← Tech Station
04 — LLMOps

LLM Fine-Tuning & LLMOps

When to fine-tune vs. prompt, LoRA, QLoRA, evaluation, and deployment.

13 guides in this track·1.6h reading·Intermediate → Advanced
What you'll learn

Know when to fine-tune, how to do it without wasting a budget, and how to run models in production — from LoRA and QLoRA to serving, monitoring, and cost control.

Who this is for

ML/AI engineers past their first prompt-engineered feature, and the platform teams supporting them.

How to use it

The guides are numbered — read in order for the curriculum path, or jump straight to the one you need. Each card is self-contained.

The guides

Ordered by recommended reading path
01
Fine-Tuning vs RAG vs Prompting

The decision framework for adapting an LLM.

8 minRead →
02
How Fine-Tuning Works

Continuing training to shift a model's weights.

6 minRead →
03
Preparing a Fine-Tuning Dataset

Quality, format, size and common pitfalls.

7 minRead →
04
Full Fine-Tuning vs LoRA & QLoRA

Parameter-efficient methods that cut cost.

7 minRead →
05
A Practical Guide to Fine-Tuning an Open Model

Hands-on walkthrough on Llama/Mistral-class models.

7 minRead →
06
Instruction Tuning & RLHF Explained

How models are aligned to be helpful.

8 minRead →
07
Evaluating a Fine-Tuned Model

Metrics, benchmarks, avoiding overfitting.

9 minRead →
08
LLMOps vs MLOps

What changes when LLMs go to production.

7 minRead →
09
Deploying & Serving LLMs

Hosting, APIs, throughput and reliability.

8 minRead →
10
Monitoring LLMs: Cost, Latency, Drift & Quality

Watching production systems that can degrade silently.

7 minRead →
11
Evaluation Pipelines & Golden Test Sets

Automated regression testing for LLM apps.

8 minRead →
12
Guardrails & Safety Layers

Input/output filtering and abuse prevention.

7 minRead →
13
Cost Optimisation: Caching, Routing & Model Selection

Practical levers to cut the token bill.

9 minRead →

Related tracks

Continue in the same group
01 — Start hereAI PrimerThe concepts every LLM practitioner needs before writing a single prompt.02 — Core skillPrompt EngineeringZero-shot to few-shot, chain-of-thought, and structured output patterns.03 — ArchitectureRAGChunking, embeddings, vector search, and fixing hallucinations in production.05 — AdvancedAgentic AIPlanning loops, tool use, memory, and multi-agent coordination.