Datasetpublic

tardelr/worldcup2026

World Cup 2026 Q&A 952 question-and-answer pairs about the 2026 FIFA World Cup, data was retrieved from Openfootbal github Dataset Summary This is a very simple adjustment from the original data adapted to Q&A format using LLMs (you should double-check results for any serious application). Dataset Structure Each line is a JSON object with a messages field: { "messages": [ {"role": "system", "content": "You are a helpful assistant with… See the full description on the dataset page: https://huggingface.co/datasets/tardelr/worldcup2026.

sourceHugging Facemitupdated 12h agoView on Hugging Face

0likes16downloads

Dataset Card

World Cup 2026 Q&A

952 question-and-answer pairs about the 2026 FIFA World Cup, data was retrieved from Openfootbal github

Dataset Summary

This is a very simple adjustment from the original data adapted to Q&A format using LLMs (you should double-check results for any serious application).

Dataset Structure

Each line is a JSON object with a messages field:

json

{
  "messages": [
    {"role": "system", "content": "You are a helpful assistant with knowledge about the 2026 FIFA World Cup."},
    {"role": "user", "content": "Who won the 2026 FIFA World Cup?"},
    {"role": "assistant", "content": "Spain won the 2026 FIFA World Cup, beating Argentina in the final (Spain beat Argentina 1-0 after extra time)."}
  ]
}

The data was structured to be directly into TRL (and others similar).