Metadata-Version: 2.1
Name: TextPunctuator
Version: 1.0.3
Summary: A package to punctuate text, currently supporting French text, more languages comming soon.
Home-page: https://github.com/rouzki/TextPunctuator
Author: Zakarya ROUZKI
Author-email: zakaryarouzki@gmail.com
License: UNKNOWN
Keywords: punctuator,nlp,text,transformers,t5
Platform: UNKNOWN
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.6
Description-Content-Type: text/markdown

# Text Punctuator Based on Transformers model T5.
T5 model fine-tuned for punctuation restoration.
Model currently supports only French Language. More language supports will be added later using mT5.

Train Datasets : 
Model trained using 2 french datasets (around 500k records): 
- [orange_sum](https://huggingface.co/datasets/orange_sum) 
- [mlsum](https://huggingface.co/datasets/mlsum) (only french text)


more info will be added later.

---------------------------
## ðŸš€ Usage
**Below is a quick way to get up and running with the model.**
1. First, install the package.
```bash
pip install TextPunctuator
```
2. Sample python code.
```python

from Punctuator import TextPunctuator

punctuator = TextPunctuator(use_gpu=False)

text = "Sur la base de ces Ã©changes Blake Lemoine a donc jugÃ© que le systÃ¨me avait atteint un niveau de conscience lui permettant dâ€™Ãªtre sensible Ce dernier a ensuite envoyÃ© par email un rapport sur la sensibilitÃ© supposÃ©e de LaMDA Ã  deux cents employÃ©s de Google TrÃ¨s vite les dirigeants de lâ€™entreprise ont rejetÃ© les allÃ©gations"

text_punctuated = punctuator.punctuate(text, lang='fr')

text_punctuated
# Outputs the following:
# Sur la base de ces Ã©changes, Blake Lemoine a donc jugÃ© que le systÃ¨me avait atteint un niveau de conscience lui permettant dâ€™Ãªtre sensible. Ce dernier a ensuite envoyÃ© par email un rapport sur la sensibilitÃ© supposÃ©e de LaMDA Ã  deux cents employÃ©s de Google. TrÃ¨s vite, les dirigeants de lâ€™entreprise ont rejetÃ© les allÃ©gations.
```

-----------------------------------------------
## â˜• Contact 
Contact [Zakarya ROUZKI ](zakaryarouzki@gmail.com) or at [Linkedin](linkedin.com/in/rouzki).

-----------------------------------------------

