<!-- 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-large-gramatika-final-e8-b16
This model is a fine-tuned version of google/mt5-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1999
- Rouge1: 66.308
- Rouge2: 58.8739
- Rougel: 66.1027
- Rougelsum: 66.1039
- Gen Len: 18.5592
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adafactor
- lr_scheduler_type: linear
- num_epochs: 8
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 0.8846 | 0.37 | 300 | 0.2954 | 64.6179 | 55.294 | 64.2807 | 64.2792 | 18.5597 |
| 0.3711 | 0.73 | 600 | 0.2474 | 65.6388 | 57.2663 | 65.3219 | 65.3365 | 18.5592 |
| 0.2874 | 1.1 | 900 | 0.2193 | 65.8689 | 57.6871 | 65.5424 | 65.5719 | 18.5603 |
| 0.1953 | 1.46 | 1200 | 0.2131 | 66.0438 | 57.8166 | 65.7565 | 65.7705 | 18.5409 |
| 0.1919 | 1.83 | 1500 | 0.1999 | 66.308 | 58.8739 | 66.1027 | 66.1039 | 18.5592 |
| 0.1487 | 2.2 | 1800 | 0.2034 | 66.5939 | 59.0628 | 66.3361 | 66.3475 | 18.5592 |
| 0.1132 | 2.56 | 2100 | 0.2010 | 67.0441 | 59.8117 | 66.8455 | 66.8562 | 18.5487 |
| 0.1087 | 2.93 | 2400 | 0.2001 | 67.0048 | 59.7807 | 66.7885 | 66.7972 | 18.5535 |
| 0.0681 | 3.29 | 2700 | 0.2143 | 67.2327 | 60.2527 | 67.0047 | 67.0106 | 18.5556 |
| 0.0621 | 3.66 | 3000 | 0.2093 | 67.357 | 60.51 | 67.1561 | 67.1709 | 18.5466 |
| 0.062 | 4.02 | 3300 | 0.2157 | 67.4353 | 60.7193 | 67.2526 | 67.2554 | 18.5624 |
| 0.036 | 4.39 | 3600 | 0.2208 | 67.5469 | 60.8111 | 67.3457 | 67.3472 | 18.5503 |
| 0.0351 | 4.76 | 3900 | 0.2282 | 67.3835 | 60.4009 | 67.138 | 67.1612 | 18.5561 |
| 0.0297 | 5.12 | 4200 | 0.2370 | 67.4004 | 60.5787 | 67.2004 | 67.2087 | 18.5603 |
| 0.0193 | 5.49 | 4500 | 0.2446 | 67.5339 | 60.6808 | 67.3484 | 67.3737 | 18.5577 |
| 0.0185 | 5.85 | 4800 | 0.2483 | 67.5055 | 60.8104 | 67.3217 | 67.3443 | 18.5566 |
| 0.0134 | 6.22 | 5100 | 0.2563 | 67.5748 | 60.9475 | 67.3996 | 67.4081 | 18.5597 |
| 0.0114 | 6.59 | 5400 | 0.2585 | 67.6337 | 61.0146 | 67.4553 | 67.472 | 18.5482 |
| 0.0099 | 6.95 | 5700 | 0.2622 | 67.6613 | 61.037 | 67.4761 | 67.4843 | 18.5498 |
| 0.0067 | 7.32 | 6000 | 0.2728 | 67.7996 | 61.2206 | 67.6194 | 67.6282 | 18.5561 |
| 0.0052 | 7.68 | 6300 | 0.2802 | 67.8009 | 61.2862 | 67.6178 | 67.6357 | 18.5545 |
Framework versions
- Transformers 4.30.1
- Pytorch 1.11.0a0+b6df043
- Datasets 2.12.0
- Tokenizers 0.13.3
