ner_roberta_model / README.md
pritmanvar's picture
End of training
4c71718 verified
metadata
license: mit
base_model: xlm-roberta-large
tags:
  - generated_from_trainer
datasets:
  - shipping_label_ner
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: ner_roberta_model
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: shipping_label_ner
          type: shipping_label_ner
          config: shipping_label_ner
          split: validation
          args: shipping_label_ner
        metrics:
          - name: Precision
            type: precision
            value: 0.5272727272727272
          - name: Recall
            type: recall
            value: 0.7837837837837838
          - name: F1
            type: f1
            value: 0.6304347826086956
          - name: Accuracy
            type: accuracy
            value: 0.7796610169491526

ner_roberta_model

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

  • Loss: 2.0623
  • Precision: 0.5273
  • Recall: 0.7838
  • F1: 0.6304
  • Accuracy: 0.7797

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: 4
  • eval_batch_size: 2
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 1.0 14 1.1206 0.3125 0.4054 0.3529 0.6610
No log 2.0 28 0.7363 0.5128 0.5405 0.5263 0.7119
No log 3.0 42 0.6219 0.5333 0.6486 0.5854 0.7542
No log 4.0 56 0.7328 0.4727 0.7027 0.5652 0.7627
No log 5.0 70 0.8181 0.5 0.7297 0.5934 0.7542
No log 6.0 84 0.8485 0.5185 0.7568 0.6154 0.7627
No log 7.0 98 0.9692 0.5 0.7027 0.5843 0.7542
No log 8.0 112 0.9842 0.4915 0.7838 0.6042 0.7458
No log 9.0 126 1.1196 0.5 0.7838 0.6105 0.7542
No log 10.0 140 1.2147 0.5 0.7838 0.6105 0.7542
No log 11.0 154 1.4110 0.5 0.7568 0.6022 0.7712
No log 12.0 168 1.2104 0.5370 0.7838 0.6374 0.7881
No log 13.0 182 1.4145 0.5283 0.7568 0.6222 0.7797
No log 14.0 196 1.4939 0.5179 0.7838 0.6237 0.7712
No log 15.0 210 1.5558 0.5273 0.7838 0.6304 0.7797
No log 16.0 224 1.5639 0.5273 0.7838 0.6304 0.7797
No log 17.0 238 1.5208 0.5179 0.7838 0.6237 0.7712
No log 18.0 252 1.4787 0.5918 0.7838 0.6744 0.7966
No log 19.0 266 1.3946 0.5283 0.7568 0.6222 0.7797
No log 20.0 280 1.6672 0.5370 0.7838 0.6374 0.7881
No log 21.0 294 1.5746 0.5185 0.7568 0.6154 0.7712
No log 22.0 308 1.8881 0.5091 0.7568 0.6087 0.7712
No log 23.0 322 1.5084 0.5370 0.7838 0.6374 0.7881
No log 24.0 336 1.7922 0.5091 0.7568 0.6087 0.7712
No log 25.0 350 1.7265 0.5273 0.7838 0.6304 0.7797
No log 26.0 364 1.7467 0.5273 0.7838 0.6304 0.7797
No log 27.0 378 2.0162 0.5 0.7568 0.6022 0.7627
No log 28.0 392 1.9460 0.5 0.7568 0.6022 0.7627
No log 29.0 406 1.8957 0.5091 0.7568 0.6087 0.7712
No log 30.0 420 1.9941 0.5 0.7568 0.6022 0.7627
No log 31.0 434 1.9095 0.5 0.7568 0.6022 0.7712
No log 32.0 448 1.8920 0.5273 0.7838 0.6304 0.7797
No log 33.0 462 1.9310 0.5091 0.7568 0.6087 0.7712
No log 34.0 476 1.9830 0.5091 0.7568 0.6087 0.7712
No log 35.0 490 2.0445 0.5091 0.7568 0.6087 0.7712
0.2599 36.0 504 2.1138 0.5091 0.7568 0.6087 0.7712
0.2599 37.0 518 2.0024 0.5091 0.7568 0.6087 0.7797
0.2599 38.0 532 2.0004 0.5091 0.7568 0.6087 0.7712
0.2599 39.0 546 2.0725 0.5091 0.7568 0.6087 0.7712
0.2599 40.0 560 2.0507 0.5091 0.7568 0.6087 0.7712
0.2599 41.0 574 2.0548 0.5091 0.7568 0.6087 0.7712
0.2599 42.0 588 2.1176 0.5091 0.7568 0.6087 0.7712
0.2599 43.0 602 2.0946 0.5091 0.7568 0.6087 0.7712
0.2599 44.0 616 2.1211 0.5 0.7568 0.6022 0.7627
0.2599 45.0 630 2.1103 0.5091 0.7568 0.6087 0.7712
0.2599 46.0 644 2.0876 0.5 0.7568 0.6022 0.7627
0.2599 47.0 658 2.0910 0.5179 0.7838 0.6237 0.7712
0.2599 48.0 672 2.0800 0.5179 0.7838 0.6237 0.7712
0.2599 49.0 686 2.0584 0.5273 0.7838 0.6304 0.7797
0.2599 50.0 700 2.0623 0.5273 0.7838 0.6304 0.7797

Framework versions

  • Transformers 4.39.1
  • Pytorch 2.2.1+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2