kmkarakaya/nutuk_soru_cevap

--- language: - tr license: apache-2.0 task_categories: - question-answering - text-generation tags: - nutuk - ataturk - turkish - llm-generated - gemini - instruct-tuning - rag pretty_name: Nutuk QA Fine-Tuning Dataset size_categories: - n<1K # (Update this based on your actual row count, e.g., 1K<n<10K) --- # 🇹🇷 Nutuk Question-Answer Fine-Tuning Dataset ## Dataset Description This dataset provides high-quality question-answer pairs derived from **Nutuk** (The Great Speech), the foundational historical text delivered by Mustafa Kemal Atatürk. It is meticulously designed for fine-tuning Large Language Models (LLMs) on Turkish language tasks, specifically focusing on instruction-following, reading comprehension, and historical question-answering. To ensure high cognitive diversity, the data was synthetically generated using **Gemini 2.5 Flash**, categorized into 13 distinct reasoning types. - **Language:** Turkish (`tr`) - **License:** Apache 2.0 - **Primary Tasks:** Question Answering, Instruction Fine-Tuning, RAG Evaluation --- ## Potential Use Cases * **Instruction Fine-Tuning:** Training Turkish LLMs to respond accurately to history-based prompts. * **RAG (Retrieval-Augmented Generation):** Serving as a benchmark or training set for Turkish RAG systems over historical documents. * **Educational AI:** Developing intelligent tutoring systems or chatbots focused on the history of the Turkish Republic. --- ## Dataset Structure ### Data Fields The dataset consists of a single split (`train`) with the following features: * **`question`** `(string)`: A specific question generated from a paragraph within Nutuk. * **`answer`** `(string)`: The detailed, context-grounded answer to the question. * **`question_type`** `(string)`: The cognitive category of the question (e.g., Analytical, Factual, Hypothetical). ### Example Data Row ```json { "question": "Mustafa Kemal Paşa Samsun'a hangi tarihte ve hangi görevle çıkmıştır?", "answer": "Mustafa Kemal Paşa, 19 Mayıs 1919'da 9. Ordu Müfettişi göreviyle Samsun'a çıkmıştır.", "question_type": "Factual" }