ruanchaves
commited on
Commit
•
4ce28bd
1
Parent(s):
fe74db9
Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
inference: false
|
3 |
+
language: pt
|
4 |
+
datasets:
|
5 |
+
- assin
|
6 |
+
---
|
7 |
+
|
8 |
+
|
9 |
+
# BERTimbau base for Semantic Textual Similarity
|
10 |
+
|
11 |
+
This is the [neuralmind/bert-base-portuguese-cased](https://huggingface.co/neuralmind/bert-base-portuguese-cased) model finetuned for
|
12 |
+
Semantic Textual Similarity with the [ASSIN](huggingface.co/datasets/assin) dataset.
|
13 |
+
This model is suitable for Portuguese.
|
14 |
+
|
15 |
+
- Git Repo: [Evaluation of Portuguese Language Models](https://github.com/ruanchaves/eplm).
|
16 |
+
- Demo: [Portuguese Semantic Similarity](https://ruanchaves-portuguese-semantic-similarity.hf.space)
|
17 |
+
|
18 |
+
## Full regression example
|
19 |
+
|
20 |
+
```python
|
21 |
+
from transformers import AutoModelForSequenceClassification, AutoTokenizer, AutoConfig
|
22 |
+
import numpy as np
|
23 |
+
import torch
|
24 |
+
|
25 |
+
model_name = "neuralmind/bert-base-portuguese-cased"
|
26 |
+
s1 = "A gente faz o aporte financeiro, é como se a empresa fosse parceira do Monte Cristo."
|
27 |
+
s2 = "Fernando Moraes afirma que não tem vínculo com o Monte Cristo além da parceira."
|
28 |
+
model = AutoModelForSequenceClassification.from_pretrained(model_name)
|
29 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
30 |
+
config = AutoConfig.from_pretrained(model_name)
|
31 |
+
model_input = tokenizer(*([s1], [s2]), padding=True, return_tensors="pt")
|
32 |
+
with torch.no_grad():
|
33 |
+
output = model(**model_input)
|
34 |
+
score = output[0][0].detach().numpy().item()
|
35 |
+
print(f"Similarity Score: {np.round(float(score), 4)}")
|
36 |
+
```
|
37 |
+
|
38 |
+
Output:
|
39 |
+
|
40 |
+
```
|
41 |
+
Similarity Score: 1.9335
|
42 |
+
```
|
43 |
+
|
44 |
+
## Citation
|
45 |
+
|
46 |
+
Our research is ongoing, and we are currently working on describing our experiments in a paper, which will be published soon.
|
47 |
+
In the meanwhile, if you would like to cite our work or models before the publication of the paper, please cite our [GitHub repository](https://github.com/ruanchaves/eplm):
|
48 |
+
|
49 |
+
```
|
50 |
+
@software{Chaves_Rodrigues_eplm_2023,
|
51 |
+
author = {Chaves Rodrigues, Ruan and Tanti, Marc and Agerri, Rodrigo},
|
52 |
+
doi = {10.5281/zenodo.7781848},
|
53 |
+
month = {3},
|
54 |
+
title = {{eplm}},
|
55 |
+
url = {https://github.com/ruanchaves/eplm},
|
56 |
+
version = {1.0.0},
|
57 |
+
year = {2023}
|
58 |
+
}
|
59 |
+
```
|