codeswitching hindi-english language-identification

codeswitch-hineng-lid-lince

This is a pretrained model for language identification of hindi-english code-mixed data used from LinCE

This model is trained for this below repository.

https://github.com/sagorbrur/codeswitch

To install codeswitch:

pip install codeswitch

Identify Language


from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline

tokenizer = AutoTokenizer.from_pretrained("sagorsarker/codeswitch-hineng-lid-lince")

model = AutoModelForTokenClassification.from_pretrained("sagorsarker/codeswitch-hineng-lid-lince")
lid_model = pipeline('ner', model=model, tokenizer=tokenizer)

lid_model("put any hindi english code-mixed sentence")


from codeswitch.codeswitch import LanguageIdentification
lid = LanguageIdentification('hin-eng') 
text = "" # your code-mixed sentence 
result = lid.identify(text)
print(result)