<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
polish_transliterator_T5
This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.0705
- Rouge1: 15.1042
- Rouge2: 0.0
- Rougel: 15.1042
- Rougelsum: 15.625
- Gen Len: 4.0938
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: 3e-06
- 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: 20
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
---|---|---|---|---|---|---|---|---|
3.0242 | 1.0 | 572 | 1.8076 | 3.5937 | 0.0 | 3.75 | 3.75 | 1.25 |
2.8296 | 2.0 | 1144 | 1.6997 | 4.6875 | 0.0 | 4.6875 | 4.6875 | 0.7031 |
2.4707 | 3.0 | 1716 | 1.5717 | 6.0417 | 0.0 | 6.25 | 6.3542 | 1.1719 |
2.4367 | 4.0 | 2288 | 1.4617 | 6.4062 | 0.0 | 6.875 | 6.875 | 0.9688 |
2.296 | 5.0 | 2860 | 1.3847 | 8.4375 | 0.0 | 8.125 | 8.4375 | 1.3906 |
2.0905 | 6.0 | 3432 | 1.3177 | 8.4375 | 0.0 | 8.125 | 8.4375 | 1.9688 |
1.8223 | 7.0 | 4004 | 1.2645 | 9.375 | 0.0 | 9.375 | 9.375 | 2.3125 |
1.6881 | 8.0 | 4576 | 1.2157 | 10.625 | 0.0 | 10.625 | 10.9375 | 2.7969 |
1.6655 | 9.0 | 5148 | 1.1841 | 12.5 | 0.0 | 12.2917 | 12.5 | 3.1562 |
1.5736 | 10.0 | 5720 | 1.1582 | 13.4896 | 0.0 | 13.3333 | 13.3333 | 3.25 |
1.4754 | 11.0 | 6292 | 1.1382 | 13.4896 | 0.0 | 13.3333 | 13.3333 | 3.6562 |
1.4927 | 12.0 | 6864 | 1.1176 | 13.4896 | 0.0 | 13.3333 | 13.3333 | 4.1406 |
1.3628 | 13.0 | 7436 | 1.1069 | 13.4896 | 0.0 | 13.3333 | 13.3333 | 4.1719 |
1.3288 | 14.0 | 8008 | 1.0968 | 13.4896 | 0.0 | 13.3333 | 13.3333 | 4.2344 |
1.313 | 15.0 | 8580 | 1.0889 | 14.7917 | 0.0 | 14.7917 | 15.1042 | 4.2188 |
1.3215 | 16.0 | 9152 | 1.0820 | 14.7917 | 0.0 | 14.7917 | 15.1042 | 4.2188 |
1.2772 | 17.0 | 9724 | 1.0771 | 14.7917 | 0.0 | 14.7917 | 15.1042 | 4.2188 |
1.1895 | 18.0 | 10296 | 1.0735 | 15.1042 | 0.0 | 15.1042 | 15.625 | 4.0938 |
1.3394 | 19.0 | 10868 | 1.0712 | 15.1042 | 0.0 | 15.1042 | 15.625 | 4.0938 |
1.2656 | 20.0 | 11440 | 1.0705 | 15.1042 | 0.0 | 15.1042 | 15.625 | 4.0938 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.3