roberta-base-biomedical-clinical-es-ner-breast-cancer

This model is a fine-tuned version of PlanTL-GOB-ES/roberta-base-biomedical-clinical-es on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2843
• Precision: 0.8858
• Recall: 0.8799
• F1: 0.8829
• Accuracy: 0.9477

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_ratio: 0.2
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
2.0684	1.0	213	2.5125	0.0	0.0	0.0	0.4888
1.1248	2.0	426	1.2611	0.5215	0.4616	0.4897	0.7386
0.5784	3.0	639	0.6768	0.7367	0.7785	0.7571	0.8910
0.3367	4.0	852	0.4469	0.7996	0.8359	0.8174	0.9227
0.2784	5.0	1065	0.3739	0.8410	0.8646	0.8526	0.9328
0.1799	6.0	1278	0.3285	0.8709	0.8686	0.8697	0.9393
0.1392	7.0	1491	0.3132	0.8758	0.8659	0.8708	0.9397
0.1399	8.0	1704	0.3047	0.8798	0.8739	0.8768	0.9427
0.1207	9.0	1917	0.3080	0.8755	0.8773	0.8764	0.9400
0.0968	10.0	2130	0.3021	0.8757	0.8739	0.8748	0.9395
0.1218	11.0	2343	0.2862	0.8835	0.8753	0.8794	0.9431
0.088	12.0	2556	0.2894	0.8807	0.8819	0.8813	0.9429
0.0808	13.0	2769	0.2891	0.8818	0.8759	0.8788	0.9451
0.1002	14.0	2982	0.2829	0.8837	0.8766	0.8801	0.9453
0.0617	15.0	3195	0.2840	0.8820	0.8773	0.8796	0.9460
0.0757	16.0	3408	0.2843	0.8858	0.8799	0.8829	0.9477
0.0758	17.0	3621	0.2869	0.8845	0.8786	0.8815	0.9462
0.0617	18.0	3834	0.2844	0.8835	0.8799	0.8817	0.9463
0.0719	19.0	4047	0.2842	0.8852	0.8793	0.8822	0.9467
0.0717	19.9088	4240	0.2842	0.8852	0.8793	0.8822	0.9467

Framework versions

• Transformers 4.48.2
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0