Markov: Build a Text Generator
Feed it a book and it writes new sentences in the same voice. You build the great-grandparent of every language model, in about fifteen lines.
What you'll be able to build
Feed it a book and it writes new sentences in the same voice. You build the great-grandparent of every language model, in about fifteen lines. Along the way you pick up real, transferable Ruby skills, not just this one project:
- a Hash with a default for transitions
- each_cons for pairwise iteration
- seeding random for reproducibility
- sample and weighted choice
- a train-then-generate workflow
- accumulating with each_with_object
A course like this one
Yours is built from your own placement, so module count and depth will differ. This map shows what a advanced-level Ruby learner building Markov actually gets.
- Module 1: Idiomatic Ruby: symbols, ranges, and string methods5 lessons
Builds the production-ready version of the script for your markov.
- Module 2: Blocks, procs, lambdas, and yield5 lessons
Builds the production-ready version of the reusable class for your markov.
- Module 3: Mixins, duck typing, and Comparable/Enumerable5 lessons
Builds the production-ready version of the collection pipeline for your markov.
- Module 4: Enumerable mastery: map/reduce/group_by/each_with_object5 lessons
Builds the production-ready version of the module workflow for your markov.
- Module 5: Truthiness, case/when, and the safe-navigation operator5 lessons
Builds the production-ready version of the method that powers your markov.
- Module 6: Error contracts, dependency-free packaging, documented output3 lessons
Builds the production-ready version of the command-line tool for your markov.
How the lessons actually work
Every lesson has you predict what a piece of Ruby code will output before you run it, then run it for real in your browser and fix what you got wrong. Each module ends in a challenge gate with hidden tests, so you can't advance until your code actually works. The course closes with a capstone that assembles everything into Markov, and a runnable proof page tied to your own code.
Common questions
How long does the Markov: Build a Text Generator course take?
about 7 hours, across 6 modules and 28 lessons, at roughly 15 minutes per lesson. Your own course may run shorter or longer, since it's sized to your placement result, not a fixed template.
Do I need experience?
Yes. This is an advanced-tier Ruby project, so it assumes you're already comfortable writing and reading Ruby before you start.
How much does it cost?
$15 one-time, no subscription. The first module is free, so you can see exactly how the course teaches before you pay for the rest.