Pretrained checkpoint: bert-base-uncased
Traning hyperparameters:
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 24
- eval_batch_size: 24
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- prompt_format: sentence aspect - sentiment
Training results
| Epoch | Train loss | Subtask 3 f1 | Subtask 3 precision | Subtask 3 recall | Subtask4 accuracy |
|---|---|---|---|---|---|
| 1 | 305.5731324516237 | 0.8653648509763618 | 0.9142236699239956 | 0.8214634146341463 | 0.7921951219512195 |
| 2 | 160.19575848057866 | 0.8591029023746701 | 0.9356321839080459 | 0.7941463414634147 | 0.8009756097560976 |
| 3 | 101.52328581456095 | 0.8882175226586102 | 0.9177939646201873 | 0.8604878048780488 | 0.8321951219512195 |
| 4 | 63.44610589882359 | 0.8818737270875764 | 0.9222577209797657 | 0.8448780487804878 | 0.8282926829268292 |
| 5 | 43.48708916385658 | 0.8917835671342685 | 0.9165808444902163 | 0.8682926829268293 | 0.8214634146341463 |
