Roberta Base Turkish Uncased

🚀 RoBERTaTurkish

This is a Turkish RoBERTa base model, which provides powerful language processing capabilities for Turkish text by pre - training on a large - scale Turkish corpus.

🚀 Quick Start

This Turkish RoBERTa base model is pre - trained on Turkish Wikipedia, Turkish OSCAR, and some news websites. The final training corpus has a size of 38 GB and contains 329,720,508 sentences. Turkcell trained the model on an Intel(R) Xeon(R) Gold 6230R CPU @ 2.10GHz with 256GB RAM and 2 x GV100GL [Tesla V100 PCIe 32GB] GPU for 2.5M steps.

💻 Usage Examples

Basic Usage

Load the transformers library with the following code:

from transformers import AutoTokenizer, AutoModelForMaskedLM
  
tokenizer = AutoTokenizer.from_pretrained("TURKCELL/roberta-base-turkish-uncased")
model = AutoModelForMaskedLM.from_pretrained("TURKCELL/roberta-base-turkish-uncased")

Advanced Usage

Here is an example of using the fill - mask pipeline:

from transformers import pipeline

fill_mask = pipeline(
    "fill-mask",
    model="TURKCELL/roberta-base-turkish-uncased",
    tokenizer="TURKCELL/roberta-base-turkish-uncased"
)

fill_mask("iki ülke arasında <mask> başladı")

[{'sequence': 'iki ülke arasında savaş başladı',
  'score': 0.3013845384120941,
  'token': 1359,
  'token_str': ' savaÅŸ'},
 {'sequence': 'iki ülke arasında müzakereler başladı',
  'score': 0.1058429479598999,
  'token': 30439,
  'token_str': ' müzakereler'},
 {'sequence': 'iki ülke arasında görüşmeler başladı',
  'score': 0.07718811184167862,
  'token': 4916,
  'token_str': ' görüşmeler'},
 {'sequence': 'iki ülke arasında kriz başladı',
  'score': 0.07174749672412872,
  'token': 3908,
  'token_str': ' kriz'},
 {'sequence': 'iki ülke arasında çatışmalar başladı',
  'score': 0.05678590387105942,
  'token': 19346,
  'token_str': ' çatışmalar'}]

📄 License

This project is licensed under the MIT license.