NERDE-base

This model is a fine-tuned version of pierreguillou/bert-base-cased-pt-lenerbr on the nerde dataset. It achieves the following results on the evaluation set:

• Loss: 0.1246
• Precision: 0.9119
• Recall: 0.9153
• F1: 0.9136
• Accuracy: 0.9842

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.2466	1.0	541	0.1003	0.8515	0.8822	0.8666	0.9782
0.0608	2.0	1082	0.0855	0.8990	0.9083	0.9036	0.9837
0.0411	3.0	1623	0.1006	0.9078	0.9103	0.9090	0.9837
0.0266	4.0	2164	0.1052	0.9023	0.9163	0.9092	0.9828
0.0191	5.0	2705	0.1060	0.9112	0.9183	0.9147	0.9847
0.0153	6.0	3246	0.1152	0.9052	0.9098	0.9075	0.9831
0.0124	7.0	3787	0.1209	0.9029	0.9185	0.9107	0.9835
0.0083	8.0	4328	0.1176	0.9072	0.9163	0.9117	0.9844
0.0077	9.0	4869	0.1240	0.9080	0.9201	0.9140	0.9844
0.0051	10.0	5410	0.1246	0.9119	0.9153	0.9136	0.9842

Framework versions

• Transformers 4.20.1
• Pytorch 1.12.0+cu113
• Datasets 2.3.2
• Tokenizers 0.12.1