Fine-tuned Flair Model on Finnish NewsEye NER Dataset (HIPE-2022)
This Flair model was fine-tuned on the Finnish NewsEye NER Dataset using hmBERT 64k as backbone LM.
The NewsEye dataset is comprised of diachronic historical newspaper material published between 1850 and 1950 in French, German, Finnish, and Swedish. More information can be found here.
The following NEs were annotated: PER, LOC, ORG and HumanProd.
Results
We performed a hyper-parameter search over the following parameters with 5 different seeds per configuration:
- Batch Sizes:
[4, 8] - Learning Rates:
[3e-05, 5e-05]
And report micro F1-score on development set:
| Configuration | Seed 1 | Seed 2 | Seed 3 | Seed 4 | Seed 5 | Average |
|---|---|---|---|---|---|---|
bs8-e10-lr3e-05 |
0.7527 | 0.7732 | 0.7849 | 0.7702 | 0.7702 | 0.7702 ± 0.0115 |
bs4-e10-lr5e-05 |
0.7522 | 0.7642 | 0.787 | 0.7532 | 0.7832 | 0.768 ± 0.0164 |
bs4-e10-lr3e-05 |
0.7716 | 0.7419 | 0.7716 | 0.7722 | 0.7638 | 0.7642 ± 0.013 |
bs8-e10-lr5e-05 |
0.7484 | 0.7811 | 0.7706 | 0.7516 | 0.7521 | 0.7608 ± 0.0143 |
The training log and TensorBoard logs (not available for hmBERT Base model) are also uploaded to the model hub.
More information about fine-tuning can be found here.
Acknowledgements
We thank Luisa März, Katharina Schmid and Erion Çano for their fruitful discussions about Historic Language Models.
Research supported with Cloud TPUs from Google's TPU Research Cloud (TRC). Many Thanks for providing access to the TPUs ❤️
