distilbert-base-spanish-uncased-finetuned-text-intelligence

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

• Loss: 0.6945
• Accuracy: 0.8834
• F1: 0.8827

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
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.0148	1.0	235	0.7880	0.7138	0.6551
0.6349	2.0	470	0.5415	0.8516	0.8500
0.4709	3.0	705	0.4505	0.8587	0.8613
0.3727	4.0	940	0.4156	0.8905	0.8900
0.3163	5.0	1175	0.4262	0.8905	0.8910
0.2695	6.0	1410	0.5090	0.8869	0.8874
0.2332	7.0	1645	0.5014	0.8869	0.8865
0.1811	8.0	1880	0.5735	0.8834	0.8827
0.1542	9.0	2115	0.5626	0.8940	0.8932
0.1192	10.0	2350	0.5680	0.8905	0.8900
0.124	11.0	2585	0.6291	0.8869	0.8857
0.0988	12.0	2820	0.6424	0.8834	0.8835
0.0933	13.0	3055	0.7085	0.8693	0.8668
0.0813	14.0	3290	0.6560	0.8905	0.8893
0.0599	15.0	3525	0.7175	0.8799	0.8793
0.0632	16.0	3760	0.6862	0.8799	0.8786
0.0489	17.0	3995	0.7064	0.8869	0.8858
0.0449	18.0	4230	0.7046	0.8834	0.8830
0.039	19.0	4465	0.6997	0.8799	0.8790
0.0388	20.0	4700	0.6945	0.8834	0.8827

Framework versions

• Transformers 4.40.1
• Pytorch 2.1.0
• Datasets 2.19.0
• Tokenizers 0.19.1