README.md

---
language: 
  - pt
thumbnail: "Portugues SBERT for the Legal Domain"
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- transformers
datasets:
- assin
- assin2
---

# rufimelo/Legal-SBERTimbau-large

This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 1024 dimensional dense vector space and can be used for tasks like clustering or semantic search.

<!--- Describe your model here -->

## Usage (Sentence-Transformers)

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('rufimelo/Legal-SBERTimbau-large')
embeddings = model.encode(sentences)
print(embeddings)
```



## Usage (HuggingFace Transformers)
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.

```python
from transformers import AutoTokenizer, AutoModel
import torch


#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)


# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']

# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('rufimelo/Legal-SBERTimbau-large')
model = AutoModel.from_pretrained('rufimelo/Legal-SBERTimbau-large}')

# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')

# Compute token embeddings
with torch.no_grad():
    model_output = model(**encoded_input)

# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])

print("Sentence embeddings:")
print(sentence_embeddings)
```


## Evaluation Results STS

| Model| Dataset | PearsonCorrelation        | 
| ---------------------------------------- | ---------- |
| Legal-SBERTimbau-large| Assin | 0.766293861 |
| Legal-SBERTimbau-large| Assin2| 0.823565322 |
| ---------------------------------------- | ---------- |
| paraphrase-multilingual-mpnet-base-v2| Assin | 0.743740222 |
| paraphrase-multilingual-mpnet-base-v2| Assin2| 0.823565322 |
| paraphrase-multilingual-mpnet-base-v2 Fine tuned with assin(s)| Assin | 0.77641  |
| paraphrase-multilingual-mpnet-base-v2 Fine tuned with assin(s)| Assin2| 0.79831 |


## Training

Legal-SBERTimbau-large is based on Legal-BERTimbau-large whioch derives from [BERTimbau](https://huggingface.co/neuralmind/bert-base-portuguese-cased) Large.
It was trained using the multilingual knowledge distillation process, meaning it was trained as a multilingual model. This was chosen due to the lack of Portuguese available data.
In addition to that, it was submitted to a fine tuning stage with the [assin](https://huggingface.co/datasets/assin) and [assin2](https://huggingface.co/datasets/assin2) datasets.

## Full Model Architecture
```
SentenceTransformer(
  (0): Transformer({'max_seq_length': 75, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False})
)
```

## Citing & Authors

<!--- Describe where people can find more information -->