Edit model card

robertalex-ptbr-ulyssesner

This model is a fine-tuned version of eduagarcia/RoBERTaLexPT-base on the ulysses_ner_br dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0711
  • Data: {'precision': 0.96, 'recall': 1.0, 'f1': 0.9795918367346939, 'number': 72}
  • Evento: {'precision': 0.6666666666666666, 'recall': 0.4, 'f1': 0.5, 'number': 5}
  • Fundamento: {'precision': 0.7967479674796748, 'recall': 0.9158878504672897, 'f1': 0.8521739130434782, 'number': 107}
  • Local: {'precision': 0.950354609929078, 'recall': 0.9241379310344827, 'f1': 0.9370629370629371, 'number': 145}
  • Organizacao: {'precision': 0.75, 'recall': 0.8888888888888888, 'f1': 0.8135593220338982, 'number': 81}
  • Pessoa: {'precision': 0.823076923076923, 'recall': 0.9385964912280702, 'f1': 0.8770491803278688, 'number': 114}
  • Produtodelei: {'precision': 0.6470588235294118, 'recall': 0.717391304347826, 'f1': 0.6804123711340206, 'number': 46}
  • Overall Precision: 0.8368
  • Overall Recall: 0.9088
  • Overall F1: 0.8713
  • Overall Accuracy: 0.9860

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: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Data Evento Fundamento Local Organizacao Pessoa Produtodelei Overall Precision Overall Recall Overall F1 Overall Accuracy
0.4776 1.0 71 0.2170 {'precision': 1.0, 'recall': 0.4166666666666667, 'f1': 0.5882352941176471, 'number': 72} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 5} {'precision': 0.5714285714285714, 'recall': 0.5607476635514018, 'f1': 0.5660377358490566, 'number': 107} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 145} {'precision': 0.13445378151260504, 'recall': 0.5925925925925926, 'f1': 0.2191780821917808, 'number': 81} {'precision': 0.16793893129770993, 'recall': 0.19298245614035087, 'f1': 0.17959183673469387, 'number': 114} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 46} 0.2528 0.2807 0.2660 0.9344
0.124 2.0 142 0.0854 {'precision': 0.8666666666666667, 'recall': 0.9027777777777778, 'f1': 0.8843537414965987, 'number': 72} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 5} {'precision': 0.7165354330708661, 'recall': 0.8504672897196262, 'f1': 0.7777777777777777, 'number': 107} {'precision': 0.8187919463087249, 'recall': 0.8413793103448276, 'f1': 0.8299319727891157, 'number': 145} {'precision': 0.6078431372549019, 'recall': 0.7654320987654321, 'f1': 0.6775956284153005, 'number': 81} {'precision': 0.8303571428571429, 'recall': 0.8157894736842105, 'f1': 0.8230088495575222, 'number': 114} {'precision': 0.6590909090909091, 'recall': 0.6304347826086957, 'f1': 0.6444444444444444, 'number': 46} 0.7586 0.8105 0.7837 0.9783
0.0463 3.0 213 0.0699 {'precision': 0.9210526315789473, 'recall': 0.9722222222222222, 'f1': 0.9459459459459458, 'number': 72} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 5} {'precision': 0.7404580152671756, 'recall': 0.9065420560747663, 'f1': 0.8151260504201681, 'number': 107} {'precision': 0.9236111111111112, 'recall': 0.9172413793103448, 'f1': 0.9204152249134949, 'number': 145} {'precision': 0.7156862745098039, 'recall': 0.9012345679012346, 'f1': 0.7978142076502731, 'number': 81} {'precision': 0.8048780487804879, 'recall': 0.868421052631579, 'f1': 0.8354430379746836, 'number': 114} {'precision': 0.6304347826086957, 'recall': 0.6304347826086957, 'f1': 0.6304347826086957, 'number': 46} 0.8055 0.8789 0.8406 0.9838
0.0277 4.0 284 0.0709 {'precision': 0.9473684210526315, 'recall': 1.0, 'f1': 0.972972972972973, 'number': 72} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 5} {'precision': 0.8347826086956521, 'recall': 0.897196261682243, 'f1': 0.8648648648648648, 'number': 107} {'precision': 0.9246575342465754, 'recall': 0.9310344827586207, 'f1': 0.9278350515463917, 'number': 145} {'precision': 0.7553191489361702, 'recall': 0.8765432098765432, 'f1': 0.8114285714285715, 'number': 81} {'precision': 0.796875, 'recall': 0.8947368421052632, 'f1': 0.8429752066115702, 'number': 114} {'precision': 0.6481481481481481, 'recall': 0.7608695652173914, 'f1': 0.7000000000000001, 'number': 46} 0.8336 0.8965 0.8639 0.9833
0.0165 5.0 355 0.0640 {'precision': 0.9473684210526315, 'recall': 1.0, 'f1': 0.972972972972973, 'number': 72} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 5} {'precision': 0.8448275862068966, 'recall': 0.9158878504672897, 'f1': 0.8789237668161435, 'number': 107} {'precision': 0.9640287769784173, 'recall': 0.9241379310344827, 'f1': 0.943661971830986, 'number': 145} {'precision': 0.8111111111111111, 'recall': 0.9012345679012346, 'f1': 0.8538011695906432, 'number': 81} {'precision': 0.7969924812030075, 'recall': 0.9298245614035088, 'f1': 0.8582995951417005, 'number': 114} {'precision': 0.673469387755102, 'recall': 0.717391304347826, 'f1': 0.6947368421052631, 'number': 46} 0.8543 0.9053 0.8790 0.9848
0.0087 6.0 426 0.0612 {'precision': 0.9594594594594594, 'recall': 0.9861111111111112, 'f1': 0.9726027397260274, 'number': 72} {'precision': 0.5, 'recall': 0.2, 'f1': 0.28571428571428575, 'number': 5} {'precision': 0.8048780487804879, 'recall': 0.9252336448598131, 'f1': 0.8608695652173913, 'number': 107} {'precision': 0.9574468085106383, 'recall': 0.9310344827586207, 'f1': 0.9440559440559441, 'number': 145} {'precision': 0.8131868131868132, 'recall': 0.9135802469135802, 'f1': 0.8604651162790699, 'number': 81} {'precision': 0.8333333333333334, 'recall': 0.9210526315789473, 'f1': 0.875, 'number': 114} {'precision': 0.7083333333333334, 'recall': 0.7391304347826086, 'f1': 0.723404255319149, 'number': 46} 0.8579 0.9105 0.8834 0.9873
0.0057 7.0 497 0.0691 {'precision': 0.9473684210526315, 'recall': 1.0, 'f1': 0.972972972972973, 'number': 72} {'precision': 1.0, 'recall': 0.2, 'f1': 0.33333333333333337, 'number': 5} {'precision': 0.784, 'recall': 0.9158878504672897, 'f1': 0.8448275862068965, 'number': 107} {'precision': 0.9375, 'recall': 0.9310344827586207, 'f1': 0.9342560553633218, 'number': 145} {'precision': 0.8202247191011236, 'recall': 0.9012345679012346, 'f1': 0.8588235294117647, 'number': 81} {'precision': 0.8106060606060606, 'recall': 0.9385964912280702, 'f1': 0.8699186991869918, 'number': 114} {'precision': 0.5789473684210527, 'recall': 0.717391304347826, 'f1': 0.6407766990291262, 'number': 46} 0.8317 0.9105 0.8693 0.9866
0.0042 8.0 568 0.0701 {'precision': 0.96, 'recall': 1.0, 'f1': 0.9795918367346939, 'number': 72} {'precision': 0.3333333333333333, 'recall': 0.2, 'f1': 0.25, 'number': 5} {'precision': 0.8181818181818182, 'recall': 0.9252336448598131, 'f1': 0.868421052631579, 'number': 107} {'precision': 0.9640287769784173, 'recall': 0.9241379310344827, 'f1': 0.943661971830986, 'number': 145} {'precision': 0.7634408602150538, 'recall': 0.8765432098765432, 'f1': 0.8160919540229884, 'number': 81} {'precision': 0.828125, 'recall': 0.9298245614035088, 'f1': 0.8760330578512396, 'number': 114} {'precision': 0.6538461538461539, 'recall': 0.7391304347826086, 'f1': 0.693877551020408, 'number': 46} 0.8462 0.9070 0.8755 0.9863
0.0029 9.0 639 0.0713 {'precision': 0.96, 'recall': 1.0, 'f1': 0.9795918367346939, 'number': 72} {'precision': 0.6666666666666666, 'recall': 0.4, 'f1': 0.5, 'number': 5} {'precision': 0.8448275862068966, 'recall': 0.9158878504672897, 'f1': 0.8789237668161435, 'number': 107} {'precision': 0.9436619718309859, 'recall': 0.9241379310344827, 'f1': 0.9337979094076655, 'number': 145} {'precision': 0.7708333333333334, 'recall': 0.9135802469135802, 'f1': 0.8361581920903954, 'number': 81} {'precision': 0.8294573643410853, 'recall': 0.9385964912280702, 'f1': 0.8806584362139916, 'number': 114} {'precision': 0.6415094339622641, 'recall': 0.7391304347826086, 'f1': 0.6868686868686867, 'number': 46} 0.8485 0.9140 0.8801 0.9860
0.0025 10.0 710 0.0711 {'precision': 0.96, 'recall': 1.0, 'f1': 0.9795918367346939, 'number': 72} {'precision': 0.6666666666666666, 'recall': 0.4, 'f1': 0.5, 'number': 5} {'precision': 0.7967479674796748, 'recall': 0.9158878504672897, 'f1': 0.8521739130434782, 'number': 107} {'precision': 0.950354609929078, 'recall': 0.9241379310344827, 'f1': 0.9370629370629371, 'number': 145} {'precision': 0.75, 'recall': 0.8888888888888888, 'f1': 0.8135593220338982, 'number': 81} {'precision': 0.823076923076923, 'recall': 0.9385964912280702, 'f1': 0.8770491803278688, 'number': 114} {'precision': 0.6470588235294118, 'recall': 0.717391304347826, 'f1': 0.6804123711340206, 'number': 46} 0.8368 0.9088 0.8713 0.9860

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.4.0+cu121
  • Datasets 3.0.0
  • Tokenizers 0.19.1
Downloads last month
35
Safetensors
Model size
124M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for giliardgodoi/robertalex-ptbr-ulyssesner

Finetuned
(1)
this model