79 lines
2.8 KiB
Markdown
79 lines
2.8 KiB
Markdown
---
|
|
language:
|
|
- sk
|
|
pipeline_tag: sentence-similarity
|
|
tags:
|
|
- sentence-transformers
|
|
- feature-extraction
|
|
- sentence-similarity
|
|
- sts
|
|
license: cc-by-4.0
|
|
datasets:
|
|
- glue
|
|
metrics:
|
|
- spearmanr
|
|
widget:
|
|
- source_sentence: Izrael uskutočnil letecké údery v blízkosti Damasku.
|
|
sentences:
|
|
- Izrael uskutočnil vzdušný útok na Sýriu.
|
|
- Pes leží na gauči a má hlavu na bielom vankúši.
|
|
base_model:
|
|
- gerulata/slovakbert
|
|
library_name: sentence-transformers
|
|
---
|
|
|
|
|
|
# Sentence similarity model based on SlovakBERT
|
|
|
|
This is a sentence similarity model based on [SlovakBERT](https://huggingface.co/gerulata/slovakbert). The model was fine-tuned using [STSbenchmark](https://ixa2.si.ehu.eus/stswiki/index.php/STSbenchmark) [Cer et al 2017] translated to Slovak using [M2M100](https://huggingface.co/facebook/m2m100_1.2B). The model can be used as an universal sentence encoder for Slovak sentences.
|
|
|
|
## Results
|
|
|
|
The model was evaluated in [our paper](https://arxiv.org/abs/2109.15254) [Pikuliak et al 2021, Section 4.3]. It achieves \\(0.791\\) Spearman correlation on STSbenchmark test set.
|
|
|
|
|
|
## Usage
|
|
|
|
Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
|
|
|
|
```
|
|
pip install -U sentence-transformers
|
|
```
|
|
|
|
Then you can use the model like this:
|
|
|
|
```python
|
|
from sentence_transformers import SentenceTransformer
|
|
sentences = ["This is an example sentence", "Each sentence is converted"]
|
|
|
|
model = SentenceTransformer('kinit/slovakbert-sts-stsb')
|
|
embeddings = model.encode(sentences)
|
|
print(embeddings)
|
|
```
|
|
|
|
|
|
## Cite
|
|
|
|
```
|
|
@inproceedings{pikuliak-etal-2022-slovakbert,
|
|
title = "{S}lovak{BERT}: {S}lovak Masked Language Model",
|
|
author = "Pikuliak, Mat{\'u}{\v{s}} and
|
|
Grivalsk{\'y}, {\v{S}}tefan and
|
|
Kon{\^o}pka, Martin and
|
|
Bl{\v{s}}t{\'a}k, Miroslav and
|
|
Tamajka, Martin and
|
|
Bachrat{\'y}, Viktor and
|
|
Simko, Marian and
|
|
Bal{\'a}{\v{z}}ik, Pavol and
|
|
Trnka, Michal and
|
|
Uhl{\'a}rik, Filip",
|
|
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2022",
|
|
month = dec,
|
|
year = "2022",
|
|
address = "Abu Dhabi, United Arab Emirates",
|
|
publisher = "Association for Computational Linguistics",
|
|
url = "https://aclanthology.org/2022.findings-emnlp.530",
|
|
pages = "7156--7168",
|
|
abstract = "We introduce a new Slovak masked language model called \textit{SlovakBERT}. This is to our best knowledge the first paper discussing Slovak transformers-based language models. We evaluate our model on several NLP tasks and achieve state-of-the-art results. This evaluation is likewise the first attempt to establish a benchmark for Slovak language models. We publish the masked language model, as well as the fine-tuned models for part-of-speech tagging, sentiment analysis and semantic textual similarity.",
|
|
}
|
|
``` |