t5-base-finetuned-en-to-it-lrs-back
This model is a fine-tuned version of t5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.2426
- Bleu: 27.5339
- Gen Len: 50.356
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 40
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 1.997 | 1.0 | 1125 | 1.8350 | 10.7997 | 58.974 |
| 1.8185 | 2.0 | 2250 | 1.7013 | 13.8533 | 56.2727 |
| 1.6898 | 3.0 | 3375 | 1.6177 | 16.2932 | 53.914 |
| 1.5706 | 4.0 | 4500 | 1.5542 | 18.0621 | 53.4973 |
| 1.5151 | 5.0 | 5625 | 1.5086 | 19.4227 | 53.6307 |
| 1.468 | 6.0 | 6750 | 1.4707 | 20.8101 | 52.2427 |
| 1.4313 | 7.0 | 7875 | 1.4438 | 21.7975 | 51.8 |
| 1.383 | 8.0 | 9000 | 1.4182 | 22.4852 | 51.7193 |
| 1.3555 | 9.0 | 10125 | 1.3991 | 23.0898 | 51.448 |
| 1.3259 | 10.0 | 11250 | 1.3802 | 23.5277 | 51.6953 |
| 1.3041 | 11.0 | 12375 | 1.3656 | 23.9072 | 51.1547 |
| 1.2781 | 12.0 | 13500 | 1.3518 | 24.1772 | 51.3293 |
| 1.2577 | 13.0 | 14625 | 1.3394 | 24.4547 | 51.6307 |
| 1.24 | 14.0 | 15750 | 1.3312 | 24.9846 | 50.9827 |
| 1.2198 | 15.0 | 16875 | 1.3180 | 25.1942 | 51.042 |
| 1.2071 | 16.0 | 18000 | 1.3098 | 25.5082 | 50.6113 |
| 1.1948 | 17.0 | 19125 | 1.3024 | 25.523 | 50.782 |
| 1.1795 | 18.0 | 20250 | 1.2961 | 25.8367 | 50.7987 |
| 1.1691 | 19.0 | 21375 | 1.2904 | 25.9142 | 50.6667 |
| 1.159 | 20.0 | 22500 | 1.2824 | 26.2538 | 50.602 |
| 1.1504 | 21.0 | 23625 | 1.2794 | 26.2023 | 50.5987 |
| 1.1389 | 22.0 | 24750 | 1.2746 | 26.464 | 50.4593 |
| 1.1309 | 23.0 | 25875 | 1.2694 | 26.4899 | 50.5353 |
| 1.1185 | 24.0 | 27000 | 1.2676 | 26.8721 | 50.468 |
| 1.1126 | 25.0 | 28125 | 1.2635 | 26.8721 | 50.4693 |
| 1.1093 | 26.0 | 29250 | 1.2603 | 27.0334 | 50.334 |
| 1.104 | 27.0 | 30375 | 1.2569 | 27.2444 | 50.554 |
| 1.0988 | 28.0 | 31500 | 1.2535 | 27.2597 | 50.5367 |
| 1.0893 | 29.0 | 32625 | 1.2509 | 27.3268 | 50.42 |
| 1.0883 | 30.0 | 33750 | 1.2519 | 27.4412 | 50.4253 |
| 1.0802 | 31.0 | 34875 | 1.2479 | 27.3715 | 50.4247 |
| 1.0697 | 32.0 | 36000 | 1.2476 | 27.3871 | 50.5567 |
| 1.0729 | 33.0 | 37125 | 1.2457 | 27.4333 | 50.418 |
| 1.065 | 34.0 | 38250 | 1.2451 | 27.4571 | 50.4287 |
| 1.0649 | 35.0 | 39375 | 1.2451 | 27.5448 | 50.3393 |
| 1.0644 | 36.0 | 40500 | 1.2438 | 27.5387 | 50.2813 |
| 1.0624 | 37.0 | 41625 | 1.2423 | 27.5011 | 50.402 |
| 1.0617 | 38.0 | 42750 | 1.2434 | 27.5414 | 50.336 |
| 1.0606 | 39.0 | 43875 | 1.2429 | 27.5247 | 50.3387 |
| 1.0508 | 40.0 | 45000 | 1.2426 | 27.5339 | 50.356 |
Framework versions
- Transformers 4.22.1
- Pytorch 1.12.1
- Datasets 2.5.1
- Tokenizers 0.11.0