bert-small-eurlex
This model is a fine-tuned version of google/bert_uncased_L-4_H-512_A-8 on the eurlex dataset. It achieves the following results on the evaluation set:
- Loss: 1.4260
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: 10
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 80
- 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 |
|---|---|---|---|
| 2.9536 | 1.5 | 1000 | 2.0670 |
| 2.0331 | 3.0 | 2000 | 1.7540 |
| 1.8046 | 4.5 | 3000 | 1.5993 |
| 1.678 | 6.0 | 4000 | 1.5039 |
| 1.6074 | 7.5 | 5000 | 1.4544 |
| 1.5664 | 8.99 | 6000 | 1.4260 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1