Credits & Attribution

Everything in this curriculum uses free, open-source tools and properly licensed content. This page tracks what's used and where.

Libraries & CDN resources

All libraries are loaded via CDN and imported only as needed. No accounts, no API keys, no build step.

KaTeX Math rendering in all lesson pages — $x^2$, $\sum$, matrices, etc. MIT

Canvas API Built-in HTML5 — used for matrix transform, gradient descent, CLT, and distribution visuals Web standard

Jupyter Notebook format for code-heavy exercises (Module 1 gradient descent, and later modules) BSD-3

Python ecosystem (used in exercises)

NumPy Numerical arrays and linear algebra primitives BSD-3

scikit-learn Classical ML workhorse — used heavily from Module 3 onward BSD-3

matplotlib Plotting in notebooks matplotlib (BSD-style)

pandas DataFrame manipulation BSD-3

PyTorch Deep learning (Module 5 onward) BSD-3

Hugging Face Transformers Transformers, pretrained models, tokenizers (Module 6) Apache 2.0

External media

No external images, videos, or audio are embedded in the current lessons. All visualizations are generated live on the HTML canvas. If future lessons embed external media (e.g., CC-licensed diagrams from Wikimedia), they will be credited here with creator, source, and license.

Datasets (used in exercises)

Planned datasets for Module 3 onward:

California Housing — scikit-learn's built-in regression dataset (CC0 equivalent, sourced from the StatLib repository at CMU).
MNIST — public domain, hosted by Yann LeCun's lab.
Tiny Shakespeare — public domain text used for language-model training demos.
User-chosen data for capstones — bring your own from work or personal interests.

Inspiration

Curriculum shape draws on standard ML engineering tracks — Andrew Ng's course sequencing, fast.ai's top-down approach, the "make a GPT" Karpathy lectures, and the scikit-learn / PyTorch official tutorials. No content is reproduced; the debt is stylistic.

If you spot missing attribution or a license concern, note it in progress/journal.md and flag it in the next tutoring session.