t5-small-finetuned-kernewek-plus-br-cy
This model is a fine-tuned version of Jendersen/t5-small-finetuned-kernewek-plus-br-cy on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1670
- Rouge1: 27.318
- Rouge2: 13.5923
- Rougel: 25.1685
- Rougelsum: 25.3722
- Bleu: 0.0151
- Chrf++: 14.341
- F1: 41.5514
- Gen Len: 19.9586
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 45
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Bleu | Chrf++ | F1 | Gen Len |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1.5062 | 1.0 | 1485 | 1.3429 | 25.6029 | 11.7283 | 23.5488 | 23.753 | 0.0124 | 13.0823 | 40.4386 | 19.9619 |
| 1.4881 | 2.0 | 2970 | 1.3335 | 25.7116 | 11.8205 | 23.6827 | 23.8737 | 0.0125 | 13.1671 | 40.5548 | 19.9614 |
| 1.4774 | 3.0 | 4455 | 1.3220 | 25.8298 | 11.9618 | 23.7697 | 23.9725 | 0.0127 | 13.2465 | 40.5981 | 19.9635 |
| 1.4594 | 4.0 | 5940 | 1.3118 | 25.8411 | 12.0124 | 23.8195 | 24.021 | 0.0126 | 13.2928 | 40.6417 | 19.9631 |
| 1.4513 | 5.0 | 7425 | 1.3006 | 25.9926 | 12.1855 | 23.9235 | 24.1263 | 0.0129 | 13.4023 | 40.7798 | 19.9621 |
| 1.4255 | 6.0 | 8910 | 1.2946 | 26.26 | 12.3361 | 24.1914 | 24.3947 | 0.0134 | 13.5309 | 40.9229 | 19.9614 |
| 1.4205 | 7.0 | 10395 | 1.2832 | 26.1618 | 12.3798 | 24.1252 | 24.3218 | 0.0132 | 13.5239 | 40.8967 | 19.9608 |
| 1.4065 | 8.0 | 11880 | 1.2788 | 26.4143 | 12.4893 | 24.3089 | 24.5138 | 0.0135 | 13.6449 | 41.0338 | 19.9608 |
| 1.4022 | 9.0 | 13365 | 1.2693 | 26.3817 | 12.4848 | 24.283 | 24.4877 | 0.0135 | 13.6422 | 40.9885 | 19.9614 |
| 1.3947 | 10.0 | 14850 | 1.2635 | 26.4912 | 12.6371 | 24.4041 | 24.6144 | 0.0137 | 13.7143 | 41.0899 | 19.9619 |
| 1.3807 | 11.0 | 16335 | 1.2556 | 26.5174 | 12.6435 | 24.4404 | 24.6327 | 0.0136 | 13.7252 | 41.1222 | 19.9609 |
| 1.3696 | 12.0 | 17820 | 1.2488 | 26.5883 | 12.7175 | 24.5004 | 24.6991 | 0.0138 | 13.7714 | 41.1291 | 19.9633 |
| 1.3591 | 13.0 | 19305 | 1.2430 | 26.626 | 12.7825 | 24.5489 | 24.7576 | 0.014 | 13.8226 | 41.0921 | 19.9606 |
| 1.3570 | 14.0 | 20790 | 1.2388 | 26.6863 | 12.8661 | 24.5939 | 24.7962 | 0.014 | 13.8611 | 41.1619 | 19.9603 |
| 1.3530 | 15.0 | 22275 | 1.2333 | 26.7387 | 12.9319 | 24.6663 | 24.8517 | 0.0141 | 13.9063 | 41.1669 | 19.9592 |
| 1.3343 | 16.0 | 23760 | 1.2272 | 26.7814 | 12.9765 | 24.6644 | 24.86 | 0.0143 | 13.9511 | 41.1985 | 19.9614 |
| 1.3246 | 17.0 | 25245 | 1.2227 | 26.8 | 13.0579 | 24.6996 | 24.9041 | 0.0142 | 13.9546 | 41.1806 | 19.9601 |
| 1.3228 | 18.0 | 26730 | 1.2171 | 26.8662 | 13.0906 | 24.7478 | 24.9428 | 0.0143 | 14.0091 | 41.264 | 19.9599 |
| 1.3139 | 19.0 | 28215 | 1.2156 | 26.9931 | 13.1569 | 24.8493 | 25.0564 | 0.0146 | 14.0802 | 41.2823 | 19.9599 |
| 1.3125 | 20.0 | 29700 | 1.2106 | 27.0149 | 13.2193 | 24.9087 | 25.1058 | 0.0145 | 14.1095 | 41.3381 | 19.9586 |
| 1.3190 | 21.0 | 31185 | 1.2067 | 27.0084 | 13.228 | 24.8653 | 25.0618 | 0.0146 | 14.0888 | 41.3169 | 19.9603 |
| 1.3155 | 22.0 | 32670 | 1.2026 | 27.0349 | 13.2667 | 24.9248 | 25.1266 | 0.0146 | 14.1182 | 41.3522 | 19.9591 |
| 1.3001 | 23.0 | 34155 | 1.1996 | 27.0917 | 13.3104 | 24.9344 | 25.1363 | 0.0147 | 14.144 | 41.3756 | 19.9591 |
| 1.3043 | 24.0 | 35640 | 1.1957 | 27.1726 | 13.3511 | 25.0214 | 25.2124 | 0.0148 | 14.1805 | 41.4131 | 19.9592 |
| 1.2931 | 25.0 | 37125 | 1.1951 | 27.1423 | 13.3543 | 24.9929 | 25.1952 | 0.0149 | 14.1989 | 41.4103 | 19.9581 |
| 1.2898 | 26.0 | 38610 | 1.1901 | 27.0899 | 13.3727 | 24.9498 | 25.1611 | 0.0148 | 14.1824 | 41.3881 | 19.9594 |
| 1.2863 | 27.0 | 40095 | 1.1878 | 27.0954 | 13.4075 | 24.9747 | 25.1778 | 0.0148 | 14.183 | 41.3753 | 19.9581 |
| 1.2884 | 28.0 | 41580 | 1.1862 | 27.1579 | 13.4647 | 25.0357 | 25.228 | 0.0149 | 14.2228 | 41.401 | 19.9579 |
| 1.2828 | 29.0 | 43065 | 1.1824 | 27.1882 | 13.4747 | 25.0503 | 25.2599 | 0.015 | 14.2522 | 41.4186 | 19.9579 |
| 1.2751 | 30.0 | 44550 | 1.1812 | 27.1544 | 13.4621 | 25.0335 | 25.2313 | 0.0149 | 14.246 | 41.449 | 19.9586 |
| 1.2754 | 31.0 | 46035 | 1.1797 | 27.244 | 13.5217 | 25.0988 | 25.3078 | 0.0151 | 14.2867 | 41.4518 | 19.9579 |
| 1.2727 | 32.0 | 47520 | 1.1781 | 27.2461 | 13.5148 | 25.0875 | 25.2872 | 0.015 | 14.286 | 41.4687 | 19.9603 |
| 1.2707 | 33.0 | 49005 | 1.1768 | 27.249 | 13.5417 | 25.1162 | 25.3072 | 0.0151 | 14.2962 | 41.5131 | 19.9598 |
| 1.2678 | 34.0 | 50490 | 1.1749 | 27.2825 | 13.5632 | 25.138 | 25.3386 | 0.0152 | 14.317 | 41.4794 | 19.9587 |
| 1.2679 | 35.0 | 51975 | 1.1732 | 27.3079 | 13.5821 | 25.1551 | 25.3552 | 0.0151 | 14.3222 | 41.5345 | 19.9591 |
| 1.2604 | 36.0 | 53460 | 1.1714 | 27.275 | 13.5615 | 25.1178 | 25.316 | 0.0151 | 14.3055 | 41.5094 | 19.9591 |
| 1.2615 | 37.0 | 54945 | 1.1711 | 27.3122 | 13.5727 | 25.1475 | 25.3549 | 0.0152 | 14.3259 | 41.5286 | 19.9591 |
| 1.2599 | 38.0 | 56430 | 1.1708 | 27.3298 | 13.5941 | 25.1612 | 25.3621 | 0.0152 | 14.3451 | 41.54 | 19.9592 |
| 1.2570 | 39.0 | 57915 | 1.1691 | 27.322 | 13.5882 | 25.1582 | 25.3653 | 0.0152 | 14.3341 | 41.5336 | 19.9598 |
| 1.2601 | 40.0 | 59400 | 1.1688 | 27.332 | 13.6108 | 25.1875 | 25.3938 | 0.0152 | 14.3418 | 41.5489 | 19.9591 |
| 1.2646 | 41.0 | 60885 | 1.1684 | 27.3288 | 13.6272 | 25.1822 | 25.3878 | 0.0152 | 14.3557 | 41.5653 | 19.9599 |
| 1.2608 | 42.0 | 62370 | 1.1677 | 27.3472 | 13.5869 | 25.1854 | 25.3888 | 0.0152 | 14.3419 | 41.5599 | 19.9589 |
| 1.2592 | 43.0 | 63855 | 1.1671 | 27.3276 | 13.5815 | 25.179 | 25.3812 | 0.0151 | 14.3379 | 41.5377 | 19.9586 |
| 1.2568 | 44.0 | 65340 | 1.1670 | 27.2974 | 13.5784 | 25.147 | 25.3529 | 0.0151 | 14.3338 | 41.5502 | 19.9586 |
| 1.2546 | 45.0 | 66825 | 1.1670 | 27.318 | 13.5923 | 25.1685 | 25.3722 | 0.0151 | 14.341 | 41.5514 | 19.9586 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2