README.md

---
license: apache-2.0
datasets:
- Kundyzka/informatics_kaz
language:
- kk
metrics:
- name: F1 (Before Training)
  type: F1 Score
  value: 31.405
- name: Exact Match (Before Training)
  type: Exact Match
  value: 14.675
- name: F1 (After Training)
  type: F1 Score
  value: 56.819
- name: Exact Match (After Training)
  type: Exact Match
  value: 35.454
base_model:
- Kyrmasch/t5-kazakh-qa
new_version: Kundyzka/t5-kazakh-qa-informatics-kaz
pipeline_tag: question-answering
library_name: adapter-transformers
---

# Description

This model was developed by **Kundyz Maksutova**, PhD Candidate, as part of research on improving question-answering systems in the Kazakh language. It is a fine-tuned version of `Kyrmasch/t5-kazakh-qa` on the `Kundyzka/informatics_kaz` dataset. The model is specifically optimized for question-answering tasks in Kazakh, focusing on the domain of computer science and related fields.

### Key Features:
- **Developer**: Kundyz Maksutova, PhD Candidate
- **Base Model**: `Kyrmasch/t5-kazakh-qa`
- **Dataset**: `Kundyzka/informatics_kaz`
- **Language**: Kazakh (`kk`)
- **Task**: Question Answering

### Performance:
This model demonstrates significant improvements after fine-tuning, as shown by the following metrics:

- **Before Training**:
  - F1 Score: 31.405
  - Exact Match (EM): 14.675
- **After Training**:
  - F1 Score: 56.819
  - Exact Match (EM): 35.454

These metrics highlight the enhanced ability of the model to handle domain-specific questions after training on the `Kundyzka/informatics_kaz` dataset.

### Dataset:
The `Kundyzka/informatics_kaz` dataset is curated to provide a diverse set of questions and answers in Kazakh, primarily targeting topics in computer science. This dataset ensures the model handles domain-specific terminology effectively.

### Intended Use:
This model is designed for answering questions in the Kazakh language, with applications in:
- **Educational Platforms**: Supporting students in learning computer science.
- **Research Projects**: Facilitating studies in Kazakh natural language processing.
- **Applications**: Powering intelligent systems like chatbots or question-answering assistants.

### Limitations and Ethical Considerations:
- **Domain-Specific Bias**: Performance may drop on topics outside computer science.
- **Dataset Bias**: Potential biases from the dataset can influence model outputs.
- **Language Support**: The model is optimized for Kazakh and does not support other languages.

### Tags:
- `computerscience`
- `question-answering`
- `Kazakh`

This model represents a significant step toward advancing natural language processing tools for low-resource languages like Kazakh. For further details or customization, refer to the model repository.