Edit model card

TajBERTo: RoBERTa-like Language model trained on Tajik

First ever Tajik NLP model 🔥

Dataset:

This model was trained on filtered and merged version of Leipzig Corpora https://wortschatz.unileipzig.de/en/download/Tajik

Intended use

You can use the raw model for masked text generation or fine-tune it to a downstream task.

Example pipeline

from transformers import pipeline
fill_mask = pipeline(
    "fill-mask",
    model="muhtasham/TajBERTo",
    tokenizer="muhtasham/TajBERTo"
)
fill_mask("Пойтахти <mask> Душанбе")

# This is the beginning of a beautiful <mask>.

{'score': 0.1952248513698578,
  'sequence': 'Пойтахти шаҳри Душанбе',
  'token': 710,
  'token_str': ' шаҳри'},
 {'score': 0.029092855751514435,
  'sequence': 'Пойтахти дар Душанбе',
  'token': 310,
  'token_str': ' дар'},
 {'score': 0.020065447315573692,
  'sequence': 'Пойтахти Душанбе Душанбе',
  'token': 717,
  'token_str': ' Душанбе'},
 {'score': 0.016725927591323853,
  'sequence': 'Пойтахти Тоҷикистон Душанбе',
  'token': 424,
  'token_str': ' Тоҷикистон'},
 {'score': 0.011400512419641018,
  'sequence': 'Пойтахти аз Душанбе',
  'token': 335,
  'token_str': ' аз'}
  
Downloads last month
22
Safetensors
Model size
83.5M params
Tensor type
I64
·
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Collection including muhtasham/TajBERTo