|
|
|
--- |
|
language: en |
|
datasets: |
|
- cuad |
|
--- |
|
|
|
# Model Card for RoBERTa Large Model fine-tuned with CUAD dataset |
|
|
|
This model is the fine-tuned version of "RoBERTa Large" using CUAD dataset |
|
|
|
|
|
|
|
|
|
|
|
# Model Details |
|
|
|
## Model Description |
|
|
|
The [Contract Understanding Atticus Dataset (CUAD)](https://www.atticusprojectai.org/cuad), pronounced "kwad", a dataset for legal contract review curated by the Atticus Project. |
|
|
|
Contract review is a task about "finding needles in a haystack." |
|
We find that Transformer models have nascent performance on CUAD, but that this performance is strongly influenced by model design and training dataset size. Despite some promising results, there is still substantial room for improvement. As one of the only large, specialized NLP benchmarks annotated by experts, CUAD can serve as a challenging research benchmark for the broader NLP community. |
|
|
|
- **Developed by:** TheAtticusProject |
|
- **Shared by [Optional]:** HuggingFace |
|
- **Model type:** Language model |
|
- **Language(s) (NLP):** en |
|
- **License:** More information needed |
|
- **Related Models:** RoBERTA |
|
- **Parent Model:**RoBERTA Large |
|
- **Resources for more information:** |
|
- [GitHub Repo](https://github.com/TheAtticusProject/cuad) |
|
- [Associated Paper](https://arxiv.org/abs/2103.06268) |
|
|
|
# Uses |
|
|
|
## Direct Use |
|
|
|
Legal contract review |
|
|
|
## Downstream Use [Optional] |
|
|
|
More information needed |
|
|
|
## Out-of-Scope Use |
|
|
|
|
|
The model should not be used to intentionally create hostile or alienating environments for people. |
|
|
|
# Bias, Risks, and Limitations |
|
|
|
Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups. |
|
|
|
|
|
## Recommendations |
|
|
|
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recomendations. |
|
|
|
|
|
# Training Details |
|
|
|
## Training Data |
|
See [cuad dataset card](https://huggingface.co/datasets/cuad) for further details |
|
|
|
## Training Procedure |
|
|
|
More information needed |
|
|
|
### Preprocessing |
|
|
|
More information needed |
|
|
|
### Speeds, Sizes, Times |
|
|
|
More information needed |
|
|
|
# Evaluation |
|
|
|
|
|
|
|
## Testing Data, Factors & Metrics |
|
|
|
### Testing Data |
|
#### Extra Data |
|
Researchers may be interested in several gigabytes of unlabeled contract pretraining data, which is available [here](https://drive.google.com/file/d/1of37X0hAhECQ3BN_004D8gm6V88tgZaB/view?usp=sharing). |
|
|
|
### Factors |
|
|
|
More information needed |
|
|
|
### Metrics |
|
|
|
More information needed |
|
|
|
## Results |
|
|
|
|
|
|
|
|
|
We [provide checkpoints](https://zenodo.org/record/4599830) for three of the best models fine-tuned on CUAD: RoBERTa-base (~100M parameters), RoBERTa-large (~300M parameters), and DeBERTa-xlarge (~900M parameters). |
|
|
|
|
|
|
|
|
|
# Model Examination |
|
|
|
More information needed |
|
|
|
# Environmental Impact |
|
|
|
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). |
|
|
|
- **Hardware Type:** More information needed |
|
- **Hours used:** More information needed |
|
- **Cloud Provider:** More information needed |
|
- **Compute Region:** More information needed |
|
- **Carbon Emitted:** More information needed |
|
|
|
# Technical Specifications [optional] |
|
|
|
## Model Architecture and Objective |
|
|
|
More information needed |
|
|
|
## Compute Infrastructure |
|
|
|
More information needed |
|
|
|
### Hardware |
|
|
|
More information needed |
|
|
|
### Software |
|
|
|
The HuggingFace [Transformers](https://huggingface.co/transformers) library. It was tested with Python 3.8, PyTorch 1.7, and Transformers 4.3/4.4. |
|
|
|
# Citation |
|
|
|
|
|
**BibTeX:** |
|
|
|
@article{hendrycks2021cuad, |
|
title={CUAD: An Expert-Annotated NLP Dataset for Legal Contract Review}, |
|
author={Dan Hendrycks and Collin Burns and Anya Chen and Spencer Ball}, |
|
journal={NeurIPS}, |
|
year={2021} |
|
} |
|
|
|
|
|
|
|
# Glossary [optional] |
|
|
|
More information needed |
|
|
|
# More Information [optional] |
|
|
|
For more details about CUAD and legal contract review, see the [Atticus Project website](https://www.atticusprojectai.org/cuad). |
|
|
|
# Model Card Authors [optional] |
|
|
|
TheAtticusProject |
|
|
|
# Model Card Contact |
|
|
|
[TheAtticusProject](https://www.atticusprojectai.org/), in collaboration with the Ezi Ozoani and the HuggingFace Team |
|
|
|
|
|
# How to Get Started with the Model |
|
|
|
Use the code below to get started with the model. |
|
|
|
<details> |
|
<summary> Click to expand </summary> |
|
|
|
```python |
|
|
|
from transformers import AutoTokenizer, AutoModelForQuestionAnswering |
|
|
|
tokenizer = AutoTokenizer.from_pretrained("akdeniz27/roberta-large-cuad") |
|
|
|
model = AutoModelForQuestionAnswering.from_pretrained("akdeniz27/roberta-large-cuad") |
|
``` |
|
|
|
|
|
</details> |
|
|
|
|
