flan-t5-base-turkish-summarisation
This model is a fine-tuned version of google/flan-t5-base on the mlsum dataset. It achieves the following results on the evaluation set:
- Loss: 1.2759
- Rouge1: 17.7215
- Rouge2: 11.6449
- Rougel: 17.1215
- Rougelsum: 17.0317
- 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: 5e-05
- train_batch_size: 4
- eval_batch_size: 4
- 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
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 1.5856 | 0.0802 | 200 | 1.3164 | 18.0769 | 11.7482 | 17.3132 | 17.32 | 20.0 |
| 1.4888 | 0.1604 | 400 | 1.2901 | 17.6893 | 11.6682 | 16.9148 | 16.8964 | 20.0 |
| 1.4787 | 0.2407 | 600 | 1.2827 | 17.5252 | 11.5143 | 16.8586 | 16.8281 | 20.0 |
| 1.488 | 0.3209 | 800 | 1.2637 | 17.8913 | 11.7712 | 17.1369 | 17.0949 | 20.0 |
| 1.4105 | 0.4011 | 1000 | 1.2759 | 17.7215 | 11.6449 | 17.1215 | 17.0317 | 20.0 |
Framework versions
- Transformers 4.48.0
- Pytorch 2.2.2
- Datasets 3.2.0
- Tokenizers 0.21.0