---
language: en
tags:
- AMRBART
license: mit
---

## AMRBART-large-finetuned-AMR3.0-AMRParsing

This model is a fine-tuned version of [AMRBART-large](https://huggingface.co/xfbai/AMRBART-large) on an AMR3.0 dataset. It achieves a Smatch of 84.2 on the evaluation set: More details are introduced in the paper: [Graph Pre-training for AMR Parsing and Generation](https://arxiv.org/pdf/2203.07836.pdf) by bai et al. in ACL 2022.

## Model description
Same with AMRBART.

## Training data

The model is finetuned on [AMR3.0](https://catalog.ldc.upenn.edu/LDC2020T02), a dataset consisting of 55,635
training instances, 1,722 validation instances, and 1,898 test instances.

## Intended uses & limitations

You can use the model for AMR parsing, but it's mostly intended to be used in the domain of News. 

## How to use
Here is how to initialize this model in PyTorch:

```python
from transformers import BartForConditionalGeneration
model = BartForConditionalGeneration.from_pretrained("xfbai/AMRBART-large-finetuned-AMR3.0-AMRParsing")
```
Please refer to [this repository](https://github.com/muyeby/AMRBART) for tokenizer initialization and data preprocessing.


## BibTeX entry and citation info
Please cite this paper if you find this model helpful

```bibtex
@inproceedings{bai-etal-2022-graph,
    title = "Graph Pre-training for {AMR} Parsing and Generation",
    author = "Bai, Xuefeng  and
      Chen, Yulong and
      Zhang, Yue",
    booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = may,
    year = "2022",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "todo",
    doi = "todo",
    pages = "todo"
}
```