<!-- 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. -->
synpre_set_1M_token_3000_t5-small
This model is a fine-tuned version of t5-small on the tyzhu/synpre_set_1M_token_3000 dataset. It achieves the following results on the evaluation set:
- Loss: 0.0128
- Bleu: 99.5308
- Gen Len: 58.2779
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: 0.0001
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 10000
- training_steps: 80000
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 7.7 | 0.64 | 5000 | 7.5576 | 0.013 | 36.8765 |
| 6.7262 | 1.28 | 10000 | 6.4589 | 0.1615 | 27.7415 |
| 5.8711 | 1.92 | 15000 | 5.2029 | 2.6094 | 37.903 |
| 4.2308 | 2.56 | 20000 | 3.1763 | 9.3873 | 40.2111 |
| 0.4897 | 3.2 | 25000 | 0.0550 | 98.3671 | 58.3008 |
| 0.0911 | 3.84 | 30000 | 0.0290 | 99.5062 | 58.3274 |
| 0.0473 | 4.48 | 35000 | 0.0192 | 99.6581 | 58.289 |
| 0.0329 | 5.12 | 40000 | 0.0163 | 99.6899 | 58.2881 |
| 0.0237 | 5.76 | 45000 | 0.0138 | 99.7129 | 58.2863 |
| 0.02 | 6.4 | 50000 | 0.0123 | 99.6935 | 58.2544 |
| 0.0175 | 7.04 | 55000 | 0.0116 | 99.7395 | 58.2788 |
| 0.0152 | 7.68 | 60000 | 0.0107 | 99.746 | 58.2793 |
| 0.0153 | 8.32 | 65000 | 0.0096 | 99.7601 | 58.2851 |
| 0.0149 | 8.96 | 70000 | 0.0093 | 99.7478 | 58.2843 |
| 0.0123 | 9.6 | 75000 | 0.0086 | 99.7795 | 58.2769 |
| 0.0109 | 10.24 | 80000 | 0.0128 | 99.5308 | 58.2779 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.1.0+cu121
- Datasets 2.14.5
- Tokenizers 0.14.1
