---
license: apache-2.0
language:
- bg
size_categories:
- n<1K
tags:
- function_calling
- MCP
- tool_use
---
# Tucan-BG-Eval-v1.0 🇧🇬

**Paper: https://arxiv.org/abs/2506.23394**

Official evaluation dataset for [Tucan models](https://huggingface.co/collections/s-emanuilov/tucan-tool-using-and-function-calling-in-bulgarian-684546a02b297f30768993dd) - a series of Bulgarian language models fine-tuned for function calling and tool use.

[![License](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
[![Dataset Size](https://img.shields.io/badge/Size-120_samples-blue.svg)](https://huggingface.co/datasets/s-emanuilov/Tucan-BG-Eval-v1.0)

## 📊 Dataset Overview

This dataset contains 120 curated samples designed to evaluate function-calling capabilities in Bulgarian language models. Each sample includes:

- User queries in Bulgarian
- Available functions with their parameters
- Expected behavior (function calls or text responses)
- Evaluation criteria
- Scenario types for different use cases

## 🎯 Evaluation Scenarios

The dataset covers various function-calling scenarios:

- `function_call_required` - Cases where a function must be called
- `multiple_functions_selection` - Choosing between multiple available functions
- `irrelevant_question_with_functions` - Questions that should be answered without function calls
- `no_functions_available` - General knowledge questions
- `missing_required_parameters` - Cases requiring parameter clarification

## 🔧 Usage

```python
from datasets import load_dataset

# Load the dataset
dataset = load_dataset("s-emanuilov/Tucan-BG-Eval-v1.0")

# Access the samples
samples = dataset["train"]
```

For detailed evaluation instructions, see the [Tucan Evaluation Framework](https://github.com/s-emanuilov/tucan).

## 📝 Dataset Format

Each sample contains:

```json
{
    "user_message": "Bulgarian user query",
    "functions": [
        {
            "name": "function_name",
            "description": "Function description",
            "parameters": {
                "properties": {...},
                "required": [...]
            }
        }
    ],
    "expected_behavior": {
        "expected_function": "function_name",
        "expected_parameters": {...},
        "should_call_function": true/false
    },
    "evaluation_criteria": {...},
    "scenario_type": "scenario_category"
}
```

## 🔗 Links

- [Tucan Models Collection](https://huggingface.co/collections/s-emanuilov/tucan-tool-using-and-function-calling-in-bulgarian-684546a02b297f30768993dd)
- [Evaluation Framework](https://github.com/s-emanuilov/tucan)