<!-- 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-qg-context-dataset-2-hard-medium
This model is a fine-tuned version of Deigant/t5-base-finetuned-qg-context-dataset-2 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.1877
- Rouge1: 27.9067
- Rouge2: 6.8779
- Rougel: 24.6502
- Rougelsum: 24.7749
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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 73 | 2.1134 | 27.571 | 8.3183 | 25.3973 | 25.2743 |
| No log | 2.0 | 146 | 2.0800 | 28.4972 | 9.7451 | 26.9093 | 26.7337 |
| No log | 3.0 | 219 | 2.0406 | 21.4309 | 5.817 | 19.4819 | 19.8555 |
| No log | 4.0 | 292 | 2.0391 | 27.2786 | 8.283 | 24.3314 | 24.3751 |
| No log | 5.0 | 365 | 2.0367 | 26.3524 | 7.6263 | 23.9034 | 23.8929 |
| No log | 6.0 | 438 | 2.0270 | 26.3718 | 6.7074 | 22.995 | 23.0177 |
| 1.3439 | 7.0 | 511 | 2.0106 | 27.8601 | 10.5485 | 26.8103 | 26.4962 |
| 1.3439 | 8.0 | 584 | 2.0292 | 27.1811 | 7.1941 | 23.9117 | 24.0093 |
| 1.3439 | 9.0 | 657 | 2.0462 | 25.6595 | 8.3529 | 23.0955 | 23.1946 |
| 1.3439 | 10.0 | 730 | 2.0600 | 27.1996 | 9.0098 | 25.7921 | 25.8295 |
| 1.3439 | 11.0 | 803 | 2.0754 | 25.3094 | 7.6857 | 23.5524 | 23.6875 |
| 1.3439 | 12.0 | 876 | 2.0532 | 27.2136 | 9.0147 | 24.7405 | 24.8211 |
| 1.3439 | 13.0 | 949 | 2.0742 | 26.298 | 8.6826 | 24.6878 | 24.9118 |
| 0.8957 | 14.0 | 1022 | 2.0975 | 22.9575 | 4.2021 | 20.6208 | 20.6539 |
| 0.8957 | 15.0 | 1095 | 2.0941 | 26.778 | 7.1756 | 24.4053 | 24.4951 |
| 0.8957 | 16.0 | 1168 | 2.1025 | 28.9102 | 10.5549 | 25.912 | 25.9433 |
| 0.8957 | 17.0 | 1241 | 2.1265 | 27.8301 | 9.7377 | 25.3236 | 25.3889 |
| 0.8957 | 18.0 | 1314 | 2.1403 | 26.1619 | 7.8019 | 23.5346 | 23.351 |
| 0.8957 | 19.0 | 1387 | 2.1396 | 26.664 | 6.8261 | 24.2991 | 24.328 |
| 0.8957 | 20.0 | 1460 | 2.1481 | 29.8898 | 9.8211 | 27.0922 | 27.2485 |
| 0.69 | 21.0 | 1533 | 2.1466 | 26.3418 | 5.7845 | 24.0772 | 24.3122 |
| 0.69 | 22.0 | 1606 | 2.1559 | 27.5789 | 7.7653 | 25.9896 | 25.8088 |
| 0.69 | 23.0 | 1679 | 2.1624 | 27.9455 | 7.4094 | 25.3163 | 25.3905 |
| 0.69 | 24.0 | 1752 | 2.1633 | 27.5236 | 8.1967 | 24.9498 | 24.974 |
| 0.69 | 25.0 | 1825 | 2.1698 | 26.899 | 6.4382 | 24.2075 | 24.1523 |
| 0.69 | 26.0 | 1898 | 2.1745 | 28.7721 | 8.872 | 24.8299 | 24.9028 |
| 0.69 | 27.0 | 1971 | 2.1818 | 25.8046 | 6.0655 | 23.156 | 23.1971 |
| 0.5965 | 28.0 | 2044 | 2.1854 | 25.4431 | 4.6566 | 22.2794 | 22.4561 |
| 0.5965 | 29.0 | 2117 | 2.1858 | 24.7881 | 6.4357 | 22.8869 | 22.8331 |
| 0.5965 | 30.0 | 2190 | 2.1877 | 27.9067 | 6.8779 | 24.6502 | 24.7749 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
