GPU labs where every student has a budget and an end time.
Teach fine-tuning and model deployment without buying servers. You set the lab once, and each student gets a GPU session that stops by itself.
Lab: fine-tune an open model
- Student 01
- Student 02
- Student 03
- Student 04
- Student 05
How a lab runs
Pick a course lab
Start from a template, such as fine-tuning an open model or building a document assistant, and adjust it to your course.
Invite your cohort
Each student gets a workspace with notebooks, course data and a GPU budget.
Sessions stop on their own
A session ends at the budget limit or at the lab end time. You see usage for every student.
Course labs
- Fine-tune an open model
- Adapt a small open-weight model to a domain dataset with LoRA or QLoRA, then evaluate the result against a test set.
- Build a document assistant
- Add retrieval over course materials and measure how well the assistant answers.
- Deploy a model API
- Turn a model into an endpoint with keys and limits, and learn how serving is managed.
- Evaluate and compare models
- Run the same test set on several open models and compare quality, speed and cost.
For instructors
Budget per student
Set GPU hours for each student or for the whole cohort.
Automatic stop
No forgotten sessions and no surprise bills at the end of term.
Usage reports
See who used what, and when, for every lab.
Shared starter material
Distribute notebooks and datasets to the whole cohort at once.
Who it is for
Universities, colleges, learning providers and bootcamps, and research groups that teach or study with open models.
QazTech Cloud is run by Digital Qazaqstan Ltd, a UK-registered learning provider (UKPRN 10102276). Pilot access is limited, and capacity is provided through infrastructure partners.
Request a pilot
Fill in the form and your email app opens with the message ready to send. Nothing is stored by this site. We reply within two business days.