⚙️ NLP Approach

Radio

Zero-Shot — Prompt only. No examples, no data. The LLM relies entirely on its own knowledge.

Few-Shot — 2 hand-crafted story examples guide the writing style. Still uses LLM knowledge, no dataset.

RAG — Retrieves real match data from the WC dataset (1930–2014) before generating. Most factually grounded.

✍️ Your Prompt

Try an example

📖 Story


Data: FIFA World Cup dataset (1930–2014) · LLM: Llama 3 8B via Groq · Embeddings: all-MiniLM-L6-v2 · Built for NLP HW4 — ARI 525