martinkorelic/slovenian-places-instruct

---
dataset_info:
features:
- name: id
dtype: string
- name: template_id
dtype: string
- name: group
dtype: string
- name: instruction
dtype: string
- name: output
dtype: string
- name: meta_kraj
dtype: string
- name: meta_vrsta
dtype: string
- name: meta_razred
dtype: string
splits:
- name: train
num_bytes: 44551100
num_examples: 200219
download_size: 3151745
dataset_size: 44551100
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: cc-by-4.0
task_categories:
- text-generation
- question-answering
language:
- sl
tags:
- geographical
- instruction
pretty_name: Slovenian Places Instruction Dataset
size_categories:
- 100K<n<1M
---

# 🇸🇮 Slovenian Places Instruction Dataset

A Slovenian-language instruction-tuning dataset of question–answer pairs about Slovenian geographical places. All examples are generated automatically from structured data — no hallucinated facts.

---

## Template Groups

### Group 1 — Type & Region
Basic factual questions about a place's geographic type and administrative region.

### Group 2 — Neighbours
Questions about spatially nearby places.

### Group 3 — Region Details
Aggregate queries listing all places of a given type within a region.

### Group 4 — Neighbour Types
Type-filtered neighbourhood queries (e.g. closest river, closest settlement).

### Group 5 — Mixed / Multi-hop
Questions combining location, type, and region in a single hop.

### Group 6 — Structured Output
The same facts rendered as JSON, Markdown tables, bullet lists, and key-value summaries — useful for teaching structured output generation.

---

## Data Fields

| Field | Type | Description |
|---|---|---|
| `id` | string | Unique example identifier |
| `template_id` | string | Template that generated this example |
| `group` | string | Template group label |
| `instruction` | string | Input question or prompt (Slovenian) |
| `output` | string | Expected answer or structured output (Slovenian) |
| `meta_*` | string | Additional metadata fields (place name, region, type, …) |

---