bert_uncased_L-4_H-128_A-2_emotion
This model is a fine-tuned version of google/bert_uncased_L-4_H-128_A-2 on the emotion dataset. It achieves the following results on the evaluation set:
- Loss: 0.2776
- Accuracy: 0.916
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: 64
- eval_batch_size: 64
- seed: 33
- distributed_type: multi-GPU
- 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 |
|---|---|---|---|---|
| 1.4323 | 1.0 | 250 | 1.1059 | 0.6495 |
| 0.9588 | 2.0 | 500 | 0.7596 | 0.788 |
| 0.6816 | 3.0 | 750 | 0.5526 | 0.8505 |
| 0.5074 | 4.0 | 1000 | 0.4233 | 0.895 |
| 0.4131 | 5.0 | 1250 | 0.3628 | 0.906 |
| 0.3536 | 6.0 | 1500 | 0.3261 | 0.9015 |
| 0.3128 | 7.0 | 1750 | 0.2960 | 0.911 |
| 0.2905 | 8.0 | 2000 | 0.2894 | 0.9125 |
| 0.2765 | 9.0 | 2250 | 0.2776 | 0.916 |
| 0.2682 | 10.0 | 2500 | 0.2728 | 0.915 |
Framework versions
- Transformers 4.34.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.14.5
- Tokenizers 0.14.1