synpre_set_1M_token_1000_t5-small
This model is a fine-tuned version of t5-small on the tyzhu/synpre_set_1M_token_1000 dataset. It achieves the following results on the evaluation set:
- Loss: 0.0218
- Bleu: 99.3901
- Gen Len: 56.5391
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 |
|---|---|---|---|---|---|
| 6.6752 | 0.64 | 5000 | 6.5130 | 0.021 | 33.1488 |
| 5.6329 | 1.28 | 10000 | 5.3242 | 0.2865 | 41.2393 |
| 4.0493 | 1.92 | 15000 | 3.2439 | 5.9293 | 39.8638 |
| 0.1673 | 2.56 | 20000 | 0.0218 | 99.3901 | 56.5391 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.1.0+cu121
- Datasets 2.14.5
- Tokenizers 0.14.1