| --- |
| license: apache-2.0 |
| language: |
| - sr |
| task_categories: |
| - question-answering |
| --- |
| |
| # OZ Eval |
|  |
|
|
| ## Dataset Description |
| OZ Eval (_sr._ Opšte Znanje Evaluacija) dataset was created for the purposes of evaluating General Knowledge of LLM models in Serbian language. |
| Data consists of 1k+ high-quality questions and answers which were used as part of entry exams at the Faculty of Philosophy and Faculty of Organizational Sciences, University of Belgrade. |
| The exams test the General Knowledge of students and were used in the enrollment periods from 2003 to 2024. |
|
|
| This is a joint work with [@Stopwolf](https://huggingface.co/Stopwolf)! |
|
|
|
|
| ## Evaluation process |
| Models are evaluated by the following principle using HuggingFace's library `lighteval`. We supply the model with the following template: |
| ``` |
| Pitanje: {question} |
| |
| Ponuđeni odgovori: |
| A. {option_a} |
| B. {option_b} |
| C. {option_c} |
| D. {option_d} |
| E. {option_e} |
| |
| Krajnji odgovor: |
| ``` |
| We then compare likelihoods of each letter (`A, B, C, D, E`) and calculate the final accuracy. All evaluations are ran in a 0-shot manner using a **chat template**. |
|
|
| GPT-like models were evaluated by taking top 20 probabilities of the first output token, which were further filtered for letters `A` to `E`. Letter with the highest probability was taken as a final answer. |
|
|
| Exact code for the task will be posted [here](). |
|
|
|
|
| ## Evaluation results |
| | Model |Accuracy| |Stderr| |
| |-------|-------:|--|-----:| |
| |GPT-4-0125-preview|0.9199|±|0.002| |
| |GPT-4o-2024-05-13|0.9196|±|0.0017| |
| |GPT-3.5-turbo-0125|0.8245|±|0.0016| |
| |[Tito-7B-slerp](https://huggingface.co/Stopwolf/Tito-7B-slerp)|0.7099|±|0.0101| |
| [Qwen2-7B-instruct](https://huggingface.co/Qwen/Qwen2-7B-Instruct)|0.6730|±|0.0105| |
| |[Yugo60-GPT](https://huggingface.co/datatab/Yugo60-GPT)|0.6411|±|0.0107| |
| |[Llama3-70B-Instruct (4bit)](https://huggingface.co/unsloth/llama-3-70b-Instruct-bnb-4bit)|0.5942|±| 0.011| |
| |[Hermes-2-Theta-Llama-3-8B](https://huggingface.co/NousResearch/Hermes-2-Theta-Llama-3-8B)|0.5852|±|0.011| |
| |[Llama3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct)|0.5274|±|0.0111| |
| |[Starling-7B-beta](https://huggingface.co/Nexusflow/Starling-LM-7B-beta)||0.5244|±|0.0112| |
| |[Hermes-2-Pro-Mistral-7B](https://huggingface.co/NousResearch/Hermes-2-Pro-Mistral-7B)|0.5145|±|0.0112| |
| |[Qwen2-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2-1.5B-Instruct)|0.4506|±|0.0111| |
| |[Perucac-7B-slerp](https://huggingface.co/Stopwolf/Perucac-7B-slerp)|0.4247|±|0.011| |
| |[SambaLingo-Serbian-Chat](https://huggingface.co/sambanovasystems/SambaLingo-Serbian-Chat)|0.2802|±|0.01| |
| |[Gemma-2-9B-it](https://huggingface.co/google/gemma-2-9b-it)|0.2193|±|0.0092| |
|
|
|
|
| ### Citation |
| ``` |
| @article{oz-eval, |
| author = "Stanivuk Siniša & Đorđević Milena", |
| title = "Opšte znanje LLM Eval", |
| year = "2024" |
| howpublished = {\url{https://huggingface.co/datasets/DjMel/oz-eval}}, |
| } |
| ``` |
