|
--- |
|
tags: |
|
- pegasus |
|
|
|
--- |
|
# Model Card for brio-xsum-cased |
|
|
|
|
|
# Model Details |
|
|
|
## Model Description |
|
|
|
BRIO: Bringing Order to Abstractive Summarization |
|
|
|
- **Developed by:** Yale LILY Lab |
|
- **Shared by [Optional]:** Hugging Face |
|
- **Model type:** PEGASUS |
|
- **Language(s) (NLP):** Text2Text Generation |
|
- **License:** More information needed |
|
- **Related Models:** |
|
- **Parent Model:** PEGASUS |
|
- **Resources for more information:** |
|
- [Github Repo](https://github.com/Yale-LILY/BRIO) |
|
- [Associated Paper](https://arxiv.org/abs/2203.16804) |
|
- [Associated Space](https://huggingface.co/spaces/darveen/text_summarizer) |
|
|
|
|
|
# Uses |
|
|
|
## Direct Use |
|
This model can be used for the task of Text2Text Generation |
|
|
|
## Downstream Use [Optional] |
|
|
|
The model creators note in the [associated paper](https://arxiv.org/abs/2203.16804) |
|
> It is possible to apply our method in a reinforcement learning setting, where the candidate summaries are dynamically generated. |
|
|
|
## 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 recommendations. |
|
|
|
|
|
# Training Details |
|
|
|
## Training Data |
|
The model creators note in the [associated paper](https://arxiv.org/abs/2203.16804) |
|
> CNNDM4: is a large scale news dataset. |
|
Nallapati et al: we treat the news articles as the source documents and the associated highlights as the summaries. |
|
XSum5: is a highly abstractive dataset of articles from the British Broadcasting Corporation (BBC). NYT6: contains articles from the New York Times and the associated summaries |
|
|
|
## Training Procedure |
|
|
|
|
|
### Preprocessing |
|
The model creators note in the [associated paper](https://arxiv.org/abs/2203.16804) |
|
> We follow Kedzie et al. (2018) for data preprocessing and splitting, and use the associated archival abstracts as the summaries |
|
|
|
### Speeds, Sizes, Times |
|
|
|
More information needed |
|
|
|
# Evaluation |
|
|
|
|
|
## Testing Data, Factors & Metrics |
|
|
|
### Testing Data |
|
|
|
More information needed |
|
|
|
### Factors |
|
|
|
More information needed |
|
|
|
### Metrics |
|
|
|
More information needed |
|
|
|
## Results |
|
|
|
|
|
|
|
### CNNDM |
|
| | ROUGE-1 | ROUGE-2 | ROUGE-L | |
|
|----------|---------|---------|---------| |
|
| BART | 44.16 | 21.28 | 40.90 | |
|
| Ours | 47.78 | 23.55 | 44.57 | |
|
|
|
|
|
### XSum |
|
| | ROUGE-1 | ROUGE-2 | ROUGE-L | |
|
|----------|---------|---------|---------| |
|
| Pegasus | 47.21 | 24.56 | 39.25 | |
|
| Ours | 49.07 | 25.59 | 40.40 | |
|
|
|
|
|
### NYT |
|
| | ROUGE-1 | ROUGE-2 | ROUGE-L | |
|
|----------|---------|---------|---------| |
|
| BART | 55.78 | 36.61 | 52.60 | |
|
| Ours | 57.75 | 38.64 | 54.54 | |
|
|
|
|
|
|
|
# Model Examination |
|
The model creators note in the [associated paper](https://arxiv.org/abs/2203.16804) |
|
We attribute BRIO-Ctr’s superior performance to its use of the same model architecture (BART) for both candidate generation and scoring, while SimCLS uses RoBERTa as the evaluation model. As a result, BRIO-Ctr maximizes the parameter sharing between the two stages, and preserves the power of the Seq2Seq model pre-trained on the same dataset. |
|
|
|
# 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 |
|
The model creators note in the [associated paper](https://arxiv.org/abs/2203.16804) |
|
|
|
> Formulate summarization as a sequence-to-sequence (Seq2Seq) problem |
|
|
|
## Compute Infrastructure |
|
|
|
More information needed |
|
|
|
### Hardware |
|
|
|
More information needed |
|
|
|
### Software |
|
|
|
More information needed |
|
|
|
# Citation |
|
|
|
|
|
**BibTeX:** |
|
``` |
|
@misc{https://doi.org/10.48550/arxiv.2203.16804, |
|
doi = {10.48550/ARXIV.2203.16804}, |
|
|
|
url = {https://arxiv.org/abs/2203.16804}, |
|
|
|
author = {Liu, Yixin and Liu, Pengfei and Radev, Dragomir and Neubig, Graham}, |
|
|
|
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences}, |
|
|
|
title = {BRIO: Bringing Order to Abstractive Summarization}, |
|
``` |
|
|
|
|
|
|
|
# Glossary [optional] |
|
|
|
More information needed |
|
|
|
# More Information [optional] |
|
|
|
More information needed |
|
|
|
# Model Card Authors [optional] |
|
|
|
Yale LILY Lab in collaboration with Ezi Ozoani and the Hugging Face team |
|
|
|
# Model Card Contact |
|
|
|
More information needed |
|
|
|
# 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, AutoModelForSeq2SeqLM |
|
|
|
tokenizer = AutoTokenizer.from_pretrained("Yale-LILY/brio-xsum-cased") |
|
|
|
model = AutoModelForSeq2SeqLM.from_pretrained("Yale-LILY/brio-xsum-cased") |
|
|
|
``` |
|
</details> |
|
|