xlm-roberta-large-finetuned-lener-br

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

• Loss: nan
• Precision: 0.8762
• Recall: 0.8966
• F1: 0.8863
• Accuracy: 0.9795

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: 2e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.0785	1.0	3914	nan	0.7119	0.8410	0.7711	0.9658
0.076	2.0	7828	nan	0.8397	0.8679	0.8536	0.9740
0.0434	3.0	11742	nan	0.8545	0.8666	0.8605	0.9693
0.022	4.0	15656	nan	0.8293	0.8573	0.8431	0.9652
0.0284	5.0	19570	nan	0.8789	0.8571	0.8678	0.9776
0.029	6.0	23484	nan	0.8521	0.8788	0.8653	0.9771
0.0227	7.0	27398	nan	0.7648	0.8873	0.8215	0.9686
0.0219	8.0	31312	nan	0.8609	0.9026	0.8813	0.9780
0.0121	9.0	35226	nan	0.8746	0.8979	0.8861	0.9812
0.0087	10.0	39140	nan	0.8829	0.8827	0.8828	0.9808
0.0081	11.0	43054	nan	0.8740	0.8749	0.8745	0.9765
0.0058	12.0	46968	nan	0.8838	0.8842	0.8840	0.9788
0.0044	13.0	50882	nan	0.869	0.8984	0.8835	0.9788
0.002	14.0	54796	nan	0.8762	0.8966	0.8863	0.9795
0.0017	15.0	58710	nan	0.8729	0.8982	0.8854	0.9791

Framework versions

• Transformers 4.23.1
• Pytorch 1.12.1+cu113
• Datasets 2.6.1
• Tokenizers 0.13.1