Skip to content

Latest commit

 

History

History
34 lines (25 loc) · 1.05 KB

File metadata and controls

34 lines (25 loc) · 1.05 KB

Notebooks

Interactive Jupyter notebooks demonstrating the LLM framework capabilities.

Available Notebooks

Notebook Description
quick_start.ipynb Complete tutorial: model building, training, inference, advanced features

Quick Start

# Install Jupyter if needed
uv pip install jupyter

# Launch Jupyter
cd notebooks
jupyter notebook

Topics Covered

quick_start.ipynb

  1. Building a Model - Create a Decoder-only Transformer
  2. Training - Train on synthetic data with loss visualization
  3. Inference - Text generation (streaming and non-streaming)
  4. Advanced Features:
    • Gradient Checkpointing (memory-efficient training)
    • Grouped Query Attention (GQA)
    • SwiGLU Activation
    • Mixture of Experts (MoE)
  5. E2E Pipeline - Full train → evaluate → inference workflow