---
language:
  - kk
license: cc-by-sa-4.0
library_name: transformers
tags:
  - gec
  - grammatical-error-correction
  - kazakh
  - mt5
  - seq2seq
base_model: stukenov/kazakh-gec-mt5-base-run12-kazsandra-new
pipeline_tag: text2text-generation
---

# kazakh-gec-mt5-base-run13-finetune

Run 13: Latest and best mT5-base GEC model — final fine-tuning.

## Overview

| Property | Value |
|----------|-------|
| **Task** | Kazakh Grammatical Error Correction |
| **Architecture** | mt5-base (seq2seq) |
| **Base model** | [stukenov/kazakh-gec-mt5-base-run12-kazsandra-new](https://huggingface.co/stukenov/kazakh-gec-mt5-base-run12-kazsandra-new) |
| **Training data** | [kazakh-synthetic-gec-datasets](https://huggingface.co/datasets/stukenov/kazakh-synthetic-gec-datasets) |
| **Language** | Kazakh (kk) |
| **License** | CC-BY-SA-4.0 |

Best mT5-base variant. Final fine-tuning stage.

## Usage

```python
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("stukenov/kazakh-gec-mt5-base-run13-finetune")
model = AutoModelForSeq2SeqLM.from_pretrained("stukenov/kazakh-gec-mt5-base-run13-finetune")

input_text = "gec: " + "Мен кеше мектепке бардым"
inputs = tokenizer(input_text, return_tensors="pt", max_length=128, truncation=True)
outputs = model.generate(**inputs, max_new_tokens=128)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```

## Training Details

- Fine-tuned from [stukenov/kazakh-gec-mt5-base-run12-kazsandra-new](https://huggingface.co/stukenov/kazakh-gec-mt5-base-run12-kazsandra-new)
- Training data: 1M+ synthetic GEC pairs (correct Kazakh with introduced errors)
- Task prefix: "gec: "

## Project

Part of the [Kazakh GEC](https://huggingface.co/collections/stukenov) project, building grammatical error correction models for Kazakh.

## Citation

```bibtex
@misc{tukenov2026gec,
  title={Kazakh Grammatical Error Correction with mT5},
  author={Tukenov, Saken},
  year={2026},
  url={https://huggingface.co/stukenov/kazakh-gec-mt5-base-run13-finetune}
}
```

## License

CC-BY-SA-4.0

## Benchmark Results

Evaluated on **100-example custom GEC test** (pure model inference, no pre/post pipeline).

| Category | Score |
|----------|-------|
| Орфография (емле) | 0/30 (0%) |
| Грамматика | 2/20 (10%) |
| Пунктуация | 0/15 (0%) |
| Смешанный | 0/20 (0%) |
| Identity preservation | 3/15 (20%) |
| **Total** | **5/100 (5%)** |


## Leaderboard (100-example custom benchmark)

| Модель | Total | Емле/30 | Грамм/20 | Пункт/15 | Смеш/20 | Ident/15 |
|--------|-------|---------|----------|----------|---------|---------|
| **[sozkz-core-llama-600m-kk-gec-v1](https://huggingface.co/stukenov/sozkz-core-llama-600m-kk-gec-v1)** | **47%** | 15 | 12 | 3 | 2 | 15/15 |
| [sozkz-fix-qwen-500m-kk-gec-v3](https://huggingface.co/stukenov/sozkz-fix-qwen-500m-kk-gec-v3) | 38% | 0 | 16 | 9 | 0 | 13/15 |
| [sozkz-core-llama-300m-kk-gec-v4](https://huggingface.co/stukenov/sozkz-core-llama-300m-kk-gec-v4) | 37% | 9 | 6 | 4 | 3 | 15/15 |
| [sozkz-fix-qwen-500m-kk-gec-v1](https://huggingface.co/stukenov/sozkz-fix-qwen-500m-kk-gec-v1) | 35% | 0 | 12 | 8 | 0 | 15/15 |
| [sozkz-fix-qwen-500m-kk-gec-v2](https://huggingface.co/stukenov/sozkz-fix-qwen-500m-kk-gec-v2) | 30% | 0 | 11 | 7 | 0 | 12/15 |
| [sozkz-core-llama-1b-kk-gec-v1](https://huggingface.co/stukenov/sozkz-core-llama-1b-kk-gec-v1) | 16% | 2 | 6 | 1 | 0 | 7/15 |
| [sozkz-fix-qwen-500m-kk-gec-v4](https://huggingface.co/stukenov/sozkz-fix-qwen-500m-kk-gec-v4) | 5% | 0 | 1 | 4 | 0 | 0/15 |
| [sozkz-fix-mt5b-kk-gec-run13-v1](https://huggingface.co/stukenov/sozkz-fix-mt5b-kk-gec-run13-v1) | 5% | 0 | 2 | 0 | 0 | 3/15 |
| [sozkz-nllb-1b-kk-gec-v1](https://huggingface.co/stukenov/sozkz-nllb-1b-kk-gec-v1) | 1% | 0 | 1 | 0 | 0 | 0/15 |
| [sozkz-nllb-1b-kk-pretrain-v1](https://huggingface.co/stukenov/sozkz-nllb-1b-kk-pretrain-v1) | 1% | 0 | 1 | 0 | 0 | 0/15 |
| [sozkz-core-llama-300m-kk-gec-v3](https://huggingface.co/stukenov/sozkz-core-llama-300m-kk-gec-v3) | 1% | 0 | 1 | 0 | 0 | 0/15 |
| sozkz-core-llama-300m-kk-gec-v1/v2a/v2b | 0–1% | 0 | 0 | 0 | 0 | 0–1 |
| sozkz-fix-mt5-50m-kk-gec-v1 | 0% | 0 | 0 | 0 | 0 | 0/15 |