xlm-roberta-large-finetuned-ner-lenerBr

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

• Loss: nan
• Precision: 0.9166
• Recall: 0.9289
• F1: 0.9227
• Accuracy: 0.9854

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
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• optimizer: Use 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.9995	489	nan	0.8191	0.8167	0.8179	0.9751
0.163	1.9990	978	nan	0.8600	0.9080	0.8833	0.9790
0.0427	2.9985	1467	nan	0.8736	0.9163	0.8944	0.9814
0.0279	4.0	1957	nan	0.8688	0.9191	0.8932	0.9801
0.019	4.9995	2446	nan	0.9123	0.9196	0.9159	0.9840
0.0143	5.9990	2935	nan	0.9008	0.9346	0.9174	0.9842
0.0112	6.9985	3424	nan	0.9063	0.9250	0.9156	0.9843
0.0072	8.0	3914	nan	0.8954	0.9315	0.9131	0.9841
0.0065	8.9995	4403	nan	0.9226	0.9245	0.9236	0.9857
0.0048	9.9949	4890	nan	0.9166	0.9289	0.9227	0.9854

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3