sum_model_0318_15_20epochs
This is the model trained for 20 epochs!! (actually it does not goe improvements since epoch 15)
This model is a fine-tuned version of weny22/sum_model_0318_20epochs on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.0783
- Rouge1: 0.1997
- Rouge2: 0.0704
- Rougel: 0.1603
- Rougelsum: 0.1603
- Gen Len: 18.9507
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: 20
- eval_batch_size: 20
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
---|---|---|---|---|---|---|---|---|
2.3596 | 1.0 | 1071 | 2.0975 | 0.1985 | 0.0698 | 0.1599 | 0.1598 | 18.944 |
2.3513 | 2.0 | 2142 | 2.1044 | 0.1987 | 0.0694 | 0.1596 | 0.1596 | 18.9507 |
2.3632 | 3.0 | 3213 | 2.0884 | 0.2005 | 0.0714 | 0.1609 | 0.1609 | 18.9447 |
2.3713 | 4.0 | 4284 | 2.0795 | 0.2001 | 0.0712 | 0.1608 | 0.1607 | 18.9513 |
2.3908 | 5.0 | 5355 | 2.0783 | 0.1997 | 0.0704 | 0.1603 | 0.1603 | 18.9507 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 4