Edit model card

lilt-ruroberta

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4919
  • Comment: {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 6}
  • Date: {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 3}
  • Labname: {'precision': 0.5833333333333334, 'recall': 0.6666666666666666, 'f1': 0.6222222222222222, 'number': 21}
  • Laboratory: {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1}
  • Measure: {'precision': 0.5833333333333334, 'recall': 0.7777777777777778, 'f1': 0.6666666666666666, 'number': 9}
  • Ref Value: {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 8}
  • Result: {'precision': 0.25, 'recall': 0.25, 'f1': 0.25, 'number': 12}
  • Overall Precision: 0.4528
  • Overall Recall: 0.4
  • Overall F1: 0.4248
  • Overall Accuracy: 0.8698

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • training_steps: 25
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Comment Date Labname Laboratory Measure Ref Value Result Overall Precision Overall Recall Overall F1 Overall Accuracy
2.4398 5.0 5 1.5928 {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 6} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 3} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 21} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 9} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 8} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 12} 0.0 0.0 0.0 0.5850
1.4788 10.0 10 1.1857 {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 6} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 3} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 21} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 9} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 8} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 12} 0.0 0.0 0.0 0.6512
0.9806 15.0 15 0.8188 {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 6} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 3} {'precision': 0.21875, 'recall': 0.3333333333333333, 'f1': 0.2641509433962264, 'number': 21} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1} {'precision': 0.5, 'recall': 0.1111111111111111, 'f1': 0.1818181818181818, 'number': 9} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 8} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 12} 0.1667 0.1333 0.1481 0.7660
0.6358 20.0 20 0.5763 {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 6} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 3} {'precision': 0.41935483870967744, 'recall': 0.6190476190476191, 'f1': 0.5, 'number': 21} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1} {'precision': 0.7, 'recall': 0.7777777777777778, 'f1': 0.7368421052631577, 'number': 9} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 8} {'precision': 0.42857142857142855, 'recall': 0.25, 'f1': 0.3157894736842105, 'number': 12} 0.4182 0.3833 0.4 0.8675
0.4712 25.0 25 0.4919 {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 6} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 3} {'precision': 0.5833333333333334, 'recall': 0.6666666666666666, 'f1': 0.6222222222222222, 'number': 21} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1} {'precision': 0.5833333333333334, 'recall': 0.7777777777777778, 'f1': 0.6666666666666666, 'number': 9} {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 8} {'precision': 0.25, 'recall': 0.25, 'f1': 0.25, 'number': 12} 0.4528 0.4 0.4248 0.8698

Framework versions

  • Transformers 4.25.1
  • Pytorch 1.12.1
  • Datasets 2.8.0
  • Tokenizers 0.13.2
Downloads last month
82
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.