File size: 3,922 Bytes
3ecb39b
233f898
968454d
 
a48af2f
2e133f8
b406843
2e133f8
233f898
2e133f8
 
 
 
d9e2467
 
68d075c
d9e2467
 
968454d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
68d075c
968454d
 
 
 
 
98d19af
968454d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
---
language:
- multilingual
- pt
tags:
- xlm-roberta-large
- semantic role labeling
- finetuned
license: apache-2.0
datasets:
- PropBank.Br
metrics:
- F1 Measure
---

# XLM-R large fine-tuned on Portuguese semantic role labeling

## Model description

This model is the [`xlm-roberta-large`](https://huggingface.co/xlm-roberta-large) fine-tuned on Portuguese semantic role labeling data. This is part of a project from which resulted the following models:

* [liaad/srl-pt_bertimbau-base](https://huggingface.co/liaad/srl-pt_bertimbau-base)
* [liaad/srl-pt_bertimbau-large](https://huggingface.co/liaad/srl-pt_bertimbau-large)
* [liaad/srl-pt_xlmr-base](https://huggingface.co/liaad/srl-pt_xlmr-base)
* [liaad/srl-pt_xlmr-large](https://huggingface.co/liaad/srl-pt_xlmr-large)
* [liaad/srl-pt_mbert-base](https://huggingface.co/liaad/srl-pt_mbert-base)
* [liaad/srl-en_xlmr-base](https://huggingface.co/liaad/srl-en_xlmr-base)
* [liaad/srl-en_xlmr-large](https://huggingface.co/liaad/srl-en_xlmr-large)
* [liaad/srl-en_mbert-base](https://huggingface.co/liaad/srl-en_mbert-base)
* [liaad/srl-enpt_xlmr-base](https://huggingface.co/liaad/srl-enpt_xlmr-base)
* [liaad/srl-enpt_xlmr-large](https://huggingface.co/liaad/srl-enpt_xlmr-large)
* [liaad/srl-enpt_mbert-base](https://huggingface.co/liaad/srl-enpt_mbert-base)
* [liaad/ud_srl-pt_bertimbau-large](https://huggingface.co/liaad/ud_srl-pt_bertimbau-large)
* [liaad/ud_srl-pt_xlmr-large](https://huggingface.co/liaad/ud_srl-pt_xlmr-large)
* [liaad/ud_srl-enpt_xlmr-large](https://huggingface.co/liaad/ud_srl-enpt_xlmr-large)

For more information, please see the accompanying article (See BibTeX entry and citation info below) and the [project's github](https://github.com/asofiaoliveira/srl_bert_pt).


## Intended uses & limitations

#### How to use

To use the transformers portion of this model:
```python
from transformers import AutoTokenizer, AutoModel

tokenizer = AutoTokenizer.from_pretrained("liaad/srl-pt_xlmr-large")
model = AutoModel.from_pretrained("liaad/srl-pt_xlmr-large")
```

To use the full SRL model (transformers portion + a decoding layer), refer to the [project's github](https://github.com/asofiaoliveira/srl_bert_pt).


#### Limitations and bias

- This model does not include a Tensorflow version. This is because the "type_vocab_size" in this model was changed (from 1 to 2) and, therefore, it cannot be easily converted to Tensorflow.

## Training procedure

The model was trained on the PropBank.Br datasets, using 10-fold Cross-Validation. The 10 resulting models were tested on the folds as well as on a smaller opinion dataset "Buscapé". For more information, please see the accompanying article (See BibTeX entry and citation info below) and the [project's github](https://github.com/asofiaoliveira/srl_bert_pt).

## Eval results


| Model Name | F<sub>1</sub> CV PropBank.Br (in domain) | F<sub>1</sub> Buscapé (out of domain) |
| --------------- | ------ | ----- | 
| `srl-pt_bertimbau-base` | 76.30 | 73.33 | 
| `srl-pt_bertimbau-large` | 77.42 | 74.85 | 
| `srl-pt_xlmr-base` | 75.22 | 72.82 | 
| `srl-pt_xlmr-large` | 77.59 | 73.84 | 
| `srl-pt_mbert-base` | 72.76 | 66.89 | 
| `srl-en_xlmr-base` | 66.59 | 65.24 | 
| `srl-en_xlmr-large` | 67.60 | 64.94 | 
| `srl-en_mbert-base` | 63.07 | 58.56 | 
| `srl-enpt_xlmr-base` | 76.50 | 73.74 | 
| `srl-enpt_xlmr-large` | **78.22** | 74.55 | 
| `srl-enpt_mbert-base` | 74.88 | 69.19 | 
| `ud_srl-pt_bertimbau-large` | 77.53 | 74.49 | 
| `ud_srl-pt_xlmr-large` | 77.69 | 74.91 |
| `ud_srl-enpt_xlmr-large` | 77.97 | **75.05** | 


### BibTeX entry and citation info

```bibtex
@misc{oliveira2021transformers,
      title={Transformers and Transfer Learning for Improving Portuguese Semantic Role Labeling}, 
      author={Sofia Oliveira and Daniel Loureiro and Alípio Jorge},
      year={2021},
      eprint={2101.01213},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}
```