hasoc19-microsoft-mdeberta-v3-base-targinsult1
This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9030
- Accuracy: 0.6958
- Precision: 0.6536
- Recall: 0.6464
- F1: 0.6493
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: 1e-05
- train_batch_size: 32
- eval_batch_size: 32
- 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 | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 263 | 0.5647 | 0.6858 | 0.6528 | 0.6609 | 0.6556 |
| 0.5745 | 2.0 | 526 | 0.5520 | 0.6987 | 0.6612 | 0.6626 | 0.6619 |
| 0.5745 | 3.0 | 789 | 0.6162 | 0.7139 | 0.6764 | 0.6733 | 0.6747 |
| 0.4908 | 4.0 | 1052 | 0.6262 | 0.7063 | 0.6698 | 0.6715 | 0.6706 |
| 0.4908 | 5.0 | 1315 | 0.6799 | 0.7067 | 0.6661 | 0.6568 | 0.6604 |
| 0.4069 | 6.0 | 1578 | 0.7646 | 0.7044 | 0.6635 | 0.6550 | 0.6584 |
| 0.4069 | 7.0 | 1841 | 0.7791 | 0.7029 | 0.6646 | 0.6633 | 0.6639 |
| 0.338 | 8.0 | 2104 | 0.8646 | 0.7029 | 0.6623 | 0.6554 | 0.6582 |
| 0.338 | 9.0 | 2367 | 0.9417 | 0.7006 | 0.6556 | 0.6317 | 0.6374 |
| 0.2847 | 10.0 | 2630 | 0.9030 | 0.6958 | 0.6536 | 0.6464 | 0.6493 |
Framework versions
- Transformers 4.24.0.dev0
- Pytorch 1.11.0+cu102
- Datasets 2.6.1
- Tokenizers 0.13.1