not-ner-v2

This model is a fine-tuned version of dccuchile/tulio-chilean-spanish-bert on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.0838
Accuracy: 0.9727
Precision: 0.9723
Recall: 0.9727
F1: 0.9724

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: 32
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.1679	0.1595	250	0.1431	0.9555	0.9546	0.9555	0.9549
0.1291	0.3191	500	0.1328	0.9595	0.9583	0.9595	0.9586
0.1055	0.4786	750	0.0925	0.9653	0.9648	0.9653	0.9650
0.1044	0.6382	1000	0.1415	0.9630	0.9619	0.9630	0.9615
0.1094	0.7977	1250	0.1030	0.9630	0.9624	0.9630	0.9612
0.0927	0.9572	1500	0.0878	0.9710	0.9706	0.9710	0.9707
0.0836	1.1168	1750	0.1265	0.9663	0.9666	0.9663	0.9665
0.0651	1.2763	2000	0.1025	0.9709	0.9702	0.9709	0.9704
0.0637	1.4359	2250	0.0998	0.9676	0.9668	0.9676	0.9667
0.0713	1.5954	2500	0.0838	0.9727	0.9723	0.9727	0.9724
0.0561	1.7549	2750	0.0905	0.9722	0.9722	0.9722	0.9722
0.058	1.9145	3000	0.1030	0.9707	0.9701	0.9707	0.9702
0.0531	2.0740	3250	0.1066	0.9714	0.9710	0.9714	0.9711
0.0398	2.2336	3500	0.1035	0.9722	0.9721	0.9722	0.9721
0.0444	2.3931	3750	0.1009	0.9728	0.9725	0.9728	0.9726
0.037	2.5526	4000	0.1068	0.9725	0.9721	0.9725	0.9722
0.0261	2.7122	4250	0.1192	0.9735	0.9731	0.9735	0.9732
0.0266	2.8717	4500	0.1191	0.9732	0.9727	0.9732	0.9729

Framework versions

Transformers 4.42.4
Pytorch 2.3.1+cu121
Datasets 2.21.0
Tokenizers 0.19.1

lmaccarini
/

not-ner-v2

not-ner-v2

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Model tree for lmaccarini/not-ner-v2

Evaluation results