Notes Library

Served by the Ramiyar MD Engine running on the Raspberry Pi (hramiyar), always on. Every .md file under this folder is rendered here — with KaTeX math, Mermaid, Chart.js and NeuralFlow diagrams — and reachable from anywhere on the Tailscale network.

Browse everything: the file browser lists every note found under the library, including anything you add later.

Contents

Start here

How to reach these notes, and how to add more

Machine Learning Notes — six notes, read in order

Start here: the reading path — the map, why that order, and a topic lookup
1 · Linear regression · 2 · Logistic regression
3 · Neural networks — one neuron to a trained network (Parts I–IV)
4 · Multiclass and the softmax · 5 · Developing a model · 6 · Decision trees
Appendices: backpropagation, step by step · forward/backward, entry by entry

Guides

Getting started — the engine in one page
Authoring template — the house style, every building block
NeuralFlow guide — the diagram language
NeuralFlow showcase
Feature showcase · Charts

Other

FFT and templates — see the file browser

Adding a note

Drop the .md file anywhere under ~/mdengine/notes/ on the Pi. It appears immediately — no restart, no rebuild. To add a note with its pictures, copy the whole folder:

~/mdengine/notes/
└── My Topic/
    ├── my-note.md
    └── images/
        └── figure.png

Inside my-note.md, reference the picture either way — both work:

![A figure](images/figure.png)          <!-- relative to the note -->
<img src="/images/figure.png">          <!-- the house style -->

The engine resolves images/ next to the note itself, so a folder copied in from anywhere keeps working exactly as it did where it came from.

From the Mac

mdpush ~/Documents/md-engine/"Machine Learning Notes"    # push a folder to the Pi library
mdpush ~/some/note.md                            # or a single file