---
license: mit
language:
- es
metrics:
- accuracy
pipeline_tag: fill-mask
widget:
- text: Vamos a comer unos [MASK]
  example_title: "Vamos a comer unos tacos"
tags:
- code
- nlp
- custom
- bilma
tokenizer:
- yes
---
# BILMA (Bert In Latin aMericA)

Bilma is a BERT implementation in tensorflow and trained on the Masked Language Model task under the https://sadit.github.io/regional-spanish-models-talk-2022/ datasets.

The accuracy of the models trained on the MLM task for different regions are:

![bilma-mlm-comp](https://user-images.githubusercontent.com/392873/163045798-89bd45c5-b654-4f16-b3e2-5cf404e12ddd.png)

# Pre-requisites

You will need TensorFlow 2.4 or newer.

# Quick guide

You can see the demo notebooks for a quick guide on how to use the models.

Clone this repository and then run 
```
bash download-emoji15-bilma.sh
```

to download the MX model. Then to load the model you can use the code:
```
from bilma import bilma_model
vocab_file = "vocab_file_All.txt"
model_file = "bilma_small_MX_epoch-1_classification_epochs-13.h5"
model = bilma_model.load(model_file)
tokenizer = bilma_model.tokenizer(vocab_file=vocab_file,
max_length=280)
```

Now you will need some text:
```
texts = ["Tenemos tres dias sin internet ni senal de celular en el pueblo.",
         "Incomunicados en el siglo XXI tampoco hay servicio de telefonia fija",
         "Vamos a comer unos tacos",
         "Los del banco no dejan de llamarme"]
toks = tokenizer.tokenize(texts)
```

With this, you are ready to use the model
```
p = model.predict(toks)
tokenizer.decode_emo(p[1])
```

which produces the output: ![emoji-output](https://user-images.githubusercontent.com/392873/165176270-77dd32ca-377e-4d29-ab4a-bc5f75913241.jpg)
each emoji correspond to each entry in `texts`.