ner-bert-large-cased-pt-lenerbr-finetuned-ner

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

• Loss: nan
• Precision: 0.7549
• Recall: 0.8115
• F1: 0.7822
• Accuracy: 0.8832

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: 4
• eval_batch_size: 4
• 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
No log	1.0	91	nan	0.6987	0.7433	0.7203	0.8620
No log	2.0	182	nan	0.7040	0.7564	0.7292	0.8624
No log	3.0	273	nan	0.7317	0.7929	0.7611	0.8731
No log	4.0	364	nan	0.7501	0.8097	0.7788	0.8838
No log	5.0	455	nan	0.7504	0.8332	0.7897	0.8857
0.3495	6.0	546	nan	0.7551	0.8103	0.7817	0.8799
0.3495	7.0	637	nan	0.7533	0.8215	0.7859	0.8824
0.3495	8.0	728	nan	0.7578	0.7991	0.7779	0.8785
0.3495	9.0	819	nan	0.7520	0.8196	0.7843	0.8840
0.3495	10.0	910	nan	0.7549	0.8115	0.7822	0.8832

Framework versions

• Transformers 4.36.0
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0