alimkacar/Turkish-Llama-8B-STEM-QLoRA
Text Generation • Updated • 1
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Türkçe K-12 STEM ve kodlama eğitimi için instruction-tuning veri seti.
Bu veri seti, Türkiye'deki K-12 seviyesinde STEM ve kodlama eğitimi için hazırlanmış 1.000 instruction-output çiftinden oluşur.
Hibrit pipeline ile derlenmiştir:
{
"instruction": "Arduino ile LED yakıp söndüren kodu yaz.",
"input": "",
"output": "Detaylı açıklama ve kod...",
"category": "arduino",
"difficulty": "ortaokul"
}
| Split | Örnek Sayısı |
|---|---|
| Train | 53 |
| Validation | 7 |
| Test | 8 |
from datasets import load_dataset
dataset = load_dataset("sehinsahfanboy/eding-stem-tr-instruct-1k")
print(dataset["train"][0])
Apache 2.0
@misc{eding-stem-tr-2026,
title={Eding STEM TR Instruct 1K: Türkçe K-12 STEM Eğitimi Veri Seti},
author={Kacar, Alim},
year={2026},
publisher={HuggingFace},
url={https://huggingface.co/datasets/sehinsahfanboy/eding-stem-tr-instruct-1k}
}