Expert Curated Dataset 2026
Collection
The dataset is used for fine-tuning and aligning LLMs with Kazakh culture. • 13 items • Updated
🇰🇿 Kazakh Multi-Step Planning and Tool Composition Dataset
Dataset Summary
Kazakh Multi-Step Planning and Tool Composition Dataset is a Kazakh-language dataset for training and evaluating Large Language Models (LLMs) in agentic AI workflows that require multi-step planning, tool composition, and structured function calling.
The dataset contains user requests, available tool schemas, expected tool calls, simulated tool responses, and complete multi-turn interaction traces. It focuses on scenarios where a model must break down a user request into several connected actions, call the appropriate tools in the correct order, combine outputs from different tools, and generate a final answer in Kazakh.
This dataset is useful for developing Kazakh-language AI agents capable of handling practical workflows such as travel planning, ticket search, currency conversion, scheduling, information lookup, and other multi-tool tasks.
📊 Dataset Statistics
General Metrics
| Metric | Count |
|---|---|
| Total Samples | 3,025 |
| Total Words (approx.) | 943,485 |
| Avg. Words per Sample | 311 |
Word Count Distribution Per Field
The following table details the distribution of word counts across different fields in the dataset.
| Field | Mean | Median | Min | Max | Total Words |
|---|---|---|---|---|---|
| answers | 32.5 | 27.0 | 13 | 134 | 98,398 |
| category | 1.0 | 1.0 | 1 | 1 | 3,025 |
| difficulty | 1.0 | 1.0 | 1 | 1 | 3,025 |
| id | 1.0 | 1.0 | 1 | 1 | 3,025 |
| query | 15.0 | 14.0 | 3 | 46 | 45,357 |
| tools | 91.5 | 90.0 | 50 | 200 | 276,832 |
| turns | 169.9 | 163.0 | 103 | 417 | 513,823 |
Dataset Structure
Each dataset instance represents a complete agentic workflow. A sample usually contains a Kazakh user request, the tools available to the assistant, the expected tool calls, mock tool responses, and the final assistant answer.
Data Fields
id: A unique identifier for the sample.query: The original user request in Kazakh. The request often requires several connected steps, such as searching for transport information and then converting prices into another currency.category: The task category. For example,03_planning_multistep_compositionindicates that the sample focuses on planning, multi-step reasoning, and tool composition.tools: A list of tools available to the assistant. Each tool includes:name: the tool name;description: a short explanation of the tool’s function;parameters: the required or optional input schema;mock_response: the expected response structure.
difficulty: The difficulty level of the sample. In this dataset, multi-step composition tasks may involve harder reasoning because the assistant must coordinate several tools.answers: The target tool calls expected from the assistant. This field usually contains multiple tool calls with serialized JSON arguments.turns: The full conversation trajectory, including:- user request;
- assistant planning step;
- assistant tool calls;
- mock tool outputs;
- final answer grounded in the tool responses.
Data Instance
Below is one representative example from the dataset.
{
"id": "aikerim_id_1",
"query": "Алматыдан Астанаға 2024-07-01 күні жүретін пойыздардың бағасын тауып, купе билетін доллармен есептеп бер.",
"category": "03_planning_multistep_composition",
"tools": [
{
"name": "trains.search",
"description": "Search train schedules",
"parameters": {
"from": {
"type": "string",
"description": "Departure station",
"required": true
},
"to": {
"type": "string",
"description": "Arrival station",
"required": true
},
"date": {
"type": "string",
"description": "Travel date YYYY-MM-DD",
"required": true
}
},
"mock_response": {
"trains": [
{
"number": "",
"type": "",
"departure": "",
"arrival": "",
"price_coupe": null,
"price_platz": null
}
]
}
},
{
"name": "forex.rate",
"description": "Get currency exchange rate",
"parameters": {
"from": {
"type": "string",
"description": "Source currency code",
"required": true
},
"to": {
"type": "string",
"description": "Target currency code",
"required": true
}
},
"mock_response": {
"from": "",
"to": "",
"rate": null,
"updated": ""
}
}
],
"difficulty": "hard",
"answers": [
{
"name": "trains.search",
"arguments": "{\"from\": \"Алматы\", \"to\": \"Астана\", \"date\": \"2024-07-01\"}"
},
{
"name": "forex.rate",
"arguments": "{\"from\": \"KZT\", \"to\": \"USD\"}"
}
],
"turns": [
{
"role": "user",
"content": "Алматыдан Астанаға 2024-07-01 күні жүретін пойыздардың бағасын тауып, купе билетін доллармен есептеп бер."
},
{
"role": "assistant",
"content": "Алдымен пойыз бағасын іздеп, кейін доллар бағамын анықтаймын.",
"meta": {
"plan": "trains.search + forex.rate connection fetching parameters to query",
"tags": [],
"difficulty": "hard"
}
},
{
"role": "assistant",
"content": "",
"tool_call": {
"name": "trains.search",
"arguments": "{\"from\": \"Алматы\", \"to\": \"Астана\", \"date\": \"2024-07-01\"}"
}
},
{
"role": "tool",
"content": "{\"trains\": [{\"number\": \"001Т\", \"type\": \"Тальго\", \"departure\": \"2024-07-01 20:00\", \"arrival\": \"2024-07-02 08:00\", \"price_coupe\": 18000, \"price_platz\": 9500}]}"
},
{
"role": "assistant",
"content": "",
"tool_call": {
"name": "forex.rate",
"arguments": "{\"from\": \"KZT\", \"to\": \"USD\"}"
}
},
{
"role": "tool",
"content": "{\"from\": \"KZT\", \"to\": \"USD\", \"rate\": 450.5, \"updated\": \"2024-06-25T10:00:00Z\"}"
},
{
"role": "assistant",
"content": "Алматыдан Астанаға 2024-07-01 күнгі \"001Т Тальго\" пойызының купе билеті 18000 теңге тұрады. Бұл шамамен 40 доллар (айырбас бағамы 1 USD = 450.5 KZT) болады. (Дерек көзі: trains.search + forex.rate)"
}
]
}
Funding
This dataset was developed as part of the project funded by the Ministry of Science and Higher Education of the Republic of Kazakhstan under Grant No. BR24993001, “Creation of a Large Language Model (LLM) to Support the Kazakh Language and Advance Technological Development.”
Citation
If you use this dataset in your research, please cite the following article:
APA
Kadyrbek, N., Tuimebayev, Z., Mansurova, M., & Viegas, V. (2025). The development of small-scale language models for low-resource languages, with a focus on Kazakh and direct preference optimization. Big Data and Cognitive Computing, 9(5), 137. https://doi.org/10.3390/bdcc9050137
BibTeX
@article{kadyrbek2025development,
title = {The Development of Small-Scale Language Models for Low-Resource Languages, with a Focus on Kazakh and Direct Preference Optimization},
author = {Kadyrbek, Nurgali and Tuimebayev, Zhanseit and Mansurova, Madina and Viegas, Vitor},
journal = {Big Data and Cognitive Computing},
volume = {9},
number = {5},
pages = {137},
year = {2025},
publisher = {MDPI},
doi = {10.3390/bdcc9050137},
url = {https://www.mdpi.com/2504-2289/9/5/137}
}
- Downloads last month
- 12