CNEC2_0_Supertypes_xlm-roberta-large
This model is a fine-tuned version of FacebookAI/xlm-roberta-large on the cnec dataset. It achieves the following results on the evaluation set:
- Loss: 0.1551
- Precision: 0.8389
- Recall: 0.8649
- F1: 0.8517
- Accuracy: 0.9678
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: 16
- eval_batch_size: 16
- 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 | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 1.0 | 450 | 0.1584 | 0.7594 | 0.7993 | 0.7788 | 0.9572 |
0.3769 | 2.0 | 900 | 0.1327 | 0.8143 | 0.8476 | 0.8306 | 0.9653 |
0.1646 | 3.0 | 1350 | 0.1236 | 0.8242 | 0.8658 | 0.8445 | 0.9674 |
0.1242 | 4.0 | 1800 | 0.1384 | 0.8143 | 0.8641 | 0.8385 | 0.9654 |
0.1044 | 5.0 | 2250 | 0.1379 | 0.8346 | 0.8798 | 0.8566 | 0.9675 |
0.0824 | 6.0 | 2700 | 0.1354 | 0.8254 | 0.8496 | 0.8374 | 0.9679 |
0.0748 | 7.0 | 3150 | 0.1436 | 0.8304 | 0.8637 | 0.8467 | 0.9676 |
0.0656 | 8.0 | 3600 | 0.1540 | 0.8286 | 0.8587 | 0.8434 | 0.9662 |
0.0592 | 9.0 | 4050 | 0.1515 | 0.8407 | 0.8653 | 0.8528 | 0.9675 |
0.0508 | 10.0 | 4500 | 0.1551 | 0.8389 | 0.8649 | 0.8517 | 0.9678 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0
- Downloads last month
- 22
Finetuned from
Evaluation results
- Precision on cnecvalidation set self-reported0.839
- Recall on cnecvalidation set self-reported0.865
- F1 on cnecvalidation set self-reported0.852
- Accuracy on cnecvalidation set self-reported0.968