Course · 7 chapters

Fine-Tuning & Distilling Open Models

From we should fine-tune to a governed loop: LoRA on open weights, distillation from frontier models, and the evals that prove it worked.

Paidadvanced7 chapters124 minEnglish + 6 languagesCertificate on completion

What you'll be able to do

  • A decision tree for when fine-tuning earns its cost, and the lab you'll use to prove it on your own task.
  • Turning a pile of scraped examples into the few hundred curated, deduplicated, and licensed rows actually worth training on.
  • Sizing the adapter to the task, sizing the GPU to the adapter, and picking the checkpoint that actually generalizes.
  • Turning a frontier model's judgment into training signal for your open model, without crossing the line its terms of service draws.
  • Turning a single passing score into a before/after harness, a capability drift set, and a ship/no-ship gate anyone else could rerun.
  • Turning a checkpoint that passed the gate into a live adapter with a name, a canary, and a rollback plan.

What's inside

  1. 1
    Fine-Tuning Open Models: Start Here

    Six chapters that take an open-weight checkpoint from a fine-tuning decision to a served adapter with a rollback plan.

    10 min
  2. 2
    The Fine-Tuning Decision

    A decision tree for when fine-tuning earns its cost, and the lab you'll use to prove it on your own task.

    18 min
  3. 3
    Dataset Engineering for Tuning

    Turning a pile of scraped examples into the few hundred curated, deduplicated, and licensed rows actually worth training on.

    18 min
  4. 4
    LoRA & QLoRA in Practice

    Sizing the adapter to the task, sizing the GPU to the adapter, and picking the checkpoint that actually generalizes.

    20 min
  5. 5
    Distillation from Frontier Models

    Turning a frontier model's judgment into training signal for your open model, without crossing the line its terms of service draws.

    20 min
  6. 6
    Evals That Prove It Worked

    Turning a single passing score into a before/after harness, a capability drift set, and a ship/no-ship gate anyone else could rerun.

    18 min
  7. 7
    Serving Your Tuned Model

    Turning a checkpoint that passed the gate into a live adapter with a name, a canary, and a rollback plan.

    20 min

Earn a certificate

Complete all chapters to receive your certificate of completion.