Mini-project
Logistics & Supply Chain
Fine-tuned delay reason classifier
Fine-tune a small model to classify shipment delay notes by cause, and compare it with a large model.
The brief
Drivers and hubs write delay notes in their own words, so nobody can count the causes. Fine-tune a small model that classifies delay notes (vehicle breakdown, traffic, weather, address issue, customer unavailable), and compare it with prompting a large hosted model. Good looks like: a labelled set, accuracy per cause, and a recommendation with cost.
What you will build
- Write or collect labelled delay notes
- Fine-tune a small model with LoRA
- Prompt a large model on the same test set
- Compare accuracy per cause, speed and cost
- Summarise delay causes by route
Free, open data to build it with
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.