Fine-tuned outage report classifier

The brief

Field crews write outage reports in their own words, and nobody counts the causes. Fine-tune a small model that classifies reports by cause (tree fall, transformer failure, cable fault and so on) and extracts the area, and compare it with a large hosted model. Good looks like: a labelled set, accuracy per cause, and a recommendation with cost and privacy.

What you will build

1. Write or collect labelled outage reports
2. Fine-tune a small model with LoRA
3. Prompt a large model on the same test set
4. Compare accuracy per cause, speed and cost
5. Summarise outage causes by area

Free, open data to build it with

• Central Electricity Authority reports ↗ Central Electricity Authority

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.