StivenLancheros
/

Roberta-base-bne-NER-EN-ES

Token Classification PyTorch TensorBoard Transformers

apache-2.0 roberta generated_from_trainer Eval Results AutoTrain Compatible

Model card Files Files and versions Metrics Training metrics Community

Edit model card

roberta-base-bne-finetuned-ner-finetuned2-ner

This model is a fine-tuned version of StivenLancheros/roberta-base-bne-finetuned-ner on the conll2002 dataset. It achieves the following results on the evaluation set:

Loss: 0.1067
Precision: 0.8698
Recall: 0.8794
F1: 0.8745
Accuracy: 0.9809

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: 3e-05
train_batch_size: 5
eval_batch_size: 5
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.0582	1.0	1665	0.0852	0.8697	0.8759	0.8728	0.9800
0.0297	2.0	3330	0.0919	0.8841	0.8867	0.8854	0.9817
0.0121	3.0	4995	0.0950	0.8751	0.8807	0.8779	0.9812
0.0056	4.0	6660	0.1067	0.8698	0.8794	0.8745	0.9809

Framework versions

Transformers 4.12.3
Pytorch 1.10.0+cu111
Datasets 1.15.1
Tokenizers 0.10.3

Downloads last month: 3

Hosted inference API

Token Classification

Examples

Examples

This model can be loaded on the Inference API on-demand.

Dataset used to train StivenLancheros/Roberta-base-bne-NER-EN-ES

Evaluation results

Precision on conll2002
self-reported

0.870
Recall on conll2002
self-reported

0.879
F1 on conll2002
self-reported

0.875
Accuracy on conll2002
self-reported

0.981

View leaderboard (Papers With Code)