This repository contains supporting materials and human feedback structures for the Kayra language model series.
Kayra is an experimental Turkish small language model project focused on understanding instruction tuning, hallucination behavior, and limits of low-parameter LLMs.
This repository is intentionally minimal. Model weights and training details live on Hugging Face.
See:
- CONTRIBUTORS.md for human feedback involvement
- https://huggingface.co/sixfingerdev/kayra-1