my_wikilingua_model_mbart_sup
This model is a fine-tuned version of facebook/mbart-large-50 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.6924
- Rouge1: 0.202
- Rouge2: 0.054
- Rougel: 0.1637
- Rougelsum: 0.1636
- Gen Len: 55.06
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: 1
- eval_batch_size: 1
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 2.5419 | 1.0 | 1600 | 2.4122 | 0.1341 | 0.0278 | 0.1199 | 0.12 | 38.345 |
| 5.6466 | 2.0 | 3200 | 9.1834 | 0.0 | 0.0 | 0.0 | 0.0 | 19.0425 |
| 1.5485 | 3.0 | 4800 | 2.6034 | 0.1779 | 0.0415 | 0.1443 | 0.1445 | 60.155 |
| 1.1824 | 4.0 | 6400 | 2.6924 | 0.202 | 0.054 | 0.1637 | 0.1636 | 55.06 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3