Token classification with BERT-like models

This repository contains code for token classification with BERT-like models.

• train_camembert_span_classification.ipynb: CamemBERT fine-tuning for span classification.
• evaluate_camembert_span_classification.ipynb: Evaluation.
• camembert_span_classification.ipynb: Usage examples.

The fine-tuned GeoEDdA CamemBERT model is available on the HuggingFace Hub: hhttps://huggingface.co/GEODE/camembert-base-edda-span-classification 🤗

For training and evaluation we used the GeoEDdA dataset available also on HuggingFace: https://huggingface.co/datasets/GEODE/GeoEDdA 🤗

For more details about the dataset and tagset used, see: https://github.com/GEODE-project/ner-spancat-edda

Acknowledgements

The authors are grateful to the ASLAN project (ANR-10-LABX-0081) of the Université de Lyon, for its financial support within the French program "Investments for the Future" operated by the National Research Agency (ANR).