final-ft__bert-base-spanish-wwm-uncased__70k-ultrasounds-ner

This model is a fine-tuned version of manucos/final-ft__bert-base-spanish-wwm-uncased__70k-ultrasounds on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3276
• Precision: 0.8874
• Recall: 0.9111
• F1: 0.8991
• Accuracy: 0.9459

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	280	0.2520	0.8345	0.8593	0.8467	0.9271
0.3534	2.0	560	0.2560	0.8431	0.8919	0.8668	0.9279
0.3534	3.0	840	0.2567	0.8622	0.8711	0.8666	0.9322
0.1033	4.0	1120	0.2605	0.8571	0.8978	0.8770	0.9410
0.1033	5.0	1400	0.2829	0.8775	0.9022	0.8897	0.9405
0.0566	6.0	1680	0.2683	0.8780	0.9067	0.8921	0.9432
0.0566	7.0	1960	0.2757	0.8775	0.9022	0.8897	0.9448
0.032	8.0	2240	0.2767	0.8739	0.9141	0.8936	0.9480
0.0235	9.0	2520	0.3582	0.8681	0.9067	0.8870	0.9354
0.0235	10.0	2800	0.3272	0.8841	0.9037	0.8938	0.9464
0.016	11.0	3080	0.3200	0.8846	0.9081	0.8962	0.9440
0.016	12.0	3360	0.3277	0.8831	0.9067	0.8947	0.9418
0.0126	13.0	3640	0.3175	0.8858	0.9081	0.8969	0.9475
0.0126	14.0	3920	0.3294	0.8890	0.9141	0.9014	0.9461
0.0121	15.0	4200	0.3276	0.8874	0.9111	0.8991	0.9459

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1