clasificador-ser-estar-bert-base

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

• Loss: 0.3628
• F1 Score: 0.9015
• Recall: 0.8996
• Precision: 0.9034
• Roc Auc: 0.9416

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: 16
• seed: 42
• 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: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Score	Recall	Precision	Roc Auc
No log	1.0	100	0.3752	0.8898	0.9289	0.8538	0.9078
No log	2.0	200	0.4126	0.9061	0.9289	0.8845	0.9282
No log	3.0	300	0.3628	0.9015	0.8996	0.9034	0.9416
No log	4.0	400	0.3830	0.9072	0.9205	0.8943	0.9462
0.3105	5.0	500	0.6145	0.8894	0.8577	0.9234	0.9383
0.3105	6.0	600	0.6179	0.8952	0.9289	0.8638	0.9356
0.3105	7.0	700	0.7010	0.8975	0.9163	0.8795	0.9406
0.3105	8.0	800	0.7734	0.8819	0.8745	0.8894	0.9378
0.3105	9.0	900	1.0186	0.8833	0.8870	0.8797	0.9164
0.0377	10.0	1000	0.8641	0.8921	0.8996	0.8848	0.9086
0.0377	11.0	1100	0.9488	0.8836	0.8577	0.9111	0.9192
0.0377	12.0	1200	0.8875	0.8996	0.9372	0.8649	0.9115
0.0377	13.0	1300	0.8985	0.8934	0.9121	0.8755	0.9270
0.0377	14.0	1400	0.9961	0.8986	0.9456	0.8561	0.9182
0.0135	15.0	1500	0.8714	0.9026	0.9498	0.8598	0.8975
0.0135	16.0	1600	0.9774	0.9018	0.9414	0.8654	0.9056
0.0135	17.0	1700	1.0255	0.8974	0.9331	0.8643	0.9023
0.0135	18.0	1800	1.0403	0.8996	0.9372	0.8649	0.9060
0.0135	19.0	1900	1.0448	0.9018	0.9414	0.8654	0.9085
0.0041	20.0	2000	1.0479	0.9018	0.9414	0.8654	0.9104

Framework versions

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