fine_tuned_t5_small_model_sec_5_v4
This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.8764
- Rouge1: 0.355
- Rouge2: 0.1153
- Rougel: 0.2143
- Rougelsum: 0.2145
- Gen Len: 78.3077
- Bert F1: 0.8693
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len | Bert F1 |
|---|---|---|---|---|---|---|---|---|---|
| 3.6662 | 0.6667 | 20 | 3.4084 | 0.3269 | 0.1126 | 0.216 | 0.215 | 57.9487 | 0.8683 |
| 3.5242 | 1.3333 | 40 | 3.1961 | 0.3466 | 0.1162 | 0.2206 | 0.2198 | 64.5983 | 0.8691 |
| 3.2821 | 2.0 | 60 | 3.0743 | 0.3538 | 0.1216 | 0.2251 | 0.2246 | 68.8718 | 0.8692 |
| 3.215 | 2.6667 | 80 | 3.0090 | 0.3558 | 0.1177 | 0.2198 | 0.2189 | 77.4786 | 0.8688 |
| 3.0862 | 3.3333 | 100 | 2.9625 | 0.3572 | 0.1138 | 0.2156 | 0.2154 | 82.1197 | 0.8679 |
| 3.1459 | 4.0 | 120 | 2.9354 | 0.3607 | 0.1192 | 0.2201 | 0.2201 | 82.4872 | 0.8684 |
| 3.1067 | 4.6667 | 140 | 2.9168 | 0.3565 | 0.1157 | 0.2132 | 0.2131 | 80.3675 | 0.8686 |
| 3.0274 | 5.3333 | 160 | 2.9061 | 0.3554 | 0.1147 | 0.2134 | 0.2132 | 80.8974 | 0.8687 |
| 3.0695 | 6.0 | 180 | 2.8974 | 0.356 | 0.1135 | 0.2147 | 0.2146 | 80.9316 | 0.8686 |
| 3.0547 | 6.6667 | 200 | 2.8897 | 0.3528 | 0.1123 | 0.2119 | 0.2114 | 79.3248 | 0.8684 |
| 3.0037 | 7.3333 | 220 | 2.8846 | 0.3522 | 0.1122 | 0.2108 | 0.211 | 79.265 | 0.8685 |
| 3.0316 | 8.0 | 240 | 2.8806 | 0.354 | 0.1149 | 0.2134 | 0.2134 | 78.6581 | 0.8689 |
| 3.0027 | 8.6667 | 260 | 2.8786 | 0.354 | 0.1151 | 0.2129 | 0.213 | 78.1624 | 0.8693 |
| 3.0372 | 9.3333 | 280 | 2.8769 | 0.355 | 0.1153 | 0.2143 | 0.2145 | 78.3077 | 0.8693 |
| 3.0197 | 10.0 | 300 | 2.8764 | 0.355 | 0.1153 | 0.2143 | 0.2145 | 78.3077 | 0.8693 |
Framework versions
- Transformers 4.46.3
- Pytorch 2.4.0
- Datasets 3.1.0
- Tokenizers 0.20.3