results_t5small_first25
This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.5455
- Rouge1: 0.1758
- Rouge2: 0.0585
- Rougel: 0.1389
- Rougelsum: 0.1388
- Gen Len: 20.0
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: 0.0005
- 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
- lr_scheduler_warmup_steps: 500
- num_epochs: 2
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 2.9172 | 0.32 | 1000 | 2.7575 | 0.1643 | 0.051 | 0.1293 | 0.1293 | 20.0 |
| 2.7163 | 0.64 | 2000 | 2.6575 | 0.1738 | 0.0568 | 0.1377 | 0.1376 | 20.0 |
| 2.8039 | 0.96 | 3000 | 2.6067 | 0.1722 | 0.0576 | 0.137 | 0.1369 | 20.0 |
| 2.6519 | 1.28 | 4000 | 2.5771 | 0.1758 | 0.0589 | 0.1384 | 0.1382 | 20.0 |
| 2.6342 | 1.6 | 5000 | 2.5627 | 0.1771 | 0.0603 | 0.1401 | 0.1401 | 20.0 |
| 2.5756 | 1.92 | 6000 | 2.5455 | 0.1758 | 0.0585 | 0.1389 | 0.1388 | 20.0 |
Framework versions
- Transformers 4.47.0
- Pytorch 2.5.1+cu121
- Datasets 3.3.1
- Tokenizers 0.21.0