How long is the Fine-Tuning and Open-Weights Bootcamp?

12 weeks at about 15 hours a week. The weeks run as six two-week blocks; each block ends in a project you deploy and a gate you must pass before the next block opens.

Is the Fine-Tuning and Open-Weights Bootcamp live or self-paced?

Live. It is taught on Zoom by one instructor, with a 30-minute one-to-one every week. Nova, Square 1's AI tutor, grades every exercise and project between sessions against a rubric you can read.

When are the live sessions for the Fine-Tuning and Open-Weights Bootcamp?

Live sessions run on weekday evenings from 5:30 PM Sri Lanka and India time, with full-day sessions at weekends. That is 5:00 PM in Pakistan, 5:45 PM in Nepal and 6:00 PM in Bangladesh, so the sessions fit around a working day.

How much does the Fine-Tuning and Open-Weights Bootcamp cost?

A$200 for the founding cohort, paid once; the standard price is A$400. It is the same price in every country. Founding students keep their founding price on any second programme.

What will I build in the Fine-Tuning and Open-Weights Bootcamp?

6 deployed projects: Baseline, Training data, SFT, Preference tuning, Production and Employer brief. Each is graded against a published rubric and kept on a record an employer can open at /verify.

What skills does the Fine-Tuning and Open-Weights Bootcamp teach?

Python, PyTorch, LoRA, DPO, evals, serving and data curation. It prepares you for roles such as ML engineer (LLM), Applied scientist and AI engineer (fine-tuning).

Does the Fine-Tuning and Open-Weights Bootcamp guarantee a job?

No. No placement rate is published because the first cohorts have not graduated. What you leave with is six deployed projects, a recorded defence of your code and a hiring plan worked through with your instructor, including a two-week hiring sprint at the end.

Live bootcampTaught live on Zoom, one instructor

Fine-Tuning and Open-Weights Bootcamp

Post-train, evaluate and deploy your own model for a real task.

Fine-Tuning and Open-Weights Bootcamp is a 12-week live online bootcamp, taught on Zoom by one instructor at about 15 hours a week, with 6 deployed projects and a one-to-one every week. The founding price is A$200, down from A$400. Live sessions run on weekday evenings from 5:30 PM Sri Lanka and India time, with full-day sessions at weekends.

Taught with

PythonPyTorchHugging FaceJupyterAWS

Not sure it is for you?Take the free Generative AI skill check

By week 12, you will have shipped

Week 1Model LoaderA Jupyter notebook that loads an open model from Hugging Face and runs it on sample prompts
Weeks 1–2Open weightsA baseline eval of three open models on your task, published
Week 3Dataset CardA data card for a small dataset with sources, licences and a cleaning script
Weeks 3–4DataA curated, documented training set with provenance, published
Weeks 5–6Supervised fine-tuningA fine-tuned model that beats the baseline, deployed
Week 7Preference PairsA small set of chosen and rejected answer pairs built from your model's outputs
Weeks 7–8Preference tuningA preference-tuned model with a preference eval, deployed
Week 9Quantised BuildA quantised version of your model with a script that compares speed and quality
Weeks 9–10DeploymentThe capstone: your model in production with monitoring and a cost comparison
Weeks 11–12Employer brief and hiring sprintThe employer brief shipped, and demo day

weeks, live on Zoom: 12
hours a week: 15
projects to build: 10
minutes 1-1, every week: 30

Who it is for, and what you receive.

Engineers who need a model they own, for cost, latency, privacy or a task hosted APIs do badly. A GPU budget of about A$150 over the twelve weeks is stated up front.

Before you start

Python, some PyTorch, and a GPU budget (we say how much).

Skills

PythonPyTorchLoRADPOevalsservingdata curation

A week, about 15 hours

Live on Zoom: 4h
Set reading and exercises: 6h
The project: 5h

1
Six deployed projects
Each graded by Nova against a published rubric and signed off by the instructor at the gate.
2
The twelve-week plan
Every week has its theme, its topics and what you build, published before you apply and taught live by your instructor.
3
Live on Zoom every week
A 90-minute live code review, a 60-minute squad lab, 60 minutes of office hours, and a 30-minute 1-1 with your instructor.
4
Your hiring plan
Opened in week 1 and reviewed in every 1-1: target roles, the gap map, the proof list, weekly actions, an interview log, the outcome.
5
The record
Every grade and every project with its repository at square1ai.com/u/{handle}, and the recording of your viva, verifiable by any employer.
6
The certificate
A credential ID that resolves at square1ai.com/verify to the real completion.
7
The hiring sprint
Weeks 11 and 12: CV and portfolio from graded work, applications, mock interviews scored against the role, demo day.
8
Nova for twelve weeks
A tutor with every submission of yours in its memory, at 2am as well as in class.

Twelve weeks in six parts.

Each part teaches for a week, then you build and deploy a project, then a gate checks it before the next part opens. The project and its gate are the part's entry on your record.

Live time each week on Zoom: a 90-minute class where the instructor reviews real submissions, a 60-minute squad lab, 60 minutes of office hours, and your own 30-minute 1-1. Nothing is lectured live; the recorded lessons do that.

Key skillsThe open model landscapeRunning and serving open weightsBaselines and benchmarks
Week 1
Open weights
- The open model landscape
- Running and serving open weights
- Benchmarks and baselines
- Licences
You build: Baseline eval
Week 2
Project 1: the baseline
- Reproducibility
- Choosing the task
You build: Reproducible baseline of three models
Project 1 · the gate at week 2
Baseline
A baseline eval of three open models on your task, published
You hand in
- Eval harness in a repository
- Three model baselines
- Licence notes per model
- Short baseline report
The gate
The baseline results reproduce from the repository alone.

What you can do by week 12.

A post-training method from data to serving, and a model you trained and own, with the numbers to show when it beats the hosted API.

1
Choose among open-weight models with reproducible baseline evals.
2
Build a training set with provenance, licensing and filtering, including synthetic data.
3
Fine-tune with LoRA and full fine-tuning, with checkpoints and evals during training.
4
Preference-tune with DPO and measure win rate against the SFT model.
5
Quantise, serve and monitor a custom model and compare its cost to a hosted API.
6
Defend a model, its data and its cost.

Roles this prepares you for

ML engineer (LLM)Applied scientistAI engineer (fine-tuning)

No placement rate is shown, because there are no graduates to count yet. The roles above are what the projects are built for.

Your record at week 12.

Every exercise and project is graded by Nova against a rubric you can read, and every grade is kept on one page an employer can open and run. This is what the programme writes to it.

Graded, line by line: Nova reads every submission against the brief and the rubric and returns a score, what you did well and what to fix.
Six gates: A block does not open until the previous project passes its gate. You always know where you are and what is next.
A weekly 1-1: Thirty minutes with your instructor, who has already read your code before the call.
One page an employer can run: Every grade, project and the viva recording at /verify. An employer opens it, runs the code and watches you defend it.

Record, Fine-Tuning and Open-Weights BootcampExample

Week 2
Baseline
The baseline is reproducible from the repo
Graded, gate signed
Week 4
Training data
Every example has a source and a licence
Graded, gate signed
Week 6
SFT
The gain is measured on a held-out set
Graded, gate signed
Week 8
Preference tuning
Win rate against the SFT model is measured
Graded, gate signed

The entries, not the grades: those are yours to earn. The page lives at /verify and an employer needs no account to open it.

8 entries, twelve weeks. Reserving a spot holds the founding price.

Reserve a seat

How we help you find a job.

The last block is not curriculum. It is the hiring sprint, and the proof you built in the ten weeks before it.

The course booklet

The whole plan on a few pages: the parts, the projects, the gates and the price. We email it to you.

Get the booklet

1
Your hiring plan, from week 1
Six parts you and your instructor keep: target roles, the gap map from real postings, the proof to send, weekly actions, an interview log, the outcome. Read before every 1-1.
2
The hiring sprint
Weeks 11 and 12: CV, portfolio, applications, mock interviews, and demo day in front of hiring partners.
3
A record an employer can run
Your deployed projects and the recorded viva on /verify. An employer opens it, runs the code and watches you defend it.
4
The career agent
Paste a real job posting at /career and it maps the role to your graded work and what to do next.
5
The roles directory
Every role we prepare people for, what it pays and what it asks, at /roles.

Questions people ask.

About Fine-Tuning and Open-Weights Bootcamp, answered from the plan on this page.

One cohort, one instructor. Hold your spot.

A cohort of 25, one instructor, 12 weeks, six things you ship. Spot holders hear the start date first, and nothing is charged before you confirm.

Six deployed models and datasets
A post-training method from data to serving
A recorded viva
Your record on /verify

A$200~~A$400~~