datos-ner

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

• Loss: 0.0751
• Precision: 0.9516
• Recall: 0.9219
• F1: 0.9365
• Accuracy: 0.9805

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	38	0.6419	0.8947	0.2656	0.4096	0.8357
No log	2.0	76	0.2665	0.8511	0.625	0.7207	0.9276
No log	3.0	114	0.1322	0.9508	0.9062	0.9280	0.9749
No log	4.0	152	0.0907	0.9524	0.9375	0.9449	0.9805
No log	5.0	190	0.0760	0.9683	0.9531	0.9606	0.9833
No log	6.0	228	0.0644	0.9531	0.9531	0.9531	0.9861
No log	7.0	266	0.0728	0.9365	0.9219	0.9291	0.9805
No log	8.0	304	0.0690	0.9365	0.9219	0.9291	0.9805
No log	9.0	342	0.0709	0.9365	0.9219	0.9291	0.9805
No log	10.0	380	0.0781	0.9516	0.9219	0.9365	0.9805
No log	11.0	418	0.0654	0.9365	0.9219	0.9291	0.9805
No log	12.0	456	0.0746	0.9516	0.9219	0.9365	0.9805
No log	13.0	494	0.0721	0.9516	0.9219	0.9365	0.9805
0.1874	14.0	532	0.0739	0.9516	0.9219	0.9365	0.9805
0.1874	15.0	570	0.0765	0.9516	0.9219	0.9365	0.9805
0.1874	16.0	608	0.0777	0.9516	0.9219	0.9365	0.9805
0.1874	17.0	646	0.0756	0.9516	0.9219	0.9365	0.9805
0.1874	18.0	684	0.0765	0.9516	0.9219	0.9365	0.9805
0.1874	19.0	722	0.0758	0.9516	0.9219	0.9365	0.9805
0.1874	20.0	760	0.0751	0.9516	0.9219	0.9365	0.9805

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2