not-ner-v3_batch16

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: 0.1059
• Accuracy: 0.9709
• Precision: 0.9698
• Recall: 0.9709
• F1: 0.9699

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: 16
• 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
0.0452	0.2646	200	0.2503	0.9596	0.9578	0.9596	0.9558
0.1058	0.5291	400	0.1167	0.9686	0.9673	0.9686	0.9668
0.1106	0.7937	600	0.1059	0.9709	0.9698	0.9709	0.9699
0.103	1.0582	800	0.1179	0.9662	0.9657	0.9662	0.9659
0.0629	1.3228	1000	0.1298	0.9719	0.9709	0.9719	0.9705
0.0633	1.5873	1200	0.1176	0.9732	0.9728	0.9732	0.9730
0.0436	1.8519	1400	0.1257	0.9725	0.9721	0.9725	0.9723
0.0487	2.1164	1600	0.1206	0.9729	0.9724	0.9729	0.9726
0.0309	2.3810	1800	0.1232	0.9705	0.9705	0.9705	0.9705
0.0235	2.6455	2000	0.1379	0.9742	0.9738	0.9742	0.9740
0.0272	2.9101	2200	0.1327	0.9729	0.9724	0.9729	0.9726

Framework versions

• Transformers 4.42.4
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1