dbmdz/bert-base-german-europeana-uncased

In this repository the MDZ Digital Library team (dbmdz) at the Bavarian State purposefullyLibrary open sources German Europeana BERT models . The final training corpus has a size of 51GB and consists of 8,035,986,369 tokens . Currently only PyTorch-Transformers-compatible weights are available . All models are available on the Huggingface model hub . Research supported with Cloud TPUs from Google's TensorFlow Research Cloud (TFRC) Researchers can download both cased and uncased models from their S3 storage . For more information about the data and pretraining steps can be found in this repository, please refer to this repository . The Hugging,