rest15_50
This model is a fine-tuned version of google-t5/t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.0418
- Precision: 0.2766
- Recall: 0.1873
- F1: 0.2234
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 |
|---|---|---|---|---|---|---|
| No log | 1.0 | 7 | 3.6671 | 0 | 0.0 | 0 |
| No log | 2.0 | 14 | 2.2951 | 0.0 | 0.0 | 0 |
| No log | 3.0 | 21 | 1.7160 | 0.1111 | 0.0202 | 0.0341 |
| No log | 4.0 | 28 | 1.4462 | 0.1301 | 0.0548 | 0.0771 |
| No log | 5.0 | 35 | 1.2381 | 0.1789 | 0.098 | 0.1266 |
| No log | 6.0 | 42 | 1.1546 | 0.2047 | 0.1268 | 0.1566 |
| No log | 7.0 | 49 | 1.1470 | 0.2242 | 0.1441 | 0.1754 |
| No log | 8.0 | 56 | 1.1186 | 0.2385 | 0.1499 | 0.1841 |
| No log | 9.0 | 63 | 1.1021 | 0.2511 | 0.17 | 0.2027 |
| No log | 10.0 | 70 | 1.0770 | 0.2533 | 0.1671 | 0.2014 |
| No log | 11.0 | 77 | 1.0839 | 0.2596 | 0.1556 | 0.1946 |
| No log | 12.0 | 84 | 1.0612 | 0.2667 | 0.1729 | 0.2098 |
| No log | 13.0 | 91 | 1.0618 | 0.2739 | 0.1816 | 0.2184 |
| No log | 14.0 | 98 | 1.0613 | 0.2641 | 0.1758 | 0.2111 |
| No log | 15.0 | 105 | 1.0557 | 0.2727 | 0.1816 | 0.218 |
| No log | 16.0 | 112 | 1.0416 | 0.2573 | 0.1787 | 0.2109 |
| No log | 17.0 | 119 | 1.0418 | 0.2746 | 0.1931 | 0.2267 |
| No log | 18.0 | 126 | 1.0438 | 0.2797 | 0.1902 | 0.2264 |
| No log | 19.0 | 133 | 1.0432 | 0.2754 | 0.1873 | 0.223 |
| No log | 20.0 | 140 | 1.0418 | 0.2766 | 0.1873 | 0.2234 |
Framework versions
- Transformers 4.39.1
- Pytorch 2.7.0+cu118
- Datasets 2.12.0
- Tokenizers 0.15.2