jordyvl's picture
update model card README.md
d4c8d64
metadata
license: mit
tags:
  - generated_from_trainer
datasets:
  - harem
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: bert-base-portuguese-cased_harem-selective-lowC-sm-first-ner
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: harem
          type: harem
          args: selective
        metrics:
          - name: Precision
            type: precision
            value: 0.8
          - name: Recall
            type: recall
            value: 0.8764044943820225
          - name: F1
            type: f1
            value: 0.8364611260053619
          - name: Accuracy
            type: accuracy
            value: 0.9764089121887287

bert-base-portuguese-cased_harem-selective-lowC-sm-first-ner

This model is a fine-tuned version of neuralmind/bert-base-portuguese-cased on the harem dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1160
  • Precision: 0.8
  • Recall: 0.8764
  • F1: 0.8365
  • Accuracy: 0.9764

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
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 3

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
0.055 1.0 2517 0.0934 0.81 0.9101 0.8571 0.9699
0.0236 2.0 5034 0.0883 0.8307 0.8820 0.8556 0.9751
0.0129 3.0 7551 0.1160 0.8 0.8764 0.8365 0.9764

Framework versions

  • Transformers 4.18.0
  • Pytorch 1.10.2+cu102
  • Datasets 2.2.2
  • Tokenizers 0.12.1