Bert Base Turkish Ner Cased

🚀 Easy-to-use NER Application for Turkish Language

This is an easy-to-use NER (Named Entity Recognition) application for the Turkish language. It's a Python NER model (Bert + Transfer Learning) for Turkish.

📚 Citation

Please cite the following works if you use this model in your study:

@misc{yildirim2024finetuning,
      title={Fine-tuning Transformer-based Encoder for Turkish Language Understanding Tasks}, 
      author={Savas Yildirim},
      year={2024},
      eprint={2401.17396},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

@book{yildirim2021mastering,
  title={Mastering Transformers: Build state-of-the-art models from scratch with advanced natural language processing techniques},
  author={Yildirim, Savas and Asgari-Chenaghlu, Meysam},
  year={2021},
  publisher={Packt Publishing Ltd}
}

📦 Installation

Thanks to @stefan-it, the following steps can be used for training:

cd tr-data
for file in train.txt dev.txt test.txt labels.txt
do
  wget https://schweter.eu/storage/turkish-bert-wikiann/$file
done
cd ..

This will download the pre-processed datasets with training, development, and test splits and put them in a tr-data folder.

🔧 Pre-training

After downloading the dataset, pre-training can be started. Just set the following environment variables:

export MAX_LENGTH=128
export BERT_MODEL=dbmdz/bert-base-turkish-cased 
export OUTPUT_DIR=tr-new-model
export BATCH_SIZE=32
export NUM_EPOCHS=3
export SAVE_STEPS=625
export SEED=1

Then run pre-training:

python3 run_ner_old.py --data_dir ./tr-data3 \
--model_type bert \
--labels ./tr-data/labels.txt \
--model_name_or_path $BERT_MODEL \
--output_dir $OUTPUT_DIR-$SEED \
--max_seq_length $MAX_LENGTH \
--num_train_epochs $NUM_EPOCHS \
--per_gpu_train_batch_size $BATCH_SIZE \
--save_steps $SAVE_STEPS \
--seed $SEED \
--do_train \
--do_eval \
--do_predict \
--fp16

💻 Usage Examples

Basic Usage

from transformers import pipeline, AutoModelForTokenClassification, AutoTokenizer
model = AutoModelForTokenClassification.from_pretrained("savasy/bert-base-turkish-ner-cased")
tokenizer = AutoTokenizer.from_pretrained("savasy/bert-base-turkish-ner-cased")
ner = pipeline('ner', model=model, tokenizer=tokenizer)
ner("Mustafa Kemal Atatürk 19 Mayıs 1919'da Samsun'a ayak bastı.")

📊 Some Results

Data1: For the data above

Eval Results:

• precision = 0.916400580551524
• recall = 0.9342309684101502
• f1 = 0.9252298787412536
• loss = 0.11335893666411284

Test Results:

• precision = 0.9192058759362955
• recall = 0.9303010230367262
• f1 = 0.9247201697271198
• loss = 0.11182546521618497

Data2:

https://github.com/stefan-it/turkish-bert/files/4558187/nerdata.txt The performance for the data given by @kemalaraz is as follows:

savas@savas-lenova:~/Desktop/trans/tr-new-model-1$ cat eval_results.txt
* precision = 0.9461980692049029
* recall = 0.959309358847465
* f1 = 0.9527086063783312
* loss = 0.037054269206847804

savas@savas-lenova:~/Desktop/trans/tr-new-model-1$ cat test_results.txt
* precision = 0.9458370635631155
* recall = 0.9588201928530913
* f1 = 0.952284378344882
* loss = 0.035431676572445225