<!-- 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. -->
mt5-base-gecid23-e3
This model is a fine-tuned version of google/mt5-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2913
- Rouge1: 64.5987
- Rouge2: 58.284
- Rougel: 64.5263
- Rougelsum: 64.5192
- Gen Len: 18.7512
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: 0.001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adafactor
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 1.5553 | 0.25 | 221 | 0.5815 | 58.7873 | 48.3787 | 58.6622 | 58.6428 | 18.7486 |
| 0.6944 | 0.5 | 442 | 0.5010 | 60.225 | 50.8407 | 60.1109 | 60.0966 | 18.7418 |
| 0.5891 | 0.75 | 663 | 0.4477 | 61.4891 | 53.2811 | 61.4099 | 61.4089 | 18.7588 |
| 0.5145 | 1.0 | 884 | 0.3926 | 62.3704 | 54.3562 | 62.255 | 62.252 | 18.7520 |
| 0.3682 | 1.25 | 1105 | 0.3805 | 62.4976 | 54.8233 | 62.4265 | 62.4327 | 18.7622 |
| 0.3332 | 1.5 | 1326 | 0.3471 | 63.2736 | 56.0263 | 63.1982 | 63.1901 | 18.7495 |
| 0.3097 | 1.75 | 1547 | 0.3173 | 63.5672 | 56.5358 | 63.4813 | 63.4756 | 18.7541 |
| 0.2958 | 2.0 | 1768 | 0.3219 | 63.8092 | 57.1715 | 63.7764 | 63.7692 | 18.7512 |
| 0.1901 | 2.25 | 1989 | 0.3053 | 64.1292 | 57.5296 | 64.052 | 64.0478 | 18.7533 |
| 0.1861 | 2.5 | 2210 | 0.3018 | 64.4658 | 58.0416 | 64.3975 | 64.3918 | 18.7537 |
| 0.1696 | 2.75 | 2431 | 0.2928 | 64.5337 | 58.1328 | 64.4735 | 64.4619 | 18.7507 |
| 0.1691 | 3.0 | 2652 | 0.2913 | 64.5987 | 58.284 | 64.5263 | 64.5192 | 18.7512 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
