<!-- 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. -->
t5-base-finetuned-unam-es-to-pua
This model is a fine-tuned version of t5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1424
- Bleu: 5.6655
- Gen Len: 18.3375
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 100
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
---|---|---|---|---|---|
No log | 1.0 | 35 | 3.4462 | 0.0 | 15.525 |
No log | 2.0 | 70 | 2.7253 | 0.3534 | 17.6375 |
No log | 3.0 | 105 | 2.3461 | 0.532 | 18.475 |
No log | 4.0 | 140 | 2.1536 | 0.5158 | 18.2625 |
No log | 5.0 | 175 | 2.0331 | 0.6926 | 18.5375 |
No log | 6.0 | 210 | 1.9242 | 0.7148 | 18.5 |
No log | 7.0 | 245 | 1.8474 | 0.5904 | 17.6125 |
No log | 8.0 | 280 | 1.7931 | 0.5887 | 17.525 |
No log | 9.0 | 315 | 1.7359 | 0.6126 | 17.7375 |
No log | 10.0 | 350 | 1.6828 | 0.5915 | 17.775 |
No log | 11.0 | 385 | 1.6440 | 0.677 | 17.5 |
No log | 12.0 | 420 | 1.6004 | 0.6503 | 17.6375 |
No log | 13.0 | 455 | 1.5656 | 1.1308 | 18.0375 |
No log | 14.0 | 490 | 1.5399 | 1.2924 | 18.1875 |
2.2426 | 15.0 | 525 | 1.5116 | 1.2398 | 18.05 |
2.2426 | 16.0 | 560 | 1.4969 | 1.5902 | 18.075 |
2.2426 | 17.0 | 595 | 1.4712 | 1.5022 | 18.0875 |
2.2426 | 18.0 | 630 | 1.4510 | 1.5548 | 18.175 |
2.2426 | 19.0 | 665 | 1.4385 | 2.4875 | 18.4 |
2.2426 | 20.0 | 700 | 1.4255 | 1.8098 | 18.25 |
2.2426 | 21.0 | 735 | 1.4175 | 1.3967 | 18.2875 |
2.2426 | 22.0 | 770 | 1.3840 | 1.4478 | 18.4125 |
2.2426 | 23.0 | 805 | 1.3794 | 2.5712 | 18.3875 |
2.2426 | 24.0 | 840 | 1.3651 | 1.6709 | 18.2625 |
2.2426 | 25.0 | 875 | 1.3516 | 2.7181 | 18.2125 |
2.2426 | 26.0 | 910 | 1.3438 | 1.9771 | 18.0625 |
2.2426 | 27.0 | 945 | 1.3204 | 3.3283 | 18.175 |
2.2426 | 28.0 | 980 | 1.3148 | 4.0754 | 18.2 |
1.4238 | 29.0 | 1015 | 1.3111 | 3.831 | 18.325 |
1.4238 | 30.0 | 1050 | 1.2939 | 4.2328 | 18.2375 |
1.4238 | 31.0 | 1085 | 1.2896 | 3.5101 | 18.3125 |
1.4238 | 32.0 | 1120 | 1.2833 | 3.6533 | 18.4125 |
1.4238 | 33.0 | 1155 | 1.2771 | 3.3986 | 18.5125 |
1.4238 | 34.0 | 1190 | 1.2727 | 3.3589 | 18.3125 |
1.4238 | 35.0 | 1225 | 1.2622 | 3.24 | 18.1625 |
1.4238 | 36.0 | 1260 | 1.2512 | 3.2334 | 18.45 |
1.4238 | 37.0 | 1295 | 1.2506 | 3.3842 | 18.5125 |
1.4238 | 38.0 | 1330 | 1.2407 | 3.2457 | 18.375 |
1.4238 | 39.0 | 1365 | 1.2315 | 3.4157 | 18.3875 |
1.4238 | 40.0 | 1400 | 1.2343 | 3.2254 | 18.3375 |
1.4238 | 41.0 | 1435 | 1.2334 | 3.3025 | 18.3375 |
1.4238 | 42.0 | 1470 | 1.2290 | 3.3478 | 18.3375 |
1.1647 | 43.0 | 1505 | 1.2209 | 3.3824 | 18.4 |
1.1647 | 44.0 | 1540 | 1.2088 | 3.6032 | 18.375 |
1.1647 | 45.0 | 1575 | 1.2077 | 3.549 | 18.425 |
1.1647 | 46.0 | 1610 | 1.2058 | 3.5507 | 18.3625 |
1.1647 | 47.0 | 1645 | 1.2030 | 3.8134 | 18.4125 |
1.1647 | 48.0 | 1680 | 1.2100 | 3.7041 | 18.4 |
1.1647 | 49.0 | 1715 | 1.1968 | 3.5977 | 18.4 |
1.1647 | 50.0 | 1750 | 1.1911 | 3.7133 | 18.3 |
1.1647 | 51.0 | 1785 | 1.1874 | 3.7578 | 18.3 |
1.1647 | 52.0 | 1820 | 1.1920 | 3.7871 | 18.2875 |
1.1647 | 53.0 | 1855 | 1.1867 | 3.7594 | 18.4 |
1.1647 | 54.0 | 1890 | 1.1880 | 3.7163 | 18.3375 |
1.1647 | 55.0 | 1925 | 1.1807 | 4.7929 | 18.375 |
1.1647 | 56.0 | 1960 | 1.1832 | 4.0148 | 18.3375 |
1.1647 | 57.0 | 1995 | 1.1789 | 4.7512 | 18.3875 |
1.0097 | 58.0 | 2030 | 1.1819 | 4.9173 | 18.35 |
1.0097 | 59.0 | 2065 | 1.1742 | 5.0857 | 18.3875 |
1.0097 | 60.0 | 2100 | 1.1771 | 4.006 | 18.4 |
1.0097 | 61.0 | 2135 | 1.1677 | 3.812 | 18.325 |
1.0097 | 62.0 | 2170 | 1.1683 | 4.1118 | 18.3625 |
1.0097 | 63.0 | 2205 | 1.1653 | 3.7104 | 18.3625 |
1.0097 | 64.0 | 2240 | 1.1578 | 3.785 | 18.4375 |
1.0097 | 65.0 | 2275 | 1.1616 | 3.8943 | 18.4 |
1.0097 | 66.0 | 2310 | 1.1617 | 4.0885 | 18.325 |
1.0097 | 67.0 | 2345 | 1.1689 | 5.0819 | 18.4 |
1.0097 | 68.0 | 2380 | 1.1602 | 5.0775 | 18.3375 |
1.0097 | 69.0 | 2415 | 1.1581 | 5.0943 | 18.375 |
1.0097 | 70.0 | 2450 | 1.1590 | 5.2458 | 18.3125 |
1.0097 | 71.0 | 2485 | 1.1605 | 4.925 | 18.425 |
0.9105 | 72.0 | 2520 | 1.1604 | 5.4352 | 18.35 |
0.9105 | 73.0 | 2555 | 1.1547 | 5.157 | 18.2875 |
0.9105 | 74.0 | 2590 | 1.1488 | 5.0934 | 18.3625 |
0.9105 | 75.0 | 2625 | 1.1487 | 5.1848 | 18.4 |
0.9105 | 76.0 | 2660 | 1.1530 | 5.1413 | 18.3625 |
0.9105 | 77.0 | 2695 | 1.1541 | 5.1413 | 18.4 |
0.9105 | 78.0 | 2730 | 1.1445 | 5.1848 | 18.3375 |
0.9105 | 79.0 | 2765 | 1.1509 | 5.2391 | 18.3375 |
0.9105 | 80.0 | 2800 | 1.1512 | 5.444 | 18.3375 |
0.9105 | 81.0 | 2835 | 1.1532 | 5.2204 | 18.4125 |
0.9105 | 82.0 | 2870 | 1.1551 | 5.3439 | 18.3375 |
0.9105 | 83.0 | 2905 | 1.1504 | 4.6498 | 18.4 |
0.9105 | 84.0 | 2940 | 1.1497 | 4.7896 | 18.4 |
0.9105 | 85.0 | 2975 | 1.1513 | 5.3543 | 18.35 |
0.8502 | 86.0 | 3010 | 1.1523 | 5.4728 | 18.35 |
0.8502 | 87.0 | 3045 | 1.1509 | 5.8559 | 18.35 |
0.8502 | 88.0 | 3080 | 1.1504 | 5.4819 | 18.35 |
0.8502 | 89.0 | 3115 | 1.1496 | 5.3352 | 18.35 |
0.8502 | 90.0 | 3150 | 1.1451 | 5.6849 | 18.2875 |
0.8502 | 91.0 | 3185 | 1.1436 | 5.6849 | 18.2875 |
0.8502 | 92.0 | 3220 | 1.1449 | 5.6473 | 18.3625 |
0.8502 | 93.0 | 3255 | 1.1425 | 5.6562 | 18.3 |
0.8502 | 94.0 | 3290 | 1.1445 | 5.6655 | 18.2875 |
0.8502 | 95.0 | 3325 | 1.1441 | 5.6655 | 18.2875 |
0.8502 | 96.0 | 3360 | 1.1436 | 5.6655 | 18.2875 |
0.8502 | 97.0 | 3395 | 1.1426 | 5.6655 | 18.2875 |
0.8502 | 98.0 | 3430 | 1.1428 | 5.6655 | 18.2875 |
0.8502 | 99.0 | 3465 | 1.1422 | 5.6655 | 18.3375 |
0.819 | 100.0 | 3500 | 1.1424 | 5.6655 | 18.3375 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.0
- Datasets 2.10.1
- Tokenizers 0.13.2