<!-- 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-small-finetuned-19jan-9
This model is a fine-tuned version of google/mt5-small on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.6105
- Rouge1: 7.7
- Rouge2: 0.1667
- Rougel: 7.5759
- Rougelsum: 7.6113
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: 3e-05
- train_batch_size: 12
- eval_batch_size: 12
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 60
Training results
Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
---|---|---|---|---|---|---|---|
16.5659 | 1.0 | 50 | 5.8214 | 2.2745 | 0.2338 | 2.2506 | 2.2815 |
10.2143 | 2.0 | 100 | 3.8680 | 4.4303 | 0.7671 | 4.4208 | 4.483 |
7.4492 | 3.0 | 150 | 3.2448 | 3.6533 | 0.6857 | 3.6861 | 3.6791 |
5.8239 | 4.0 | 200 | 3.0981 | 5.6287 | 0.8679 | 5.642 | 5.627 |
4.9377 | 5.0 | 250 | 3.0326 | 6.1068 | 1.1621 | 5.9651 | 6.0361 |
4.4824 | 6.0 | 300 | 2.9802 | 6.6496 | 1.3443 | 6.5332 | 6.5395 |
4.2193 | 7.0 | 350 | 2.9484 | 6.0845 | 1.2364 | 6.1077 | 6.0827 |
4.0234 | 8.0 | 400 | 2.9076 | 6.0958 | 1.3299 | 6.0806 | 6.0413 |
3.9046 | 9.0 | 450 | 2.8460 | 5.6462 | 1.1644 | 5.6397 | 5.6103 |
3.8087 | 10.0 | 500 | 2.8036 | 5.7538 | 1.1644 | 5.774 | 5.7442 |
3.6872 | 11.0 | 550 | 2.7727 | 6.5993 | 1.3311 | 6.5773 | 6.6049 |
3.6338 | 12.0 | 600 | 2.7285 | 6.0417 | 1.0778 | 6.047 | 6.089 |
3.574 | 13.0 | 650 | 2.7132 | 8.7833 | 0.25 | 8.803 | 8.6985 |
3.548 | 14.0 | 700 | 2.7023 | 8.9393 | 0.75 | 8.9619 | 8.8679 |
3.49 | 15.0 | 750 | 2.6943 | 9.1778 | 1.0 | 9.1537 | 9.0722 |
3.4098 | 16.0 | 800 | 2.6856 | 8.9167 | 0.75 | 8.9477 | 8.8597 |
3.3776 | 17.0 | 850 | 2.6827 | 8.3503 | 0.1667 | 8.3179 | 8.2614 |
3.3493 | 18.0 | 900 | 2.6899 | 8.6983 | 0.4524 | 8.6503 | 8.602 |
3.3309 | 19.0 | 950 | 2.6833 | 8.2433 | 0.4524 | 8.1185 | 8.1429 |
3.2833 | 20.0 | 1000 | 2.6785 | 8.2194 | 0.4524 | 8.106 | 8.1227 |
3.2491 | 21.0 | 1050 | 2.6817 | 8.2194 | 0.4524 | 8.106 | 8.1227 |
3.22 | 22.0 | 1100 | 2.6697 | 8.3829 | 0.4524 | 8.2852 | 8.3167 |
3.2433 | 23.0 | 1150 | 2.6522 | 8.2194 | 0.4524 | 8.106 | 8.1227 |
3.1882 | 24.0 | 1200 | 2.6493 | 8.2194 | 0.4524 | 8.106 | 8.1227 |
3.1622 | 25.0 | 1250 | 2.6630 | 8.3593 | 0.4524 | 8.2859 | 8.3167 |
3.1396 | 26.0 | 1300 | 2.6523 | 8.3593 | 0.4524 | 8.2859 | 8.3167 |
3.121 | 27.0 | 1350 | 2.6565 | 8.3593 | 0.4524 | 8.2859 | 8.3167 |
3.1095 | 28.0 | 1400 | 2.6385 | 8.5833 | 0.4524 | 8.45 | 8.516 |
3.1113 | 29.0 | 1450 | 2.6378 | 7.6135 | 0.3333 | 7.5385 | 7.5885 |
3.0661 | 30.0 | 1500 | 2.6415 | 8.2734 | 0.3333 | 8.1583 | 8.25 |
3.0316 | 31.0 | 1550 | 2.6435 | 7.6135 | 0.3333 | 7.5385 | 7.5885 |
3.0468 | 32.0 | 1600 | 2.6342 | 7.6135 | 0.3333 | 7.5385 | 7.5885 |
3.0323 | 33.0 | 1650 | 2.6330 | 7.8333 | 0.4167 | 7.7551 | 7.8317 |
3.0031 | 34.0 | 1700 | 2.6332 | 8.1192 | 0.4167 | 8.0718 | 8.1167 |
2.9904 | 35.0 | 1750 | 2.6291 | 8.2734 | 0.3333 | 8.1583 | 8.25 |
2.9765 | 36.0 | 1800 | 2.6364 | 7.8667 | 0.4167 | 7.8167 | 7.8269 |
2.9872 | 37.0 | 1850 | 2.6267 | 7.9984 | 0.4167 | 7.875 | 7.9843 |
2.976 | 38.0 | 1900 | 2.6252 | 7.9984 | 0.4167 | 7.875 | 7.9843 |
2.9528 | 39.0 | 1950 | 2.6319 | 7.701 | 0.3333 | 7.7167 | 7.6769 |
2.9385 | 40.0 | 2000 | 2.6279 | 7.8667 | 0.4167 | 7.8167 | 7.8269 |
2.9371 | 41.0 | 2050 | 2.6227 | 7.4658 | 0.4167 | 7.4167 | 7.4397 |
2.9214 | 42.0 | 2100 | 2.6172 | 8.1355 | 0.4167 | 8.0537 | 8.1329 |
2.9472 | 43.0 | 2150 | 2.6133 | 8.1355 | 0.4167 | 8.0537 | 8.1329 |
2.9215 | 44.0 | 2200 | 2.6101 | 7.4516 | 0.1667 | 7.3718 | 7.3647 |
2.9188 | 45.0 | 2250 | 2.6097 | 7.4516 | 0.1667 | 7.3718 | 7.3647 |
2.9003 | 46.0 | 2300 | 2.6089 | 7.4516 | 0.1667 | 7.3718 | 7.3647 |
2.8926 | 47.0 | 2350 | 2.6137 | 7.7769 | 0.1667 | 7.6692 | 7.7272 |
2.8872 | 48.0 | 2400 | 2.6118 | 7.7769 | 0.1667 | 7.6692 | 7.7272 |
2.8809 | 49.0 | 2450 | 2.6089 | 7.247 | 0.1667 | 7.151 | 7.1897 |
2.8676 | 50.0 | 2500 | 2.6027 | 7.2881 | 0.1667 | 7.1551 | 7.1947 |
2.8792 | 51.0 | 2550 | 2.6131 | 7.1382 | 0.1667 | 7.0703 | 7.0476 |
2.8705 | 52.0 | 2600 | 2.6144 | 7.7 | 0.1667 | 7.5759 | 7.6113 |
2.8887 | 53.0 | 2650 | 2.6130 | 7.7 | 0.1667 | 7.5759 | 7.6113 |
2.872 | 54.0 | 2700 | 2.6080 | 7.7 | 0.1667 | 7.5759 | 7.6113 |
2.8593 | 55.0 | 2750 | 2.6093 | 7.2881 | 0.1667 | 7.1551 | 7.1947 |
2.868 | 56.0 | 2800 | 2.6091 | 7.8387 | 0.1667 | 7.6729 | 7.7334 |
2.8729 | 57.0 | 2850 | 2.6096 | 7.8387 | 0.1667 | 7.6729 | 7.7334 |
2.8526 | 58.0 | 2900 | 2.6100 | 7.1382 | 0.1667 | 7.0703 | 7.0476 |
2.8671 | 59.0 | 2950 | 2.6105 | 7.7 | 0.1667 | 7.5759 | 7.6113 |
2.8544 | 60.0 | 3000 | 2.6105 | 7.7 | 0.1667 | 7.5759 | 7.6113 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.1+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2