nuevo-breast_cancer-biomedical-ner-sp-24-02-2026

This model is a fine-tuned version of FacebookAI/xlm-roberta-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2563
• Precision: 0.8930
• Recall: 0.9116
• F1: 0.9022
• Accuracy: 0.9527

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
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
16.7479	1.0	96	1.7735	0.3413	0.0271	0.0503	0.6238
4.9427	2.0	192	0.5038	0.6915	0.7203	0.7056	0.8806
2.4504	3.0	288	0.2948	0.8211	0.8434	0.8321	0.9309
1.6450	4.0	384	0.2615	0.8195	0.8801	0.8487	0.9340
1.0866	5.0	480	0.2503	0.8325	0.8914	0.8610	0.9396
0.9843	6.0	576	0.2386	0.8634	0.8782	0.8707	0.9459
0.7857	7.0	672	0.2428	0.8772	0.8971	0.8870	0.9480
0.6685	8.0	768	0.2348	0.8720	0.9072	0.8892	0.9496
0.5669	9.0	864	0.2505	0.8710	0.9040	0.8872	0.9507
0.3850	10.0	960	0.2552	0.8795	0.9122	0.8956	0.9519
0.3826	11.0	1056	0.2517	0.8837	0.9160	0.8996	0.9512
0.3398	12.0	1152	0.2561	0.8930	0.9116	0.9022	0.9527
0.2609	13.0	1248	0.2712	0.8794	0.9116	0.8952	0.9522
0.2260	14.0	1344	0.2670	0.8823	0.9179	0.8998	0.9533
0.2393	15.0	1440	0.2666	0.8888	0.9129	0.9007	0.9535
0.1719	16.0	1536	0.2693	0.8828	0.9129	0.8976	0.9536
0.2091	17.0	1632	0.2724	0.8854	0.9167	0.9007	0.9544
0.2027	18.0	1728	0.2762	0.8855	0.9179	0.9014	0.9537
0.1897	19.0	1824	0.2768	0.8837	0.9167	0.8999	0.9536
0.1859	20.0	1920	0.2767	0.8837	0.9167	0.8999	0.9537

Framework versions

• Transformers 5.2.0
• Pytorch 2.10.0+cu128
• Datasets 4.5.0
• Tokenizers 0.22.2