| license: mit | |
| language: | |
| - nl | |
| # hmByT5 - Preliminary Language Models | |
| Preliminary Historic Multilingual and Monolingual ByT5 Models. Following languages are currently covered: | |
| * Dutch (Delpher Corpus) | |
| More details can be found in [our GitHub repository](https://github.com/stefan-it/hmByT5). | |
| # Pretraining | |
| We use the official JAX/FLAX example in Hugging Face Transformers to pretrain a ByT5 model on a single v3-8 TPU. | |
| Details about the training can be found [here](https://github.com/stefan-it/hmByT5/tree/main/hmbyt5-flax). | |
| # Evaluation on Downstream Tasks (NER) | |
| We evaluated the hmByT5 model on ICDAR Europeana dataset: | |
| | Configuration | Run 1 | Avg. | | |
| |------------------------------------------|-------|-------------| | |
| | `wsFalse-bs4-e10-lr0.00015-poolingfirst` | 87.63 | 87.63 ± 0.0 | | |
| | `wsFalse-bs8-e10-lr0.00016-poolingfirst` | 87.35 | 87.35 ± 0.0 | | |
| | `wsFalse-bs8-e10-lr0.00015-poolingfirst` | 87.26 | 87.26 ± 0.0 | | |
| | `wsFalse-bs4-e10-lr0.00016-poolingfirst` | 86.31 | 86.31 ± 0.0 | | |
| We only performed fine-tuning for one epoch. Unfortunately, this ByT5 Base model shows no improvement over ByT5 Small architecture. | |
| # Acknowledgements | |
| Research supported with Cloud TPUs from Google's [TPU Research Cloud](https://sites.research.google/trc/about/) (TRC). | |
| Many Thanks for providing access to the TPUs ❤️ | |