Final-year students
Placement season is close and your resume still has no project on it.
Finish with three pieces of work you can demo in an interview.
AI Agents
Meta AI and Llama Mastery — Open Models You Can Run Yourself
Understand what open weights actually let you do, run a model on your own laptop, and build one small tool that would not have been possible through a hosted chat window.
- 5 weeksEstimated time
- 4Modules
- 15Lessons
- 3Projects
- YesCertificate
- FreePrice
No credit card. No sign-up required to preview. See the full syllabus
Who this course is for
Three people take this course. The work is the same; the reason is different.
Fresh graduates
You have the degree, but the job ads ask for proof of skill.
Build a portfolio before the next hiring round, not after it.
Career switchers
You are experienced elsewhere, new to this field, and starting from zero.
Learn the vocabulary and ship one published project that proves the switch.
Proof, not notes
What you will build
Every module ends in something finished. These are the pieces you leave with.
Build a script that walks
a folder, summarises every file through a local model, and writes one combined markdown file. It must run with the network switched off.
Skills used AI Agents
Portfolio project
A chunk-and-summarise pipeline for one
long report
Skills used AI Agents
Portfolio project
A comparison note between
local model and a hosted one, with numbers
Skills used AI Agents
Portfolio project
The brief
Your final project
A comparison note between
local model and a hosted one, with numbers
- Write down, in one sentence, what the finished thing does and who it is for.
- Build it with the tools this course taught. A borrowed project cannot be explained in an interview.
- Run it from a clean start on another machine, and fix whatever breaks.
- Publish it, and put the link at the top of your resume's project section.
What changes
What you will be able to do
Not definitions to recite — tasks you can open a laptop and complete.
Roles this supports Machine-learning and AI engineering internships · Backend roles that ship an AI feature inside a product · Data roles where the data cannot leave the building
- Explain what open weights let you do that a hosted model does not
- Run a model on your own machine and see what it costs in memory
- Write a prompt that suits a small model rather than fighting it
- Build one tool that works with no network connection
- Measure a model on speed and memory before promising it to anybody
- Chunk a long document yourself, because you own the context window
The syllabus
Course curriculum
4 modules · 15 lessons. Open a module to see its outcome and lesson titles.
Module 1 Foundations — what Llama is Meet the tool, get an account, and learn the screen before you learn the prompting.
- 01 Meet Llama — what it is and who makes it Free preview
- 02 Signing in: what is free, what is paid, and what you actually need
- 03 The screen: where every control lives
Deliverable 3 practice tasks, one per lesson
Module 2 Prompting — getting a real answer How Llama reads text, the four-part prompt, your own work, and what to do when the answer is wrong.
- 04 How Llama reads what you type
- 05 The four-part prompt: role, task, context, format
- 06 Llama for the work you actually have
- 07 When the answer is wrong: iterate instead of restarting
Deliverable 4 practice tasks, one per lesson
Module 3 Files, tools and your own data Documents and tables, Running and tuning a model on your own machine, the API, and one boring task automated.
- 08 Files, tables and long documents
- 09 Running and tuning a model on your own machine — Llama's own feature
- 10 A local API and the hosted endpoints and your first script
- 11 Automate one boring task with Llama
Deliverable 4 practice tasks, one per lesson
Module 4 Career, projects and honesty A finished portfolio project, the privacy rules, the limits, and the Meta certification paths.
- 12 Build the portfolio project: an offline summariser for your own documents
- 13 Ethics, privacy and what never to paste
- 14 Limits, hallucinations and how to check
- 15 Get certified: the official Meta paths
Deliverable 4 practice tasks, one per lesson
Lesson explanations, worked examples and practice tasks are in the learning player, not on this page — so you can read the syllabus without scrolling through the whole course.
How the course works
Every lesson follows the same four steps. No passive video playlists.
- 1
Learn the concept
A short explanation in plain English, with the one idea the lesson depends on.
- 2
See a worked example
The same task done end to end, with the reasoning behind each step.
- 3
Complete the practice task
You write it yourself. The task is the lesson; the reading is only the setup.
- 4
Build and publish a project
Each module ends in work you can open, explain and attach to an application.
Way2Fresher Academy
Courses written for the fresher hiring market — not translated from a bootcamp built for another country.
- Plain English Every explanation is written to be read once, slowly, by someone new to the field.
- Project-first A practice task in every lesson and a project at the end of every module.
- No paywall The full syllabus, all lessons and the certificate are free. Nothing is held back for a paid tier.
3Portfolio projects
15Lessons
The work is the proof
A fresher interview cannot assess a certificate. It can assess a project you can open, explain and change on the spot — which is why every course here ends in one.
- A finished project brief with an evaluation checklist
- A shareable repository, dashboard or case study
- A certificate that records the work, once every lesson is done
Frequently asked questions
The answers that decide whether this is the right six weeks.
Do I need a graphics card?
Not for a small model. A modern laptop with 16GB of memory runs a 3B model acceptably. A GPU makes the larger models usable, and renting one by the hour for a few days is enough for this course.
Is a local model good enough to use daily?
Not compared with a frontier model, and that is not the point. It is good enough to be private, free and yours, which are different advantages.
How is this different from the other AI courses here?
Every other course uses somebody else's model over the internet. This one is about owning the model — which is the only route to privacy, offline work, product integration and fine-tuning.
What will I have at the end of this course?
Three things: an offline summariser for your own documents, a saved set of prompts you wrote and tested on your own work, and a Way2Fresher certificate naming the course. Meta also publishes its own learning material for the tools it makes, and the rail on this course page links to it.
Is the course really free?
Yes. Every lesson, practice task and project brief on this page is free to read, and no card is required. A free account only saves your progress and unlocks the certificate.
Do I need coding experience?
No. This course is written for intermediate learners and every lesson starts from the tool itself. If you can use a browser and a file, you can start lesson one today.
How long will it take?
About 5 weeks. The course holds 15 lessons, each built around one practice task, so the calendar depends on how many sessions you can keep each week.
Start this course free.
No credit card. No sign-up required to preview. Lesson one is open now.
Meta AI and Llama Mastery — Open Models You Can Run Yourself
Start Course Free