Update README.md

README.md

@@ -114,95 +114,3 @@ size_categories:
 - 10K<n<100K
 license: mit
 ---
-
-# Multilingual COMET Translation Quality Scores
-
-This dataset contains COMET (Crosslingual Optimized Metric for Evaluation of Translation) quality scores for English-to-multilingual translations of the GSM8K dataset.
-
-## Dataset Description
-
-The dataset includes translation quality scores for 17 language pairs, all with English as the source language:
-
-- **English → Amharic** (አማርኛ)
-- **English → Ewe** (Eʋegbe)
-- **English → French** (Français)
-- **English → Hausa** (Harshen Hausa)
-- **English → Igbo** (Asụsụ Igbo)
-- **English → Kinyarwanda** (Ikinyarwanda)
-- **English → Lingala** (Lingála)
-- **English → Luganda** (Luganda)
-- **English → Oromo** (Afaan Oromoo)
-- **English → Shona** (chiShona)
-- **English → Southern Sotho** (Sesotho)
-- **English → Swahili** (Kiswahili)
-- **English → Twi** (Twi)
-- **English → Wolof** (Wolof)
-- **English → Xhosa** (isiXhosa)
-- **English → Yoruba** (Yorùbá)
-- **English → Zulu** (isiZulu)
-
-## Model Used
-
-The scores were generated using the **McGill-NLP/ssa-comet-qe** model, which is specifically designed for quality estimation of translations involving Sub-Saharan African languages.
-
-## Data Format
-
-Each CSV file contains three columns:
-- `english`: Original English text
-- `{target_language}`: Translated text in the target language
-- `score`: COMET quality score (higher scores indicate better translation quality)
-
-## Usage
-
-```python
-from datasets import load_dataset
-
-# Load specific language pair
-dataset = load_dataset("Tadesse/COMET_score", "english_swahili")
-
-# Load all language pairs
-configs = ["english_amharic", "english_ewe", "english_french", "english_hausa", 
-           "english_igbo", "english_kinyarwanda", "english_lingala", "english_luganda",
-           "english_oromo", "english_shona", "english_sotho", "english_swahili", 
-           "english_twi", "english_wolof", "english_xhosa", "english_yoruba", "english_zulu"]
-
-all_datasets = {}
-for config in configs:
-    all_datasets[config] = load_dataset("Tadesse/COMET_score", config)
-```
-
-## Score Statistics
-
-COMET scores typically range from 0 to 1, where:
-- **0.8-1.0**: Excellent translation quality
-- **0.6-0.8**: Good translation quality
-- **0.4-0.6**: Moderate translation quality
-- **0.2-0.4**: Poor translation quality
-- **0.0-0.2**: Very poor translation quality
-
-## Source Data
-
-The translations are based on the GSM8K dataset (`israel/translated_gsm8k`), which contains grade school math word problems translated into multiple languages.
-
-## Citation
-
-If you use this dataset in your research, please cite:
-
-```bibtex
-@dataset{multilingual_comet_scores,
-  title={Multilingual COMET Translation Quality Scores},
-  author={[Your Name]},
-  year={2025},
-  url={https://huggingface.co/datasets/Tadesse/COMET_score}
-}
-```
-
-## License
-
-This dataset is released under the MIT License.
-
-## Acknowledgments
-
-- Original GSM8K dataset creators
-- McGill NLP Lab for the COMET model
-- Hugging Face for hosting the translated GSM8K dataset