-foundation/01-normalization/experiments
python exp1_gradient_vanishing.py

Next: /en/modules/01-foundation/01-normalization/

---

Experiment 2: RoPE (10 min)

Compare absolute position encoding and learn why RoPE extrapolates better.

cd ../../02-position-encoding/experiments
python exp1_rope_basics.py

Next: /en/modules/01-foundation/02-position-encoding/

---

Experiment 3: Attention (10 min)

Understand how Q/K/V and attention weights work.

cd ../../03-attention/experiments
python exp1_attention_basics.py

Next: /en/modules/01-foundation/03-attention/

---

Where to go next

• Systematic Study: /en/docs/guide/systematic
• Deep Mastery: /en/docs/guide/mastery
• Roadmap: /en/ROADMAP