README.md

---
license: mit
datasets:
- best2009
- scb_mt_enth_2020
- oscar
- wikipedia
language:
- th
widget:
  - text: วัน ที่ _ 12 _ มีนาคม นี้ _ ฉัน จะ ไป <mask> วัดพระแก้ว _ ที่ กรุงเทพ
library_name: transformers
---
# HoogBERTa

This repository includes the Thai pretrained language representation (HoogBERTa_base) and can be used for **Feature Extraction and Masked Language Modeling Tasks**.


# Documentation


## Prerequisite
Since we use subword-nmt BPE encoding, input needs to be pre-tokenize using [BEST](https://huggingface.co/datasets/best2009) standard before inputting into HoogBERTa 
```
pip install attacut
```

## Getting Start
To initialize the model from hub, use the following commands
```python
from transformers import AutoTokenizer, AutoModel
from attacut import tokenize
import torch

tokenizer = AutoTokenizer.from_pretrained("new5558/HoogBERTa")
model = AutoModel.from_pretrained("new5558/HoogBERTa")
```

To extract token features, based on the RoBERTa architecture, use the following commands

```python
model.eval()
sentence = "วันที่ 12 มีนาคมนี้ ฉันจะไปเที่ยววัดพระแก้ว ที่กรุงเทพ"
all_sent = []
sentences = sentence.split(" ")
for sent in sentences:
    all_sent.append(" ".join(tokenize(sent)).replace("_","[!und:]"))

sentence = " _ ".join(all_sent)
tokenized_text = tokenizer(sentence, return_tensors = 'pt')
token_ids = tokenized_text['input_ids']

with torch.no_grad():
  features = model(**tokenized_text, output_hidden_states = True).hidden_states[-1]
```

For batch processing,

```python
model.eval()
sentenceL = ["วันที่ 12 มีนาคมนี้","ฉันจะไปเที่ยววัดพระแก้ว ที่กรุงเทพ"]
inputList = []
for sentX in sentenceL:
  sentences = sentX.split(" ")
  all_sent = []
  for sent in sentences:
      all_sent.append(" ".join(tokenize(sent)).replace("_","[!und:]"))

  sentence = " _ ".join(all_sent)
  inputList.append(sentence)
tokenized_text = tokenizer(inputList, padding = True, return_tensors = 'pt')
token_ids = tokenized_text['input_ids']

with torch.no_grad():
    features = model(**tokenized_text, output_hidden_states = True).hidden_states[-1]
```

To use HoogBERTa as an embedding layer, use

```python
with torch.no_grad():
  features = model(token_ids, output_hidden_states = True).hidden_states[-1] # where token_ids is a tensor with type "long".
```

# Huggingface Models
1. `HoogBERTaEncoder`
 - [HoogBERTa](https://huggingface.co/new5558/HoogBERTa): `Feature Extraction` and `Mask Language Modeling`
2. `HoogBERTaMuliTaskTagger`:
 - [HoogBERTa-NER-lst20](https://huggingface.co/new5558/HoogBERTa-NER-lst20): `Named-entity recognition (NER)` based on LST20
 - [HoogBERTa-POS-lst20](https://huggingface.co/new5558/HoogBERTa-POS-lst20): `Part-of-speech tagging (POS)` based on LST20
 - [HoogBERTa-SENTENCE-lst20](https://huggingface.co/new5558/HoogBERTa-SENTENCE-lst20): `Clause Boundary Classification` based on LST20


# Citation

Please cite as:

``` bibtex
@inproceedings{porkaew2021hoogberta,
  title = {HoogBERTa: Multi-task Sequence Labeling using Thai Pretrained Language Representation},
  author = {Peerachet Porkaew, Prachya Boonkwan and Thepchai Supnithi},
  booktitle = {The Joint International Symposium on Artificial Intelligence and Natural Language Processing (iSAI-NLP 2021)},
  year = {2021},
  address={Online}
}
```

Download full-text [PDF](https://drive.google.com/file/d/1hwdyIssR5U_knhPE2HJigrc0rlkqWeLF/view?usp=sharing)  
Check out the code on [Github](https://github.com/lstnlp/HoogBERTa)