t5-small-entailement-Writer-T5-small
This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5628
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 |
|---|---|---|---|
| No log | 1.0 | 83 | 1.2943 |
| No log | 2.0 | 166 | 0.9323 |
| No log | 3.0 | 249 | 0.8443 |
| No log | 4.0 | 332 | 0.7884 |
| No log | 5.0 | 415 | 0.7582 |
| No log | 6.0 | 498 | 0.7355 |
| 1.2761 | 7.0 | 581 | 0.7178 |
| 1.2761 | 8.0 | 664 | 0.7105 |
| 1.2761 | 9.0 | 747 | 0.6972 |
| 1.2761 | 10.0 | 830 | 0.6847 |
| 1.2761 | 11.0 | 913 | 0.6774 |
| 1.2761 | 12.0 | 996 | 0.6708 |
| 0.7765 | 13.0 | 1079 | 0.6609 |
| 0.7765 | 14.0 | 1162 | 0.6566 |
| 0.7765 | 15.0 | 1245 | 0.6507 |
| 0.7765 | 16.0 | 1328 | 0.6454 |
| 0.7765 | 17.0 | 1411 | 0.6438 |
| 0.7765 | 18.0 | 1494 | 0.6384 |
| 0.693 | 19.0 | 1577 | 0.6347 |
| 0.693 | 20.0 | 1660 | 0.6321 |
| 0.693 | 21.0 | 1743 | 0.6254 |
| 0.693 | 22.0 | 1826 | 0.6237 |
| 0.693 | 23.0 | 1909 | 0.6215 |
| 0.693 | 24.0 | 1992 | 0.6167 |
| 0.6504 | 25.0 | 2075 | 0.6167 |
| 0.6504 | 26.0 | 2158 | 0.6131 |
| 0.6504 | 27.0 | 2241 | 0.6120 |
| 0.6504 | 28.0 | 2324 | 0.6091 |
| 0.6504 | 29.0 | 2407 | 0.6076 |
| 0.6504 | 30.0 | 2490 | 0.6058 |
| 0.615 | 31.0 | 2573 | 0.6031 |
| 0.615 | 32.0 | 2656 | 0.6015 |
| 0.615 | 33.0 | 2739 | 0.6015 |
| 0.615 | 34.0 | 2822 | 0.6000 |
| 0.615 | 35.0 | 2905 | 0.5998 |
| 0.615 | 36.0 | 2988 | 0.5969 |
| 0.586 | 37.0 | 3071 | 0.5959 |
| 0.586 | 38.0 | 3154 | 0.5941 |
| 0.586 | 39.0 | 3237 | 0.5923 |
| 0.586 | 40.0 | 3320 | 0.5936 |
| 0.586 | 41.0 | 3403 | 0.5929 |
| 0.586 | 42.0 | 3486 | 0.5922 |
| 0.5618 | 43.0 | 3569 | 0.5910 |
| 0.5618 | 44.0 | 3652 | 0.5885 |
| 0.5618 | 45.0 | 3735 | 0.5879 |
| 0.5618 | 46.0 | 3818 | 0.5873 |
| 0.5618 | 47.0 | 3901 | 0.5877 |
| 0.5618 | 48.0 | 3984 | 0.5878 |
| 0.5418 | 49.0 | 4067 | 0.5881 |
| 0.5418 | 50.0 | 4150 | 0.5858 |
| 0.5418 | 51.0 | 4233 | 0.5847 |
| 0.5418 | 52.0 | 4316 | 0.5839 |
| 0.5418 | 53.0 | 4399 | 0.5843 |
| 0.5418 | 54.0 | 4482 | 0.5826 |
| 0.5283 | 55.0 | 4565 | 0.5843 |
| 0.5283 | 56.0 | 4648 | 0.5833 |
| 0.5283 | 57.0 | 4731 | 0.5825 |
| 0.5283 | 58.0 | 4814 | 0.5827 |
| 0.5283 | 59.0 | 4897 | 0.5830 |
| 0.5283 | 60.0 | 4980 | 0.5806 |
| 0.5135 | 61.0 | 5063 | 0.5808 |
| 0.5135 | 62.0 | 5146 | 0.5806 |
| 0.5135 | 63.0 | 5229 | 0.5807 |
| 0.5135 | 64.0 | 5312 | 0.5823 |
| 0.5135 | 65.0 | 5395 | 0.5801 |
| 0.5135 | 66.0 | 5478 | 0.5799 |
| 0.5053 | 67.0 | 5561 | 0.5808 |
| 0.5053 | 68.0 | 5644 | 0.5796 |
| 0.5053 | 69.0 | 5727 | 0.5793 |
| 0.5053 | 70.0 | 5810 | 0.5785 |
| 0.5053 | 71.0 | 5893 | 0.5790 |
| 0.5053 | 72.0 | 5976 | 0.5775 |
| 0.4985 | 73.0 | 6059 | 0.5770 |
| 0.4985 | 74.0 | 6142 | 0.5777 |
| 0.4985 | 75.0 | 6225 | 0.5780 |
| 0.4985 | 76.0 | 6308 | 0.5779 |
| 0.4985 | 77.0 | 6391 | 0.5782 |
| 0.4985 | 78.0 | 6474 | 0.5773 |
| 0.4889 | 79.0 | 6557 | 0.5787 |
| 0.4889 | 80.0 | 6640 | 0.5787 |
| 0.4889 | 81.0 | 6723 | 0.5773 |
| 0.4889 | 82.0 | 6806 | 0.5777 |
| 0.4889 | 83.0 | 6889 | 0.5759 |
| 0.4889 | 84.0 | 6972 | 0.5765 |
| 0.4806 | 85.0 | 7055 | 0.5758 |
| 0.4806 | 86.0 | 7138 | 0.5760 |
| 0.4806 | 87.0 | 7221 | 0.5758 |
| 0.4806 | 88.0 | 7304 | 0.5760 |
| 0.4806 | 89.0 | 7387 | 0.5759 |
| 0.4806 | 90.0 | 7470 | 0.5758 |
| 0.4817 | 91.0 | 7553 | 0.5753 |
| 0.4817 | 92.0 | 7636 | 0.5757 |
| 0.4817 | 93.0 | 7719 | 0.5754 |
| 0.4817 | 94.0 | 7802 | 0.5750 |
| 0.4817 | 95.0 | 7885 | 0.5753 |
| 0.4817 | 96.0 | 7968 | 0.5752 |
| 0.4767 | 97.0 | 8051 | 0.5754 |
| 0.4767 | 98.0 | 8134 | 0.5756 |
| 0.4767 | 99.0 | 8217 | 0.5755 |
| 0.4767 | 100.0 | 8300 | 0.5755 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2