Negation_Scope_Detection_NubEs_Training_Development_mBERT_fine_tuned

This model is a fine-tuned version of bert-base-multilingual-cased on NubEs (Training + Development) dataset. It achieves the following results on the evaluation set:

• Loss: 0.1712
• Precision: 0.8970
• Recall: 0.9177
• F1: 0.9072
• Accuracy: 0.9722

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 7

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.1805	1.0	1726	0.1542	0.8408	0.8600	0.8503	0.9591
0.126	2.0	3452	0.1271	0.8652	0.8774	0.8713	0.9640
0.0814	3.0	5178	0.1395	0.8715	0.8881	0.8797	0.9662
0.0447	4.0	6904	0.1326	0.8893	0.9068	0.8979	0.9696
0.0247	5.0	8630	0.1805	0.9012	0.9020	0.9016	0.9708
0.0156	6.0	10356	0.1524	0.8972	0.9120	0.9045	0.9706
0.0056	7.0	12082	0.1712	0.8970	0.9177	0.9072	0.9722

Framework versions

• Transformers 4.27.3
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2