clasificador-ser-estar-window-5-bert-base

This model is a fine-tuned version of google-bert/bert-base-multilingual-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4932
• F1 Score: 0.8401
• Recall: 0.9536
• Precision: 0.7508
• Roc Auc: 0.8601

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
• gradient_accumulation_steps: 4
• 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: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Score	Recall	Precision	Roc Auc
No log	1.0	25	0.6813	0.7465	1.0	0.5955	0.5145
No log	2.0	50	0.6746	0.7465	1.0	0.5955	0.5832
No log	3.0	75	0.5414	0.8136	0.9578	0.7072	0.7754
No log	4.0	100	0.4586	0.8296	0.9451	0.7393	0.8525
No log	5.0	125	0.4932	0.8401	0.9536	0.7508	0.8601
No log	6.0	150	0.5393	0.8281	0.8945	0.7709	0.8585
No log	7.0	175	0.5481	0.8311	0.9241	0.7552	0.8636
No log	8.0	200	0.5758	0.8336	0.9409	0.7483	0.8506
No log	9.0	225	0.6115	0.8277	0.8819	0.7799	0.8622
No log	10.0	250	0.5712	0.8333	0.9072	0.7706	0.8641
No log	11.0	275	0.6646	0.8291	0.8903	0.7757	0.8605
No log	12.0	300	0.6946	0.8397	0.9283	0.7666	0.8583
No log	13.0	325	0.7200	0.8356	0.9114	0.7714	0.8584
No log	14.0	350	0.6916	0.8330	0.9156	0.7641	0.8556
No log	15.0	375	0.6988	0.8369	0.9198	0.7676	0.8612
No log	16.0	400	0.7424	0.8308	0.9114	0.7633	0.8587
No log	17.0	425	0.7404	0.8340	0.9114	0.7687	0.8678
No log	18.0	450	0.7558	0.8356	0.9114	0.7714	0.8689
No log	19.0	475	0.7841	0.8324	0.9114	0.7660	0.8634
0.3179	20.0	500	0.7865	0.8324	0.9114	0.7660	0.8636

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0