BulBERT-cinexio-10epochs
This model is a fine-tuned version of mor40/BulBERT-finetuned-cinexio on the bgglue dataset. It achieves the following results on the evaluation set:
- Loss: 1.1866
- Accuracy: 0.6289
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: 128
- eval_batch_size: 128
- 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 | Accuracy |
|---|---|---|---|---|
| No log | 1.0 | 64 | 1.3334 | 0.5746 |
| No log | 2.0 | 128 | 1.2053 | 0.6017 |
| No log | 3.0 | 192 | 1.1826 | 0.6227 |
| No log | 4.0 | 256 | 1.1826 | 0.6252 |
| No log | 5.0 | 320 | 1.1671 | 0.6227 |
| No log | 6.0 | 384 | 1.1743 | 0.6289 |
| No log | 7.0 | 448 | 1.1795 | 0.6375 |
| 1.0262 | 8.0 | 512 | 1.1847 | 0.6178 |
| 1.0262 | 9.0 | 576 | 1.1877 | 0.6264 |
| 1.0262 | 10.0 | 640 | 1.1866 | 0.6289 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1