πŸ€— + πŸ“š dbmdz Turkish ELECTRA model

In this repository the MDZ Digital Library team (dbmdz) at the Bavarian State Library open sources a cased ELECTRA small model for Turkish πŸŽ‰

Turkish ELECTRA model

We release a small ELECTRA model for Turkish, that was trained on the same data as BERTurk.

ELECTRA is a new method for self-supervised language representation learning. It can be used to pre-train transformer networks using relatively little compute. ELECTRA models are trained to distinguish "real" input tokens vs "fake" input tokens generated by another neural network, similar to the discriminator of a GAN.

More details about ELECTRA can be found in the ICLR paper or in the official ELECTRA repository on GitHub.

Stats

The current version of the model is trained on a filtered and sentence segmented version of the Turkish OSCAR corpus, a recent Wikipedia dump, various OPUS corpora and a special corpus provided by Kemal Oflazer.

The final training corpus has a size of 35GB and 44,04,976,662 tokens.

Thanks to Google's TensorFlow Research Cloud (TFRC) we could train a cased model on a TPU v3-8 for 1M steps.

Model weights

Transformers compatible weights for both PyTorch and TensorFlow are available.

Model Downloads
dbmdz/electra-small-turkish-cased-discriminator config.json β€’ pytorch_model.bin β€’ vocab.txt

Usage

With Transformers >= 2.8 our ELECTRA small cased model can be loaded like:

from transformers import AutoModelWithLMHead, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("dbmdz/electra-small-turkish-cased-discriminator")
model = AutoModelWithLMHead.from_pretrained("dbmdz/electra-small-turkish-cased-discriminator")

Results

For results on PoS tagging or NER tasks, please refer to this repository.

Huggingface model hub

All models are available on the Huggingface model hub.

Contact (Bugs, Feedback, Contribution and more)

For questions about our ELECTRA models just open an issue here πŸ€—

Acknowledgments

Thanks to Kemal Oflazer for providing us additional large corpora for Turkish. Many thanks to Reyyan Yeniterzi for providing us the Turkish NER dataset for evaluation.

Research supported with Cloud TPUs from Google's TensorFlow Research Cloud (TFRC). Thanks for providing access to the TFRC ❀️

Thanks to the generous support from the Hugging Face team, it is possible to download both cased and uncased models from their S3 storage πŸ€—

Downloads last month
140
Safetensors
Model size
13.7M params
Tensor type
F32
Β·
Inference API
Unable to determine this model’s pipeline type. Check the docs .