README.md

---
license: apache-2.0
language:
- en
metrics:
- spearmanr
base_model:
- FacebookAI/roberta-large
---

# Model Card for GCSE

<p align="center">
    <p align="center">
        <a href="https://github.com/aleversn/GCSE">
            <img alt="Static Badge" src="https://img.shields.io/badge/GCSE-black?logo=github">
        </a>
    </p>
</p>

[Model](https://huggingface.co/aleversn/GCSE-RoBERTa-large/) | [Paper](https://arxiv.org/abs/2409.12887) | [Code](https://github.com/aleversn/GCSE)

### Model Checkpoints

We release our model checkpoints in huggingface as listed below:
| Model                                                                             | Avg. STS |
| :-------------------------------------------------------------------------------- | :------: |
| [aleversn/GCSE-BERT-base](https://huggingface.co/aleversn/GCSE-BERT-base)         |  81.98   |
| [aleversn/GCSE-BERT-large](https://huggingface.co/aleversn/GCSE-BERT-large)       |  83.07   |
| [aleversn/GCSE-RoBERTa-base](https://huggingface.co/aleversn/GCSE-RoBERTa-base)   |  82.12   |
| [aleversn/GCSE-RoBERTa-large](https://huggingface.co/aleversn/GCSE-RoBERTa-large) |  83.82   |

### Usage

```python
from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained("aleversn/GCSE-BERT-base")
model = AutoModel.from_pretrained("aleversn/GCSE-RoBERTa-large")
```