BERT BASE (cased)

Pretrained model on Bulgarian language using a masked language modeling (MLM) objective. It was introduced in this paper and first released in this repository. This model is cased: it does make a difference between bulgarian and Bulgarian.

Model description

The model was trained similarly to RuBert wherein the Multilingual Bert was adapted for the Russian language.

The training data was Bulgarian text from OSCAR, Chitanka and Wikipedia.

How to use

Here is how to use this model in PyTorch:

>>> from transformers import pipeline
>>> 
>>> model = pipeline(
>>>     'fill-mask',
>>>     model='rmihaylov/bert-base-bg',
>>>     tokenizer='rmihaylov/bert-base-bg',
>>>     device=0,
>>>     revision=None)
>>> output = model("София е [MASK] на България.")
>>> print(output)

[{'score': 0.12665307521820068,
  'sequence': 'София е  столица на България.',
  'token': 2659,
  'token_str': 'столица'},
 {'score': 0.07470757514238358,
  'sequence': 'София е Перлата на България.',
  'token': 102146,
  'token_str': 'Перлата'},
 {'score': 0.06786204129457474,
  'sequence': 'София е  Столицата на България.',
  'token': 45495,
  'token_str': 'Столицата'},
 {'score': 0.05533991754055023,
  'sequence': 'София е  Столица на България.',
  'token': 100524,
  'token_str': 'Столица'},
 {'score': 0.05485989898443222,
  'sequence': 'София е  столицата на България.',
  'token': 2294,
  'token_str': 'столицата'}]