deberta-v3-small-company-names
This model is a fine-tuned version of microsoft/deberta-v3-small on the nbroad/company_names dataset. It achieves the following results on the evaluation set:
- Loss: 0.0707
- Precision: 0.7688
- Recall: 0.7921
- F1: 0.7802
- Accuracy: 0.9766
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: 8e-05
- train_batch_size: 48
- eval_batch_size: 8
- 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: 3.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.0746 | 1.0 | 2126 | 0.0657 | 0.7415 | 0.7868 | 0.7635 | 0.9753 |
| 0.0485 | 2.0 | 4252 | 0.0651 | 0.7631 | 0.7904 | 0.7765 | 0.9764 |
| 0.044 | 3.0 | 6378 | 0.0707 | 0.7688 | 0.7921 | 0.7802 | 0.9766 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.0.1+cu117
- Datasets 2.16.1
- Tokenizers 0.14.1