ABL_trad_h

This model is a fine-tuned version of dccuchile/bert-base-spanish-wwm-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.3621
• Accuracy: 0.6867
• F1: 0.6845

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: 12
• eval_batch_size: 12
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 24

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.0316	1.0	1000	0.9725	0.5383	0.5182
0.8887	2.0	2000	0.8652	0.5925	0.5901
0.8132	3.0	3000	0.8327	0.6175	0.6143
0.7639	4.0	4000	0.8030	0.6367	0.6356
0.7233	5.0	5000	0.7969	0.6433	0.6408
0.6977	6.0	6000	0.7848	0.6508	0.6500
0.6531	7.0	7000	0.7923	0.66	0.6590
0.6149	8.0	8000	0.7879	0.6683	0.6667
0.5964	9.0	9000	0.7835	0.6708	0.6705
0.5508	10.0	10000	0.8072	0.6725	0.6701
0.5288	11.0	11000	0.8277	0.665	0.6604
0.4828	12.0	12000	0.8436	0.6833	0.6810
0.4457	13.0	13000	0.8679	0.6717	0.6681
0.432	14.0	14000	0.8695	0.6692	0.6679
0.4128	15.0	15000	0.9217	0.6858	0.6833
0.3831	16.0	16000	0.9300	0.6733	0.6714
0.3471	17.0	17000	0.9796	0.675	0.6726
0.3245	18.0	18000	1.0319	0.6742	0.6726
0.3016	19.0	19000	1.0977	0.6733	0.6692
0.2702	20.0	20000	1.1130	0.6667	0.6634
0.2602	21.0	21000	1.1475	0.6758	0.6741
0.2316	22.0	22000	1.2237	0.6725	0.6694
0.2176	23.0	23000	1.2664	0.6775	0.6758
0.2067	24.0	24000	1.3621	0.6867	0.6845

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1