← All works
Research paper

Mistral 7B

At a glance

Mistral 7B demonstrated strong language-model performance relative to larger models.

In one paragraph

Introduced an openly released language model that outperformed the larger Llama 2 13B on the reported tests. Its design reduced the computing resources needed to generate text, showing that smaller models could deliver strong performance in reasoning, mathematics and coding.

Authors 18

  1. Albert Q. Jiang
  2. Alexandre Sablayrolles
  3. Arthur Mensch
  4. Chris Bamford
  5. Devendra Singh Chaplot
  6. Diego de las Casas
  7. Florian Bressand
  8. Gianna Lengyel
  9. Guillaume Lample
  10. Lucile Saulnier
  11. Lélio Renard Lavaud
  12. Marie-Anne Lachaux
  13. Pierre Stock
  14. Teven Le Scao
  15. Thibaut Lavril
  16. Thomas Wang
  17. Timothée Lacroix
  18. William El Sayed